identifier
stringlengths 11
32
| pdf_url
stringlengths 17
4.62k
⌀ | lang
stringclasses 120
values | error
stringclasses 1
value | title
stringlengths 2
500
⌀ | source_name
stringlengths 1
435
⌀ | publication_year
float64 1.9k
2.02k
| license
stringclasses 3
values | word_count
int64 0
1.64M
| text
stringlengths 1
9.75M
|
|---|---|---|---|---|---|---|---|---|---|
https://openalex.org/W2517567155
|
http://pratsi.opu.ua/app/webroot/articles/1476296064.pdf
|
Russian
| null |
Modelling and optimization of seawater desalination process using mechanical steam compression
|
Trudy Odesskogo politehničeskogo universiteta
| 2,016
|
cc-by
| 6,325
|
MODELLING AND OPTIMIZATION OF SEAWATER
DESALINATION PROCESS USING MECHANICAL
VAPOUR COMPRESSION В.П. Кравченко, С.В. Сурков, Хуссам Ганем. Моделювання й оптимізація процесу опріснення морської води в установ-
ках з механічним стисненням пари. В умовах глобальних змін клімату нестача прісної води стає актуальною проблемою для все
більшої кількості країн. Однією з найбільш перспективних технологій опріснення морської води є механічне стиснення пари
(МСП), що забезпечує низьке енергоспоживання завдяки використанню принципу теплового насоса. Мета: Метою роботи є вияв-
лення резервів підвищення ефективності опріснювальних систем, заснованих на механічному стисненні пари, шляхом оптимізації
схеми і параметрів установок з МСП. Матеріали і методи: Запропоновано новий тип опріснювальної установки, головним
елементом якої є теплообмінник прихованої теплоти. Морська вода після попереднього нагрівання в теплообмінниках надходить у
випарник-конденсатор, де отримує основну кількість теплоти від пари, що конденсується. Частина морської води випаровується, а
концентрований розчин солі (розсіл) виходить з випарника і після охолодження скидається назад в море. Пара, що утворюється,
стискається за допомогою компресора і надходить у конденсатор. Суттєвою особливістю даної схеми є те, що конденсація відбува-
ється при вищих температурах, ніж випаровування. Завдяки цьому теплота, що виділяється при конденсації пари, використовується
для випаровування морської води. Результати: Було вирішено наступні питання: модифікована і доповнена математична модель
установок з МСП, модифіковано схему включення теплообмінників, досліджено вплив конструктивних параметрів установки на
вартість обладнання і електроенергії. Аналіз схем установки і математичної моделі дозволив визначити шляхи зниження енерго-
витрат. Проаналізовано вплив двох основних параметрів – питомої потужності компресора й питомої площі поверхні випарника-
конденсатора – на величину приведених витрат опріснювальної установки. Визначено оптимальне співвідношення цих параметрів. Ключові слова: опріснення морської води, механічне стиснення пари, чисельне моделювання. V.P. Kravchenko, S.V. Surkov, Hussam Ghanem. Modelling and optimization of seawater desalination process using mechanical
vapour compression. In the conditions of global climate changes shortage of fresh water becomes an urgent problem for an increasing
number of the countries. One of the most perspective technologies of a desalting of sea water is the mechanical vapour compression (MVC)
providing low energy consumption due to the principle of a heat pump. Aim: The aim of this research is to identify the reserves of efficiency
increasing of the desalination systems based on mechanical vapour compression by optimization of the scheme and parameters of installa-
tions with MVC. Materials and Methods: The new type of desalination installation is offered which main element is the heat exchanger of
the latent heat. Sea water after preliminary heating in heat exchangers comes to the evaporator-condenser where receives the main amount of
heat from the condensed steam. ISSN 2076-2429 (print)
ISSN 2223-3814 (online) ISSN 2076-2429 (print)
ISSN 2223-3814 (online) 47 Odes’kyi Politechnichnyi Universytet. Pratsi, Issue 2(49), 2016 MODELLING AND OPTIMIZATION OF SEAWATER
DESALINATION PROCESS USING MECHANICAL
VAPOUR COMPRESSION Effectiveness of thermal
energy transfer and the its subsequent transformation into a mechanical energy is defined by purity of
surfaces of metal contacting with water and steam. According to technical requirements, the quantity
of permeates in working water must be no more than 10 μg/l [2]. The required quality of water can be reached only by means of technologies of a thermal desalt-
ing (distillation) which have several modifications. Multistage installations of instantaneous flashing
(MSF) are mainly used in large stationary installations. In the presence of the industrial vapor sources
the installations with thermal vapour compression (TVC) are optimum. Far from the industrial vapour
sources and in mobile installations the technology of a desalting with mechanical vapour compression
(MVC) has the greatest advantages. In this research this technology is chosen for the detailed analysis. In the works of El-Dessouky [3] and Al-Juwayhel et al. [4] the mathematical models of systems
with MVC are developed. Further researches were devoted to decrease in prime cost of the water
freshened with MVC use. In particular, Lara [5, 6] investigated the prospects of use of high
performance gerotor compressors and hydrophobic evaporators-condensers, and Cherkasskiy [7]
investigated high-temperature working modes. The aim of this research is to identify the reserves of efficiency increasing of the desalination
systems based on mechanical vapour compression by optimization of the scheme and parameters of
installations with MVC. To achieve the goal, it is necessary to solve the following problems: To achieve the goal, it is necessary to solve the following problems: to modify and complete the mathematical model of installations with MVC; increase the convergence of iterative methods and the reliability the obtained results;
to modify the scheme of heat exchangers plug-in; increase the convergence of iterative methods and the reliability the obtained results; to research the influence of design parameters of desalination installation on the cost of a
ntory and the electric power. Compressor Brine
Distillate
Seawater
Latent heat
exchangers
Evaporator
Compressor
Condenser
Vapour
Fig. 1. The scheme of
desalination installation
with mechanical vapour
compression Materials and Methods. The scheme of the offered desalination instal-
lation is submitted in Fig. 1. Condenser
Vapour Evaporator Condenser The main element of this installation is a uniform block of the
evaporator-condenser, or the latent heat exchanger. MODELLING AND OPTIMIZATION OF SEAWATER
DESALINATION PROCESS USING MECHANICAL
VAPOUR COMPRESSION A part of sea water evaporates, and the strong solution of salt (brine) goes out of the evaporator, and after
cooling is dumped back in the sea. The formed steam is compressed by the compressor and comes to the condenser. An essential singularity
of this scheme is that condensation happens at higher temperature, than evaporation. Thanks to this the heat, which is comes out at devapora-
tion, is used for evaporation of sea water. Thereby, in this class of desalination installations the principle of a heat pump is implemented. Results: For achievement of a goal the following tasks were solved: the mathematical model of installations with MVC is modified and
supplemented; the scheme of heat exchangers switching is modified; influence of design data of desalination installation on the cost of an
inventory and the electric power is investigated. The detailed analysis of the main schemes of installation and mathematical model allowed
defining ways of decrease in energy consumption and the possible merit value. Influence of two key parameters a specific power of the
compressor and a specific surface area of the evaporator-condenser on a value of given expenses of desalination installation is analyzed. The optimum ratio of these parameters is defined. p
p
Keywords: desalting of sea water, mechanical vapour compression, numerical modeling. Introduction. Now the need for fresh water, both for domestic needs and for the production
enterprises, grows in many countries of the world. Especially this problem is particularly acute in the DOI 10.15276/opu.2.49.2016.08 DOI 10.15276/opu.2.49.2016.08 2016 The Authors. This is an open access article under the CC BY license (http://creativecommons.org/licenses/by/4.0/). 2016 The Authors. This is an open access article under the CC BY license (http://creativecommons.org/licenses/by/4.0/). ENERGETICS. HEAT ENGINEERING. ELECTRICAL ENGINEERING ISSN 2076-2429 (print)
ISSN 2223-3814 (online) 48 Праці Одеського політехнічного університету, 2016. Вип. 2(49) Праці Одеського політехнічного університету, 2016. Вип. 2(49)
ISSN 2076-2429 (print)
ISSN 2223-3814 (online) countries of the Arabian Peninsula and the Persian Gulf. As of 2008 the Persian Gulf countries were
producing 58% of the bulk world volume of desalinated water. Global climate changes and population
growth do a problem of fresh water more and more urgent for the majority of regions of the Globe [1]. Especially high requirements for the consumed water are placed by heat power production. On
nuclear power plants water is used as a working medium and as the coolant. MODELLING AND OPTIMIZATION OF SEAWATER
DESALINATION PROCESS USING MECHANICAL
VAPOUR COMPRESSION Seawater after preliminary heating in heat exchangers comes to the
evaporator-condenser where receives the main heat amount from the
condensed vapour. A part of seawater evaporates, and the strong solution of
salt (brine) goes out of the evaporator, and after cooling is dumped back in
the sea. The formed vapour is compressed by the compressor and comes to
the condenser. The T(s) chart of process is schematically shown in Fig. 2. Here, the
piece 1–2 shows the process of evaporation of seawater; 2–3 compression
of vapour in the compressor which can be considered as adiabatic; 3–4
isobaric cooling, 4–5 devaporation. Latent heat
exchangers Essential singularity of this scheme is that condensation takes place at
higher temperature, than evaporation. Due to this the heat evolved at
devaporation is used for evaporation of seawater. Thereby in this class of
desalination installations the principle of a heat pump is implemented. Fig. 1. The scheme of
desalination installation
with mechanical vapour
compression ЕНЕРГЕТИКА. ТЕПЛОТЕХНІКА. ЕЛЕКТРОТЕХНІКА 49 Odes’kyi Politechnichnyi Universytet. Pratsi, Issue 2(49), 2016 Т, С
s, J/(kgK)
0
1000
2000
3000
4000
5000
6000
7000
60
65
70
75
Fig. 2. The T(s) chart of desalting process with mechanical vapour compression Fig. 2. The T(s) chart of desalting process with mechanical vapour compression Now let’s write down a weight conservation law for water and for salt respectively: (1)
(2) (1) f
d
b
G
G
G
,
(1)
f
f
b
b
G X
G X
,
(2) (2) X salt concentration. Here indexes f, d and b belong to the feeding ocean water, the distillate and a brine dumped in the
sea respectively. p
y
The energy balance for the device in general is calculated according to the equation p
y
The energy balance for the device in general is calculated according to the equation
h G
WG
h G
h G
(3) cw
f
d
d
d
b
b
h G
WG
h G
h G
,
(3) (3) (3) h enthalpy,
W
specific energy transferred to the steam flow from the compressor where h enthalpy, W specific energy, transferred to the steam flow from the compressor. Using mean specific heat
pc , the equation (3) can be rewritten as W specific energy, transferred to the steam flow from the compressor. MODELLING AND OPTIMIZATION OF SEAWATER
DESALINATION PROCESS USING MECHANICAL
VAPOUR COMPRESSION Using mean specific heat
pc , the equation (3) can be rewritten as (
)
(
)
d
b
d
p
out
cw
WG
G
G
c T
T
, where
out
T
temperature of the output streams of a brine and distillate, cw
T temperature of the input seawater. From here it is possible to obtain cw
T temperature of the input se
From here it is possible to obtain (
)
d
out
cw
p
b
d
WG
T
T
c
G
G
. (4) (4) For a numerical example is set that a mass output of distillate is 1 kg/s. nt of heat needed to evaporate the second mass of vapour is calculated using the formula The amount of heat needed to evaporate the second mass of vapour is calculated using the formula 2
d b
Q
G r
, (5) 2
d b
Q
G r
, 2
d b
Q
G r
,
(5)
where br latent heat of vaporization at boiling point of a brine. where br latent heat of vaporization at boiling point of a brine. where br latent heat of vaporization at boiling point of a brine. ENERGETICS. HEAT ENGINEERING. ELECTRICAL ENGINEERING 50 Праці Одеського політехнічного університету, 2016. Вип. 2(49)
ISSN 2076-2429 (print)
ISSN 2223-3814 (online) The quantity of heat
2
Q has to be transmitted to the boiling brine through evaporator-condenser
walls. The quantity of heat
2
Q has to be transmitted to the boiling brine through evaporator-condenser
walls. e temperature of vapour which is formed during salty water boiling can be calculated as The temperature of vapour which is formed during salty water boiling can be cal v
b
T
T
BPE
, e BPE boiling point elevation, which can be calculated according to [3, 4]. The saturation pressure in an evaporator corresponding to temperature
vT is calculated by th
ula (
)
v
s
v
p
p T
. For further calculations it is necessary to set first approximation for temperature
dT at which the
vapour in the condenser is condensed. Further value of
dT is defined by method of simple iterations For further calculations it is necessary to set first approximation for temperature
dT at which the
vapour in the condenser is condensed. MODELLING AND OPTIMIZATION OF SEAWATER
DESALINATION PROCESS USING MECHANICAL
VAPOUR COMPRESSION Further value of
dT is defined by method of simple iterations
The compressor has to create pressure in the condenser that would be equal to saturated vapor For further calculations it is necessary to set first approximation for temperature
dT at which the
vapour in the condenser is condensed. Further value of
dT is defined by method of simple iterations
The compressor has to create pressure in the condenser that would be equal to saturated vapor
pressure at temperature
dT , i.e. vapour in the condenser is condensed. Further value of
dT is defined by method of simple iterations
The compressor has to create pressure in the condenser that would be equal to saturated vapor
pressure at temperature
dT , i.e. (
)
d
s
d
p
p T
. (
)
d
s
d
p
p T
. Vapor pressures on the saturation line
(
)
s
d
p T
and
(
)
s
b
p T
are also calculated according to the
ratios given in [4]. Estimation of minimum necessary intensity of vapour compression in the compressor is made
according to the equation (
)
(
)
s
d
s
b
Cr
p T
p T
. (6) (6) The specific power consumed by the compressor electric motor according to [7] is calculated
using formula 1
1
1
v
v
d
c
v
p v
p
W
p
,
(7) (7) where
vv specific volume of a saturated vapour at a temperature
vT ; e
vv specific volume of a saturated vapour at a temperature
vT ; where
vv specific volume of a saturated vapour at a temperature
vT γ heat capacity ratio; γ heat capacity ratio; η compressor efficiency. η compressor efficiency. Apparently, the following ratio is fair in adiabatic process: Apparently, the following ratio is fair in adiabatic process: Apparently, the following ratio is fair in adiabatic process: 1
s
d
v
v
T
p
T
p
. MODELLING AND OPTIMIZATION OF SEAWATER
DESALINATION PROCESS USING MECHANICAL
VAPOUR COMPRESSION (8) (8) From here it is possible to evaluate the temperature of compressed vapor From here it is possible to evaluate the temperature of compressed vapor From here it is possible to evaluate the temperature of compressed vapor 1
d
s
v
v
p
T
T
p
. The heat amount allocated during devaporation process is calculated according to the ratio
1
[
(
)]
d
d
p
s
d
Q
G r
c T
T
, The heat amount allocated during devaporation process is calculated according to the ratio (9) 1
[
(
)]
d
d
p
s
d
Q
G r
c T
T
,
(9) e
dr latent heat of vaporization at condensation temperature. latent heat of vaporization at condensation temperature. where
dr latent heat of vaporization at condensation temperature. As calculations show, this power, as a rule, is more, than
2
Q because of the “excess” values of
sT
arising at adiabatic vapour compression in the compressor. Excess of power
1
2
(
)
Q
Q
in practice
leads to boiling temperature
bT rise, and together with it the rise of all other specific temperatures. ЕНЕРГЕТИКА. ТЕПЛОТЕХНІКА. ЕЛЕКТРОТЕХНІКА ISSN 2076-2429 (print)
ISSN 2223-3814 (online) 51 Odes’kyi Politechnichnyi Universytet. Pratsi, Issue 2(49), 2016 Odes’kyi Politechnichnyi Universytet. Pratsi, Issue 2(49), 2016 However, in this work authors assume that excess of power is compensated by a thermolysis to
ambient medium, and process can be considered as conservative one. However, in this work authors assume that excess of power is compensated by a thermolysis to
ambient medium, and process can be considered as conservative one. p
The heat transfer equation can be written as The heat transfer equation can be written as The heat transfer equation can be written as (10) 2
(
)
e
e
d
b
Q
U A T
T
,
(10) where
e
U total heat transfer coefficient of the evaporator-condenser; e
e
U total heat transfer coefficient of the evaporator-condenser; total heat transfer coefficient of the evaporator-condenser; eA area of surface-heat transfer of evaporator-condenser. eA area of surface-heat transfer of evaporator-condenser. MODELLING AND OPTIMIZATION OF SEAWATER
DESALINATION PROCESS USING MECHANICAL
VAPOUR COMPRESSION At predefined values of
eA and
e
U the condensation temperature id determined uniquely 1
d
b
e
e
Q
T
T
U A
,
(11) (11) but, because
e
U depends on temperature [3], it is necessary the iterative refinement of
dT . The loop body containing the equations (6…11) repeats until change of
dT value becomes less
than the preset small value. Calculations show that the iterative cycle converges quickly. Results and Discussion. All calculations were performed for a hypothetical machine which mass
yield of distillate is 1 kg/s. Thus, all values of power and areas can be considered as appanage. The
heat transfer coefficient was calculated by the integral relations given in [3, 4], and average value of it
is 2.45 kW/(m2·K). )
In fig. 3 the dependence of power and excess of power of the compressor on a heat exchange
surface area is shown. From Fig. 3 it is evident that the necessary compressor power significantly decreases at increase
of a heat exchange surface area of evaporator-condenser. The optimum ratio between these parameters
can be found based on the economic calculation. Specific energy consumption on production of one cubic meter of desalinated water is bound to a
specific power with the ratio 3.6
W
N
, 3.6
W
N
, e W specific consumption of the electric power, (kW·h)/m3; where W specific consumption of the electric power, (kW·h)/m3; N specific power, kW/kg. N specific power, kW/kg. In Fig. 4 the dependence of specific power consumption of W on temperature of seawater boiling
at two various values of heat-transfer coefficient in evaporator-condenser is shown. N, kW
Ae, m2
400
0
350
300
250
10
20
30
40
50
W,
kWh/m3
Tb, C
100
0
90
80
70
60
50
2
4
6
8
10
Fig. 3. Dependences of necessary power
of the compressor («––») and excess of a thermal
power («---») on a surface area
of heat exchange of evaporator-condenser
Fig. 4. Dependence of specific power consumption on
temperature of seawater boiling at two various values
of heat-transfer coefficient in evaporator-condenser:
«—»
2,5
e
U
kW/(m2·K);
«---»
5,0
e
U
kW/(m2·K) In Fig. 4 the dependence of specific power consumption of W on temperature of seawater boiling
at two various values of heat-transfer coefficient in evaporator-condenser is shown. and substitute in (13). (
)
When all temperatures at input and output for each heat exchanger are determined, it is possible
to calculate the necessary areas of surfaces of heat exchange. We will calculate log mean temperature
pressures according to the following ratios: g
(
)
(
)
(
)
(
)
,
. ln
ln
lm
lm
d
b
b
x
b
x
out
cw
d
b
d
b
b
x
b
x
out
cw
T
T
T
T
T
T
T
T
T
T
T
T
T
T
T
T
T
T
(14) (14) The necessary surfaces areas of heat exchange are calculated by the formulas:
Q The necessary surfaces areas of heat exchange are calculated by the formulas:
Q d
d
lm
d
d
Q
A
U
T
,
(15)
b
b
lm
b
b
Q
A
U
T
. (16) (15) (16) The equations (12…16) give the only solution in whole studied range of temperatures and
provide heating of seawater to temperature
bT . In Fig. 6 the dependence of the necessary surfaces areas of heat exchange on boiling temperature
is shown
bT . Tb
Gb
Tb
Td
Gd
Tout
Tout
Gf
Tcw
Tx
Tb
cp1
cp4
cp3
cp2
cp5 Tb
Gb
Tb
Td
Gd
Tout
Tout
Gf
Tcw
Tx
Tb
cp1
cp4
cp3
cp2
cp5
Fig. 5. The scheme of plug-in of heat exchangers of
seawater preliminary heating with the indication of
temperatures and mass expenses of water Tb
Gb
Tb
Td
Gd
Tout
Tout
Gf
Tcw
Tx
Tb
cp1
cp4
cp3
cp2
cp5
A, m2
Tb, C
100
90
80
70
60
60
50
40
30
20
10
0
50
Fig. 5. The scheme of plug-in of heat exchangers of
seawater preliminary heating with the indication of
temperatures and mass expenses of water
Fig. 6. The necessary surfaces areas of heat exchange
for preliminary heaters of seawater:
«—»
bA ; «---»
d
A A, m2
Tb, C
100
90
80
70
60
60
50
40
30
20
10
0
50
Fig. 6. The necessary surfaces areas of heat exchange
for preliminary heaters of seawater:
«—»
bA ; «---»
d
A A, m2
Tb, C
100
90
80
70
60
60
50
40
30
20
10
0
50 Tb Fig. 5. MODELLING AND OPTIMIZATION OF SEAWATER
DESALINATION PROCESS USING MECHANICAL
VAPOUR COMPRESSION N, kW
W, evaporator-condenser is shown. W,
kWh/m3
Tb, C
100
0
90
80
70
60
50
2
4
6
8
10
Fig. 4. Dependence of specific power consumption on
temperature of seawater boiling at two various values
of heat-transfer coefficient in evaporator-condenser:
«—»
2,5
e
U
kW/(m2·K);
«---»
5,0
e
U
kW/(m2·K) W,
kWh/m3
Tb, C
100
0
90
80
70
60
50
2
4
6
8
10 N, kW
Ae, m2
400
0
350
300
250
10
20
30
40
50 Fig. 4. Dependence of specific power consumption on
temperature of seawater boiling at two various values
of heat-transfer coefficient in evaporator-condenser:
«—»
2,5
e
U
kW/(m2·K);
«---»
5,0
e
U
kW/(m2·K) Fig. 3. Dependences of necessary power
of the compressor («––») and excess of a thermal
power («---») on a surface area
of heat exchange of evaporator-condenser ENERGETICS. HEAT ENGINEERING. ELECTRICAL ENGINEERING 52 Праці Одеського політехнічного університету, 2016. Вип. 2(49)
ISSN 2076-2429 (print)
ISSN 2223-3814 (online) From Fig. 4 it is evident that at increasing of boiling temperature the specific energy consumption
and the surplus of power lost are simultaneously decrease. However, at increasing of boiling tempera-
ture above 70 ºС the speed of adjournment of a scum on device walls sharply increases that leads to
growth of operational expenses. p
p
stem of preliminary water heating consisting of two heat exchangers is offered. g
p
p
The system of preliminary water heating consisting of two heat exchangers is offered. MODELLING AND OPTIMIZATION OF SEAWATER
DESALINATION PROCESS USING MECHANICAL
VAPOUR COMPRESSION h
b l
f
h
ll
h
h
ili i
h
f di ill
b The energy balance for the smaller heat exchanger utilizing heat of distillate can be calculated
from the equation 1
2
(
)
(
)
d
d
p
d
b
f
p
b
x
Q
G c
T
T
G c
T
T
;
(12) (12) 1
2
(
)
(
)
d
d
p
d
b
f
p
b
x
Q
G c
T
T
G c
T
T
; in turn, for the second heat exchanger utilizing heat of a brine and distillate can be calculated from
the equation second heat exchanger utilizing heat of a brine and distillate can be calculated from 3
4
5
(
)
(
)
(
)
b
b
p
b
out
d
p
b
out
cw
p
x
cw
Q
G c
T
T
G c
T
T
G c
T
T
. (13) (13) The
xT we evaluate of (12) 1
2
(
)
d
p
x
b
d
b
f
p
G c
T
T
T
T
G c
and substitute in (13). and substitute in (13). where
.
el
C annual power expenses; . clean
C
costs of water purification and transfer of salts to a soluble state for prevention of
salts loss on the heating surfaces (we accept equal 0.3 USD/m3 of initial water); . clean
C
costs of water purification and transfer of salts to a soluble state for prevention of
salts loss on the heating surfaces (we accept equal 0.3 USD/m3 of initial water); . clean
C
costs of water purification and transfer of salts to a soluble state for prevention of
salts loss on the heating surfaces (we accept equal 0.3 USD/m3 of initial water); . am
C
annual costs for depreciation of capital investments and maintenance estimated at
10...20 % of depreciation expense; . am
C
annual costs for depreciation of capital investments and maintenance estimated at
10...20 % of depreciation expense; exp .l
C
operating expenses. exp .l
C
operating expenses. Annual energy costs are calculated by the formula: Annual energy costs are calculated by the formula: Annual energy costs are calculated by the formula: (19) . . . el
el en
el
C
A
P
,
(19) . . . el
el en
el
C
A
P
, el
P national electricity cost (for UAE
0.04
el
P
USD/(kW·h)); where
el
P national electricity cost (for UAE
0.04
el
P
USD/(kW·h)); where
el
P national electricity cost (for UAE
0.04
el
P
USD/(kW·h)); . . el en
A
annual energy consumption, kWh. Annual costs on depreciation of capital investment and maintenance calculated by the formula . . el en
A
annual energy consumption, kWh. Annual costs on depreciation of capital investment and maintenance calculated by the formula Annual costs on depreciation of capital investment and maintenance calculated by the formula . 0.1
1.1
am
a
c
m
a
c
a
c
a
c
C
p C
C
p C
p C
p C
,
(20) (20) ap depreciation rate (assumed equal to 0.06). e
ap depreciation rate (assumed equal to 0.06). where
ap depreciation rate (assumed equal to 0.06). m
C maintenance costs. Maintenance costs are determined according to the ratio exp . . . l
lab
oth
C
C
C
, where
. lab
C
annual salary of the staff; ISSN 2076-2429 (print)
ISSN 2223-3814 (online) 53 Odes’kyi Politechnichnyi Universytet. Pratsi, Issue 2(49), 2016 From the plot it is visible that the areas of surfaces
d
A for preliminary heat exchangers are 10
times less than the areas of surfaces of evaporator-condenser
b
A . The area
b
A is greatly rises at rise of
boiling temperature
bT . Therefore, in economic calculation the main attention needs to be paid to the
cost of evaporator-condenser. Criterion of technical and economic optimization. As criterion of comparison the annual given
costs are accepted [8, 9]: (17) n
c
op
C
e C
C
,
(17) where
ne normative rate of return (assumed equal to 0.12); e
ne normative rate of return (assumed equal to 0.12); where
ne normative rate of return (assumed equal to 0.12); where
ne normative rate of return (assumed equal to 0.12); Cc capitalized expenses; Cc capitalized expenses; Cc capitalized expenses;
C
annual expenditure op
C
annual expenditure. op
C
annual expenditure. nual expenditure
op
C
may be calculated by the formula Annual expenditure
op
C
may be calculated by the formula Annual expenditure
op
C
may be calculated by the formula ture
op
C
may be calculated by the formula (18) . . . exp . op
el
clean
am
l
C
C
C
C
C
,
(18) where
. el
C annual power expenses; where
. el
C annual power expenses; and substitute in (13). The scheme of plug-in of heat exchangers of
seawater preliminary heating with the indication of
temperatures and mass expenses of water Fig. 6. The necessary surfaces areas of heat exchange
for preliminary heaters of seawater:
«—»
bA ; «---»
d
A ЕНЕРГЕТИКА. ТЕПЛОТЕХНІКА. ЕЛЕКТРОТЕХНІКА where
.
lab
C
annual salary of the staff; . oth
C
other manufacturing expenses. The value of manufacturing expenses
. oth
C
is 20…30 % total costs for depreciation, maintenance
works and wages, i.e. operating costs are in accordance with the ratio: exp . . . . 0.2(
0.1
)
1.2
0.22
l
lab
a
c
a
c
lab
lab
a
c
C
C
p C
p C
C
C
p C
. (21) exp . . . . 0.2(
0.1
)
1.2
0.22
l
lab
a
c
a
c
lab
lab
a
c
C
C
p C
p C
C
C
p C
. (21) (21) Substituting the values of all components in the expression (18), we obtain the formula for annual
operating costs: . . . . . 1.1
1.2
0.22
1.32
1.2
op
el
clean
a
c
lab
a
c
el
a
c
lab
C
C
C
p C
C
p C
C
p C
C
. (22) (22) The expression for the given expenses can be written as
. . . . . . . . . 1.32
1.2
1.32
1.2
0.1992
1.2
. n
c
el
clean
a
c
lab
c
n
a
el
clean
lab
c
el
clean
lab
C
e C
C
C
p C
C
C
e
p
C
C
C
C
C
C
C
(23) (23) ENERGETICS. HEAT ENGINEERING. ELECTRICAL ENGINEERING 54 Праці Одеського політехнічного університету, 2016. Вип. 2(49)
ISSN 2076-2429 (print)
ISSN 2223-3814 (online Then, the final expression for the variable part of given expenses takes the form: Then, the final expression for the variable part of given expenses takes the form: (24) . 0.1992
c
el
C
C
C
. (24) . 0.1992
c
el
C
C
C
. The needs for the electric power are determined under condition of installation operating within
22 hours a day and 360 days within a year. During the choice of the equipment was accepted that
productivity of installation makes 25 m3 per day. p
y
p
y
Capital expenditure consists, mainly, of the cost of evaporator-condenser, preliminary heat
exchangers and the compressor. where
.
lab
C
annual salary of the staff; As evaporator-condenser and preliminary heaters of seawater it is expedient to use the lamellar
heat exchangers differing in high coefficient of a heat transfer. The analysis of cost of heat exchangers
shows that in approximate calculation it is possible to accept their cost of directly proportional to
surface area of heat exchange. For typical in the studied range heat exchanger Ridan NN 41 the surface area of heat exchange is
217.35 m2 at the price of 4806 USD [10]. Then the total cost of heat exchangers can be estimated
using the formula . . 22.11(
)
h e
e
b
d
C
A
A
A
. Compressor cost
compr. C
can be averagely set to proportional power of electric engine. According
to [11],
645
compr. c
C
N
. Compressor cost
compr. C
can be averagely set to proportional power of electric engine. According
to [11],
645
compr. c
C
N
. Total capital expenditure will be Total capital expenditure will be . . . h e
compr
C
C
. Ae, m2
C,
USD/year
900
0
800
700
600
500
400
2000
4000
6000
8000
10000
12000
14000
16000
Fig. 7. The given annual costs as function
of the surface area of heat exchange
of evaporator-condenser:
«—» total annual costs;
«---» reduced capital costs;
«– • –» the annual cost of electricity . . . c
h e
compr
C
C
C
. . . . c
h e
compr
C
C
C
. The given annual costs (17) have been
calculated taking into account the power of the
compressor and the areas of heat exchange
surfaces obtained higher. Results of calculation
of the specified cost for installation with a
productivity of 1 kg/s of desalinated water are
presented in Fig. 7. The analysis of the plots shows that the
minimum of the given capital expenditure is
reached at the specific area of evaporator-
condenser of 570 m2. The electricity cost
continuously drops at increase of surface of
heat exchange of evaporator-condenser
eA . Total annual costs decrease up to value
1000
eA
m2. But at such great areas the heat
exchangers have the large weight and
dimensions and therefore
eA needs to be
limited accordingly to transportation cost. Fig. 7. ISSN 2076-2429 (print)
ISSN 2223-3814 (online) ISSN 2076-2429 (print)
ISSN 2223-3814 (online) 55 Odes’kyi Politechnichnyi Universytet. Pratsi, Issue 2(49), 2016 Література 1. Escobar, I.C. Sustainable water for the future: Water recycling versus desalination / I.C. Escobar,
A.I. Schäfer. — Amsterdam; Boston: Elsevier Science, 2010. — 416 p. 1. Escobar, I.C. Sustainable water for the future: Water recycling versus desalination / I.C. Escobar,
A.I. Schäfer. — Amsterdam; Boston: Elsevier Science, 2010. — 416 p. p
2. Копылов, А.С. Водоподготовка в энергетике / А.С. Копылов, В.М. Лавыгин, В.Ф. Очков. — 2-е
изд., стер. — М.: Издат. дом МЭИ, 2006. — 309 с. р
3. El-Dessouky, H.T. Fundamentals of salt water desalination / H.T. El-Dessouky, H.M. Ettouney. — Am-
sterdam: Elsevier Science, 2002. — 670 p. 4. Al-Juwayhel, F. Analysis of single-effect evaporator desalination systems combined with vapor
compression heat pumps / F. Al-Juwayhel, H. El-Dessouky, H. Ettouney // Desalination. — 1997. —
Vol. 114, Issue 3. — PP. 253–275. 5. Lara, J.R. Advanced mechanical vapor-compression desalination system / J.R. Lara, Omorinsola
Osunsan, M.T. Holtzapple // Desalination, Trends and Technologies / ed. by M. Schorr. — Rijeka:
InTech, 2011. — PP. 129–148. 6. Lara, J.R. An investigation of high operating temperatures in mechanical vapor-compression desalina-
tion / J.R. Lara, G. Noyes, M.T. Holtzapple // Desalination. — 2008. — Vol. 227, Issues 1–3. —
PP. 217–232. 7. Черкасский, В.М. Насосы, вентиляторы, компрессоры / В.М. Черкасский. — М.: Энергоатом
издат, 1984. — 416 с. 8. Беренс, В. Руководство по подготовке промышленных технико-экономических исследований:
монография / В. Беренс, П.М. Хавранек; пер. с англ. А.О. Гридин и др.; науч.ред. Р.П. Вчерашний и др. - новое перераб. и доп. изд. — М.: Интерэксперт, 1995. — 343 с. р
р
р р
р
р
9. Preparation of a Feasibility Study for New Nuclear Power Projects / International Atomic Energy
Agency. — Vienna: IAEA, 2014. — 125 p. 10. Типовые цены на Теплообменники пластинчатые [Електронний ресурс] / ТеплоПрофи:
Комплексные поставки инженерного оборудования по России и СНГ. — Режим доступу:
http://www.teploprofi.com/ceni/ (Дата звернення: 25.05.2016). p
p p
(Д
р
)
11. Оборудование
[Електронний
ресурс]
/
ERSTEVAK
Ltd. —
Режим
доступу:
http://www.erstvak.com/equipment/ (Дата звернення: 24.05.2016). where
.
lab
C
annual salary of the staff; The given annual costs as function
of the surface area of heat exchange Conclusions. Numerical modeling has
shown that in systems with adiabatic compres-
sion of steam there are losses of power which are 26…29 % of the general expenses of energy. These
losses can be reduced when using advanced processes of compression of vapour along the line of
saturation, for example at injection of water in the compressor. « »
the annual cost of electricity Decrease in prime cost of desalinated water can be reached mainly due to increase of efficiency
of the compressor and increase of a heat transfer coefficient of evaporator-condenser. With the given productivity of desalination installation, the optimum parameters of the compres-
sor (from the economic point of view) and evaporator-condenser can be found. ЕНЕРГЕТИКА. ТЕПЛОТЕХНІКА. ЕЛЕКТРОТЕХНІКА References 1. Escobar, I.C., & Schäfer, A.I. (2010). Sustainable Water for the Future: Water Recycling Vers
Desalination. Amsterdam; Boston: Elsevier Science. 2. Kopylov, A.S., Lavygin, V.M., & Ochkov, V.F. (2006). Water Treatment in Power Engineering (2nd
Ed.). Moscow: MEI. 3. El-Dessouky, H.T., & Ettouney, H.M. (2002). Fundamentals of Salt Water Desalination. Amsterdam:
Elsevier Science. 4. Al-Juwayhel, F., El-Dessouky, H., & Ettouney, H. (1997). Analysis of single-effect evaporator desalina-
tion systems combined with vapor compression heat pumps. Desalination, 114(3), 253-275. DOI:10.1016/S0011-9164(98)00017-4 5. Lara, J.R., Omorinsola Osunsan, Holtzapple, M.T. (2011). Advanced mechanical vapor-compression
desalination system. In M. Schorr (Ed.), Desalination, Trends and Technologies (pp. 129–148). Rijeka:
InTech. DOI:10.5772/14711 6. Lara, J.R., Noyes, G., & Holtzapple, M.T. (2008). An investigation of high operating temperatures in
mechanical
vapor-compression
desalination. Desalination,
227(1–3),
217–232. DOI:10.1016/j.desal.2007.06.027 j
7. Cherkassky, V.M. (1985). Pumps, Fans, Compressors. Moscow: Mir. 7. Cherkassky, V.M. (1985). Pumps, Fans, Compressors. Moscow: Mir. 8. Behrens, W., & Hawranek, P.M. (1991). Manual for the Preparation of Industrial Feasibility Studie
Vienna: UNIDO. 9. International Atomic Energy Agency. (2014). Preparation of a Feasibility Study for New Nuclear
Power Projects. Vienna: IAEA. 10. TeploProfi. (n.d.). Prices for Lamella Heat-Exchangers. Retrieved from http://www.teploprofi.com/ceni/ 10. TeploProfi. (n.d.). Prices for Lamella Heat-Exchangers. Retrieved from http://www.teploprofi.com/ceni/
11. ERSTEVAK Ltd. (n.d.). Equipment. Retrieved from http://www.erstvak.com/equipment/ p
(
)
f
g
p
p p
11. ERSTEVAK Ltd. (n.d.). Equipment. Retrieved from http://www.erstvak.com/equipment/ 11. ERSTEVAK Ltd. (n.d.). Equipment. Retrieved from http://www.erstvak.com/equipment/ Received May 20, 2016
Accepted July 7, 2016 ENERGETICS. HEAT ENGINEERING. ELECTRICAL ENGINEERING
|
https://openalex.org/W2605182438
|
https://europepmc.org/articles/pmc5367420?pdf=render
|
English
| null |
Transmission of Hepatitis A Virus through Combined Liver–Small Intestine–Pancreas Transplantation
|
Emerging infectious diseases
| 2,017
|
cc-by
| 4,974
|
SYNOPSIS SYNOPSIS CME Editor Karen L. Foster, Technical Writer/Editor, Emerging Infectious Diseases. Disclosure: Karen L. Foster has disclosed no
relevant financial relationships. This activity has been planned and implemented through the joint providership of Medscape, LLC and
Emerging Infectious Diseases. Medscape, LLC is accredited by the American Nurses Credentialing Center (ANCC), the
Accreditation Council for Pharmacy Education (ACPE), and the Accreditation Council for Continuing Medical Education
(ACCME), to provide continuing education for the healthcare team. (
)
g
Medscape, LLC designates this Journal-based CME activity for a maximum of 1.00 AMA
Credit(s)™. Physicians should claim only the credit commensurate with the extent of their particip Medscape, LLC designates this Journal-based CME activity for a maximum of 1.00 AMA PRA Category 1
Credit(s)™. Physicians should claim only the credit commensurate with the extent of their participation in the activity. All other clinicians completing this activity will be issued a certificate of participation. To participate in this
journal CME activity: (1) review the learning objectives and author disclosures; (2) study the education content; (3) take
the post-test with a 75% minimum passing score and complete the evaluation at http://www.medscape.org/journal/eid;
and (4) view/print certificate. For CME questions, see page 729. ( )
y
y
p
p
y
All other clinicians completing this activity will be issued a certificate of participation. To participate in this
journal CME activity: (1) review the learning objectives and author disclosures; (2) study the education content; (3) take
the post-test with a 75% minimum passing score and complete the evaluation at http://www.medscape.org/journal/eid;
and (4) view/print certificate. For CME questions, see page 729. Release date: March 15, 2017; Expiration date: March 15, 2018 Learning Objectives Upon completion of this activity, participants will be able to: Upon completion of this activity, participants will be able to: Analyze the epidemiology of hepatitis A virus (H y
gy
(
)
•
Evaluate how organ donors are assessed for HAV infection •
Evaluate laboratory results of the organ recipient and caregivers exposed to HAV infec
current study. CME Author Charles P. Vega, MD, Clinical Professor of Family Medicine, University of California, Irvine. Disclosure: Charles P. Vega, MD, has disclosed the following financial relationships: served as an advisor or consultant for Allergan, Inc.;
McNeil Consumer Healthcare; served as a speaker or a member of a speakers bureau for Shire Pharmaceuticals. Transmission of Hepatitis A Virus
through Combined Liver–Small
Intestine–Pancreas Transplantation Monique A. Foster, Lauren M. Weil, Sherry Jin, Thomas Johnson, Tonya R. Hayden-Mixson,
Yury Khudyakov, Pallavi D. Annambhotla, Sridhar V. Basavaraju, Saleem Kamili,
Jana M. Ritter, Noele Nelson, George Mazariegos, Michael Green, Ryan W. Himes,
David T. Kuhar, Matthew J. Kuehnert, Jeffrey A. Miller, Rachel Wiseman, Anne C. Moorman Authors Disclosures: Monique A. Foster, MD, MPH; Lauren M. Weil, PhD, MPH; Sherry Jin, MD, MPH; Thomas Johnson,
BS; Tonya Mixson-Hayden, PhD; Yury Khudyakov, PhD; Pallavi D. Annambhotla, DrPH; Sridhar V. Basavaraju,
MD; Saleem Kamili, PhD; Jana M. Ritter, DVM; Noele P. Nelson, MD, PhD; George Mazariegos, MD; David T. Kuhar, MD; Matthew J. Kuehnert, MD; Jeffrey A. Miller, MD, MPH; Rachel Wiseman, MPH; and Anne C. Moorman, RN, MPH, have disclosed no relevant financial relationships. Michael Green, MD, has disclosed the
following relevant financial relationships: served as an advisor or consultant for Chimerix, Bristol-Myers Squibb. Ryan
W. Himes, MD, has disclosed the following relevant financial relationships: served as an advisor or consultant for
Alexion; served as a speaker or a member of a speakers bureau for Alexion; received grants for clinical research
from Alexion. 590
Emerging Infectious Diseases • www.cdc.gov/eid • Vol. 23, No.4, April 2017
Author affiliations: Centers for Disease Control and Prevention,
Atlanta, Georgia, USA (M.A. Foster, T.R. Hayden-Mixson,
Y. Khudyakov, P.D. Annambhotla, S.V. Basavaraju, S. Kamili,
J.M. Ritter, N. Nelson, D.T. Kuhar, M.J. Kuehnert, J.A. Miller,
A.C. Moorman); Texas Department of State Health Services,
Austin, Texas, USA (L.M. Weil); Harris County Public Health and
Environmental Services, Houston, Texas, USA (S. Jin); Houston
Health Department, Houston (T. Johnson); Children’s Hospital of
Pittsburgh of University of Pittsburgh Medical Center, Pittsburgh,
Pennsylvania, USA (G. Mazariegos, M. Green); Texas Children’s
Hospital, Houston, (R.W. Himes); Pennsylvania Department of
Health, Harrisburg, Pennsylvania, USA (J.A. Miller); Texas
Department of State Health Services, Austin (R. Wiseman)
DOI: http://dx.doi.org/10.3201/eid2304.161532 Health Department, Houston (T. Johnson); Children’s Hospital of
Pittsburgh of University of Pittsburgh Medical Center, Pittsburgh,
Pennsylvania, USA (G. Mazariegos, M. Green); Texas Children’s
Hospital, Houston, (R.W. Himes); Pennsylvania Department of
Health, Harrisburg, Pennsylvania, USA (J.A. Miller); Texas
Department of State Health Services, Austin (R. Wiseman)
DOI: http://dx.doi.org/10.3201/eid2304.161532 Author affiliations: Centers for Disease Control and Prevention,
Atlanta, Georgia, USA (M.A. Foster, T.R. Hayden-Mixson,
Y. Khudyakov, P.D. Annambhotla, S.V. Basavaraju, S. Kamili,
J.M. Ritter, N. Nelson, D.T. Kuhar, M.J. Kuehnert, J.A. Miller,
A.C. Moorman); Texas Department of State Health Services,
Austin, Texas, USA (L.M. Weil); Harris County Public Health and
Environmental Services, Houston, Texas, USA (S. Jin); Houston Author affiliations: Centers for Disease Control and Prevention, Author affiliations: Centers for Disease Control and Prevention,
Atlanta, Georgia, USA (M.A. Foster, T.R. Hayden-Mixson,
Y. Khudyakov, P.D. Annambhotla, S.V. Basavaraju, S. Kamili,
J.M. Ritter, N. Nelson, D.T. Kuhar, M.J. Kuehnert, J.A. Emerging Infectious Diseases • www.cdc.gov/eid • Vol. 23, No.4, April 2017 Transmission of HAV through Transplantation Transmission of HAV through Transplantation Although transmission of hepatitis A virus (HAV) through
blood transfusion has been documented, transmission
through organ transplantation has not been reported. In Au
gust 2015, state health officials in Texas, USA, were noti
fied of 2 home health nurses with HAV infection whose only
common exposure was a child who had undergone multi–
visceral organ transplantation 9 months earlier. Specimens
from the nurses, organ donor, and all organ recipients were
tested and medical records reviewed to determine a pos
sible infection source. Identical HAV RNA sequences were
detected from the serum of both nurses and the organ do
nor, as well as from the multi–visceral organ recipient’s se
rum and feces; this recipient’s posttransplant liver and intes
tine biopsy specimens also had detectable virus. The other
organ recipients tested negative for HAV RNA. Vaccination
of the donor might have prevented infection in the recipient
and subsequent transmission to the healthcare workers. home healthcare agency. The nurses did not have shared
exposures other than 1 patient and had not been vaccinated
for hepatitis A. The shared patient was a 7-year-old who underwent
multiorgan (liver, small bowel, and pancreas) transplanta
tion because of megacystis microcolon intestinal hypoperi
stalsis syndrome, a rare congenital condition characterized
by a largely dilated nonobstructed urinary bladder (mega
cystis), very small colon (microcolon), and decreased or
absent intestinal movements (intestinal peristalsis). The
transplantation occurred in December 2014. This patient’s posttransplant course was complicated
by intraabdominal abscesses, acute liver rejection, Epstein-
Barr virus enteritis, cytomegalovirus infection, and lympho
proliferative disorder. The patient was discharged to home
in March 2015 with an alanine aminotransferase (ALT)
level of 49 IU/L (reference 0–50 IU/L). By the following
month, ALT had increased to 324 IU/L, and by June 2015,
to 515 IU/L. During the time of ALT increase, the patient,
who had a colostomy and ileostomy, also had increased sto
ma output. Because of worsening clinical symptoms, a liver
biopsy was conducted on June 19, 2015; results showed
features of moderate acute cellular rejection, as well as dif
fuse, lobular acute and chronic inflammation. On June 30,
a repeat liver biopsy was performed because of persistent
ALT elevation despite increased treatment for rejection. H
epatitis A virus (HAV), the most common cause of
viral hepatitis, is a nonenveloped RNA virus belong
ing to the family Picornaviridae (1,2). Transmission of HAV through Transplantation Approximately 1.5
million clinical cases of HAV occur worldwide annually;
the virus is commonly spread person to person through
the fecal–oral route (2). Although rates of HAV infection
have declined by 95% in the United States since a vaccine
became available, infections continue to result from close
personal contact with an infected household member or
common-source outbreaks from contaminated food or wa
ter (3). HAV can cause relapsing and fulminant hepatitis,
but fatal infection is rare. H The multi–visceral organ recipient had received 2
doses of hepatitis A vaccine, as part of routine childhood
vaccinations, and was IgG HAV positive in 2013, indi
cating prior immunity against HAV, although the patient
was immune suppressed after transplantation. Because of
the other clinical conditions that could have explained the
ALT elevation and gastrointestinal symptoms, HAV infec
tion was initially not considered. When subsequent testing
prompted by the infections in the recipient’s caregivers re
vealed the multi–visceral organ recipient was positive for
HAV RNA, a laboratory and epidemiologic investigation
focused on whether infection was due to recent fecal–oral
transmission or from solid-organ transplantation. Available
transplant-related specimens were tested to rule out trans
mission through transplantation. Parenteral transmission of HAV through contaminated
blood products or needles is also rare, despite the presence
of viremia up to 30 days before symptom onset (4). No
screening tests for HAV infection are required for blood,
organ, or tissue donation in the United States (5). HAV
transmission through solid-organ transplantation has not
been reported in the literature. In August 2015, genetically identical HAV was recov
ered from 2 healthcare workers (HCWs) participating in
the care of a child recipient of multiple visceral organs. To
prevent infection in other organ recipients and contacts,
the Centers for Disease Control and Prevention (CDC),
along with state and local health departments, conducted
an investigation to determine the source of the HCW infec
tion and whether HAV was transmitted through the solid-
organ transplantation. Clinical and Epidemiologic Review The case was reported to the Organ Procurement and
Transplantation Network and reviewed by the Network’s
ad hoc Disease Transmission Advisory Committee as a
suspected donor-derived disease event. Through repre
sentation on advisory committee, CDC, with support from
state and local health departments, investigates potential
transmission to other organ recipients. To determine
whether the multi–visceral organ recipient acquired HAV Authors Miller,
A.C. Moorman); Texas Department of State Health Services,
Austin, Texas, USA (L.M. Weil); Harris County Public Health and
Environmental Services, Houston, Texas, USA (S. Jin); Houston Austin, Texas, USA (L.M. Weil); Harris County Public Health and
Environmental Services, Houston, Texas, USA (S. Jin); Houston Emerging Infectious Diseases • www.cdc.gov/eid • Vol. 23, No.4, April 2017 590 Emerging Infectious Diseases • www.cdc.gov/eid • Vol. 23, No.4, April 2017 Results of Laboratory Investigation y
g
PCR results for HAV RNA of serum specimens from the 2
home health nurses and the multi–visceral organ recipient
are shown in Table 1. The third HCW had recovered by the
time the outbreak was identified and had no available speci
men from when she was symptomatic. The specimens from
the 2 nurses and the organ recipient had detectable HAV
RNA with sequences genetically identical to those of other
isolates in the CDC HAV isolate database, thus confirm
ing the multi–visceral organ recipient as the source of the
HCW infections (Figure 2). Retrospective testing of banked liver biopsy tissue
from the multi–visceral organ recipient showed no detect
able HAV RNA in native liver but detectable HAV RNA
in all subsequent samples, as early as 5 days after trans
plantation. Sequential serum and fecal specimens from
the multi–visceral organ recipient were positive for HAV
RNA through January 2016; follow-up serum specimens
drawn monthly during March–May 2016 and a fecal speci
men from May 2016 showed clearance of HAV in serum
but persistent low-level viremia in feces (Table 2). Frozen
serum specimens from the 8-year-old organ donor also
were positive for HAV RNA (Table 1). Serum specimens
collected from the heart and kidney recipients ≈10 months
after transplant tested negative for HAV RNA at CDC and
negative for HAV IgM at their sites of clinical care (Table
1). These recipients also had no clinical signs or symptoms
consistent with HAV infection. Results of Epidemiologic Investigation Case investigations revealed that, during their incubation
and infectious periods, the home health nurses cared for a
total of 12 children. However, they had only 1 patient in
common: the multi–visceral organ recipient (8). An additional HCW with HAV infection was identi
fied during a case-contact interview with the mother of the
multi–visceral organ recipient. This nurse also provided
care to the multi–visceral organ recipient, and jaundice,
diarrhea, and arthralgia later developed that required in
patient admission. HAV infection was diagnosed by se
rologic testing during that hospitalization; this nurse also
had not received a hepatitis A vaccination. Care for the
multi–visceral organ recipient provided by all 3 nurses in
cluded managing watery feces (e.g., changing diapers and
ostomy bags). Epidemiologic investigation and the result
ing timeline (Figure 1) provide evidence that all 3 nurses
most likely were infected by exposure to the multi–visceral
organ recipient. Case Report In August 2015, the Texas Department of State Health
Services received requests from 2 separate local health de
partments for postexposure prophylaxis recommendations
for contacts of 2 nurses with confirmed HAV infection. In
both nurses, symptomatic infections developed within days
of each other; both nurses worked for the same pediatric 591 Emerging Infectious Diseases • www.cdc.gov/eid • Vol. 23, No.4, April 2017 SYNOPSIS The organ donor, an 8-year-old who died in a mo
tor vehicle collision, traveled to Guatemala, a country to
which HAV is endemic, 6 months before death. In addi
tion to the visceral organs, which were transplanted into
the index patient, the heart and both kidneys were trans
planted into 3 other recipients. The organ donor’s name
was found in a vaccination registry maintained by 1 of
the states of residence, which indicated that the hepatitis
A vaccination was not given. Serum banked from dona
tion was negative for evidence of antibody to hepatitis B
surface antigen, further indicating that the donor might
not have completed recommended childhood vaccinations
against any viral hepatitis. infection through transplantation and to identify other po
tentially infected HCWs, public health investigators re
viewed medical records, interviewed household contacts,
and conducted additional case finding among healthcare
providers and other transplant recipients. This investi
gation included review of records from the home health
agency that employed both home health nurses; consulta
tion with occupational health staff from the treating facili
ties; and review of surveillance data from the jurisdictions
in which the patient received care. Laboratory Specimen Collection and Testing y
p
g
Laboratory testing was conducted at the laboratory of
CDC’s Division of Viral Hepatitis, National Center for
HIV, Viral Hepatitis, STD, and TB Prevention (Atlanta,
GA, USA). IgG HAV, IgM HAV, and HAV RNA ex
traction of serum was conducted on both HCWs; the or
gan donor; and recipients of the visceral organs, heart,
and kidneys. When HAV RNA was detected from serum
samples, PCR and phylogenetic analysis was conducted
to determine relatedness. Similar testing was conducted
on frozen and paraffin-embedded liver and small bowel
tissue biopsy specimens from the multi–visceral organ
recipient. Methods of viral extraction from paraffin-
embedded tissue have been described previously (6). All HAV RNA–positive samples were used to sequence
the HAV VP1/P2B (viral protein 1/amino terminus of
2B) genomic region, and phylogenetic analysis was per
formed by comparing these sequences with archived
HAV sequences contained within the CDC HAV se
quence database (7). Discussion The isolation of genetically identical HAV RNA sequences
from the multi–visceral organ recipient and the organ do
nor indicates HAV infection was transmitted through organ
transplantation, even though the index patient had earlier
evidence of immunity. The 3 HCWs associated with this
investigation most likely were infected by the traditional
fecal–oral route. The infectious period of the multi–viscer
al organ recipient is also among the longest documented in
a person infected with HAV, typically an acute disease (9). 592 Emerging Infectious Diseases • www.cdc.gov/eid • Vol. 23, No.4, April 2017 Transmission of HAV through Transplantation ure 1. Polygenetic analysis of HAV isolates within the Centers for
ease Control and Prevention’s US HAV database. The genomic
ions are the 315-bp long HAV VP1/P2B (viral protein 1/amino
minus of 2B). Black squares indicate isolates from the outbreak of
V transmitted through a combined liver–small intestine–pancreas
nsplantation, Texas, USA, 2014–2015. Scale bar indicates
cleotide variation. HAV, hepatitis A virus. Figure 1. Polygenetic analysis of HAV isolates within the Centers for
Disease Control and Prevention’s US HAV database. The genomic
regions are the 315-bp long HAV VP1/P2B (viral protein 1/amino
terminus of 2B). Black squares indicate isolates from the outbreak of
HAV transmitted through a combined liver–small intestine–pancreas
transplantation, Texas, USA, 2014–2015. Scale bar indicates
nucleotide variation. HAV, hepatitis A virus. Emerging Infectious Diseases • www.cdc.gov/eid • Vol. 23, No.4, April 2017 Emerging Infectious Diseases • www.cdc.gov/eid • Vol. 23, No.4, April 2017 593 SYNOPSIS Table 1. Laboratory results for case-patients and contacts related to HAV outbreak, Texas, 2015*
Patient or source of specimen
Outcome or status
HAV rRT -PCR
Serologic testing
Vaccination status
IgM
IgG
Donor
Deceased
Detected
Detected
Not detected
Unvaccinated
Multi–visceral organ recipient
Persistent infection
Detected
Detected
Detected
Vaccinated
Heart recipient
Immune
Not detected
Not detected
Detected
Vaccinated
Left kidney recipient
Immune
Not detected
Not detected
Detected
Vaccinated
Right kidney recipient
Immune
Not detected
Not detected
Detected
Vaccinated
Home health nurse A
Recovered
Detected
Detected
Not detected
Unvaccinated
Home health nurse B
Recovered
Detected
Detected
Not detected
Unvaccinated
Inpatient nurse
Recovered
No sample
Detected
Not detected
Unvaccinated
*HAV, hepatitis A virus; rRT-PCR, real-time reverse transcription PCR. Diagnosis in the multi–visceral organ recipient prob
ably was delayed due to deferred testing because of con
current conditions that provided alternative explanations
for the recipient’s clinical presentation and history of prior
HAV immunity. Discussion The patient had history of vaccination but
immune suppression probably blunted antibody response. Symptomatic persons typically show elevated ALT levels
that coincide with onset of clinical illness. The infected or
gan recipient had elevated ALT shortly after liver trans
plantation, but it coincided with other viral infections
(Epstein-Barr virus and cytomegalovirus) and was later as
sumed to be related to possible acute liver rejection. days in immunocompromised persons (11,12). Similar
findings have been published of persistent viral shedding in
feces with median days of detection after symptom onset of
81–127 days (11). The level and length of HAV infection
and fecal shedding make the carrier a potential continuing
infectious source of the virus, which occurred in this case. Because the primary transmission of HAV is fecal–oral,
it is not surprising that diarrhea or fecal incontinence leads to
the spread of infection (13). Nosocomial outbreaks are uncom
mon because hygienic practices are generally adhered to more
consistently when the patient is symptomatic enough to be hos
pitalized (14). The use of contact precautions is recommended
for HCWs caring for patients with HAV who are diapered or
incontinent (15). Because HCWs do not have increased prev
alence of HAV infection and because nosocomial outbreaks
of HAV are rare, hepatitis A vaccination is not mandatory for
HCWs in the United States (16,17). In this case, the multi–
visceral organ recipient’s ileostomy and colostomy output
had increased, but it was difficult to determine whether these
increases represented symptoms of HAV infection because The process that results in prolonged courses of HAV
infection is unknown (10). In contrast to hepatitis B and C
viruses, HAV is not typically associated with a prolonged
infectious carrier state (11). Nonetheless, this case report,
and scant longitudinal studies in the literature, demonstrate
the potential for ongoing transmission. In published longi
tudinal studies, HAV viremia persisted for median periods
of 22–42 days in immunocompetent persons and 256–490 Figure 2. Clinical timeline of HAV infection among a multi–visceral organ transplantation recipient and infected healthcare workers,
Texas, 2014–2015. ALT, alanine aminotransferase; HAV, hepatitis A virus. Figure 2. Clinical timeline of HAV infection among a multi–visceral organ transplantation recipient and infected healthcare workers,
Texas, 2014–2015. ALT, alanine aminotransferase; HAV, hepatitis A virus. Emerging Infectious Diseases • www.cdc.gov/eid • Vol. 23, No.4, April 2017 Emerging Infectious Diseases • www.cdc.gov/eid • Vol. 23, No.4, April 2017 594 Transmission of HAV through Transplantation Table 2. Discussion Pathologic, serum, and fecal HAV RNA quantification test results of multi–visceral organ recipient, Texas, 2014–2015*
Date collected
Liver HAV RNA, IU/mL
Small bowel HAV RNA,
IU/mL
Serum HAV RNA,
IU/mL
Fecal HAV RNA, IU/mL
2014 Dec 18
Native liver, none detected
2014 Dec 23
34,000
2015 Jan 23
1,500,00
2015 Feb 10
7,000
2015 Mar 10
5,400
2015 Apr 16
>90,000,000
2015 Sep 21
>90,000,000
>90,000,000
2015 Sep 29
746,000,000
>90,000,000
2015 Oct 15
1,400,000
372,000
2015 Nov 9
27,900
2015 Nov 16
17,300
2015 Nov 23
13,000
1,500,000
2015 Dec 3
38,900
51,900
2016 Jan 11
68,000
22,600
2016 Mar 9
320
2016 Apr 6
Not detected
270
2016 May 4
Not detected
Not detected
*Blank cells indicate that no specimen was available for testing. HAV, hepatitis A virus. immunosuppression and types of organs transplanted in the
heart and kidney recipients were also factors in preventing
HAV infection after transplantation in these persons. her stoma output was typically described as a continuous liq
uid, even before hepatitis developed. Also, the home health
nurses were spending 12 hours per shift inside the patient’s
home, where they ate meals and shared space with the pa
tient’s family, and thus were considered household contacts,
which increased their risk for infection. Previous studies have
shown that length of contact with an infected patient increas
es the attack rate of HCWs, and postexposure prophylaxis
is recommended for household contacts of infected patients
(14,16). The parents of the organ recipient were tested for
HAV infection before the patient was determined to be the
source. The mother’s test results indicated immunity to HAV,
and the father was not HAV immune but received hepatitis
vaccination at testing. Once the infection was detected, con
tact precautions were instituted during subsequent hospital
izations, the local health department recommended the pa
tient’s home health nurses be fully vaccinated against HAV,
and no further transmissions to HCWs were detected during
the subsequent 8 months of continued viral shedding in the
patient’s feces. p
p
In 2006, the Advisory Committee on Immunization
Practices recommended routine hepatitis A vaccination for
all children beginning at 12–23 months of age (16). Both
inactivated whole-virus vaccines available in the United
States are well tolerated and effective, showing serologic
levels of protection for at least 17 years (21,22). This work was supported by the Division of Viral Hepatitis,
National Center for HIV, Viral Hepatitis, STD, and TB
Prevention, CDC. References 1. Ciocca M. Clinical course and consequences of hepatitis A
infection. Vaccine. 2000;18(Suppl 1):S71–4. http://dx.doi.org/
10.1016/S0264-410X(99)00470-3 2. Franco E, Meleleo C, Serino L, Sorbara D, Zaratti L. Hepatitis
A: epidemiology and prevention in developing countries. World J
Hepatol. 2012;4:68–73. http://dx.doi.org/10.4254/wjh.v4.i3.68 16. Fiore AE, Wasley A, Bell BP. Prevention of hepatitis A through
active or passive immunization: recommendations of the Advisory
Committee on Immunization Practices (ACIP). MMWR Recomm
Rep. 2006;55(RR-7):1–23. 3. Centers for Disease Control and Prevention. Hepatitis A questions
and answers for health professionals [cited 2016 Mar 14]. http://www.cdc.gov/hepatitis/hav/havfaq.htm#general 3. Centers for Disease Control and Prevention. Hepatitis A questions
and answers for health professionals [cited 2016 Mar 14]. http://www.cdc.gov/hepatitis/hav/havfaq.htm#general 4. Bower WA, Nainan OV, Han X, Margolis HS. Duration of viremia
in hepatitis A virus infection. J Infect Dis. 2000;182:12–7. http://dx.doi.org/10.1086/315701 4. Bower WA, Nainan OV, Han X, Margolis HS. Duration of viremia
in hepatitis A virus infection. J Infect Dis. 2000;182:12–7. http://dx.doi.org/10.1086/315701 17. Guturu P, Cicalese L, Duchini A. Hepatitis A vaccination in
healthcare personnel. Ann Hepatol. 2012;11:326–9.f 18. Schaffner A. Pretransplant evaluation for infections in donors and
recipients of solid organs. Clin Infect Dis. 2001;33(Suppl 1):S9–14. http://dx.doi.org/10.1086/320898 5. Hughes JA, Fontaine MJ, Gonzalez CL, Layon AG,
Goodnough LT, Galel SA. Case report of a transfusion-
associated hepatitis A infection. Transfusion. 2014;54:2202–6. http://dx.doi.org/10.1111/trf.12648 5. Hughes JA, Fontaine MJ, Gonzalez CL, Layon AG,
Goodnough LT, Galel SA. Case report of a transfusion-
associated hepatitis A infection. Transfusion. 2014;54:2202–6. http://dx.doi.org/10.1111/trf.12648 19. Arslan M, Wiesner RH, Poterucha JJ, Gross JB Jr, Zein NN. Hepatitis A antibodies in liver transplant recipients: evidence for
loss of immunity posttransplantation. Liver Transpl. 2000;6:191–5. http://dx.doi.org/10.1002/lt.500060216 6. Bhatnagar J, Blau DM, Shieh WJ, Paddock CD, Drew C, Liu L,
et al. Molecular detection and typing of dengue viruses from
archived tissues of fatal cases by RT-PCR and sequencing:
diagnostic and epidemiologic implications. Am J Trop Med Hyg. 2012;86:335–40. http://dx.doi.org/10.4269/ajtmh.2012.11-0346 20. Günther M, Stark K, Neuhaus R, Reinke P, Schröder K, Bienzle U. Rapid decline of antibodies after hepatitis A immunization in liver
and renal transplant recipients. Transplantation. 2001;71:477–9. http://dx.doi.org/10.1097/00007890-200102150-00023 7. Amon JJ, Devasia R, Xia G, Nainan OV, Hall S, Lawson B,
et al. Molecular epidemiology of foodborne hepatitis a outbreaks
in the United States, 2003. J Infect Dis. 2005;192:1323–30. http://dx.doi.org/10.1086/462425 21. Centers for Disease Control and Prevention. Hepatitis A. In:
Hamborsky J, Kroger A, Wolfe C, editors. Epidemiology and
prevention of vaccine-preventable diseases (the pink book). Discussion Her research interests include HAV
foodborne outbreaks, hepatitis B and C virus healthcare-
associated infections, and perinatal hepatitis C virus transmission
and prevention. Dr. Foster is a medical epidemiologist in the Division of Viral
Hepatitis, National Center for HIV, Viral Hepatitis, STD, and
TB Prevention, CDC. Her research interests include HAV
foodborne outbreaks, hepatitis B and C virus healthcare-
associated infections, and perinatal hepatitis C virus transmission
and prevention. 13. Krober MS, Bass JW, Brown JD, Lemon SM, Rupert KJ. Hospital outbreak of hepatitis A: risk factors for spread. Pediatr
Infect Dis. 1984;3:296–9. http://dx.doi.org/10.1097/00006454-
198407000-00003 14. Azimi PH, Roberto RR, Guralnik J, Livermore T, Hoag S,
Hagens S, et al. Transfusion-acquired hepatitis A in a premature
infant with secondary nosocomial spread in an intensive care
nursery. Am J Dis Child. 1986;140:23–7. 15. Centers for Disease Control and Prevention. Healthcare
Infection Control Practices Advisory Committee (HICPAC). 2007
Guideline for isolation precautions: preventing transmission of
infectious agents in healthcare settings, Table 2. Clinical syndromes
or conditions warranting empiric transmission-based precautions in
addition to standard precautions pending confirmation of
diagnosis [cited 2016 Jun 6]. http://www.cdc.gov/hicpac/2007IP/
2007ip_table2.html Address for correspondence: Monique A. Foster, Centers for Disease Discussion In 2014,
however, 2-dose vaccination coverage among children
19–35 months of age in the United States was only 57.5%,
the lowest vaccine coverage for a complete vaccine series
among the routine childhood vaccines (23). The 8-year-old
organ donor in this report most likely was not vaccinated
against HAV, and transmission to the recipient and HCWs
could have been prevented had vaccination occurred.fi Rapid communication between public health officials,
physicians, transplant centers, and organ procurement organi
zations made locating and testing the other organ recipients in
this investigation possible. Because of the low US incidence
of HAV infection and typically brief self-limiting course of
disease, an HAV-unvaccinated organ donor is unlikely to be
acutely infected at death and transmit HAV to a patient. The
acute nature of HAV infection, low population HAV infec
tion rate, and low rate of HAV infection–associated hospital
ization make universal vaccination of HCWs and pretrans
plant testing for patients impractical. Vaccination against
hepatitis A of the organ donor at 12–23 months of age, as
recommended by the Advisory Committee on Immunization
Practices, most likely would have prevented infection of the
multi–visceral organ recipient and exposed HCWs. Transmission of HAV after blood transfusion has been
established previously but has not been reported through
organ transplantation (5). Organ procurement organiza
tions typically complete a medical history questionnaire
focusing on prior vaccinations, infections, and exposures to
screen donors and recipients for multiple types of infection
(18). Because of the acute nature of HAV infection, pre
transplant testing is not routinely done, although hepatitis
A vaccination of transplantation candidates is recommend
ed (16). The multi–visceral organ recipient was vaccinated,
but studies have shown loss of immunity to HAV after
transplantation because of immunosuppression (19,20). The heart and kidney transplant recipients had evidence
of immunity and probably were protected through previ
ous vaccination. It is unknown whether the differences in This work was supported by the Division of Viral Hepatitis,
National Center for HIV, Viral Hepatitis, STD, and TB
Prevention, CDC. This work was supported by the Division of Viral Hepatitis,
National Center for HIV, Viral Hepatitis, STD, and TB
Prevention, CDC. 595 Emerging Infectious Diseases • www.cdc.gov/eid • Vol. 23, No.4, April 2017 SYNOPSIS Dr. Foster is a medical epidemiologist in the Division of Viral
Hepatitis, National Center for HIV, Viral Hepatitis, STD, and
TB Prevention, CDC. Address for correspondence: Monique A. Foster, Centers for Disease References 13th
ed. Atlanta: Centers for Disease Control and Prevention; 2008. p 135–48. 8. Wiseman R, Weil LM, Lozano C, Johnson TJ Jr, Jin S,
Moorman AC, et al. Notes from the field: health care–associated
hepatitis A outbreak—Texas, 2015. MMWR Morb Mortal Wkly
Rep. 2016;65:425–6. http://dx.doi.org/10.15585/mmwr.mm6516a5 22. Nolan T, Bernstein H, Blatter MM, Bromberg K, Guerra F,
Kennedy W, et al. Immunogenicity and safety of an inactivated
hepatitis A vaccine administered concomitantly with
diphtheria-tetanus-acellular pertussis and Haemophilus
influenzae type B vaccines to children less than 2 years of age. Pediatrics. 2006;118:e602–9. http://dx.doi.org/10.1542/peds. 2005-2755 9. Normann A, Jung C, Vallbracht A, Flehmig B. Time course of
hepatitis A viremia and viral load in the blood of human hepatitis
A patients. J Med Virol. 2004;72:10–6. http://dx.doi.org/10.1002/
jmv.10532 10. Dotzauer A, Heitmann A, Laue T, Kraemer L, Schwabe K,
Paulmann D, et al. The role of immunoglobulin A in prolonged and
relapsing hepatitis A virus infections. J Gen Virol. 2012;93:754–60. http://dx.doi.org/10.1099/vir.0.038406-0 23. Hill HA, Elam-Evans LD, Yankey D, Singleton JA, Kolasa M. National, state, and selected local area vaccination coverage among
children aged 19–35 months—United States, 2014. MMWR Morb
Mortal Wkly Rep. 2015;64:889–96. http://dx.doi.org/10.15585/
mmwr.mm6433a1 11. Tjon GM, Coutinho RA, van den Hoek A, Esman S, Wijkmans CJ,
Hoebe CJ, et al. High and persistent excretion of hepatitis A virus
in immunocompetent patients. J Med Virol. 2006;78:1398–405. http://dx.doi.org/10.1002/jmv.20711fi 12. Costa-Mattioli M, Allavena C, Poirier AS, Billaudel S, Raffi F, Ferré V. Prolonged hepatitis A infection in an HIV-1 seropositive patient. J Med Virol. 2002;68:7–11. http://dx.doi.org/10.1002/jmv.10163 Control and Prevention, 1600 Clifton Rd NE, Mailstop G37, Atlanta, GA
30329-4027, USA; email: ydg9@cdc.gov 596 Emerging Infectious Diseases • www.cdc.gov/eid • Vol. 23, No.4, April 2017
|
https://openalex.org/W4210728383
|
https://eprints.soton.ac.uk/420961/2/1_s2.0_S014067361830134X_main.pdf
|
English
| null |
Risk thresholds for alcohol consumption: combined analysis of individual-participant data for 599 912 current drinkers in 83 prospective studies
|
Lancet
| 2,018
|
cc-by
| 12,739
|
Lancet 2018; 391: 1513–23 Methods We did a combined analysis of individual-participant data from three large-scale data sources in 19 high-
income countries (the Emerging Risk Factors Collaboration, EPIC-CVD, and the UK Biobank). We characterised
dose–response associations and calculated hazard ratios (HRs) per 100 g per week of alcohol (12·5 units per week)
across 83 prospective studies, adjusting at least for study or centre, age, sex, smoking, and diabetes. To be eligible
for the analysis, participants had to have information recorded about their alcohol consumption amount and
status (ie, non-drinker vs current drinker), plus age, sex, history of diabetes and smoking status, at least 1 year of
follow-up after baseline, and no baseline history of cardiovascular disease. The main analyses focused on current
drinkers, whose baseline alcohol consumption was categorised into eight predefined groups according to the
amount in grams consumed per week. We assessed alcohol consumption in relation to all-cause mortality, total
cardiovascular disease, and several cardiovascular disease subtypes. We corrected HRs for estimated long-term
variability in alcohol consumption using 152 640 serial alcohol assessments obtained some years apart (median
interval 5·6 years [5th–95th percentile 1·04–13·5]) from 71 011 participants from 37 studies. Findings In the 599 912 current drinkers included in the analysis, we recorded 40 310 deaths and 39 018 incident
cardiovascular disease events during 5·4 million person-years of follow-up. For all-cause mortality, we recorded
a positive and curvilinear association with the level of alcohol consumption, with the minimum mortality risk
around or below 100 g per week. Alcohol consumption was roughly linearly associated with a higher risk of
stroke (HR per 100 g per week higher consumption 1·14, 95% CI, 1·10–1·17), coronary disease excluding
myocardial infarction (1·06, 1·00–1·11), heart failure (1·09, 1·03–1·15), fatal hypertensive disease (1·24,
1·15–1·33); and fatal aortic aneurysm (1·15, 1·03–1·28). By contrast, increased alcohol consumption was log-
linearly associated with a lower risk of myocardial infarction (HR 0·94, 0·91–0·97). In comparison to those who
reported drinking >0–≤100 g per week, those who reported drinking >100–≤200 g per week, >200–≤350 g per
week, or >350 g per week had lower life expectancy at age 40 years of approximately 6 months, 1–2 years, or
4–5 years, respectively. Interpretation In current drinkers of alcohol in high-income countries, the threshold for lowest risk of all-cause
mortality was about 100 g/week. Articles Articles Risk thresholds for alcohol consumption: combined analysis
of individual-participant data for 599 912 current drinkers in
83 prospective studies Angela M Wood, Stephen Kaptoge, Adam S Butterworth, Peter Willeit, Samantha Warnakula, Thomas Bolton, Ellie Paige, Dirk S Paul,
Michael Sweeting, Stephen Burgess, Steven Bell, William Astle, David Stevens, Albert Koulman, Randi M Selmer, W M Monique Verschuren,
Shinichi Sato, Inger Njølstad, Mark Woodward, Veikko Salomaa, Børge G Nordestgaard, Bu B Yeap, Astrid Fletcher, Olle Melander, Lewis H Kuller,
Beverley Balkau, Michael Marmot, Wolfgang Koenig, Edoardo Casiglia, Cyrus Cooper, Volker Arndt, Oscar H Franco, Patrik Wennberg,
John Gallacher, Agustín Gómez de la Cámara, Henry Völzke, Christina C Dahm, Caroline E Dale, Manuela M Bergmann, Carlos J Crespo,
Yvonne T van der Schouw, Rudolf Kaaks, Leon A Simons, Pagona Lagiou, Josje D Schoufour, Jolanda M A Boer, Timothy J Key, Beatriz Rodriguez,
Conchi Moreno-Iribas, Karina W Davidson, James O Taylor, Carlotta Sacerdote, Robert B Wallace, J Ramon Quiros, Rosario Tumino,
Dan G Blazer II, Allan Linneberg, Makoto Daimon, Salvatore Panico, Barbara Howard, Guri Skeie, Timo Strandberg, Elisabete Weiderpass,
Paul J Nietert, Bruce M Psaty, Daan Kromhout, Elena Salamanca-Fernandez, Stefan Kiechl, Harlan M Krumholz, Sara Grioni, Domenico Palli,
José M Huerta, Jackie Price, Johan Sundström, Larraitz Arriola, Hisatomi Arima, Ruth C Travis, Demosthenes B Panagiotakos, Anna Karakatsani,
Antonia Trichopoulou, Tilman Kühn, Diederick E Grobbee, Elizabeth Barrett-Connor, Natasja van Schoor, Heiner Boeing, Kim Overvad,
Jussi Kauhanen, Nick Wareham, Claudia Langenberg, Nita Forouhi, Maria Wennberg, Jean-Pierre Després, Mary Cushman, Jackie A Cooper,
Carlos J Rodriguez, Masaru Sakurai, Jonathan E Shaw, Matthew Knuiman, Trudy Voortman, Christa Meisinger, Anne Tjønneland,
Hermann Brenner, Luigi Palmieri, Jean Dallongeville, Eric J Brunner, Gerd Assmann, Maurizio Trevisan, Richard F Gillum, Ian Ford, Naveed Sattar,
Mariana Lazo, Simon G Thompson, Pietro Ferrari, David A Leon, George Davey Smith, Richard Peto, Rod Jackson, Emily Banks,
Emanuele Di Angelantonio, John Danesh, for the Emerging Risk Factors Collaboration/EPIC-CVD/UK Biobank Alcohol Study Group* Lancet 2018; 391: 1513–23
See Comment page 1460
*Other investigators of the
Emerging Risk Factors
Collaboration are listed in the
appendix p 48
Department of Public Health
and Primary Care, University of
Cambridge, Cambridge, UK
(A M Wood PhD, S Kaptoge PhD,
A S Butterworth PhD,
P Willeit MD, S Warnakula PhD,
T Bolton MMath, D S Paul PhD,
M Sweeting PhD, S Burgess PhD,
S Bell PhD, W Astle PhD,
D Stevens MSc,
Prof S G Thompson FMedSci,
E Di Angelantonio MD,
Prof J Danesh FMedSci); Medical
University Innsbruck,
Innsbruck, Austria (P Willeit,
Prof S Kiechl MD); National
Centre for Epidemiology and
Population Health, Australian
National University, Canberra,
Australia (E Paige PhD,
Prof E Banks PhD); MRC
Biostatistics Unit, Cambridge
Institute of Public Health,
University of Cambridge,
Cambridge, UK (S Burgess);
NIHR BRC Nutritional
Biomarker Laboratory,
University of Cambridge,
Cambridge, UK
(A Koulman PhD); Norwegian
Institute of Public Health, Oslo,
Norway (R M Selmer PhD);
National Institute for Public
Health and the Environment, Summary y
Background Low-risk limits recommended for alcohol consumption vary substantially across different national
guidelines. To define thresholds associated with lowest risk for all-cause mortality and cardiovascular disease, we
studied individual-participant data from 599 912 current drinkers without previous cardiovascular disease. y
Background Low-risk limits recommended for alcohol consumption vary substantially across different national
guidelines. To define thresholds associated with lowest risk for all-cause mortality and cardiovascular disease, we
studied individual-participant data from 599 912 current drinkers without previous cardiovascular disease. Lancet 2018; 391: 1513–23
See Comment page 1460
*Other investigators of the
Emerging Risk Factors
Collaboration are listed in the
appendix p 48 Introduction Alcohol consumption guidelines vary substantially across
the globe.1,2 In the USA, for example, an upper limit of
196 g per week (about 11 standard UK glasses of wine or
pints of beer per week) is recommended for men, and an
upper limit of 98 g per week is recommended for women.1
Similar recommendations apply in Canada and Sweden.2
By contrast, guidelines in Italy, Portugal, and Spain
recommend low-risk limits almost 50% higher than
these.1,2 At the other extreme, UK guidelines recommend
low-risk limits for men almost half that recommended by
US guidelines.1,2 Study design, data sources, and participants
We focused our study on current alcohol drinkers for
three main reasons. First, alcohol guidelines provide
recommendations about low-risk limits only for drinkers
(we are unaware of any guidelines that encourage
non-drinkers to consume alcohol). Second, a focus on
current drinkers should limit potential biases that are
difficult to control in observational studies (eg, reverse
causality, residual confounding, and unmeasured effect
modification) because ex-drinkers include people who
might have abstained from alcohol owing to poor health
itself,18–20 as well as those who have changed their habits
to achieve a healthier lifestyle. Third, never-drinkers
might differ systematically from drinkers in ways that
are difficult to measure, but which might be relevant to
disease causation.21 Such variation in policy might reflect ambiguity about
drinking risk thresholds associated with the lowest risk of
mortality,3–15 as well as uncertainty about the specific
consequences of alcohol consumption, including those
related to cardiovascular disease subtypes. For example,
recent studies have challenged the concept that moderate
alcohol consumption is universally associated with lower
cardiovascular disease risk,16,17 but the dose–response
associations of alcohol consumption with cardiovascular
disease subtypes remain poorly understood. Therefore, to
help in the formulation of evidence-based alcohol policy,
we analysed individual-participant data from 83 long-
term prospective studies in 19 high-income countries. Our aim was to characterise risk thresholds for all-cause
mortality and cardiovascular disease subtypes in current
drinkers of alcohol. We did a combined analysis of individual-participant
data from three large-scale data sources available to our
consortium, each constituting purpose-designed pro
spective cohort studies with quantitative information
about alcohol consumption (appendix p 21). Implications of all the available evidence The chief implication of this study for public policy is to
support reductions of alcohol consumption limits in existing
guidelines, suggesting that the threshold for lowest risk for
all-cause mortality is about 100 g per week (about
5–6 standard UK glasses of wine or pints of beer per week). The chief implication for scientific understanding is the
strengthening of evidence that the association between
alcohol consumption and total cardiovascular disease risk is
actually comprised of several distinct and opposite
dose–response curves rather than a single J-shaped
association. Funding UK Medical Research Council, British Heart Foundation, National Institute for Health Research, European
Union Framework 7, and European Research Council. Funding UK Medical Research Council, British Heart Foundation, National Institute for Health Research, European
Union Framework 7, and European Research Council. Funding UK Medical Research Council, British Heart Foundation, National Institute for Health Research, European
Union Framework 7, and European Research Council. Bilthoven, Netherlands
(Prof W M M Verschuren PhD,
J M A Boer PhD); Julius Centre for
Health Sciences and Primary
Care, University Medical Center
Utrecht, Utrecht, Netherlands
(Prof W M M Verschuren,
Prof Y T van der Schouw PhD,
Prof D E Grobbee MD); Chiba
Prefectural Institute of Public
Health, Chiba, Japan
(Prof S Sato MD); Department of
Community Medicine,
University of Tromsø, Tromsø,
Norway (Prof I Njølstad MD);
Nuffield Department of
Population Health, Medical
Sciences Division, University of
Oxford, Oxford, UK
(Prof M Woodward PhD,
Prof R Peto FRS, Prof T J Key DPhil,
R C Travis DPhil); The George
Institute for Global Health,
University of Sydney, Sydney,
NSW, Australia
(Prof M Woodward); Bloomberg
School of Public Health
(Prof M Woodward) and School
of Medicine (M Lazo MD), Johns
Hopkins University, Baltimore,
MD, USA; THL-National
Institute for Health and Welfare,
Helsinki, Finland
(Prof V Salomaa MD);
Copenhagen University
Hospital, Copenhagen, Denmark
(Prof B G Nordestgaard MD);
Department of Clinical
Medicine, University of
Copenhagen, Copenhagen,
Denmark (Prof B G Nordestgaard,
Prof A Linneberg MD); School of
Medicine, University of Western
Australia, Perth, WA, Australia
(Prof B B Yeap MBBS); Fiona
Stanley Hospital, Perth, WA,
Australia (Prof B B Yeap); Harry
Perkins Institute of Medical
Research, Perth, WA, Australia
(Prof B B Yeap); London School
of Hygiene & Tropical Medicine,
London, UK (Prof A Fletcher PhD,
Prof D A Leon PhD); Department
of Clinical Sciences, Malmö,
Lund University, Malmö,
Sweden (Prof O Melander MD);
Graduate School of Public
Health, University of
Pittsburgh, Pittsburgh, PA, USA
(Prof L H Kuller MD); CESP
INSERM UMRS 1018, Villejuif
Cedex, France (B Balkau PhD);
Department of Epidemiology
and Public Health, University
College London, London, UK
(Prof M Marmot FMedSci,
Prof E J Brunner PhD,
J A Cooper MSc); 92 Deutsches
Herzzentrum München,
Technische Universität
München Munich Germany Copyright © The Author(s). Published by Elsevier Ltd. This is an Open Access article under the CC BY 4.0 license. Introduction First, the
Emerging Risk Factors Collaboration (ERFC) is a
collaboration of prospective cohort studies with infor
mation about a variety of risk factors, cardiovascular
disease outcomes, and mortality.22 Of the 102 studies in the
ERFC with information about alcohol status, 81 contained
information about the quantity of consumption. Second, Evidence before this study First, it reduced the potentially distorting effects of reverse
causality by focusing on current drinkers without previous
cardiovascular disease who survived at least 12 months of
follow-up. Second, it enhanced generalisability by including
individual-participant data from 83 prospective studies in
19 different high-income countries. Third, it used a variety of
established and emerging risk factors, enabling investigation
of potential confounders and mediators. We searched for prospective epidemiological studies of alcohol
consumption investigating disease risk thresholds published in
any language up until March 1, 2017 (with no specified earliest
date), in PubMed, Scientific Citation Index Expanded, and
Embase using relevant terms (“alcohol”, “mortality”, “survival”,
“cardiovascular disease”, “cohort”, and “prospective”). We found
many primary reports and literature-based reviews. However,
no study had combined the following key features required to
achieve reliable estimates of dose–response associations:
availability of individual-participant data; quantitative
assessment of alcohol consumption levels using validated
instruments; periodic re-surveys of alcohol consumption levels;
recording of large numbers of deaths (eg, >20 000 deaths);
and sufficient detail and power to disaggregate incident
cardiovascular disease outcomes into subtypes
(eg, >20 000 incident total cardiovascular disease outcomes). Lancet 2018; 391: 1513–23 For cardiovascular disease subtypes other than myocardial infarction, there were no
clear risk thresholds below which lower alcohol consumption stopped being associated with lower disease risk. These
data support limits for alcohol consumption that are lower than those recommended in most current guidelines. www.thelancet.com Vol 391 April 14, 2018 1513 Articles Added value of this study The current study combined all the key study design features
mentioned above, and afforded several additional advantages. www.thelancet.com Vol 391 April 14, 2018 1514 Articles Articles EPIC-CVD, a ten-country case-cohort study nested in the
European Prospective Investigation into Cancer and
Nutrition (EPIC) prospective cohort study, had quantitative
alcohol information from 22 of its 23 contributing
centres.23 Third, UK Biobank—a single large prospective
study—had cohort-wide data about quantitative alcohol
consumption.24 Therefore, our combined analysis included
information from a total of 83 prospective studies that each
used broadly similar methods to quantify alcohol
consumption, record risk factors, and ascertain cause-
specific death and cardiovascular disease events. We
harmonised records of alcohol consumption across the
contributing studies using a conversion of 1 unit=8 g of
pure alcohol to a standard scale of grams per week
(appendix pp 1–2), enabling a common analytical approach
despite variation in the methods used (eg, self-administered
vs interview-led questionnaires; food frequency question
naires vs dietary recall surveys), and in consumption
scales over different periods of ascertainment. Details of
contributing studies are in appendix pp 3–4, 10–11. EPIC-CVD, a ten-country case-cohort study nested in the
European Prospective Investigation into Cancer and
Nutrition (EPIC) prospective cohort study, had quantitative
alcohol information from 22 of its 23 contributing
centres.23 Third, UK Biobank—a single large prospective
study—had cohort-wide data about quantitative alcohol
consumption.24 Therefore, our combined analysis included
information from a total of 83 prospective studies that each
used broadly similar methods to quantify alcohol
consumption, record risk factors, and ascertain cause-
specific death and cardiovascular disease events. We
harmonised records of alcohol consumption across the
contributing studies using a conversion of 1 unit=8 g of
pure alcohol to a standard scale of grams per week
(appendix pp 1–2), enabling a common analytical approach
despite variation in the methods used (eg, self-administered
vs interview-led questionnaires; food frequency question
naires vs dietary recall surveys), and in consumption
scales over different periods of ascertainment. Details of
contributing studies are in appendix pp 3–4, 10–11. Added value of this study EPIC-CVD’s case-cohort design (which was used because
lipids and other cardiovascular disease biomarkers were
measured only in the case-cohort subset and not the full
EPIC cohort), the Cox models for cardiovascular disease
events were adapted using Prentice weights and stratified
by centre.26 For the four case-control studies nested within
prospective cohorts of the ERFC, odds ratios were
calculated using, as appropriate, conditional or uncon
ditional logistic regression models, taking into account
relevant matching factors. Study-specific estimates were
then pooled across studies by random-effects meta-
analysis.27 We tested for violation of the proportional
hazards assumption by including time interactions with
alcohol consumption. To avoid model overfitting, studies
with fewer than five incident cases of a particular outcome
were excluded from analyses of that particular outcome. Added value of this study DZHK (German Centre for
Cardiovascular Research),
partner site Munich Heart
Alliance, Munich, Germany
(Prof W Koenig MD); University
of Ulm Medical Center, Ulm,
Germany (Prof W Koenig);
Department of Medicine,
University of Padua, Padua,
Italy (Prof E Casiglia MD); MRC
Lifecourse Epidemiology Unit,
University of Southampton,
Southampton, UK
(Prof C Cooper FMedSci);
German Cancer Research
Center (DKFZ), Heidelberg,
Germany (V Arndt MD,
T Kühn PhD, Prof H Brenner MD,
Prof R Kaaks MD); Erasmus
University Medical Center
Rotterdam, Rotterdam,
Netherlands
(Prof O H Franco MD,
J D Schoufour PhD,
T Voortman PhD); Department
of Public Health and Clinical
Medicine, Umeå University,
Umeå, Sweden
(P Wennberg MD,
M Wennberg PhD); Department
of Primary Care and Public
Health, Cardiff University,
Cardiff, UK
(Prof J Gallacher PhD);
12 de Octubre Research
Institute, CIBERESP,
Madrid, Spain
(A Gómez de la Cámara MD);
Institute for Community
Medicine, University Medicine
Greifswald, Greifswald,
Germany (Prof H Völzke MD);
Department of Public Health,
Aarhus University, Aarhus,
Denmark (C C Dahm PhD,
Prof K Overvad MD); Farr
Institute of Health Informatics
Research, UCL Institute of
Health Informatics, University
College London, London, UK
(C E Dale PhD); German
Institute of Human Nutrition,
Potsdam–Rehbrüke, Germany
(M M Bergmann PhD,
Prof H Boeing PhD); School of
Community Health, Portland
State University, Portland, OR,
USA (C J Crespo PhD);
St Vincent’s Clinical School,
University of New South
Wales, Sydney, NSW, Australia
(L A Simons MD); Hellenic
Health Foundation, Athens,
Greece (P Lagiou MD,
A Karakatsani MD,
Prof A Trichopoulou MD);
National and Kapodistrian
University of Athens, Athens,
Greece (P Lagiou, A Karakatsani,
Prof A Trichopoulou); Harvard
TH Chan School of Public y
p
To correct for measurement error and within-person
variability in alcohol consumption over time, we
estimated long-term average (henceforth, “usual”) alcohol
consumption using multi-level regression calibration and
information
from
152 640
serial
assessments
in
71 011 individuals from 37 studies. This calculation was
achieved either by regressing re-survey measurements
(for the repeat alcohol assessments available in the ERFC
studies and UK Biobank) or lifetime alcohol consumption
measurements (for calculated lifetime alcohol con
sumption measurements available in EPIC-CVD) on
baseline alcohol consumption, adjusted for duration of
follow-up and baseline age, sex, smoking status, history
of diabetes, other relevant covariate(s), and with random
effects for study and re-survey.28,29 The regression dilution
ratio (ie, the calibration slope), which measures the extent
of within-person variability,28 was extracted from the
calibration model. Added value of this study To assess the effect of
excluding known current drinkers with missing alcohol
consumption data, we did a sensitivity analysis using
multiple imputation within studies, before combining St Vincent’s Clinical School,
University of New South Wales, Sydney, NSW, Australia
(L A Simons MD); Hellenic Wales, Sydney, NSW, Australia
(L A Simons MD); Hellenic www.thelancet.com Vol 391 April 14, 2018 Added value of this study HRs in this paper relate to usual
alcohol consumption levels unless specified otherwise. To be eligible for the analysis, participants had to have
information recorded about their alcohol consumption
amount and status (ie, non-drinker vs current drinker),
plus age, sex, history of diabetes and smoking status, at
least 1 year of follow-up after baseline, and no known
baseline history of cardiovascular disease (defined as
coronary heart disease, other heart disease, stroke,
transient ischaemic attack, peripheral arterial disease, or
cardiovascular surgery); appendix p 21. The main analyses
focused on current drinkers, whose baseline alcohol
consumption was categorised into eight predefined
groups according to the amount in grams consumed per
week: >0–≤25, >25–≤50, >50–≤75, >75–≤100, >100–≤150,
>150–≤250, >250–≤350, and >350 g per week. We assessed
alcohol consumption in relation to all-cause mortality,
total cardiovascular disease, and the following car
diovascular disease subtypes (defined in appendix p 5):
fatal and non-fatal myocardial infarction; fatal and non-
fatal coronary disease excluding myocardial infarction;
fatal
and
non-fatal
stroke
(including
ischaemic,
haemorrhagic, subarachnoid, and unclassified subtypes
of stroke); fatal and non-fatal heart failure; and mortality
from other cardiovascular causes, including cardiac
dysrhythmia, hypertensive disease, sudden death, and
aortic aneurysm.7,17,25 In analyses of cardiovascular disease
subtypes, participants contributed follow-up time until
the first outcome recorded (ie, cardiovascular deaths
preceded by non-fatal outcomes were not included). Event
times were censored at the end of follow-up or death from
non-cardiovascular causes. We assessed the shapes of associations for all-cause
mortality and cardiovascular disease outcomes by
calculating study-specific HRs within the predefined
groups of baseline alcohol consumption, pooled them by
multivariate random-effects meta-analysis, and plotted
them against mean usual (and baseline) alcohol
consumption within each group. We estimated 95% CIs
for each group (including the reference group) that
corresponded to the amount of information underlying
each group.30,31 For each major outcome, we determined
the best fitting first or second order fractional polynomial32
to describe the association with baseline alcohol
consumption (using a 1% significance level as evidence
for a second order fractional polynomial over a first order
fractional polynomial) using Cox regression models
stratified by sex, study, and centre. Further analyses
assumed a linear association with alcohol consumption,
expressing results per 100 g per week (12·5 units/week)
in usual alcohol consumption. Articles ERFC
EPIC-CVD
UK Biobank
Participants with resurveys of alcohol consumption
Study level characteristics
Location
81 studies in 19 countries
22 centres in 10 European
countries
England, Scotland, and Wales
37 studies in 15 countries
Years of recruitment
1964–2008
1990–2002
2006–10
1964–2010
Year of most recent endpoint follow-up
2013
2009
2016
2016
Participant level characteristics
Total participants
356 819
30 702
358 833
89 499
Known current drinkers at baseline
247 504
26 036
326 372
71 011
Weekly baseline alcohol consumption in current drinkers
>0–≤25 g per week
53 418 (22%)
7906 (30%)
39 641 (12%)
12 301 (17% [11 g/week vs 36 g/week]‡)
>25–≤50 g per week
33 953 (14%)
3704 (14%)
39 334 (12%)
8365 (12% [38 g/week vs 56 g/week]‡)
>50–≤75 g per week
26 656 (11%)
2748 (11%)
42 907 (13%)
7322 (10% [63 g/week vs 80 g/week]‡)
>75–≤100 g per week
16 557 (7%)
2446 (9%)
36 780 (11%)
6394 (9% [87 g/week vs 98 g/week]‡)
>100–≤150 g per week
36 236 (15%)
2602 (10%)
55 815 (17%)
10 051 (14% [126 g/week vs 126 g/week]‡)
>150–≤250 g per week
31 645 (13%)
3090 (12%)
60 025 (18%)
12 255 (17% [193 g/week vs 173 g/week]‡)
>250–≤350 g per week
23 607 (10%)
1744 (7%)
26 669 (8%)
6927 (10% [303 g/week vs 248 g/week]‡)
≥350 g per week
25 432 (10%)
1796 (7%)
25 201 (8%)
7396 (10% [515 g/week vs 354 g/week]‡)
Baseline characteristics restricted to all current drinkers
Alcohol consumption (g/week), median
(5th–95th percentiles)
87·7 (2·2–522·4)
61·9 (2·6–404·0)
103·9 (11·8–420·8)
105·2 (6·0–482·8)
Age (years) at baseline
57·1 (8·7)
55·0 (9·2)
56·5 (8·0)
55·3 (8·2)
Sex
Male
162 685 (66%)
13 508 (52%)
157 809 (48%)
44 360 (62%)
Female
84 819 (34%)
12 528 (48%)
168 563 (52%)
26 651 (38%)
Smoking status
Not current
161 037 (65%)
17 608 (68%)
293 182 (90%)
50 930 (72%)
Current
86 467 (35%)
8428 (32%)
33 190 (10%)
20 081 (28%)
History of diabetes
No
237 685 (96%)
24 875 (96%)
315 090 (97%)
68 159 (96%)
Yes
9819 (4%)
1161 (4%)
11 282 (3%)
2852 (4%)
BMI, kg/m²
26·1 (3·8)
26·4 (4·1)
27·0 (4·4)
26·1 (3·8)
HDL-C, mmol/L
1·40 (0·41)
1·40 (0·42)
Not available*
1·41 (0·41)
Total cholesterol, mmol/L
5·80 (1·17)
6·11 (1·16)
Not available*
5·78 (1·08)
Systolic blood pressure, mm Hg
136·5 (19·0)
138·4 (21·3)
137·9 (18·5)
134·6 (18·4)
Major outcomes restricted to current drinkers
All-cause mortality events
32 813
784†
6720
6912
All cardiovascular disease
18 791
12 758
7469
11 597
Data are n, n (%), or mean (SD), unless otherwise indicated. Articles ERFC=Emerging Risk Factors Collaboration. EPIC-CVD=European Prospective Investigation into Cancer and Nutrition—Cardiovascular Disease. BMI=body-mass index. HDL-C=high-density-lipoprotein cholesterol. *At the time of analysis, measurements of HDL-C and total cholesterol were not available in the UK Biobank. †All-cause mortality events from
EPIC derive only from the 13 670 participants in the random sub-cohort of EPIC-CVD, rather than from the entire EPIC prospective study. ‡Mean consumption (g/week) at baseline vs resurvey. Table 1: Study-level and participant-level characteristics of the contributing data sources Data are n, n (%), or mean (SD), unless otherwise indicated. ERFC=Emerging Risk Factors Collaboration. EPIC-CVD=European Prospective Investigation into Cancer and Nutrition—Cardiovascular Disease. BMI=body-mass index. HDL-C=high-density-lipoprotein cholesterol. *At the time of analysis, measurements of HDL-C and total cholesterol were not available in the UK Biobank. †All-cause mortality events from
EPIC derive only from the 13 670 participants in the random sub-cohort of EPIC-CVD, rather than from the entire EPIC prospective study. ‡Mean consumption (g/week) at baseline vs resurvey. Table 1: Study-level and participant-level characteristics of the contributing data sources adjusted for usual levels of available potential con
founders or mediators, including body-mass index (BMI),
systolic blood pressure, high-density-lipoprotein chol
esterol (HDL-C), low-density-lipoprotein cholesterol
(LDL-C), total cholesterol, fibrinogen, and baseline
measures for smoking amount (in pack-years), level of
education reached (no schooling or primary education
only vs secondary education vs university), occupation
(not working vs manual vs office vs other), self-reported
physical activity level (inactive vs moderately inactive vs
moderately active vs active), self-reported general health
(scaled 0–1 where low scores indicate poorer health), the data in a meta-analysis. We investigated associations
with alcohol type (wine, beer, and spirits), consumption
frequency (dichotomised as drinkers who consumed
alcohol on ≤2 days per week or those who consumed
alcohol on >2 days per week) and episodic heavy drink
ing (dichotomised as binge drinkers who consumed
≥100 g per drinking occasion or non-binge drinkers who
consumed <100 g per drinking occasion). Table 1: Study-level and participant-level characteristics of the contributing data sources www.thelancet.com Vol 391 April 14, 2018 Statistical analysis Hazard ratios (HRs) for alcohol consumption were
calculated separately within each study using Cox
regression models, stratified by sex and with adjustment
for known confounders: age, smoking status (current vs
non-current) and history of diabetes. To account for www.thelancet.com Vol 391 April 14, 2018 1515 Articles Articles Articles Center, New York, NY, USA
(K W Davidson PhD); East Boston
Neighborhood Health Center,
Boston, MA, USA
(J O Taylor MD); Città della
Salute e della Scienza di Torino
Hospital, Turin, Italy
(C Sacerdote PhD); College of
Public Health, The University of
Iowa, Iowa City, IA, USA
(Prof R B Wallace MD); Consejería
de Sanidad del Principado de
Asturias, Oviedo, Asturias,
Spain (J Ramon Quiros MD);
Civic - M. Arezzo Hospital, ASP
Ragusa, Italy
(Prof R Tumino MD); Duke
Divinity School, Duke
University, Durham, NC, USA
(D G Blazer II MD); Department
of Endocrinology and
Metabolism, Hirosaki
University, Hirosaki, Japan
(M Daimon MD); Department of
Clinical and Experimental
Medicine, Federico II University,
Naples, Italy (S Panico MD);
Department of Biology,
Tuskegee University, AL, USA
(B Howard PhD); Department of
Community Medicine, UiT The
Arctic University of Norway,
Tromsø, Norway (G Skeie PhD,
Prof E Weiderpass PhD);
University of Helsinki and
Helsinki University Hospital,
Helsinki, Finland
(Prof T Strandberg MD); Center
for Life Course Health Research,
University of Oulu, Oulu,
Finland (Prof T Strandberg);
Cancer Registry of Norway,
Institute of Population-Based Center, New York, NY, USA
(K W Davidson PhD); East Boston
Neighborhood Health Center,
Boston, MA, USA
(J O Taylor MD); Città della
Salute e della Scienza di Torino
Hospital, Turin, Italy
(C Sacerdote PhD); College of
Public Health, The University of
Iowa, Iowa City, IA, USA
(Prof R B Wallace MD); Consejería
de Sanidad del Principado de
Asturias, Oviedo, Asturias,
Spain (J Ramon Quiros MD);
Civic - M. Arezzo Hospital, ASP
Ragusa, Italy Center, New York, NY, USA
(K W Davidson PhD); East Boston
Neighborhood Health Center,
Boston, MA, USA Baseline alcohol consumption varied substantially across
studies, was generally lower in more recent calendar
periods of recruitment, and was positively skewed (median
96 g/week [5th–95th percentiles 6–448]; appendix p 22). It
was weakly and positively correlated with male sex,
smoking status and amount, systolic blood pressure,
HDL-C level, fibrinogen, and lower socioeconomic status
(appendix pp 23–24). 152 640 serial assessments of alcohol
consumption were available for 71 011 participants from
37 studies (median interval between baseline and serial
measurements 5·6 years [5th–95th percentiles 1·04–13·5]). Participants with serial measurements were younger, had
slightly higher baseline alcohol consumption, and were
more likely to be men than those without serial
measurements (table 1, appendix p 14). Articles Health, Boston, MA, USA
(P Lagiou); Office of Public
Health Studies, University of
Hawaii, Honolulu, HI, USA
(Prof B Rodriguez MD); Instituto
de Salud Pública de Navarra,
IdiSNA - Navarra Institute for
Health Research, Pamplona,
Spain (C Moreno-Iribas PhD);
Red de Investigación en
Servicios de Salud en
Enfermedades Crónicas
(REDISSEC), Pamplona, Spain
(C Moreno-Iribas); Columbia
University Irving Medical Health, Boston, MA, USA
(P Lagiou); Office of Public
Health Studies, University of
Hawaii, Honolulu, HI, USA
(Prof B Rodriguez MD); Instituto
de Salud Pública de Navarra,
IdiSNA - Navarra Institute for
Health Research, Pamplona,
Spain (C Moreno-Iribas PhD);
Red de Investigación en
Servicios de Salud en
Enfermedades Crónicas
(REDISSEC), Pamplona, Spain
(C Moreno-Iribas); Columbia
University Irving Medical We used regression calibration methods similar to
those described above to estimate and adjust for long-
term levels of potential confounding factors or mediators
in individuals with available information. HRs were www.thelancet.com Vol 391 April 14, 2018 1516 Articles 0
100
200
300
All-cause mortality (40 310 deaths)
Cardiovascular disease (39 018 events)
400
0
100
200
300
400
0·9
1·0
1·1
1·2
1·3
1·4
1·5
1·6
Hazard ratio (95% Cl)
Usual alcohol consumption (g per week)
Usual alcohol consumption (g per week)
Figure 1: Associations of usual alcohol consumption with all-cause mortality and the aggregate of
cardiovascular disease in current drinkers
Cardiovascular disease was defined as an aggregate of myocardial infarction, coronary heart disease, and stroke. Hazard ratios are adjusted for age, smoking, and history of diabetes, and stratified by sex and EPIC centre. The reference category is the lowest baseline alcohol consumption category (between 0 and 25 g/week). HRs are
plotted against the mean usual alcohol consumption in each category. Sizes of the boxes are proportional to the
inverse of the variance of the log-transformed hazard ratios. Vertical lines represent 95% CIs. self-reported red meat consumption, and self-reported
use of anti-hypertensive drugs. We investigated effect
modification with formal tests for interaction, using a
0·1% significance threshold to make some allowance for
multiple testing. Articles The regression
dilution ratio for alcohol consumption was 0·50 (95% CI
0·47–0·52), similar to that for systolic blood pressure
(0·52, 0·50–0·55) but lower than that for HDL-C
concentration (0·74, 0·72–0·76) in a common set
of participants. gi
y
Analyses used Stata (version 14.2 and 15.1). All p values
presented are for 2-sided tests. Role of the funding source The funders of the study did not have any role in the study
design, data analysis, or reporting of this manuscript. AMW and SK had full access to the combined dataset,
and, together with EDA and JD, had responsibility for the
decision to submit the manuscript for publication. www.thelancet.com Vol 391 April 14, 2018 Articles Heterogeneity was investigated by
grouping studies according to recorded characteristics
and through meta-regression, assessed by the I² statistic.33
Evidence of small study effects was assessed visually with
funnel plots and by Begg and Mazumdar’s test34 and
Egger’s test.35 0
100
200
300
All-cause mortality (40 310 deaths)
Cardiovascular disease (39 018 events)
400
0
100
200
300
400
0·9
1·0
1·1
1·2
1·3
1·4
1·5
1·6
Hazard ratio (95% Cl)
Usual alcohol consumption (g per week)
Usual alcohol consumption (g per week) Methods we used to estimate reductions in life
expectancy (years of life lost) are described in the appendix
(pp 6–7). Briefly, estimates of cumulative survival from
40 years of age onwards in different categories of baseline
alcohol consumption were calculated by applying
estimated HRs (specific to age-at-risk) for cause-specific
mortality to the detailed mortality component of the
US Centers for Disease Control and Prevention’s
WONDER database,36 which recorded 10 million deaths
(from all causes) in more than 305 million individuals in
the USA during 2007–10.37,38 Results were modelled from
age 40 years and enabled estimation of years of life lost
between light drinkers (defined as those consuming
>0–≤100 g/week of alcohol) and pre-defined groups of
>100–≤200, >200–≤350, and >350 g per week. This method
does not make use of the survival estimates from the
modelled data; instead, it makes inferences by estimating
age-at-risk specific HRs, which are then combined with
external population age-specific mortality rates.39 Figure 1: Associations of usual alcohol consumption with all-cause mortality and the aggregate of
cardiovascular disease in current drinkersi Cardiovascular disease was defined as an aggregate of myocardial infarction, coronary heart disease, and stroke. Hazard ratios are adjusted for age, smoking, and history of diabetes, and stratified by sex and EPIC centre. The reference category is the lowest baseline alcohol consumption category (between 0 and 25 g/week). HRs are
plotted against the mean usual alcohol consumption in each category. Sizes of the boxes are proportional to the
inverse of the variance of the log-transformed hazard ratios. Vertical lines represent 95% CIs. and 39 018 first incident cardiovascular disease out
comes, including 12 090 stroke events, 14 539 myocardial
infarction events, 7990 coronary disease events exclud
ing myocardial infarction, 2711 heart failure events,
and 1121 deaths from other cardiovascular diseases
(appendix p 13). Center, New York, NY, USA
(K W Davidson PhD); East Boston
Neighborhood Health Center,
Boston, MA, USA
(J O Taylor MD); Città della
Salute e della Scienza di Torino
Hospital, Turin, Italy
(C Sacerdote PhD); College of
Public Health, The University of
Iowa, Iowa City, IA, USA
(Prof R B Wallace MD); Consejería
de Sanidad del Principado de
Asturias, Oviedo, Asturias,
Spain (J Ramon Quiros MD);
Civic - M. Arezzo Hospital, ASP
Ragusa, Italy
(Prof R Tumino MD); Duke
Divinity School, Duke
University, Durham, NC, USA
(D G Blazer II MD); Department
of Endocrinology and
Metabolism, Hirosaki
University, Hirosaki, Japan
(M Daimon MD); Department of
Clinical and Experimental
Medicine, Federico II University,
Naples, Italy (S Panico MD);
Department of Biology,
Tuskegee University, AL, USA
(B Howard PhD); Department of
Community Medicine, UiT The
Arctic University of Norway,
Tromsø, Norway (G Skeie PhD,
Prof E Weiderpass PhD);
University of Helsinki and
Helsinki University Hospital,
Helsinki, Finland
(Prof T Strandberg MD); Center
for Life Course Health Research,
University of Oulu, Oulu,
Finland (Prof T Strandberg);
Cancer Registry of Norway,
Institute of Population-Based Articles Hazard ratios are plotted against the mean usual alcohol consumption in each category. Studies with fewer than
five events of any outcome were excluded from the analysis of that outcome. Sizes of the boxes are proportional to the inverse of the variance of the log-transformed
hazard ratios. Vertical lines represent 95% CIs. Deaths from other cardiovascular disease include the following outcomes: cardiac dysrhythmia, hypertensive disease,
sudden death, and aortic aneurysm. 0·8
1·0
1·2
1·4
1·6
1·8
2·0
2·2
2·4
Hazard ratio (95% Cl)
All stroke (12 090 events)
Heart failure (2711 events) Myocardial infarction (14 539 events)
Deaths from othertypes of cardiovascular
0
100
200
300
400
Usual alcohol consumption (g per week)
Coronary disease excluding myocardial
infarction (7990 events) Deaths from other types of cardiovascular
disease (1121 deaths) Usual alcohol consumption (g per week) 0
100
200
300
400
Usual alcohol consumption (g per week)
Deaths from other types of cardiovascular
disease (1121 deaths) alcohol consumption with cardiovascular subtypes in alcohol drinkersi Figure 2: Associations of usual alcohol consumption with cardiovascular subtypes in alcohol drinkers
Hazard ratios are adjusted for age, smoking, and history of diabetes, and stratified by sex and EPIC centre. The reference category is the lowest baseline alcohol
consumption category (between 0 and 25g/week). Hazard ratios are plotted against the mean usual alcohol consumption in each category. Studies with fewer than
five events of any outcome were excluded from the analysis of that outcome. Sizes of the boxes are proportional to the inverse of the variance of the log-transformed
hazard ratios. Vertical lines represent 95% CIs. Deaths from other cardiovascular disease include the following outcomes: cardiac dysrhythmia, hypertensive disease,
sudden death, and aortic aneurysm. HRs (table 2, appendix pp 15, 30). First, adjustment for
HDL-C level weakened the inverse association between
alcohol consumption and myocardial infarction, but
strengthened the positive association between alcohol
consumption and both coronary disease and heart
failure. Second, adjustment for systolic blood pressure
strengthened the inverse association between alcohol
consumption and myocardial infarction, but weakened
the positive associations between alcohol consumption
and all other cardiovascular disease outcomes. Our
analysis confirmed the established association of alcohol
consumption with cancers of the digestive system, which
did not change after additional adjustment for the factors
listed above (appendix p 16). Articles Cancer Research, Oslo, Norway
(Prof E Weiderpass);
Department of Medical
Epidemiology and
Biostatistics, Karolinska
Institutet, Stockholm, Sweden
(Prof E Weiderpass); Genetic
Epidemiology Group,
Folkhälsan Research Center,
Faculty of Medicine, University
of Helsinki, Helsinki, Finland
(Prof E Weiderpass); Medical
University of South Carolina,
Charleston, SC, USA
(Prof P J Nietert PhD);
Cardiovascular Health Research
Unit, Departments of
Medicine, Epidemiology, and
Health Services, University of
Washington, Seattle, WA, USA
(Prof B M Psaty MD); Kaiser
Permanente Washington
Health Research Institute,
Seattle, WA, USA
(Prof B M Psaty); Department of
Agrotechnology and Food
Sciences, University of
Wageningen, Wageningen,
Netherlands
(Prof D Kromhout PhD); Faculty
of Medical Sciences, University
of Groningen, Groningen,
Netherlands (Prof D Kromhout);
Escuela Andaluza de Salud
Pública, Instituto de
Investigación Biosanitaria ibs. GRANADA, Hospitales
Universitarios de Granada/
Universidad de Granada,
Granada, Spain
(E Salamanca-Fernandez MSc);
CIBER de Epidemiología y Salud
Pública (CIBERESP), Madrid,
Spain (E Salamanca-Fernandez,
L Arriola MD, J M Huerta PhD);
School of Medicine, Yale
University, New Haven, CT,
USA (Prof H M Krumholz MD);
Epidemiology and Prevention
Unit, Fondazione IRCCS Istituto
Nazionale dei Tumori, Milan,
Italy (S Grioni BSc); Cancer
Research and Prevention
Institute (ISPO), Florence, Italy
(D Palli MD); Murcia Regional
Health Council, IMIB-Arrixaca,
Murcia, Spain (J M Huerta);
Usher Institute, University of
Edinburgh, Edinburgh, UK
(Prof J Price MD); Department of
Medical Sciences, Uppsala
University, Uppsala, Sweden
(Prof J Sundström MD);
Instituto BIO-Donostia, Basque
Government, San Sebastian,
Spain (L Arriola); The University
of Sydney and Royal Prince
Alfred Hospital, Sydney, NSW,
Australia (Prof H Arima MD);
Department of Preventive
Medicine and Public Health, 0·8
1·0
1·2
1·4
1·6
1·8
2·0
2·2
2·4
Hazard ratio (95% Cl)
Hazard ratio (95% Cl)
All stroke (12 090 events)
Myocardial infarction (14 539 events)
0
100
200
300
400
0·8
1·0
1·2
1·4
1·6
1·8
2·0
2·2
2·4
Usual alcohol consumption (g per week)
Heart failure (2711 events)
0
100
200
300
400
Usual alcohol consumption (g per week)
Deaths from other types of cardiovascular
disease (1121 deaths)
0
100
200
300
400
Usual alcohol consumption (g per week)
Coronary disease excluding myocardial
infarction (7990 events)
Figure 2: Associations of usual alcohol consumption with cardiovascular subtypes in alcohol drinkers
Hazard ratios are adjusted for age, smoking, and history of diabetes, and stratified by sex and EPIC centre. The reference category is the lowest baseline alcohol
consumption category (between 0 and 25g/week). www.thelancet.com Vol 391 April 14, 2018 Results Of the 786 787 participants with sufficient information for
inclusion in this consortium, 186 875 (19%) reported not
drinking at baseline, leaving 599 912 current drinkers
without a history of cardiovascular disease at base
line who were eligible for the prespecified principal
analysis. The current drinkers were derived from ERFC
(247 504 participants), EPIC-CVD (26 036), and the UK
Biobank (326 372; table 1). Baseline year of recruitment
ranged from 1964 to 2010. The mean age of the participants
was 57 years (SD 9). 265 910 (44%) of 599 912 participants
were women, and 128 085 (21%) were current smokers
(appendix p 12). About 50% reported drinking more than
100 g of alcohol per week, and 8·4% drank more than
350 g per week (table 1). During 5·4 million person-years
(median 7·5 years of follow-up [5th–95th percentiles
5·0–18·4]), there were 40 310 deaths from all causes,
(including 11 762 vascular and 15 150 neoplastic deaths), For all-cause mortality, there was a positive and
curvilinear association with alcohol consumption, with
the lowest risk for those consuming below 100 g per
week (figure 1, appendix p 25). Associations were similar
for men and women (appendix p 26), but weaker at older
ages (appendix p 27). There was a J-shaped association
for the aggregate of cardiovascular disease outcomes
(figure 1, appendix p 25). However, disaggregation
showed two opposing sets of associations (figure 2). www.thelancet.com Vol 391 April 14, 2018 1517 Articles Articles Articles Studies with fewer than five events of any out
excluded from the analysis of that outcome. Figure 3: Hazard ratios for subtypes of cardiovascular outcomes in current drinkers, per 100 g per week higher usual alcohol consumption
Hazard ratios are adjusted for age, smoking, and history of diabetes, and stratified by sex and centre. Studies with fewer than five events of any outcome were
excluded from the analysis of that outcome. Figure 3: Hazard ratios for subtypes of cardiovascular outcomes in current drinkers, per 100 g per week higher usual alcohol consumption
Hazard ratios are adjusted for age, smoking, and history of diabetes, and stratified by sex and centre. Studies with fewer than five events of any outcome were
excluded from the analysis of that outcome. excluded from the analysis of that outcome. All stroke
Myocardial infarction
Coronary disease
excluding myocardial
infarction
Heart failure
Deaths from other
types of cardio
vascular disease
Subset of participants with measurement of systolic blood pressure
Cohorts/events
70/11 297
73/13 519
46/7789
39/2668
44/1019
Basic adjustment*
1·16 (1·11–1·22)
0·95 (0·91–0·99)
1·06 (1·00–1·12)
1·11 (1·04–1·18)
1·16 (1·06–1·27)
Plus adjustment for systolic blood
pressure
1·10 (1·06–1·14)
0·91 (0·87–0·94)
1·03 (0·97–1·10)
1·08 (1·02–1·15)
1·14 (1·03–1·25)
Subset of participants with measurement of high-density-lipoprotein cholesterol
Cohorts/events
56/7982
61/9911
36/3608
29/1886
34/690
Basic adjustment*
1·16 (1·10–1·23)
0·93 (0·88–0·97)
1·07 (0·98–1·17)
1·09 (1·00–1·19)
1·22 (1·06–1·40)
Plus adjustment for high-density-
lipoprotein cholesterol
1·17 (1·11–1·22)
1·00 (0·96–1·04)
1·13 (1·05–1·22)
1·14 (1·01–1·27)
1·22 (1·08–1·38)
Subset of participants with measurement of body-mass index
Cohorts/events
68/11 733
71/14 217
43/7761
36/2566
42/1035
Basic adjustment*
1·15 (1·10–1·19)
0·95 (0·91–0·98)
1·06 (1·02–1·12)
1·12 (1·04–1·20)
1·16 (1·06–1·27)
Plus adjustment for body-mass index
1·14 (1·10–1·18)
0·94 (0·91–0·97)
1·06 (1·01–1·12)
1·10 (1·03–1·16)
1·16 (1·06–1·27)
Data are hazard ratio (95% CI) per 100 g per week higher usual alcohol consumption, unless otherwise indicated. Analyses were restricted to individuals with basic adjustment
variables plus the additional variable. Studies with fewer than five events were excluded from the analysis of each outcome. *Basic adjustment includes age, smoking, and history
of diabetes, and stratification by sex and centre. Table 2: Hazard ratios for major cardiovascular outcomes in current drinkers, without and with adjustment for usual levels of systolic blood pressure,
high-density-lipoprotein cholesterol, or body-mass index alcohol consumption with total cardiovascular disease and
all-cause mortality (appendix p 31). Articles Furthermore, additional
adjustment for smoking amount abolished the apparent
association of alcohol consumption with lung cancer
(appendix pp 16), in line with the accepted view that
alcohol consumption does not cause lung cancer.40 After adjustment for age, sex, smoking, and history of
diabetes, the amount of alcohol consumed had positive
and roughly linear associations with stroke (HR per
100 g/week higher consumption 1·14, 1·10–1·17),
coronary disease excluding myocardial infarction (1·06,
1·00–1·11), heart failure (1·09, 1·03–1·15), fatal
hypertensive disease (1·24, 1·15–1·33), and fatal aortic
aneurysm (1·15, 1·03–1·28; figures 2, 3). By contrast,
there was an inverse and approximately log-linear
association with myocardial infarction (0·94, 0·91–0·97;
figures 2, 3). Stroke associations were similar for fatal
and non-fatal outcomes (appendix p 28) and across
subtypes (appendix p 29). However, for coronary disease
excluding myocardial infarction, associations were
stronger for fatal than non-fatal outcomes (appendix
p 28). For myocardial infarction, inverse associations
were possibly more pronounced with non-fatal than fatal
outcomes (figure 3, appendix p 28). After adjustment for age, sex, smoking, and history of
diabetes, the amount of alcohol consumed had positive
and roughly linear associations with stroke (HR per
100 g/week higher consumption 1·14, 1·10–1·17),
coronary disease excluding myocardial infarction (1·06,
1·00–1·11), heart failure (1·09, 1·03–1·15), fatal
hypertensive disease (1·24, 1·15–1·33), and fatal aortic
aneurysm (1·15, 1·03–1·28; figures 2, 3). By contrast,
there was an inverse and approximately log-linear
association with myocardial infarction (0·94, 0·91–0·97;
figures 2, 3). Stroke associations were similar for fatal
and non-fatal outcomes (appendix p 28) and across
subtypes (appendix p 29). However, for coronary disease
excluding myocardial infarction, associations were
stronger for fatal than non-fatal outcomes (appendix
p 28). For myocardial infarction, inverse associations
were possibly more pronounced with non-fatal than fatal
outcomes (figure 3, appendix p 28). Articles p
g
When including never-drinkers and ex-drinkers, we
reproduced previously reported U-shaped associations of i
With the following notable exceptions, further adjust
ment for additional covariates did not substantially change www.thelancet.com Vol 391 April 14, 2018 1518 Articles Articles Events/participants
All stroke
Non-fatal stroke
Fatal stroke
Ischaemic stroke
Haemorrhagic stroke
Subarachnoid haemorrhage
Unclassified stroke
All myocardial infarction
Non-fatal myocardial infarction
Fatal myocardial infarction
Coronary disease excluding myocardial infarction
Non-fatal coronary disease excluding myocardial infarction
Fatal coronary disease excluding myocardial infarction
Heart failure (fatal and non-fatal)
Death from other types of cardiovascular disease
Cardiac dysrhythmia
Hypertensive disease
Sudden cardiac death
Aortic aneurysm
12 090/585588
9910/491050
2142/532204
6256/491204
1482/505948
663/412732
3215/527729
14 539/594561
11 706/515377
2748/538117
7990/523548
6000/389976
1889/510147
2711/447436
1121/488122
261/71682
178/383269
283/68002
289/423145
1·14 (1·10–1·17)
1·14 (1·10–1·18)
1·13 (1·07–1·19)
1·13 (1·09–1·18)
1·17 (1·12–1·23)
1·09 (1·00–1·19)
1·13 (1·06–1·20)
0·94 (0·91–0·97)
0·93 (0·90–0·97)
0·99 (0·93–1·05)
1·06 (1·00–1·11)
1·00 (0·97–1·03)
1·11 (1·04–1·18)
1·09 (1·03–1·15)
1·18 (1·07–1·30)
1·17 (0·86–1·60)
1·24 (1·15–1·33)
1·12 (0·90–1·41)
1·15 (1·03–1·28)
12 (0–35)
14 (0–40)
0 (0–35)
8 (0–37)
0 (0–37)
0 (0–58)
14 (0–40)
12 (0–35)
24 (0–45)
8 (0–35)
26 (0–49)
0 (0–52)
12 (0–40)
4 (0–31)
33 (2–53)
63 (35–79)
0 (0–55)
29 (0–63)
0 (0–49)
Heterogeneity l2 (95% Cl)
Hazard ratio (95% Cl)
Lower risk of disease
with higher alcohol consumption
Higher risk of disease
with higher alcohol consumption
1·0
0·8
1·2
1·6
1·4
Figure 3: Hazard ratios for subtypes of cardiovascular outcomes in current drinkers, per 100 g per week higher usual alcohol consumption
Hazard ratios are adjusted for age smoking and history of diabetes and stratified by sex and centre Studies with fewer than five events of any outcome were Events/participants
All stroke
Non-fatal stroke
Fatal stroke
Ischaemic stroke
Haemorrhagic stroke
Subarachnoid haemorrhage
Unclassified stroke
All myocardial infarction
Non-fatal myocardial infarction
Fatal myocardial infarction
Coronary disease excluding myocardial infarction
Non-fatal coronary disease excluding myocardial infarction
Fatal coronary disease excluding myocardial infarction
Heart failure (fatal and non-fatal)
Death from other types of cardiovascular disease
Cardiac dysrhythmia
Hypertensive disease
Sudden cardiac death
Aortic aneurysm
12 090/585588
9910/491050
2142/532204
6256/491204
1482/505948
663/412732
3215/527729
14 539/594561
11 706/515377
2748/538117
7990/523548
6000/389976
1889/510147
2711/447436
1121/488122
261/71682
178/383269
283/68002
289/423145
1·14 (1·10–1·17)
1·14 (1·10–1·18)
1·13 (1·07–1·19)
1·13 (1·09–1·18)
1·17 (1·12–1·23)
1·09 (1·00–1·19)
1·13 (1·06–1·20)
0·94 (0·91–0·97)
0·93 (0·90–0·97)
0·99 (0·93–1·05)
1·06 (1·00–1·11)
1·00 (0·97–1·03)
1·11 (1·04–1·18)
1·09 (1·03–1·15)
1·18 (1·07–1·30)
1·17 (0·86–1·60)
1·24 (1·15–1·33)
1·12 (0·90–1·41)
1·15 (1·03–1·28)
12 (0–35)
14 (0–40)
0 (0–35)
8 (0–37)
0 (0–37)
0 (0–58)
14 (0–40)
12 (0–35)
24 (0–45)
8 (0–35)
26 (0–49)
0 (0–52)
12 (0–40)
4 (0–31)
33 (2–53)
63 (35–79)
0 (0–55)
29 (0–63)
0 (0–49)
Heterogeneity l2 (95% Cl)
Hazard ratio (95% Cl)
Lower risk of disease
with higher alcohol consumption
Higher risk of disease
with higher alcohol consumption
1·0
0·8
1·2
1·6
1·4
Figure 3: Hazard ratios for subtypes of cardiovascular outcomes in current drinkers, per 100 g per week higher usual alcohol consumption Heterogeneity l2 (95% Cl)
Hazard ratio (95% Cl) Figure 3: Hazard ratios for subtypes of cardiovascular outcomes in current drinkers, per 100 g per week higher usual alcohol consumption
Hazard ratios are adjusted for age, smoking, and history of diabetes, and stratified by sex and centre. Articles 40
50
70
60
80
90
0
1
2
3
4
5
6
Years of life lost (95% Cl)
Age (years)
40
50
70
60
80
90
Age (years)
Men
Women
>100–≤200 g/week
>200–≤350 g/week
>350 g/week
Figure 4: Estimated future years of life lost by extent of reported baseline alcohol consumption compared
with those who reported consuming >0–≤100 g per week
The estimates of cumulative survival from 40 years of age onwards in the alcohol-drinking groups were calculated
by applying hazard ratios (specific to age at risk) for all-cause mortality associated with categorised baseline
alcohol consumption to US death rates at the age of 40 years or older. Mean usual levels of alcohol consumption
within each baseline alcohol consumption category were 56, 123, 208 and 367 g per week, respectively, for the
groups >0–≤100 g per week, >100–≤200 g per week, >200–≤350 g per week, and >350 g per week. pp 43–45). There was no evidence of small study effects
(appendix p 46). Our data showed no evidence of violation
of the proportional hazards assumption. 40
50
70
60
80
90
Age (years)
Women 40
50
70
60
80
90
0
1
2
3
4
5
6
Years of life lost (95% Cl)
Age (years)
Men
>100–≤200 g/week
>200–≤350 g/week
>350 g/week In comparison to those who reported drinking
>0–≤100 g (mean usual 56 g) alcohol per week, those who
reported drinking >100–≤200 g (mean usual 123 g) per
week, >200–≤350 g (mean usual 208 g) per week or
>350 g (mean usual 367 g) per week had shorter life
expectancy at age 40 years of approximately 6 months,
1–2 years, or 4–5 years respectively (figure 4). Similarly,
men who reported consuming above the UK upper limit
of 112 g per week had a shorter life expectancy at age
40 years of 1·6 years (95% CI 1·3–1·8), and men who
reported drinking above the US upper limit of 196 g per
week had a shorter life expectancy at age 40 years of
2·7 years (2·4–3·1) compared with men who reported
drinking below these respective upper limits. Thus, men
who reported drinking less than 100 g alcohol per week
had about a 1–2 years longer life expectancy at age
40 years than those who reported drinking 196 g per
week (appendix p 47). Articles (Prof G Assmann MD); The City
College of New York, New York,
NY, USA (M Trevisan MD);
Howard University Hospital,
Washington DC, USA
(R F Gillum MD); Institute of
Cardiovascular & Medical
Sciences, University of Glasgow,
Glasgow, UK (Prof I Ford PhD,
Prof N Sattar FMedSci);
International Agency for
Research on Cancer, Lyon, France
(P Ferrari PhD); MRC Integrative
Epidemiology Unit (IEU),
University of Bristol, Bristol, UK
(Prof G Davey Smith MD); and
School of Population Health,
The University of Auckland,
Auckland, New Zealand
(Prof R Jackson PhD) (Prof G Assmann MD); The City
College of New York, New York,
NY, USA (M Trevisan MD);
Howard University Hospital,
Washington DC, USA
(R F Gillum MD); Institute of
Cardiovascular & Medical
Sciences, University of Glasgow,
Glasgow, UK (Prof I Ford PhD,
Prof N Sattar FMedSci);
International Agency for
Research on Cancer, Lyon, France
(P Ferrari PhD); MRC Integrative
Epidemiology Unit (IEU),
University of Bristol, Bristol, UK
(Prof G Davey Smith MD); and
School of Population Health,
The University of Auckland,
Auckland, New Zealand
(Prof R Jackson PhD)
Correspondence to:
Dr Angela Wood, Emerging Risk
Factors Collaboration and
EPIC-CVD Coordinating Centres,
Department of Public Health and
Primary Care, Strangeways
Research Laboratory, University
of Cambridge, Cambridge,
CB1 8RN, UK (Prof G Assmann MD); The City
College of New York, New York,
NY, USA (M Trevisan MD);
Howard University Hospital,
Washington DC, USA
(R F Gillum MD); Institute of
Cardiovascular & Medical
Sciences, University of Glasgow,
Glasgow, UK (Prof I Ford PhD,
Prof N Sattar FMedSci);
International Agency for
Research on Cancer, Lyon, France
(P Ferrari PhD); MRC Integrative
Epidemiology Unit (IEU),
University of Bristol, Bristol, UK
(Prof G Davey Smith MD); and
School of Population Health,
The University of Auckland,
Auckland, New Zealand
(Prof R Jackson PhD)
Correspondence to:
Dr Angela Wood, Emerging Risk
Factors Collaboration and
EPIC-CVD Coordinating Centres,
Department of Public Health and
Primary Care, Strangeways
Research Laboratory, University
of Cambridge, Cambridge,
CB1 8RN, UK Professor John Danesh, Emerging
Risk Factors Collaboration and
EPIC-CVD Coordinating Centres,
Department of Public Health and
Primary Care, Strangeways
Research Laboratory, University
of Cambridge, Cambridge
CB1 8RN, UK www.thelancet.com Vol 391 April 14, 2018 Discussioni The main finding of this analysis was that the threshold
for lowest risk for all-cause mortality was about 100 g per
week. For men, we estimated that long-term reduction of
alcohol consumption from 196 g per week (the upper
limit recommended in US guidelines) to 100 g per week
or below was associated with about 1–2 years of longer
life expectancy at age 40 years. Exploratory analyses
suggested that drinkers of beer or spirits, as well as binge
drinkers, had the highest risk for all-cause mortality. g
y
Our study has highlighted the complex and diverse
potential mechanisms by which alcohol consumption
may exert cardiovascular effects.41,42 It has shown that the
association between alcohol consumption and total
cardiovascular disease risk comprises several distinct
and opposite dose–response curves, rather than a single
J-shaped association. In particular, whereas higher
alcohol consumption was roughly linearly associated
with a higher risk of all stroke subtypes, coronary dis
ease excluding myocardial infarction, heart failure, and
several less common cardiovascular disease subtypes, it
was approximately log-linearly associated with a lower
risk of myocardial infarction. Our results are concordant
with recent observational data and Mendelian ran
domisation studies.16,43–46 We noted little heterogeneity in the studies contrib
uting results for stroke (I²=12%), myocardial infarc
tion (I²=12%), coronary disease excluding myocardial
infarction (I²=26%), heart failure (I²=4%) or deaths from
other types of cardiovascular disease (I²=33%; figure 3). HRs for the cardiovascular disease outcomes we
studied were broadly similar for different geographical
regions, decade of study enrolment, by data source
(ie, ERFC, EPIC-CVD, and UK Biobank), and alcohol
assessment method (appendix pp 40–42). HRs for the
cardiovascular disease outcomes were generally higher at
younger ages, but did not vary substantially by sex,
history of diabetes, proatherogenic lipids, BMI, smoking
status, or other individual-level characteristics (appendix Articles Women who reported drinking
above either the UK threshold (112 g per week) or US
threshold (98 g per week) had about 1·3 (1·1–1·5) years
shorter life expectancy at age 40 years compared with
women who reported drinking below these thresholds
(appendix p 47). About 20% of the alcohol-related survival
difference for men (and slightly less for women) was
attributed to excess death from cardiovascular disease
(appendix p 47). Similar findings to those for the US
population were observed when modelling was based on
EU mortality rates (data not shown). Figure 4: Estimated future years of life lost by extent of reported baseline alcohol consumption compared
with those who reported consuming >0–≤100 g per week The estimates of cumulative survival from 40 years of age onwards in the alcohol-drinking groups were calculated
by applying hazard ratios (specific to age at risk) for all-cause mortality associated with categorised baseline
alcohol consumption to US death rates at the age of 40 years or older. Mean usual levels of alcohol consumption
within each baseline alcohol consumption category were 56, 123, 208 and 367 g per week, respectively, for the
groups >0–≤100 g per week, >100–≤200 g per week, >200–≤350 g per week, and >350 g per week. recorded fewer than five events for a particular outcome
(appendix p 36); provided separate analyses of men and
women (appendix p 17, appendix p 26); omitted outcomes
recorded in the initial 5 years of follow-up (appendix p 18);
excluded participants with diabetes or other known
chronic diseases at baseline (appendix p 18); and restricted
the analyses to studies that recorded both non-fatal and
fatal endpoints (appendix p 37). Associations of baseline
alcohol consumption with all-cause mortality were
stronger in drinkers of beer or spirits than of wine, and in
those drinking less frequently (when consuming the same
weekly amount), including binge drinkers (appendix
p 38). However, people showing these behaviours had
higher baseline levels of smoking and other indicators of
lower socioeconomic status, suggesting the potential for
confounding effects (appendix pp 19–20). For cardio
vascular disease subtypes, HRs tended to be higher in
beer and spirit drinkers than in wine drinkers, but not
significantly so in direct comparisons involving a common
set of participants (appendix p 39). Articles However, we observed
notable differences in baseline characteristics between
never drinkers and current drinkers (eg, in relation to sex,
ethnicity, smoking, and diabetes status; appendix p 12),
supporting the validity of focusing on current drinkers in our main analysis. We recorded similar findings to
those reported above in sensitivity analyses that involved
the following approaches: used multiple imputation
rather than complete-case analysis (appendix p 32); used
fractional polynomials (appendix p 34); used a fixed-effect
meta-analysis (appendix p 35); included studies that www.thelancet.com Vol 391 April 14, 2018 1519 Articles Declaration of interests ASB reports grants from European Commission Framework 7
(HEALTH-F2-2012-279233), the European Research Council (268834),
the British Heart Foundation (SP/09/002 and RG/08/014 and
RG13/13/30194), the UK Medical Research Council (G0800270 and
MR/L003120/1), from National Institute for Health Research (through the
NIHR Cambridge Biomedical Research Centre), during the conduct of the
study; and grants from Merck, Biogen, Bioverativ, Novartis, and Pfizer,
outside the submitted work. BMP reports that he serves on the DSMB of
a clinical trial funded by Zoll LifeCor and on the Steering Committee of
the Yale Open Data Access Project funded by Johnson & Johnson. MD
reports grants from Japan Society for the Promotion of Science, during
the conduct of the study. EDA reports grants from European Commission
Framework 7, the European Research Council, the British Heart
Foundation, the UK Medical Research Council, and the National Institute
for Health Research, during the conduct of the study; and grants from
NHS Blood and Transplant, outside the submitted work. EB reports
grants from the National Health and Medical Research Council of
Australia, during the conduct of the study. HMK reports a research
agreement (through Yale) from Johnson & Johnson (Janssen) and
Medtronic to develop methods of clinical trial data sharing; personal fees
from UnitedHealth, IBM Watson, Element Science, and Aetna; a personal
health information platform from Hugo; grants from the FDA and
Medtronic; and contracts from Centers for Medicare & Medicaid Services
to develop and maintain measures that are publicly reported, outside the
submitted work. JD reports grants from the UK Medical Research
Council, the British Heart Foundation, the UK National Institute of
Health Research, and the European Commision, during the conduct of
the study; personal fees and non-financial support from Merck Sharp and
Dohme UK Atherosclerosis, personal fees and non-financial support from
Novartis Cardiovascular and Metabolic Advisory Board, grants from the
British Heart Foundation, European Research Council, Merck, the
National Institute of Health Research, NHS Blood and Transplant,
Novartis, Pfizer, the UK Medical Research Council, the Wellcome Trust,
and AstraZeneca, and personal fees and non-financial support from Pfizer Nevertheless, our study has some potential limitations. Self-reported alcohol consumption data are prone to
bias and are challenging to harmonise across studies
conducted over different time periods that used varying
instruments and methods to record such data.20,57 We
did not, however, identify major differences in results
across studies that used differing alcohol measurement
instruments. amw79@medschl.cam.ac.uk Professor John Danesh, Emerging
Risk Factors Collaboration and
EPIC-CVD Coordinating Centres,
Department of Public Health and
Primary Care, Strangeways
Research Laboratory, University
of Cambridge, Cambridge
CB1 8RN, UK jd292@medschl.cam.ac.uk
See Online for appendix www.thelancet.com Vol 391 April 14, 2018 1520 Articles Our results contribute toward understanding of the
basis for these directionally divergent cardiovascular
disease associations. For example, our data have
suggested that elevated systolic blood pressure could
mediate alcohol consumption’s positive association
with stroke and coronary disease excluding myocardial
infarction.44,47,48 By contrast, pathways related to HDL-C
(but not necessarily HDL-C itself49–52) could mediate
alcohol consumption’s inverse association with myo
cardial infarction. Both blood pressure and HDL-C are
known to increase in response to alcohol consumption.50
They have contrasting associations with cardiovascular
disease outcomes: the inverse association of HDL-C
with cardiovascular disease is substantially stronger for
coronary disease than stroke,53,54 whereas the positive
association of systolic blood with cardiovascular disease
is considerably stronger for stroke than coronary
disease.55 However, we did not find convincing evidence
that other known risk factors were important mediators
or confounders. effects of reverse causation (especially since some
contributing studies did not record baseline chronic
disease other than cardiovascular disease). Therefore,
alternative study designs including randomised trials58
are needed, to control more completely for residual biases
(including those related to studying ex-drinkers and
never-drinkers). In conclusion, our study shows that among current
drinkers, the threshold for lowest risk of all-cause
mortality was about 100 g per week. For cardiovascular
disease subtypes other than myocardial infarction, there
were no clear thresholds below which lower alcohol
consumption stopped being associated with a lower
disease risk. These data support adoption of lower limits
of alcohol consumption than are recommended in most
current guidelines. Coordinating centre Coordinating centre
Thomas Bolton, Stephen Burgess, Adam S Butterworth,
Emanuele Di Angelantonio, Stephen Kaptoge, Lisa Pennells,
Catherine Perry, David Stevens, Sarah Spackman, Simon G Thompson,
Matthew Walker, Angela M Wood, and John Danesh (principal
investigator). Catherine Perry, David Stevens, Sarah Spackman, Simon G Thompson,
Matthew Walker, Angela M Wood, and John Danesh (principal
investigator). www.thelancet.com Vol 391 April 14, 2018 Contributors
ll h
h All the authors contributed to data collection, and to the design, analysis,
interpretation, and re-drafting of this report. AMW and SK had full
access to the combined data and did the statistical analysis. AMW, EDA,
and JD drafted the manuscript and had responsibility for submission of
the manuscript for publication. Our study’s access to individual-participant data avoided
limitations of previous literature-based reviews.56 To limit
reverse causality, our study focused on current drinkers
without baseline cardiovascular disease and omitted the
initial period of follow-up. To limit confounding, our
study adjusted for a variety of risk factors. To correct for
misclassification in alcohol consumption and covariates,
our study also used extensive information on serial
assessments. Our results were robust to a variety of
sensitivity analyses. Generalisability of the findings
was enhanced by inclusion of data from 83 prospective
studies based in many different high-income countries
recruited between 1964 and 2010. Although alcohol
consumption levels declined during this period, HRs
were similar over calendar time. Data management team Data management team
Thomas Bolton, Catherine Perry, Sarah Spackman, and Matthew Walker. g
Thomas Bolton, Catherine Perry, Sarah Spackman, and Matthew Walker. Thomas Bolton, Catherine Perry, Sarah Spackman, and Matthew Wa Acknowledgments The study’s coordinating centre (Emerging Risk Factors Collaboration
and EPIC-CVD Coordinating Centres, Department of Public Health and
Primary Care, University of Cambridge, Strangeways Research
Laboratory, Cambridge, UK) has been underpinned by grants from the
UK Medical Research Council (G0800270 and MR/L003120/1), British
Heart Foundation (SP/09/002, RG/08/014 and RG13/13/30194), National
Institute for Health Research (through the National Institute for Health
Research Cambridge Biomedical Research Centre), European
Commission Framework 7 (through the EPIC-CVD award;
HEALTH-F2-2012-279233), and the European Research Council
(through an Advanced Investigator Award to JD; 268834). JD holds a
BHF Professorship and NIHR Senior Investigator Award. A study
website Funding for the EPIC-InterAct project was provided by the EU
FP6 programme (grant number LSHM_CT_2006_037197). A study
website includes a list that investigators have provided of funding
agencies that have supported individual EPIC centres. A study website
includes a list that investigators have provided of funding agencies that
have supported individual cohorts of the ERFC contributing to the
present consortium. This research has been conducted using the UK
Biobank resource (application 21886). We thank Nicola Kerrison and
Stephen Sharp (both from the University of Cambridge MRC
Epidemiology Unit, Cambridge, UK) for the former’s data management
in the EPIC-InterAct subcohort and the latter’s statistical input into
d
l
t f th EPIC CVD’
l ti
l
id li 15
Xi B, Veeranki SP, Zhao M, Ma C, Yan Y, Mi J. Relationship of
alcohol consumption to all-cause, cardiovascular, and cancer-related
mortality in U.S. adults. J Am Coll Cardiol 2017; 70: 913–22. 16
Smyth A, Teo KK, Rangarajan S, et al. Alcohol consumption and
cardiovascular disease, cancer, injury, admission to hospital,
and mortality: a prospective cohort study. Lancet 2015; 386: 1945–54. 17
Bell S, Daskalopoulou M, Rapsomaniki E, et al. Association between
clinically recorded alcohol consumption and initial presentation of
12 cardiovascular diseases: population based cohort study using
linked health records. BMJ 2017; 356: j909. 18
Jackson R, Broad J, Connor J, Wells S. Alcohol and ischaemic heart
disease: probably no free lunch. Lancet 2005; 366: 1911–12. 19
Knott CS, Coombs N, Stamatakis E, Biddulph JP. All cause
mortality and the case for age specific alcohol consumption
guidelines: pooled analyses of up to 10 population based cohorts. BMJ 2015; 350: h384. 20 Emberson JR, Bennett DA. Effect of alcohol on risk of coronary
heart disease and stroke: causality, bias, or a bit of both? Vasc Health Risk Manag 2006; 2: 239–49. Articles Population Research Advisory Panel, outside the submitted work. ML reports grants from National Institutes of Health, during the conduct
of the study; grants from National Kidney Foundation, outside the
submitted work; and Funding from the National Institutes of Health,
Grant 5U10AA025286, to Johns Hopkins University. MS reports grants
from the UK Medical Research Council, the British Heart Foundation, the
National Institute for Health Research, European Commission
Framework 7, and the European Research Council, during the conduct of
the study. NvS reports grants from the Netherlands Ministry of Health
Welfare and Sports, Directorate of Long-Term Care, during the conduct of
the study. OHF reports grants from Nestle and Metagenics, outside the
submitted work. PJN reports grants from the NIH, during the conduct of
the study. SGT reports grants from the UK Medical Research Council and
the British Heart Foundation, during the conduct of the study. SKi reports
grants from FFG COMET program: “Research Center of Excellence in
Vascular Ageing—Tyrol, VASCage” (K-Project No. 843536) funded by the
BMVIT, BMWFW, Wirtschaftsagentur Wien and Standortagentur Tirol,
outside the submitted work. SKa reports grants from the UK Medical
Research Council and the British Heart Foundation, during the conduct
of the study. WK reports personal fees from AstraZeneca, Novartis, Pfizer,
The Medicines Company, GSK, DalCor, Sanofi, Berlin-Chemie, Kowa, and
Amgen; grants and non-financial support from Roche Diagnostics,
Beckmann, Singulex, and Abbott, outside the submitted work. The other
authors declare no competing interests. 7
Corrao G, Bagnardi V, Zambon A, La Vecchia C. A meta-analysis of
alcohol consumption and the risk of 15 diseases. Prev Med 2004;
38: 613–19. 8
Hvidtfeldt UA, Tolstrup JS, Jakobsen MU, et al. Alcohol intake and
risk of coronary heart disease in younger, middle-aged, and older
adults. Circulation 2010; 121: 1589–97. 8 9
Patra J, Taylor B, Irving H, et al. Alcohol consumption and the risk
of morbidity and mortality for different stroke types—a systematic
review and meta-analysis. BMC Public Health 2010; 10: 258. 10 10
Ronksley PE, Brien SE, Turner BJ, Mukamal KJ, Ghali WA. Association of alcohol consumption with selected cardiovascular
disease outcomes: a systematic review and meta-analysis. BMJ 2011;
342: d671. 11
Roerecke M, Rehm J. The cardioprotective association of average
alcohol consumption and ischaemic heart disease: a systematic
review and meta-analysis. Addiction 2012; 107: 1246–60. 12
Bergmann MM, Rehm J, Klipstein-Grobusch K, et al. For the study website of the
funding agencies that have
supported individual EPIC
centres see http://www.phpc.
cam.ac.uk/ceu/epic_cvd/
funding-sources/
For the study website of
funding agencies that have
supported individual cohorts
of the ERFC see http://www.
phpc.cam.ac.uk/ceu/erfc/list-of-
studies/ Articles The association of pattern of lifetime alcohol use and cause of death
in the European Prospective Investigation into Cancer and
Nutrition (EPIC) study. Int J Epidemiol 2013; 42: 1772–90. 13
Leong DP, Smyth A, Teo KK, et al. Patterns of alcohol consumption
and myocardial infarction risk: observations from 52 countries in
the INTERHEART case-control study. Circulation 2014; 130: 390–98. 14
Ferrari P, Licaj I, Muller DC, et al. Lifetime alcohol use and overall
and cause-specific mortality in the European Prospective
Investigation into Cancer and nutrition (EPIC) study. BMJ Open
2014; 4: e005245. Acknowledgments 21
Ng Fat L, Cable N, Marmot MG, Shelton N. Persistent long-standing
illness and non-drinking over time, implications for the use of
lifetime abstainers as a control group. J Epidemiol Community Health 2014; 68: 71–77. Epidemiology Unit, Cambridge, UK) for the former’s data management
in the EPIC-InterAct subcohort and the latter’s statistical input into
development of the EPIC-CVD’s analytical guidelines. Epidemiology Unit, Cambridge, UK) for the former’s data management
in the EPIC-InterAct subcohort and the latter’s statistical input into
development of the EPIC-CVD’s analytical guidelines. 22 Danesh J, Erqou S, Walker M, et al. The Emerging Risk Factors
Collaboration: Analysis of individual data on lipid, inflammatory
and other markers in over 1.1 million participants in
104 prospective studies of cardiovascular diseases. Eur J Epidemiol
2007; 22: 839–69. References 1
Department of Health. Alcohol Guidelines Review: Report from the
guidelines development group to the UK Chief Medical Officers. 2016. https://www.gov.uk/government/uploads/system/uploads/
attachment_data/file/545739/GDG_report-Jan2016.pdf (accessed
Feb 5, 2018). 23 Danesh J, Saracci R, Berglund G, et al. EPIC-Heart:
The cardiovascular component of a prospective study of nutritional,
lifestyle and biological factors in 520,000 middle-aged participants
from 10 European countries. Eur J Epidemiol 2007; 22: 129–41. 2
Kalinowski A, Humphreys K. Governmental standard drink
definitions and low-risk alcohol consumption guidelines in
37 countries. Addiction 2016; 111: 1293–98. 2
Kalinowski A, Humphreys K. Governmental standard drink
definitions and low-risk alcohol consumption guidelines in
37 countries. Addiction 2016; 111: 1293–98. 24
Sudlow C, Gallacher J, Allen N, et al. UK Biobank: an open access
resource for identifying the causes of a wide range of complex
diseases of middle and old age. PLoS Med 2015; 12: e1001779. 3
Stampfer MJ, Colditz GA, Willett WC, Speizer FE, Hennekens CH. A prospective study of moderate alcohol consumption and the risk
of coronary disease and stroke in women. N Engl J Med 1988;
319: 267–73. 3
Stampfer MJ, Colditz GA, Willett WC, Speizer FE, Hennekens CH. A prospective study of moderate alcohol consumption and the risk
of coronary disease and stroke in women. N Engl J Med 1988;
319: 267–73. 25 Conen D. Alcohol consumption and incident cardiovascular disease:
not just one unifying hypothesis. Eur Heart J 2015; 36: 897–98. 26 Prentice RL. A case-cohort design for epidemiologic cohort studies
and disease prevention trials. Biometrika 1986; 73: 1–11. 4
Boffetta P, Garfinkel L. Alcohol drinking and mortality among men
enrolled in an American Cancer Society prospective study. Epidemiology 1990; 1: 342–48. 27
Thompson S, Kaptoge S, White I, et al. Statistical methods for the
time-to-event analysis of individual participant data from multiple
epidemiological studies. Int J Epidemiol 2010; 39: 1345–59. 5
Thun MJ, Peto R, Lopez AD, et al. Alcohol consumption and
mortality among middle-aged and elderly U.S. adults. N Engl J Med
1997; 337: 1705–14. 5
Thun MJ, Peto R, Lopez AD, et al. Alcohol consumption and
mortality among middle-aged and elderly U.S. adults. N Engl J Med
1997; 337: 1705–14. 28 Fibrinogen Studies Collaboration, Wood AM, White I,
Thompson SG, Lewington S, Danesh J. Regression dilution
methods for meta-analysis: assessing long-term variability in
plasma fibrinogen among 27,247 adults in 15 prospective studies. Int J Epidemiol 2006; 35: 1570–8. 6
Mukamal K, Conigrave K, Mittleman M, et al. For the study website of
funding agencies that have
supported individual cohorts
of the ERFC see http://www.
phpc.cam.ac.uk/ceu/erfc/list-of-
studies/ Declaration of interests Despite our study’s access to extensive serial
alcohol re-surveys from mid-life, our study could not
investigate alcohol consumption during the entire life
course. Misclassification in outcomes would have diluted
dose-response
associations,
suggesting
that
true
underlying associations of alcohol consumption with
cardiovascular disease subtypes are stronger and more
divergent than we observed. Because we did not generally
have access to additional alcohol-related adverse out
comes (eg, non-fatal liver disease, injuries, or psychiatric
comorbidities), we probably under-estimated potential
benefits associated with lowering alcohol consumption. Because some individuals who reduced, but did not cease,
alcohol consumption due to health complications were
probably included in our analysis, we cannot exclude the 1521 Articles Articles For the study website of the
funding agencies that have
supported individual EPIC
centres see http://www.phpc.
cam.ac.uk/ceu/epic_cvd/
funding-sources/ For the study website of the
funding agencies that have
supported individual EPIC
centres see http://www.phpc.
cam.ac.uk/ceu/epic_cvd/
funding-sources/ References Roles of drinking
pattern and type of alcohol consumed in coronary heart disease in
men. N Engl J Med 2003; 348: 109–18. 6
Mukamal K, Conigrave K, Mittleman M, et al. Roles of drinking
pattern and type of alcohol consumed in coronary heart disease in
men. N Engl J Med 2003; 348: 109–18. 1522 www.thelancet.com Vol 391 April 14, 2018 Articles 29 Wood AM, Thompson SG, Kostis JB, et al. Correcting for
multivariate measurement error by regression calibration in
meta-analyses of epidemiological studies. Stat Med 2009;
28: 1067–92. 45
Au Yeung SL, Jiang C, Cheng KK, et al. Moderate alcohol use and
cardiovascular disease from Mendelian randomization. PLoS One
2013; 8: e68054. 46
Yun KE, Chang Y, Yun S-C, et al. Alcohol and coronary artery
calcification: an investigation using alcohol flushing as an
instrumental variable. Int J Epidemiol 2017; 46: 950–62. 30 Easton DF, Peto J, Babiker AG. Floating absolute risk:
an alternative to relative risk in survival and case-control analysis
avoiding an arbitrary reference group. Stat Med 1991; 10: 1025–35. 47
Taylor AE, Lu F, Carslake D, et al. Exploring causal associations of
alcohol with cardiovascular and metabolic risk factors in a Chinese
population using Mendelian randomization analysis. Sci Rep 2015;
5: 14005. 31
Plummer M. Improved estimates of floating absolute risk. Stat Med 2004; 23: 93–104. 32 Royston P, Ambler G, Sauerbrei W. The use of fractional
polynomials to model continuous risk variables in epidemiology. Int J Epidemiol 1999; 28: 964–74. 48
Chen L, Smith GD, Harbord RM, Lewis SJ. Alcohol intake and blood
pressure: a systematic review implementing a Mendelian
randomization approach. PLoS Med 2008; 5: e52. 33 Higgins JPT, Thompson SG. Quantifying heterogeneity in a
meta-analysis. Stat Med 2002; 21: 1539–58. 49
Zanoni P, Khetarpal SA, Larach DB, et al. Rare variant in scavenger
receptor BI raises HDL cholesterol and increases risk of coronary
heart disease. Science 2016; 351: 1166–71. 34
Begg CB, Mazumdar M. Operating characteristics of a rank
correlation test for publication bias. Biometrics 1994; 50: 1088. 50 Brien SE, Ronksley PE, Turner BJ, Mukamal KJ, Ghali WA. Effect of alcohol consumption on biological markers associated with
risk of coronary heart disease: systematic review and meta-analysis
of interventional studies. BMJ 2011; 342: d636. 35 Egger M, Davey Smith G, Schneider M, et al. Bias in meta-analysis
detected by a simple, graphical test. BMJ 1997; 315: 629–34. References 36 Centers for Disease Control and Prevention, National Center for
Health Statistics. Underlying cause of death 1999–2015 on CDC
WONDER Online Database (released Dec, 2016). https://wonder. cdc.gov/ucd-icd10.html (accessed Feb 5, 2018). 51
Voight BF, Peloso GM, Orho-Melander M, et al. Plasma HDL
cholesterol and risk of myocardial infarction: a mendelian
randomisation study. Lancet 2012; 380: 572–80. 37
WHO. WHO Statistical Information System (WHOSIS). Geneva,
Switzerland: World Health Organization, 2007. 52 HPS3/TIMI55-REVEAL Collaborative Group. Effects of anacetrapib
in patients with atherosclerotic vascular disease. N Engl J Med 2017;
377: 1217–27. 38 UN Population Division. World Population Prospects. New York,
NY: United Nations, 2005. 53 The Emerging Risk Factors Collaboration, Di Angelantonio E,
Sarwar N, et al. Major lipids, apolipoproteins, and risk of vascular
disease. JAMA 2009; 302: 1993–2000. 39 Emerging Risk Factors Collaboration, Di Angelantonio E,
Kaptoge S, et al. Association of cardiometabolic multimorbidity
with mortality. JAMA 2015; 314: 52–60. 54
Woodward M, Barzi F, Feigin V, et al. Associations between
high-density lipoprotein cholesterol and both stroke and coronary
heart disease in the Asia Pacific region. Eur Heart J 2007; 28: 2653–60. 40
IARC Working Group on the Evaluation of Carcinogenic Risks
to Humans. Personal habits and indoor combustions. Volume 100 E. A review of human carcinogens. IARC Monogr Eval Carcinog Risks Hum 2012; 100 (Pt E): 1–538. 55 MacMahon S, Peto R, Collins R, et al. Blood pressure, stroke, and
coronary heart disease. Part 1, prolonged differences in blood
pressure: prospective observational studies corrected for the
regression dilution bias. Lancet 1990; 335: 765–74. 41
Lazo M, Chen Y, McEvoy JW, et al. Alcohol consumption and
cardiac biomarkers: The atherosclerosis risk in communities
(ARIC) study. Clin Chem 2016; 62: 1202–10. 56 Ahmed I, Sutton AJ, Riley RD. Assessment of publication bias,
selection bias, and unavailable data in meta-analyses using individual
participant data: a database survey. BMJ 2012; 344: d7762. 42
Kiechl S, Willeit J. Complex association between alcohol
consumption and myocardial infarction: always good for a new
paradox. Circulation 2014; 130: 383–86. 57
Del Boca FK, Darkes J. The validity of self-reports of alcohol
consumption: state of the science and challenges for research. Addiction 2003; 98: 1–12. 43
Holmes MV, Dale CE, Zuccolo L, et al. Association between
alcohol and cardiovascular disease: Mendelian randomisation
analysis based on individual participant data. BMJ 2014;
349: g4164. 58 Mukamal KJ, Clowry CM, Murray MM, et al. www.thelancet.com Vol 391 April 14, 2018 References Moderate alcohol
consumption and chronic disease: the case for a long-term trial. Alcohol Clin Exp Res 2016; 40: 2283–91. 44
Cho Y, Shin S-Y, Won S, Relton CL, Davey Smith G, Shin M-J. Alcohol intake and cardiovascular risk factors: a Mendelian
randomisation study. Sci Rep 2015; 5: 18422. 1523 www.thelancet.com Vol 391 April 14, 2018 www.thelancet.com Vol 391 April 14, 2018
|
https://openalex.org/W3048627433
|
https://europepmc.org/articles/pmc7851003?pdf=render
|
English
| null |
Diagnosis of COVID-19 using multiple antibody assays in two cases with negative PCR results from nasopharyngeal swabs
|
Infection
| 2,020
|
cc-by
| 3,567
|
CASE REPORT CASE REPORT Diagnosis of COVID‑19 using multiple antibody assays in two cases
with negative PCR results from nasopharyngeal swabs Marianna Theresia Traugott1 · Wolfgang Hoepler1 · Tamara Seitz1 · Sebastian Baumgartner1 · Mario Karolyi1 ·
Erich Pawelka1 · Emanuela Friese1 · Stephanie Neuhold1 · Hasan Kelani1 · Florian Thalhammer2 · Alexander Zoufaly1 ·
Hermann Laferl1 · Judith Helene Aberle3 · Christoph Wenisch1 · Elisabeth Puchhammer‑Stöckl3 · Karin Stiasny3 ·
Stephan Walter Aberle3 · Lukas Weseslindtner3 Received: 27 May 2020 / Accepted: 30 July 2020 / Published online: 12 August 2020
© The Author(s) 2020 * Lukas Weseslindtner
lukas.weseslindtner@meduniwien.ac.at Abstract We report of two cases of progressed COVID-19 with negative PCR tests from nasopharyngeal swabs, in whom diagnosis
was made by different antibody assays, including a lateral flow rapid test and multiple commercial ELISAs, finally confirmed
by comprehensive serological assays. These cases highlight that commercial ELISAs and even rapid tests might significantly
aid the diagnosis of COVID-19, particularly, if a combination of serological assays is used with a specific clinical question,
in severely ill patients after seroconversion and when comprehensive serological methods are used for confirmation. Keywords SARS · Coronavirus · Antibodies · Immunoassay · Rapid test · PCR negative Abbreviations
Ab
Total antibody
COVID-19
Coronavirus disease 2019
CLIA
Chemiluminescent immunoassay
CRP
C-reactive protein
CT
Computer tomography
PCR
Polymerase chain reac
WBC
White blood cell count
Introduction Abbreviations
Ab
Total antibody
COVID-19
Coronavirus disease 2019
CLIA
Chemiluminescent immunoassay
CRP
C-reactive protein
CT
Computer tomography
ELISA
Enzyme-linked immunosorbent assay
HFNC
High-flow nasal cannula
ICU
intensive care unit
Ig
Immunoglobulin
SARS-CoV-2
Severe acute respiratory syndrome corona-
virus 2
PC
W
In
SA
cau
fir
20
cau
he Infection (2021) 49:171–175
https://doi.org/10.1007/s15010-020-01497-2 Infection (2021) 49:171–175
https://doi.org/10.1007/s15010-020-01497-2 CASE REPORT Introduction SARS-coronavirus 2 (SARS-CoV-2) is a novel coronavirus
causing coronavirus disease 2019 (COVID-19) [1]. After
first cases were described in Wuhan, China in December
2019, the virus has been rapidly spreading all over the world,
causing a global pandemic with massive consequences to the
health-care systems worldwide [2]. Immediately after discovery of the virus, polymerase
chain reaction (PCR) from the upper airway (nasopharyn-
geal swabs) became the main diagnostic test method [3]. However, false-negative PCR results have been described by
several studies [4–6]. Additionally, chest computer tomogra-
phy (CT) scans have diagnostic value in advanced disease,
which even caused a change of the original case definition
in China, using clinical symptoms and radiological signs
rather than PCR as main diagnostic criteria [7, 8]. Typical
radiological signs for COVID-19 are peripheral, subpleural
ground glass opacities, often located in the lower lobes [7]. Electronic supplementary material The online version of this
article (https://doi.org/10.1007/s15010-020-01497-2) contains
supplementary material, which is available to authorized users. * Lukas Weseslindtner
lukas.weseslindtner@meduniwien.ac.at
1
4th Medical Department, Department of Infectious Diseases
and Tropical Medicine, Kaiser-Franz-Josef Hospital, Vienna,
Austria
2
Division of Infectious Diseases and Tropical Medicine,
Department of Medicine I, Medical University Vienna,
Vienna, Austria
3
Center for Virology, Medical University of Vienna,
Kinderspitalgasse 15, 1090 Vienna, Austria * Lukas Weseslindtner
lukas.weseslindtner@meduniwien.ac.at 1
4th Medical Department, Department of Infectious Diseases
and Tropical Medicine, Kaiser-Franz-Josef Hospital, Vienna,
Austria 2
Division of Infectious Diseases and Tropical Medicine,
Department of Medicine I, Medical University Vienna,
Vienna, Austria As additional diagnostic tool, serological assays for
SARS-CoV-2-specific antibodies have recently become
available. While there is a high public demand to use these
tests in the common population to identify individuals with 3
Center for Virology, Medical University of Vienna,
Kinderspitalgasse 15, 1090 Vienna, Austria (0123
1 3456789)
3 172 M. T. Traugott et al. swab for SARS-CoV-2-PCR was performed, but again gave
a negative result (this time using an in-house SARS-CoV-2
real-time TaqMan PCR with WHO recommended prim-
ers, a probe located in the E-gene and an internal spiked-in
control to exclude PCR inhibition, as described previously)
[3]. No antibiotics treatment had been initiated before initial
PCR testing and no gurgling with antiseptic substances was
performed.l probable immunity, we recently showed that a combination
of multiple serological assays might have excellent perfor-
mances in diagnosing SARS-CoV-2 infections in sympto-
matic, hospitalized, and severely ill patients [9]. Introduction Here, we report of two cases, in which COVID-19 was
diagnosed only by serological assays (including a lateral
flow rapid test) in patients who displayed progressed disease
and PCR negative swabs from the upper respiratory tract. p
Since the patient remained stable on high-flow nasal
cannula (HFNC) and inflammation markers decreased, he
was transferred to a non-COVID-ICU at the 12th day after
onset of symptoms (during the stay on the ICU he was never
intubated). Still struggling for a diagnosis, PCR tests for
SARS-CoV-2 were repeated on the 19th and the 20th day
post-onset of symptoms. Both PCRs out of deep sputum
(induced by NaCl inhalation) tested negative. A serum sam-
ple drawn at the 20th day post-symptom onset had arrived
too late for ELISA testing at the same day. Therefore, a
Wantai rapid test (Wantai, Beijing, China) was performed,
which showed a positive result. On the next day, the positive
test result was confirmed by the detection of SARS-CoV-
2-specific IgM (ratio: 35.83), total antibodies (ratio: 8.55),
IgA (ratio: > 12.00), and IgG (ratio: 8.52). It was the 10th
day after hospital admission and the 20th day after onset of
symptoms. Since all PCR tests continuously tested negative,
antibody cross-reaction with other coronaviruses was not
excluded, although seroconversion of anti SARS-CoV-2 was
clearly identified by four ELISAs. To rule out infection with
other coronaviruses, the last sputum sample was retested
for coronaviruses OC43, 229E, HKU-1, and NL63 by PCR
(using specific primers) giving negative results. Finally,
an immunoblot microarray (Viramed, Munich, Germany),
which concordantly detected high levels of IgM, IgA, and
IgG against all viral proteins S1, S2, and nucleocapsid con-
firmed the presence of SARS-CoV-2-specific antibodies, and
thereby SARS-CoV-2 infection. Case 1 A 71 year old, male patient was transferred to our COVID-
19 intensive care unit (ICU) because of rapid respiratory
deterioration with high suspicion of COVID-19 based on a
chest CT-scan revealing multiple, bilateral, partially conflu-
ent ground glass opacities, thickening of the interlobar sep-
tum, and the bronchial wall and small focal consolidations. The result of his SARS-CoV-2 nasopharyngeal swab was
still pending at this time. The patient had a medical history
of prostatic hyperplasia and arterial hypertension (treated
with bisoprolol 5 mg q.d) and reported a high fever and
strong retrobulbar headache for about 10 days but no cough. He had been on a cruise ship journey in the Caribbean at the
end of February (he could not recall the exact dates), from
where he returned via the Dominican Republic. A malaria
rapid dipstick test and smear and a dengue rapid test gave
negative results. While the patient could be stabilized in our ICU, PCR
testing for SARS-CoV-2 from a nasopharyngeal swab gave
a negative test result (Roche® COBAS 6800, CE-IVD Assay;
Roche Diagnostics, Rotkreuz, Switzerland). Isolation meas-
ures, however, were maintained due to unchanged high sus-
picion of COVID-19. The initial laboratory results showed
a normal white blood cell count (WBC) of 7500 G/l with
lymphocytopenia (9%), thrombocytopenia (149 G/l), and
an elevated C-reactive protein (CRP) of 140 mg/l. Antibi-
otic therapy with intravenous cefuroxime and azithromycin
was started. A PCR out of sputum (pneumonia panel plus,
Biomerieux®) targeting 33 different kinds of bacteria and
viruses (incl. influenza A/B, Legionella, Mycoplasma) was
all negative, as was the HIV test. At the 11th day post-onset
of symptoms, a blood sample was sent to the Center for
Virology for PCR and serology testing for tropical infec-
tious diseases (chikungunya IgM, IgG and PCR, dengue
hemagglutination inhibition assay, and Zika virus IgM and
neutralization assay: all negative), as well as for SARS-
CoV-2-specific serology. However, SARS-CoV-2-specific
antibodies tested negative for IgM and total antibodies
(both Wantai, Beijing, China) as well as for IgA and IgG
(both Euroimmun, Lübeck, Germany) using enzyme-linked
immunosorbent assays (ELISAs). A second nasopharyngeal Conclusion With respect to COVID-19 diagnosis, these two cases high-
light important aspects: First, they demonstrate limitations
of PCR-based tests from nasopharyngeal swabs even in
advanced disease stages. Indeed, patients with progressed
pneumonia may display lower virus concentrations in the
upper respiratory tract (possibly causing false-negative PCR
results from pharyngeal swabs) [10]. It has already been
shown for SARS-coronavirus that virus concentrations in
nasopharyngeal swabs and saliva may be significantly lower
than in sputum [11, 12]. Also in SARS-CoV-2 infections,
viral RNA may be still detected by PCR in sputum when
nasopharyngeal swabs already test negative [13]. Therefore,
(deep) sputum should be preferably used for PCR analysis
when swabs test negative in patients with clinical suspicion
of COVID-19. On the day of ICU admission (day 9 after symptom onset,
3 days after hospital admission) laboratory results showed
leukopenia with a WBC of 2.99 G/l with lymphocytopenia
(0.64 G/l), aneosinophilia, thrombocytopenia (117 G/l), and
a slightly elevated CRP of 30 mg/l. Lopinavir/ritonavir had
been started already on the admission day and was contin-
ued. Oxygen was administered by HFNC with increasing
FiO2. 2
On the 10th day after symptom onset, another PCR was
performed with a swab from the upper respiratory tract and
again gave a negative result. The respiratory situation, how-
ever, deteriorated, CRP increased to 119 mg/l, and IL-6 was
256 pg/ml. In this typical constellation, cytokine storm was
suspected and Tocilizumab combined with cortisone was
administered. Until that day, the treating physicians had
never seen any positive SARS-CoV-2 test of the patient,
but for further patient management (including antiviral
therapy, cortisone therapy as well as convalescent plasma
administration), a definitive written positive COVID-19
result was crucial. The patient further deteriorated on the
12th day post-onset of symptoms and had to be intubated in
the early morning. The same morning, deep tracheal secre-
tion (tracheal aspirate), a nasopharyngeal swab, and a serum
sample were taken and sent to the Center for Virology. Due
to the urgent need of a diagnosis, a Wantai Rapid test (Wan-
tai, Beijing, China) was immediately performed and gave a
weakly positive result. With this preliminary result indicat-
ing SARS-CoV-2 infection (especially with association of
negative serology on the 7th day post-onset of symptoms),
convalescent plasma could be transfused and therapy with
remdesivir was started. Case 2 A 61 year old male patient was transferred from the normal
COVID-19 ward to our ICU because of rapidly increasing
oxygen demand. The patient reported onset of symptoms
with fever, cough, and dyspnoea 9 days earlier. The patient
had been tested positive for COVID-19 by nasopharyngeal
swab 6 days after onset of symptoms, performed at home
by the Austrian mobile outreach COVID-19-testing-service. The same day, he was admitted to our hospital, not bringing
a copy of the positive PCR result. The patient had a medi-
cal history of Hashimoto thyroiditis (treated with Euthy-
rox 150 µg q.d.), arterial hypertension (treated with can-
desartan 4 mg q.d.), and status post disc prolapse surgery. One day later (7th day post-onset of symptoms), another 1 3 3 COVID19 diagnosis using antibody assays in PCR negative cases 173 nasopharyngeal swab was taken, yielding a negative result
(in-house SARS-CoV-2 real -time TaqMan PCR with the
primers described above and using an internal spiked-in
control to exclude PCR inhibition), and also SARS-CoV-
2-specific IgA and IgG antibodies tested negative (both
Euroimmun, Lübeck, Germany). Also in this case, there
was no gurgling with antiseptic substances and no previous
treatment with antibiotics. Conclusion On the following day, the same
serum sample (drawn before the plasma transfusion and
already tested with the rapid test) was tested using ELISA,
and the positive test result by the rapid test was confirmed
(ratio IgA: 3.97, ratio IgG: 1.75). Finally, an in-house neu-
tralization assay confirmed the specificity of those antibod-
ies. The protocol of this assay is described in Supplemental
Material and Methods. PCR from the deep tracheal secretion
(obtained at the 12th day post-onset of symptoms) finally
tested positive (1.1 × 105 copies/ml) for SARS-CoV-2, while
in the nasopharyngeal swab from the upper respiratory tract,
viral RNA was still undetectable In severely ill patients with SARS-CoV-2 infection, how-
ever, specific antibodies may also be detectable and signifi-
cantly aid the diagnosis, as demonstrated in our cases [9]. Indeed, we detected significant antibody titres at the 20th
and 12th day post -onset of symptoms with five different
assays in both of our cases. Second, both cases highlight the caveat that serology is
not able to identify infections early after disease onset and
that the time span until significant antibody levels are pro-
duced may individually vary. Of note, antibody tests gave
negative results in the patients at the 11th and the 7th day
post-onset of disease, respectively, but these early acquired
samples, nonetheless, facilitated us to subsequently observe
seroconversion, thus providing evidence for the course of the
infection. Importantly, serological results should always be
interpreted in conjunction with the date of symptom onset. However, the detection rates of different immunoglobulins,
in relation to the interval since symptom onset, may vary
among different assays depending on the configuration of
the respective tests [13–15]. Third, these two cases demonstrate that even rapid tests
have potential to aid COVID-19 diagnosis. Importantly,
there are large numbers of lateral flow rapid tests currently
on the market, and each of these tests has to be thoroughly
evaluated for the specific setting of its possible use [16]. While certain rapid tests may not have sufficient sensitivity
and specificity for a widespread use in the population and
for large seroprevalence studies (especially when used as
single tests), other tests may be particularly useful in distinc-
tive patient cohorts under distinct circumstances. References Data availability Data will be made transparent upon publication. Data availability Data will be made transparent upon publication. Code availability Not applicable. in hospitalized patients with SARS-CoV-2 infection, while
the specificity of the test was 98% [9]. Since patients with
severe symptoms may display higher antibody levels than
patients with mild or asymptomatic SARS-CoV-2 infections,
the same rapid test may, thus, not have the same diagnostic
power in patients with mild disease (and lower antibody lev-
els). Thus, also the ELISA could display significantly lower
performances in seroprevalence studies covering a large part
of individuals with mild symptoms not requiring hospitali-
zation [15]. Open Access This article is licensed under a Creative Commons Attri-
bution 4.0 International License, which permits use, sharing, adapta-
tion, distribution and reproduction in any medium or format, as long
as you give appropriate credit to the original author(s) and the source,
provide a link to the Creative Commons licence, and indicate if changes
were made. The images or other third party material in this article are
included in the article’s Creative Commons licence, unless indicated
otherwise in a credit line to the material. If material is not included in
the article’s Creative Commons licence and your intended use is not
permitted by statutory regulation or exceeds the permitted use, you will
need to obtain permission directly from the copyright holder. To view a
copy of this licence, visit http://creativecommons.org/licenses/by/4.0/. The presented cases also highlight that comprehensive
serological methods (immunoblot and/or neutralization
assays) are required to finally confirm positive test results
by SARS-CoV-2 ELISAs/chemiluminescent immunoas-
says (CLIAs) or lateral flow assays (rapid tests), particu-
larly when the diagnosis of COVID-19 is solely based on
serology. Although the early evaluation studies demonstrate
acceptable specificities for commercial ELISA/CLIA test
systems, the possibility of unspecific or cross-reactions with
other coronaviruses has to excluded in clinically relevant
cases by antibody tests with highest specificity [17–19]. References 1. Zhu N, Zhang D, Wang W, Li X, Yang B, Song J, et al. A novel
coronavirus from patients with pneumonia in China, 2019. N Engl
J Med. 2020;382:727–33. https://doi.org/10.1056/NEJMoa2001
017. 2. Eurosurveillance Editorial Team. Note from the editors: World
Health Organization declares novel coronavirus (2019-nCoV)
sixth public health emergency of international concern. Euro
Surveill. 2020;25:200131e. https://doi.org/10.2807/1560-7917. ES.2020.25.5.200131e. i
In summary, these cases demonstrate the potential of
serological assays for diagnosing acute SARS-CoV-2 infec-
tions, especially when there is a high grade of clinical sus-
picion for COVID-19 and antibody testing is repeatedly per-
formed during progressing infection. Furthermore, our cases
indicate that antibody tests may additionally be a useful tool
for fast decision-making in patient management. Reactive
antibody tests indicate that the sensitivity of PCR-based tests
from nasopharyngeal swabs may be limited in progressed
COVID-19 cases and that antibody assays may provide
negative results early after the infection, important aspects
that should be considered in the diagnosis of COVID-19. 3. Corman VM, Landt O, Kaiser M, Molenkamp R, Meijer A, Chu
DKW, et al. Detection of 2019 novel coronavirus (2019-nCoV)
by real-time RT-PCR. Euro Surveill. 2020;25:2000045. https://
doi.org/10.2807/1560-7917.ES.2020.25.3.2000045. g
4. Liu L, Liu W, Zheng Y, Jiang X, Kou G, Ding J, et al. A pre-
liminary study on serological assay for severe acute respiratory
syndrome coronavirus 2 (SARS-CoV-2) in 238 admitted hos-
pital patients. Microbes Infect. 2020. https://doi.org/10.1016/j. micinf.2020.05.008. 5. Liu R, Han H, Liu F, Lv Z, Wu K, Liu Y, et al. Positive rate
of RT-PCR detection of SARS-CoV-2 infection in 4880 cases
from one hospital in Wuhan, China, from Jan to Feb 2020. Clin Chim Acta. 2020;505:172–5. https://doi.org/10.1016/j. cca.2020.03.009. Acknowledgements Open access funding provided by Medical Uni-
versity of Vienna. Acknowledgements Open access funding provided by Medical Uni-
versity of Vienna. 6. Ye G, Li Y, Lu M, Chen S, Luo Y, Wang S, et al. Experience
of different upper respiratory tract sampling strategies for detec-
tion of COVID-19. J Hosp Infect. 2020;105:1–2. https://doi. org/10.1016/j.jhin.2020.03.012. Funding This study was funded by the Medical Scientific Fund of the
Mayor of the City of Vienna. Funding This study was funded by the Medical Scientific Fund of the
Mayor of the City of Vienna. 7. Fang Y, Zhang H, Xie J, Lin M, Ying L, Pang P, et al. Sensitivity
of chest CT for COVID-19: comparison to RT-PCR. Radiology. 2020;296:E115–E11717. https://doi.org/10.1148/radiol.20202
00432. Conclusion For the
rapid test which we applied in our two cases, we recently
demonstrated a sensitivity of 80% between the 6th and 10th
day and of 100% after the 11th day post-onset of symptoms 1 3 174 M. T. Traugott et al. M. T. Traugott et al. Compliance with ethical standards 8. Tsang TK, Wu P, Lin Y, Lau EHY, Leung GM, Cowling BJ. Effect
of changing case definitions for COVID-19 on the epidemic curve
and transmission parameters in mainland China: a modelling
study. Lancet Public Health. 2020;5:e289–e296296. https://doi. org/10.1016/S2468-2667(20)30089-X. Conflict of interest The authors state that they do not have a commer-
cial or other association that might pose a conflict of interest. Conflict of interest The authors state that they do not have a commer-
cial or other association that might pose a conflict of interest. Ethics approval Informed, written consent was obtained from the
patients, and the study was approved by the ethics committee of the
Medical University of Vienna (EK 2156/2019, EK 2283/2019). g
9. Traugott M, Aberle SW, Aberle JH, Griebler H, Karolyi M,
Pawelka E, et al. Performance of SARS-CoV-2 antibody assays
in different stages of the infection: comparison of commercial
ELISA and rapid tests. J Infect Dis. 2020. https://doi.org/10.1093/
infdis/jiaa305. Consent to participate and for publication Patients gave written con-
sent to participate in studies and consented for publication of study
results. 10. To KK, Tsang OT, Leung WS, Tam AR, Wu TC, Lung DC, et al. Temporal profiles of viral load in posterior oropharyngeal saliva 1 3 175 COVID19 diagnosis using antibody assays in PCR negative cases samples and serum antibody responses during infection by SARS-
CoV-2: an observational cohort study. Lancet Infect Dis. 2020. https://doi.org/10.1016/S1473-3099(20)30196-1. 15. Zhao J, Yuan Q, Wang H, Liu W, Liao X, Su Y, et al. Antibody
responses to SARS-CoV-2 in patients of novel coronavirus disease
2019. Clin Infect Dis. 2020. https://doi.org/10.1093/cid/ciaa344. 11. Peiris JS, Chu CM, Cheng VC, Chan KS, Hung IF, Poon LL,
et al. Clinical progression and viral load in a community out-
break of coronavirus-associated SARS pneumonia: a prospective
study. Lancet. 2003;361:1767–72. https://doi.org/10.1016/s0140
-6736(03)13412-5. 16. Demey B, Daher N, Francois C, Lanoix JP, Duverlie G, Castelain
S, et al. Dynamic profile for the detection of anti-SARS-CoV-2
antibodies using four immunochromatographic assays. J Infect. 2020. https://doi.org/10.1016/j.jinf.2020.04.033. p
g
j j
17. Okba NMA, Muller MA, Li W, Wang C, GeurtsvanKessel CH,
Corman VM, et al. Severe acute respiratory syndrome coro-
navirus 2-specific antibody responses in coronavirus disease
2019 patients. Emerg Infect Dis. 2020;26:1478–88. https://doi. org/10.3201/eid2607.200841. 12. Drosten C, Chiu LL, Panning M, Leong HN, Preiser W, Tam
JS, et al. Compliance with ethical standards Evaluation of advanced reverse transcription-PCR
assays and an alternative PCR target region for detection of
severe acute respiratory syndrome-associated coronavirus. J Clin Microbiol. 2004;42:2043–7. https://doi.org/10.1128/
jcm.42.5.2043-2047.2004. 18. Bryan A, Pepper G, Wener MH, Fink SL, Morishima C, Chaud-
hary A, et al. Performance characteristics of the Abbott architect
SARS-CoV-2 IgG assay and seroprevalence in boise. Idaho J Clin
Microbiol. 2020. https://doi.org/10.1128/JCM.00941-20. j
13. Wolfel R, Corman VM, Guggemos W, Seilmaier M, Zange S,
Muller MA, et al. Virological assessment of hospitalized patients
with COVID-2019. Nature. 2020. https://doi.org/10.1038/s4158
6-020-2196-x. 19. Jaaskelainen AJ, Kekalainen E, Kallio-Kokko H, Mannonen
L, Kortela E, Vapalahti O, et al. Evaluation of commercial
and automated SARS-CoV-2 IgG and IgA ELISAs using
coronavirus disease (COVID-19) patient samples. Euro Sur-
veill. 2020;25:2000603. https://doi.org/10.2807/1560-7917. ES.2020.25.18.2000603. 14. Jin Y, Wang M, Zuo Z, Fan C, Ye F, Cai Z, et al. Diagnostic value
and dynamic variance of serum antibody in coronavirus disease
2019. Int J Infect Dis. 2020;94:49–52. https://doi.org/10.1016/j. ijid.2020.03.065. 1 3
|
https://openalex.org/W4283823244
|
https://bcpublication.org/index.php/BM/article/download/1004/1007
|
English
| null |
Horizontal construction: Research on the Proper sub-brand marketing benefit the main-brand
|
BCP business & management
| 2,022
|
cc-by
| 5,053
|
BCP Business & Management
Volume 20 (2022) BCP Business & Management
Volume 20 (2022) IEMSS 2022 1.1 Research background With the development of the Internet and social media platforms, some products will become
popular because of being aware of by people. We call this kind of product an Internet celebrity product,
such as Kendall Jenner’s brand Tequila 818 and Kylie Jenner’s Kylie Cosmetics. As long as
companies make good use of the power of the Internet, Internet celebrity products can become a
powerful marketing tool, which can help improve the competitive advantage of enterprises, increase
brand awareness and brand recognition, and increase customer loyalty to the brand [1]. One essential way for brands to create a popular product is to co-brand with influencers. Co-
branding is a marketing strategy that establishes partnerships with multiple brands on a single product
or service. Its significance lies in combining the market strength and brand awareness of two or more
brands and using the influence of different brands to drive customer consumption. Because co-
branded products contain the personal opinions of celebrities on fashion, the product itself is a
concrete embodiment of celebrity aesthetics; such products are more likely to stimulate consumers’
purchase behavior. At the same time, the brand has a cooperative relationship with celebrities,
resulting in higher product prices, which forces consumers to pay a higher premium for these products
[2]. Internet celebrities include not only ordinary people who become popular through major social
media platforms, but also contain well-known designers, celebrities, and singers who are active on
social media. Using the celebrity effect to open the market can turn influencers’ huge fan base into
loyal brand customers. The strong stickiness and loyalty of fans can bring a steady stream of business
income [3]. Horizontal construction: Research on the Proper sub-brand
marketing benefit the main-brand
Leyi Chen1, *, †, Wanni He2, †, Yihan Huang3, †
1Shekou International School, Shenzhen, 518000, China
2Ningbo University, Ningbo, 315211, China
3Changzhou Senior High School of Jiangsu Province, Changzhou, 213003, China
*Corresponding author: 24mc02@sis.org.cn
†These authors contributed equally. Abstract; Nowadays, people’s life is permeated by the Internet. Internet development has been
seized as a commercial opportunity by businesses. An increasing number of enterprises started to
use the Internet to spread awareness. This is known as the network effect, and the products which
became famous through this way are called internet celebrity products. It is believed that the Internet
is providing and will be creating a huge business opportunity as it becomes ever more prevalent. Yeezy has masterly used exciting, word-of-mouth, and hunger marketing as a typical example of
enterprises that have successfully generated network effects. Its marketing strategy is worthy of
studying and valuable for reference. In the exploration, Yeezy’s marketing strategy will be studied. In order to discover its marketing limitations and the perspectives of the enterprise by its biggest
potential consumer——Generation Z, the survey will be handled and analyzed. Keywords: Sub-brand; Main-brand; Marketing; Horizontal construction. 1.2 Literature review According to the official definition on Google, Yeezy is a fashion collaboration created by the
German sportswear company Adidas and the famous American rapper, designer, and entrepreneur
Kanye West. This collaboration has been a great success since it was released. 500,000 dollars worth 341 BCP Business & Management
Volume 20 (2022) IEMSS 2022 of products were sold in 30 minutes when the shoes were first launched [4]. Several serious of
products were sold out as soon as they were available. of products were sold in 30 minutes when the shoes were first launched [4]. Several serious of
products were sold out as soon as they were available. According to data released by Wikipedia, in 2019, the turnover of Yeezy series sneakers reached
1.3 billion dollars. In 2020, the annual revenue climbed up to 1.7 billion dollars, netting Yeezy 191
million dollars in royalties. This thesis is based on the co-branding success of Yeezy [5]. This study
aims to determine the influence of popular product marketing methods on consumers’ behavior. Why
did Yeezy shoes sell out as soon as they were released? What’s the customer’s psychology behind
this phenomenon? What determines the success of Yeezy? Studying these questions can help
companies understand the marketing methods of internet celebrity products better and fully
understand target consumer groups’ psychology and behavior. 2. Method Marketing methods are an important factor in determining whether a brand is successful. The
study’s goal is to determine the influence of Internet celebrities marketing methods on consumers’
purchasing behavior. To delve deeper into the subject, four methods will be used: textual analysis
method, survey method, case study, and comparative analysis 1.3 Research gap Most researchers focused on one of the marketing strategies Yeezy has successfully generated
through the Internet——scarcity. According to The Psychology of Persuasion by Robert Cialdini,
“people want more of those things they can have less of” [6]. Yeezy has masterly used consumer
psychology that the scarcer an item is, the more aspired people want to have it. For instance, Yeezy
released a new tinct with a semi-frozen yellow hue with red and grey highlights. The newly released
sneaker is not a new model but still Yeezy Boost 350 V2. The price of the shoe is in the same range
as other sneakers. However, the shoe is special because it was the most limited released Yeezy sneaker
to date. Yeezy was excited about the new sneaker and created a buzz on the Internet. The hype reached
its crescendo at ten in the morning when the sneakers were available for consumers to purchase. Within fifteen to thirty seconds, the shoe sold out [7]. Through case studies like this, researchers
discovered Yeezy’s marketing strategy among scarcity and how the enterprise has successfully used
network effects, including buzz words and hype, to spread awareness of the insufficient new products. On the other hand, most researchers neglected Kanye West’s celebrity effect on social media and the
strength of co-branded products. This will be further discussed in further sections. 1.4 Research framework The paper will be talking about how internet celebrity products became popular and the marketing
strategies used behind them. The passage will first introduce marketing strategies, including celebrity
and influencer effects, network effects, and advantages of co-branding brands and products. Using
data and graphs, the case study of the cooperation between Kanye West and Yeezy will then be
analyzed and discussed in the ways marketing strategies mentioned above are used. Also, it will be
compared with another brand. Further, Yeezy’s marketing strategies will be evaluated, and Yeezy’s
marketing suggestions will be put out. Finally, conclusions will be made about the article. 3.1.1 Internet celebrity product marketing Reducing the supply of products and adjusting the
relationship between supply and demand makes people more likely to believe that the brand is popular
once the supply is less than the demand. This may lead to people’s queuing behavior, which is
undoubtedly conducive to brand image forming so that the product will become popular on the
Internet [12]. 2.1 Textual analysis method Textual analysis is a broad term for various research methods used to describe, interpret, and
understand texts. All kinds of information can be gleaned from a text – from its literal meaning to the
subtext, symbolism, assumptions, and values it reveals. The methods used to conduct textual analysis
depend on the field and the research aims. It often aims to connect the text to a broader social, political,
cultural, or artistic context [8]. Through browsing websites, marketing methods of Internet celebrity
products will be found. By reading papers and reports, the marketing method of internet celebrities
affects consumers’ behavior, and consumer psychology will be discovered. 342 BCP Business & Management
Volume 20 (2022) IEMSS 2022 2.3 Case study A case study is a detailed study of a specific subject, such as a person, group, place, event,
organization, or phenomenon. Case studies are commonly used in social, educational, clinical, and
business research [10]. The marketing strategies that Yeezy generated will be discovered. Yeezy is a
successful example of an Internet celebrity. The principles behind its success will be explored. 2.2 Survey method The survey method gathers data by asking questions to people who are thought to have desired
information. A formal list of questionnaires is prepared. Generally, a non-disguised approach is used. The respondents are asked questions on their demographic interest opinion [9].100 questionnaires
will be handed out about what channels they mainly use to obtain commodity information and
purchase commodities. By distributing and analyzing the questionnaire, consumers’ preferences will
be summarized and concluded. It is useful for us to study the effects of marketing methods on
consumers in the next step. 3.1.1 Internet celebrity product marketing At present, the most common ways to create an internet celebrity product are: 1. Cooperate with influencers or celebrities and establish a partnership with them. This uses
influencer marketing, which lets influencers spread the word through various social media platforms. Around 40% of people will buy relevant products after they see them on social media, according to
the data given by Twitter and Annalect [11]. 1. Cooperate with influencers or celebrities and establish a partnership with them. This uses
influencer marketing, which lets influencers spread the word through various social media platforms. Around 40% of people will buy relevant products after they see them on social media, according to
the data given by Twitter and Annalect [11]. 2. Enhance brand or product awareness through word-of-mouth marketing strategy on well-known
social media platforms. Many brands will choose social media such as YouTube and Instagram to
promote their products, mainly because these platforms have more users than other apps do. What’s
more, some Internet celebrities are relatively more active on these social media platforms. 2. Enhance brand or product awareness through word-of-mouth marketing strategy on well-known
social media platforms. Many brands will choose social media such as YouTube and Instagram to
promote their products, mainly because these platforms have more users than other apps do. What’s
more, some Internet celebrities are relatively more active on these social media platforms. 3. Create a brand story. When an unfamiliar brand appears in front of the public, an appealing
story is needed to impress customers. Sellers need some differentiated highlights to attract customers
and show their uniqueness. 3. Create a brand story. When an unfamiliar brand appears in front of the public, an appealing
story is needed to impress customers. Sellers need some differentiated highlights to attract customers
and show their uniqueness. q
4. Use the method of hunger marketing. Reducing the supply of products and adjusting the
relationship between supply and demand makes people more likely to believe that the brand is popular
once the supply is less than the demand. This may lead to people’s queuing behavior, which is
undoubtedly conducive to brand image forming so that the product will become popular on the
Internet [12]. 4. Use the method of hunger marketing. BCP Business & Management
Volume 20 (2022) BCP Business & Management
Volume 20 (2022) IEMSS 2022 Take the Asia-Pacific region, for example, and there are some latest consumer trends among
Generation Z in the Asia Pacific. 1. Personalization 1. Personalization 40% of youngsters tend to buy products from brands that are popular with their peers [14]. They
prefer brands that can set them apart and highlight their individuality. In addition to the uniqueness
of the brand image, Generation Z also expects brands to offer services and features that are
personalized, customized, exclusive, and have limited-edition quality. Once the brand can align with
the values and needs of Generation Z, they are more likely to become a long-term customer. Personalization is an expectation with generation Z [15]. 40% of youngsters tend to buy products from brands that are popular with their peers [14]. They
prefer brands that can set them apart and highlight their individuality. In addition to the uniqueness
of the brand image, Generation Z also expects brands to offer services and features that are
personalized, customized, exclusive, and have limited-edition quality. Once the brand can align with
the values and needs of Generation Z, they are more likely to become a long-term customer. Personalization is an expectation with generation Z [15]. p
2. Influence of Videos 2. Influence of Videos Generation Z spends much time online. Their screen time can be as high as 6 hours or more per
day; data shows that Gen Z is most affected by videos on social media. These videos can influence
their brand awareness and product choices [16]. Generation Z spends much time online. Their screen time can be as high as 6 hours or more per
day; data shows that Gen Z is most affected by videos on social media. These videos can influence
their brand awareness and product choices [16]. 3.1.2 Consumer behavior and psychology of Generation Z According to a report released by McKinsey, in 2021, social media has already influenced
58%respondents’ purchasing decisions. Consumers claimed that the social media platforms that
influenced their purchasing decisions are: Facebook (67%), YouTube (57%), and Instagram (52%). Mckinsey noted that many apparel retailers use TikTok to help their turnover grow. According to the
data released by Bridgepearl.com in October, more than two-thirds of Gen Z shoppers plan to shop
in a non-traditional way. The study found that Z Generation typically browses Facebook, WhatsApp,
Instagram, YouTube, and live shopping on these apps [13]. Gen Z has been surrounded by various
social media and diverse information from a very young age; people of this generation like to build
their image by buying trendy items. They always want to keep up with the trend and hope to express
their individuality with their outfits and accessories. Young people will use these platforms to search
for inspiration, and they are more likely to follow the trend and buy products recommended on social
media platforms by famous bloggers they follow. 343 3.2 Results of the questionnaire Questionnaires are released to investigate the conditions of Gen Z purchasing popular products
and the psychology behind their buying behavior. Most of the respondents are Generation Z. When
it comes to the question, “ how do you get the information about Internet celebrity products”, more
than half of the respondents chose the option “commended by blogger,” and nearly 30% of the
respondents chose the option “advertising”. Regarding the frequency of their purchase behavior,
nearly half of the respondents denoted that they never buy Internet celebrity products. More than half
of the respondents said they would purchase a product once every three months. The factors that
stimulate their buying behavior are shown in the pie chart below: affordable price (54.55%), new
technology (48%), good appearance (30.3%), and people they know have bought it (27%), and good
service (21.2%). As shown in Figure 1. Figure 1. Factors attracting consumers Figure 1. Factors attracting consumers But when they were asked, “Do you often repurchase Internet celebrity products,” only about 20%
of the respondents said they would buy these products again. Good quality was the main reason
(71.43%), followed by low prices (42.86%), and several respondents chose the option “enable me to
discuss with others about i”(also known as social currency). Also, the drawbacks of the internet
celebrity marketing method are quite clear. The questionnaire also reflects that most people believe
that advertisements about Internet celebrity products are over-descriptive. As shown in Figure 2. 344 BCP Business & Management
Volume 20 (2022) IEMSS 2022 IEMSS 2022 Figure 2. Drawbacks of the Internet celebrity marketing method Figure 2. Drawbacks of the Internet celebrity marketing method Still, these advertisements have not caused general psychological disgust among customers. In the
Yeezy section, the respondents learned about the brand through various channels such as friends’,
recommendations by bloggers, advertisements, etc. As shown in Figure 3. Figure 3. Ways of Learning about Yeezy for the first time Figure 3. Ways of Learning about Yeezy for the first time
Nearly 80% of the respondents said that they would not buy Yeezy in the future. Only 18% of
respondents have already purchased or will purchase its products. The main reasons for purchasing
are comfort (50%), high popularity (50%), and good looks (16.67%), As shown in Figure 4. 3.2 Results of the questionnaire But at
the same time, they think that Yeezy is expensive (39.39%), The price/performance ratio is not high
(45.45%), and it is not so popular now (33.33%), As shown in Figure 5. Figure 3. Ways of Learning about Yeezy for the first time Nearly 80% of the respondents said that they would not buy Yeezy in the future. Only 18% of
respondents have already purchased or will purchase its products. The main reasons for purchasing
are comfort (50%), high popularity (50%), and good looks (16.67%), As shown in Figure 4. But at
the same time, they think that Yeezy is expensive (39.39%), The price/performance ratio is not high
(45.45%), and it is not so popular now (33.33%), As shown in Figure 5. Figure 4. Reasons of Purchasing Yeezy Figure 4. Reasons of Purchasing Yeezy 345 BCP Business & Management
Volume 20 (2022) IEMSS 2022 IEMSS 2022 Figure 5. Disadvantages of Yeezy Figure 5. Disadvantages of Yeezy
Judging from the results above, it is not difficult to find that Yeezy, as an Internet celebrity product
popularized by Kanye West, is facing a difficult situation. The questionnaire shows that although
people from the Z generation know this brand well, it does not stimulate their purchase behavior. The
essential reason is this product is too expensive. In addition, it also reflects that many customers of
the Z generation think Yeezy is not as popular as it used to be, which may also lead to the situation
Yeezy is facing. Since youngsters like to follow the trend, this brand seems to have lost its appeal to
the Z generation, undoubtedly a serious problem for Yeezy. Figure 5. Disadvantages of Yeezy Judging from the results above, it is not difficult to find that Yeezy, as an Internet celebrity product
popularized by Kanye West, is facing a difficult situation. The questionnaire shows that although
people from the Z generation know this brand well, it does not stimulate their purchase behavior. The
essential reason is this product is too expensive. In addition, it also reflects that many customers of
the Z generation think Yeezy is not as popular as it used to be, which may also lead to the situation
Yeezy is facing. Since youngsters like to follow the trend, this brand seems to have lost its appeal to
the Z generation, undoubtedly a serious problem for Yeezy. 3.3 Yeezy’s marketing strategy The two most important marketing methods of Yeezy are influencer marketing and hunger
marketing. As a successful rapper and designer, Kanye West is well-known worldwide. He
cooperated with Adidas’s well-known sports brand to create this “Yeezy” series of shoes. The news
was posted on social media, and in a very short time, the youngster that always kept up with the trend
became crazy about this brand. Kanye’s products are very popular among young people as an
influential designer. Kanye is very aware of the power of hunger marketing, and he can make good
use of the scarcity of products to stimulate customers buying behavior. When he and Adidas teamed
up to launch the Yeezy boost 350, he only rolled out this series in limited quantities, and only
thousands of pairs of shoes were available for sale. When these shoes were launched, they were sold
out immediately. Kanye and his team created the scarcity of the product and ensured that the product
remained popular [17]. 3.4 Problem identification On the other hand, since Yeezy has re-saled lots of different tincts and the number of shoes is not
as limited, consumers’ enthusiasm for buying Yeezy shoe decreases. Since the sale of Yeezy shoes
was exclusive and limited, Yeezy gained loyal consumers who scrambled and collected every Yeezy
tinct. However, since Yeezy has re-saled and released the gargantuan number of each tinct, it doused
consumers’ desire, which therefore decreases its consumer loyalty will suffice. According to Stockx,
one of the biggest re-sale platforms, the price of Yeezy Boost 350 V2 Zebra was $1,536 as it was first
released in 2017. However, as Yeezy re-published this tinct over time, its price dropped $349 on
average at the end of 2020. The shoe price on re-sale platforms reflects the clients’ fading enthusiasm,
which illustrates the recession of Yeezy shoes. 346 BCP Business & Management
Volume 20 (2022) IEMSS 2022 4.2 Using some unique techniques on shoes Unique techniques can add a high value to the product. Yeezy already used a new material called
boost to make its shoe sole. Other Adidas shoes that have the same material are also expensive. The
minimum price of a pair of boost shoes is 190 dollars [19]. So, a unique technology on the shoes is
quite attractive. Yeezy can try to invent more specific technologies to make the shoes comfortable. This is not only a good point that can be widely publicized but also increases the cost performance. 4.3 Inviting more famous people to advertise Inviting more famous to advertise can brings many advantages. Firstly, it can enhance the
reliability of Yeezy. Some famous people on social software will test it in all aspects. An expert
suggestion can affect consumers’ behavior. Authority effects are very real because factors can compel
obedience as much as a more formal authority [20]. Besides, having a celebrity endorsement can
improve a business’ reputation. It is important to establish the corporate image because it needs to be
accepted by the target group. Finally, there is a psychologic called groupthink. It is a psychological
phenomenon where people strive for consensus within a group [21]. So, finding more people to
advertise to can also affect the consumers’ behavior. 4.1 Cooperating with well-known e-commerce companies Cooperating with e-commerce apps would help Yeezy with spreading awareness. Well-known e-
commerce companies possess stable consumers. For instance, Taobao is the biggest e-commerce
platform in China, in which it will have 863 million annual active consumers by the end of 2021 [18]. By cooperating with e-commerce apps like Taobao, Yeezy will spread awareness and expand its
market, which will help the enterprise attract new consumers. When people want to buy shoes, they
will search keywords in the app. Because the number of consumers is so large in the app, there are
more opportunities for a group of people who have never seen Yeezy to buy it. In addition, selling in
the app can also provide a reliable way for consumers to buy it. 4.5 Providing a service that consumers can let the shop wash Yeezy for free in one year After-sales is an important part when people buy a thing. It will affect the consumers’ happiness
of the whole purchasing experience. A good attitude toward after-sales can increase the probability
that consumers buy the brand again or introduce the product to others. Offering this washing service
can satisfy consumers’ needs. It can also increase the contact between the shops and consumers. Because when they need to wash their Yeezy, they need to come to the shop. Then they can see more
products and buy some if they like them. 4.4 Having collaboration oration with other famous brands This market method can also attract more new consumers. Because the other brand also has many
fans. The jointly signed is meaningful for two friends. Some people will collect this type of shoes. Air Jordan shoes are also a celebrity product. A normal type of AJ is 150 dollars [22]. However, AJ
2 was jointly signed with Off-White. The price of the shoes is 250 dollars [23]. In addition, the amount
of shoes is limited, so the second-hand price is higher. From that case, the conclusion is that
appropriate collaboration with other famous brands can increase sales and price. 4.7 Punishing shops which sell the fake Yeezy Many shops are now selling fake Yeezy. Some of these shoes are as real as the original ones. But
its price is very low. So, a group of people will buy fake shoes. Piracies infringe on consumers’ and
producers’ benefits. People who decide to buy the piracy will no longer buy the real one. Those who
want to buy real ones may not identify which is real. Many people are deceived, and they use a high
price to buy piracy. Yeezy producers can add more invisible signs on the shoes, so it can help people
to identify whether it is a real or a fake. Shops that sell piracies also need to be punished severely. 4.6 Using technology that consumers can try Yeezy on the Internet China is a large market. But because of the COVID-19, many Chinese cannot go to the Adidas
shop. So, they can only buy things on the Internet. It is difficult for them to see what the shoes look
like in reality. They also do not know which color is more suitable for them. So Yeezy can provide a
function that consumers can use the camera. It can identify feet and out the model of shoes. As a 347 BCP Business & Management
Volume 20 (2022) IEMSS 2022 result, consumers can see what they look like when they wear the shoes. It is a good way to improve
consumers’ purchasing experience. result, consumers can see what they look like when they wear the shoes. It is a good way to improve
consumers’ purchasing experience. 5. Conclusion In conclusion, the paper analyzed Yeezy’s marketing strategy and limitations. Possible solutions
and suggestions for Yeezy’s marketing are given. By craftily using the Internet as an intermediate,
enterprises successfully conducted a network effect and gained internet celebrity products. By
contraposing generation Z’s consumer psychology and behavior, it is suggested that they are the
potential long-term consumer and targeted audience of the network effect. Moreover, Yeezy used
exciting, word-of-mouth marketing, net-work effect, and hunger marketing. Yeezy generated these
strategies well, but it is currently facing a difficult situation among the biggest potential clients—the
Z generation. It is suggested from the questionnaire that although Yeezy is well-known by Generation
Z, however, due to its high price and recession trend, generation Z is not likely to purchase Yeezy
items. This is further reflected through the declining price of Yeezy shoes on re-sale websites. In
contrast, this can be solved by cooperating with e-commerce platforms with abundant loyal
consumers, inventing distinctive techniques used on the shoes, hype, and inventing more celebrities
to advertise its products. Moreover, cooperating with other successful brands expands brand
awareness, provides extra services, and punishes those who manufacture fake Yeezy shoes. By analyzing the Z generation’s purchasing behaviors, it is suggested that the Z generation is the
biggest potential client for many companies in the future due to its strong association with the Internet. This is valuable to be referenced by companies because recognizing the importance of network effect
and potential clients helps enterprises expand their market. The article analyzed Yeezy’s marketing
strategies and suggested possible improvements they could make, and all companies could reference
this. On the other hand, the number of participants who have done the questionnaire is limited, resulting
in inaccurate results. To specify, there were only 30 questionnaires handed in and analyzed. 30
questionnaires are not enough for the investigators to analyze, and only part-of-view are analyzed. Participants are likely to be in the same age group due to the investigators’ similar ages. In this case,
the investigators were only able to analyze a limited view of perspectives. Additionally, due to the region in which all investigators live being big urban cities, they could not
interpret perspectives from people who live in rural and suburban areas. This will cause a huge
difference due to their different consumption concepts and abilities. Thus, the questionnaires are not
comprehensive enough for the investigators to analyze. [1] G.K.Grant, A Case Study of Brand Associations for Yeezy Brand, 2018 [3] L. F. Xu, Research on marketing Model and Characteristics of “Internet celebrity”, Modern Business,
2019, No.18, pp. 22-23. DOI:10.14097/j.cnki.5392/2019.18.010. [4] G. Derosier, K.Chester, J.Dally, Adidas’ s partnership with Kanye West: How celebrity collaborations
affect their affiliated brands, 2010 [2] W.Kenton, Co-Branding, Investopedia, 2000. References [1] G.K.Grant, A Case Study of Brand Associations for Yeezy Brand, 2018 [2] W.Kenton, Co-Branding, Investopedia, 2000. [4] G. Derosier, K.Chester, J.Dally, Adidas’ s partnership with Kanye West: How celebrity collaborations
affect their affiliated brands, 2010 [4] G. Derosier, K.Chester, J.Dally, Adidas’ s partnership with Kanye West: How celebrity collaborations
affect their affiliated brands, 2010 348 BCP Business & Management
Volume 20 (2022) BCP Business & Management
Volume 20 (2022) IEMSS 2022 [5] Wikipedia， Adidas Yeezy [5] Wikipedia， Adidas Yeezy [5] Wikipedia， Adidas Yeezy [6] T. McKinnon, How supreme & yeezy use scarcity to drive sales, 2019, Indigo9 Digital Inc. Retrieved
March 25, 2022, from https://www.indigo9digital.com/blog//how-the-use-of-scarcity-is-driving-the-
growth-of-popular-retail-brands [7] D.Fowler, The hype machine: Streetwear and the business of scarcity, 2018BBC Worklife. Retrieved
March 25, 2022, from https://www.bbc.com/worklife/article/20180205-the-hype-machine-streetwear-
and-the-business-of-scarcity [8] J. Caulfield, A quick guide to textual analysis. Scribbr,2019, Retrieved March 25, 2022, from
https://www.scribbr.com/methodology/textual-
analysis/#:~:text=Textual%20analysis%20is%20the%20most%20important%20method%20in,this%20c
ontext%2C%20usually%20novels%2C%20poems%2C%20stories%20or%20plays. https://www.scribbr.com/methodology/textual
analysis/#:~:text=Textual%20analysis%20is%20the%20most%20important%20method%20in,this%20c
ontext%2C%20usually%20novels%2C%20poems%2C%20stories%20or%20plays. [9] MSG Management Study Guide. Survey Method. (n.d.). Retrieved March 25, 2022, from
https://www.managementstudyguide.com/survey_method.htm#:~:text=The%20Survey%20method%20i
s%20the%20technique%20of%20gathering,are%20asked%20questions%20on%20their%20demographi
c%20interest%20opinion. [10] McCombes, S. (2022, February 7). How to do a case study. Scribbr. Retrieved March 25, 2022, from
https://www.scribbr.com/methodology/case-study/ [11] A. Virmani, Winning with Social Media-How social platforms are changing the face of Business.2021 [12] L.Corbitt, How you can build a powerful influencer marketing strategy in 2022 [13] T.Walk-Morris, McKinsey: 87% of Gen Z will use social media for holiday shopping insp [14] A.Insyirah, 5 ways to appeal to the Gen Z Market in Asia Pacific,2021 5] The shelf team, What makes them buy: generation Z buyer behaviors(updated for 2021),2020 [16] A.Insyirah, 5 ways to appeal to the Gen Z Market in Asia Pacific,2021 7] E. McCullough, T McKinnon, Kanye West’s Yeezy Strategy: How to Create a $1.5 billion br [18] Yihan Ma, Number of annual active consumers acroAlibaba’s online shopping properties from 2nd
quarter 2016 to 4th quarter 2021, 2022, Statista, https://www.statista.com/statistics/226927/alibaba-
cumulative-active-online-buyers-taobao-tmall/ [19] 2022, Adidas, https://www.adidas.com/us/ultraboost [20] Shawn P. Wilbur,Authority and Authority-effects, THE LIBERTARIAN LABYRINTH, 2018,
https://www.libertarian-labyrinth.org/author/libertarianlabyrinth_085jen/ [21] Kendra Cherry, What Is Groupthink?, very well mind, 2020, https://www.verywellmind.com/what-is-
groupthink-2795213 [22] Nike, 2022, https://www.nike.com/cn/w/jordan-shoes-37eefzy7ok [23] Mario Briguglio, Official Photos of the Off-White x Air Jordan 2 Low “White/Varsity Red”, SBD, 2021,
https://sneakerbardetroit.com/off-white-air-jordan-2-low-release-date/ [23] Mario Briguglio, Official Photos of the Off-White x Air Jordan 2 Low “White/Varsity Red”, SBD, 2021,
https://sneakerbardetroit.com/off-white-air-jordan-2-low-release-date/ 349
|
https://openalex.org/W2055816704
|
https://zenodo.org/records/2399978/files/article.pdf
|
English
| null |
TWO CASES OF SYNAESTHESIA
|
British journal of psychology
| 1,914
|
public-domain
| 2,191
|
1 The latter difficulty was iiiuch reduced by the subsequent cooperation of 11. Alexander
Briantshaninov, to whom I gladly take tliis opportllnity of expressing my obligation. 1. DURING his recent visit to England, the well-known Russian
composer, Alexander Scriabin, kindly allowed me to carry out an
examination of his coloured hearing. Unfortunately the time at his
disposal and the necessity of communicating in a language (French)
foreign to both of us1 made a very thorough investigation impossible. But the results obtained form perhaps a sufficiently interesting con-
tribution to the psychology of music to merit publication in their
present form. Colours form for Scriabin so important a part of the total effect of
sounds that he desires his Prometheus to be performed to the accom-
paniment of concealed lanips which shall flood the concert-hall with
a light of ever-changing colour; the music of his Mystery, when
completed, will be presented with a similar p1a.y of colours, and with
odours. Scriabin’s attention was first seriously drawn to his coloured
hearing owing to an experience at a concert in Paris, where, sitting
next to his fellow-countryman and composer Rimsky Korsnkov, he
remarked that the piece to which they were listening (in D m?jor)
seemed to him yellow ; whereupon his neighboiir replied that to him,
too, the colour seemed golden. Scriabin has since compared with his
compatriot and with other musicians the colour effects of other keys,
especially B, C major and F# major, and believes a general agreement
to exist in this respect. He admits, however, that whereas to him the
key of Flf major appears violet, to Rimsky Korsakov it appears green;
but this deviation he attributes to an accidental association with the
colour of leaves and grass arising from the frequent use of this key for 113 pastoral masic. He also allows that there is some disagreement
as to the colour-effect of the key of G major. Nevertheless, as is so
universally the case with the subjects of synaesthesia, he believes that
the particular coloiirs which he obtains must be shared by all who are
endowed with coloured hearing. Doubtless, on rnore systematic inquiry,
the same disagreements which, as is well known, occur in the case of
coloured vowel sounds, coloured names, coloured days of the week, and
(as I have pointed out’) with coloured pitch and coloured timbre, will
bc discovered in the case of coloured tonality2. As will have been gathered by now, Scriabin’s chromuesthesia refers
to the towlity of the music. As the tonality changes in a piece,
so the colour changes. ‘‘ A Case of Synaesthesia,” This Journal, 1911, IV. 228-238.
* Cf. Th. Flournoy, Des ph6nodnes de sympsir, Paris et GenBve, 1893, 101; also
3 L O C . eit.
Henri Laures, Les Synesthbies, Paris, 1908, 30.
J. of Psych. VII
8 1. Scriabin explains that “ the colour icnderlines
the tonality; it makes the tonality more evident.” The colour or a
change of colour sometimes appears to him before he is aware of thc
tonality or of a change of tonality. For such reasons he believes that
the musical effects are enhanced by the simultaneous prescntation to
the eye of the appropriate colour. In general, when listening to music, he has only ;i ‘ feeling’ of
colour ; only in cases where the feeling is very intense does it pass over
to give an ( image ’ of colour. ‘ The older music, with its infrequent changes of tonality, gives him
a colour changing in intensity instead of in quality ; (‘ it has not the
psychological basis of modern music.”
Certain compositions, and
most of Beethoven’s Symphonies, are not of a kind to need colour;
they arc “ too intellectual in character.” In thc case of coloured hearing which I have prcviously published3,
the colour is dependent on the pitch of the note ; in rnany other cases,
already described, it is dependent on the name of the note. But for
Scriabin a single note has in itself no colour; it has the colour of its
tonality. Indeed, he insists that whether in or apart from musical
compositions, a single tone cannot occur alone ; even outside music it
is accompanied by overtones, which, in many cases, especially in the
case of the sounds of nature, include inharmonic overtones, in addition
to the harmonic series. Thus it is, that if we imitate the sound
of a bell by adding an appropriate series of weaker tones to a
louder fundamental tone on the pianoforte, we may find on analysis 8 8 114 a combination of two or more tonalities, and consequently an inter-play
or fusion of the corresponding colours. So, too, any simultaricous com-
bination of tones produces on Scriabin a simple or composite colour
effect according as it suggcsts one or several tonalities. g
gg
The strongest colours for Scriabin appear to be those relating
to the keys of C major, D major, B major and F# major, placed
respectively in the red, orange, yellow, blue and violet. Loe. eit. 238. 1. Starting,
however, from C at the red end of the spectrum, Scriabin finds that Red
Orange
Yellow
Green
Blue
Violet
u c ;
D
A
E
F# as he passes from hue to hue, the successive colours correspond to
tonalities rising by a series of fifths. Thus the key of C is red, of G
red to orange-red, of D orange to yellow, of A yellow to green, of E
green to blue, of B blue to violet, and of F# violet. The colours of the
remaining keys Db, Ab, Eb, Bb and F are believed by Scriabin to
be extra-spectral,-either
ultra-violet or infra-red. Thus the key of F
is ‘‘ on the verge of red,” giving often the effect of a metallic lustre. g
g
g
Into Scriabin’s objections to the terms major and minor in tonality
it is impossible to enter here. The harmonies, in his Prometha: are
based upon a scale of six notes, which represent the seventh, eighth,
ninth, tenth, twelfth and thirteenth overtones of a fundamental,
yielding the tones (in the scale of C) c d e flt u bb approximately. A series of such tones, e.g. clfl#
P b @ u2 d3, simultaneously played in
true (untempered) intonation, produces what he terms a “single sound,”
one that is not easily analysable. Scriabin believes that there is always
a psychical struggle between a tone and its overtones, just as a rivalry
may be conceived between the various tones emitted at any moment by
an orchestra,-a
contest for the pitch of the single fundamental sound
which the listener shall hear. In my previous paper on synaesthesia I concluded that ‘‘for the full
development of synaesthesia, a strong tendency to a certain kind of
association is requisite-a
tendency to form associations between
corresponding members of two homologous series’.” With the subject
then under investigation, any letter (e.g. Y) immediately tended to call
up a number (25) expressing its position in the alphabet. Such a
tendency may readily yield the abovc diagram, where consecutive scales
are associated with consecutive spectral coloiirs, and may also result in 115 CHARLES
S. MYERS tht. strong inclination towards mysticisin which is characteristic of
Scriabin. For him the (red) key of C relates to matter, and is redolent
with the odour of the soil, whereas the (violet) key of F# is spiritual
and ethereal. 1. He believcs that colours have their over-colonrs, as tones
have thcir overtones. The only exception was a failure at the second sitting to obtain gold or flame colours
Thus 500
for the region 500-600.
suggested a streaky green, 513 an intense green, ti00 a streaky black and clear light blue.
It appeared to behave like the region 700-800. The other case of synaesthesia here recorded is in a lady (Subject B)
who is an accomplished painter and takes a keen enjoyment in hearing
iiiusic although she does not play on any instrument. Like the case (Subject A ) described in my previous paper, this
subject has very poor visual imagery, and the colours which she obtains
vary with the pitch and with the timbre of the tone. A visual image only comes to her when she has taken special care
to attend to the previous perception of the object. Thus, when painting,
she can visualise her ‘ sitter’ if she has attended particularly to the pose. But she cannot get a visual image of her breakfast table, “because she
has never attended to it.” Nor does she ‘see’ colours in sounds; she
cxplains her coloured hearing on the ground that sounds ‘. give her the
same mental sensations” as colours. The flow of colours she experiences in listening to music ;ifford her
“ enormous pleasure.” They vary with the composer ; the works of
Chopin, for example, yield “ very translucent colours such as green
leaves in the spring,” whereas those of Schumunn “ never give ‘ primary’
colours, they give purples and the like,-not
transparent colours.” The
colours cotne more reliably when they are not specially looked for : (‘ it
is so difficult to be truthful when one is watching.” Individual tones have each a colour dependent on their pitch ; but
the colours, ;is given in the folloN-ing table, show a. sequence very
different from that described in the case of A in my previous paper. Pitch of tone
256
300
400
500
600
700
800
no0
1200
2048
Colour
Prusaian blue, clear blue. A clear mixed colour, a suggestion of streakiness, dark blue streaked
Clear dark violet-clear purple. Deeper than red, very deep golden, transparent. No definite oolour, opaque, streaky, perhaps black and flame colour. No definite colour, uninteresting, perhaps light green. Blue. Rather like 800. Might be yellow, something of that nature, very translucent. Getting yellow. with violet. Deeper than red, very deep golden, transparent. No definite oolour, opaque, streaky, perhaps black and flame colour. 8-2 116 Two Cases of &pnexth,enin The above colours were obtained for the tones of tuning forks,
sounded in irregular order. Several of the tones were repeated at the
same and at a later sitting, and gave confirmatory results’. For higher tones, the whistles supplied with the Edelmann tone
series were employed. When the whistle giving tones from about 1704
down to 1024 vibrations per second (aS to cd) was sounded, B replied
that the highest notes were a faint green, or yellow-green, while the
lower became “ more blue, less yellow.” The tones of the next whistle,
ascending from about 1704 to about 3408 vibrations per second (a3 to c4)
were described as becoming “ more and more emerald-green.” The
Galton whistle tone of 6000 vibrations per second (about g5) appeared
green, ascending from which still higher tones became increasingly
colourless. That is to say, in this subject we appear to have (i) a change in hue
from blue through purple and reddish-yellow to green, for tones varying
from about 200 (256) to 700, and (ii) a broadly similar repetition of this
flow of colours-from
blue through yellow to green--for tones ascending
from 800 towards the upper audible limit. The blue of 200 became more
and more purple as the lower limit was approached, e.g. “ much darker
and more purple than the gentian” (128), “still niorc purple“ (64), etc. It was remarked by the subject that the colours almost invariably
appear to her as uniform or streaky (streaked with ‘ black’ or ‘ light’),
clear or opaque, smooth or rough. Yet there was “no form in the
colours” ; hence she could not explain, for example, in what direction
the streaks were running. As in the case of Subject A, described in my previous paper, traces
of the coloiirs obtained by two single tones were often also obtainable
for two simultaneously sounding tones. Thus the simultaneous tones,
300, 600, were described as ‘( something very dark,-either
blue or
black,-and
something clear and light” ; for the simultaneous tones
300, 500, she reported-“that
has that horrid Prussian blue in it,-
a colour I loathe.” The simultaneous tones, 600, 900, were described as “ an acid mixed
sound. For the moment I can’t
name them. One is black. They all seem very clear colours. (You
can have a clear black as in looking into a pond at night.) There is
a streakiness in the colours-a
black and two things bright and crude.”
To me there are three colours in it. 117 C‘IIAHT~EH
S. MYERS Like A, B is more ‘alive’ to the colour components than to the
tonal components of sounds. She “may be aware of the presence of
several colours where the number of simultaneous tones is not attended
to.” Again as with A , the effect of increasing the richness of a tone by
adding to its overtoncs rcsults in ;I ‘ risc’ of the colour of the tone. Sonrce of sound
-
Tuning fork (with resonntor)
Towiwnsrr
Tone variator
~
~
~~~~~~~~~~~~
Horn
(.’ = 256
= 51‘2
_
_
_
_
~
’
Clear hlnr
Rose coloor in blue
Reddish hlac
Briglit yellow
Indigo blne
Greenish
~
~
~~~
~
~-
A horrid red colour
|
https://openalex.org/W4232285104
|
https://hal.archives-ouvertes.fr/hal-00329127/document
|
English
| null |
MUREX: a land-surface field experiment to study the annual cycle of the energy and water budgets
|
Annales geophysicae
| 1,999
|
cc-by
| 14,518
|
MUREX: a land-surface field experiment to study the
annual cycle of the energy and water budgets
J.-C. Calvet, P. Bessemoulin, J. Noilhan, C. Berne, Isabelle Braud, Dominique
Courault, N. Fritz, E. Gonzalez-Sosa, J.-P. Goutorbe, R. Haverkamp, et al. MUREX: a land-surface field experiment to study the
annual cycle of the energy and water budgets
J.-C. Calvet, P. Bessemoulin, J. Noilhan, C. Berne, Isabelle Braud, Dominique
Courault, N. Fritz, E. Gonzalez-Sosa, J.-P. Goutorbe, R. Haverkamp, et al. To cite this version: J.-C. Calvet, P. Bessemoulin, J. Noilhan, C. Berne, Isabelle Braud, et al.. MUREX: a land-surface
field experiment to study the annual cycle of the energy and water budgets. Annales Geophysicae,
1999, 17 (6), pp.838-854. hal-00329127 Distributed under a Creative Commons Attribution 4.0 International License MUREX: a land-surface ®eld experiment to study
the annual cycle of the energy and water budgets J.-C. Calvet1, P. Bessemoulin1, J. Noilhan1, C. Berne1, I. Braud2, D. Courault3, N. F
J.-P. Goutorbe1, R. Haverkamp2, G. Jaubert1, L. Kergoat4, G. Lachaud1, J.-P. Laur
A. Olioso3, P. Pe ris1, J.-L. Roujean1, J.-L. Thony2, C. Tosca5, M. Vauclin2, D. Vig
1 Me te o-France, CNRM/GAME, URA CNRS 1357, 31057 Toulouse Cedex 1, France
2 LTHE, UMR CNRS 5564, INPG, ORSTOM, UJF, BP53 38041 Grenoble Cedex 9, France
3 INRA Bioclimatologie, 84914 Avignon Cedex 9, France
4 Laboratoire d'Ecologie Terrestre, UMR CNRS 5552, 31405 Toulouse Cedex 4, France
5 CESBIO, UMR CNRS 141, 31401 Toulouse Cedex 4, France 1 Me te o-France, CNRM/GAME, URA CNRS 1357, 31057 Toulouse Cedex 1, France
2 LTHE, UMR CNRS 5564, INPG, ORSTOM, UJF, BP53 38041 Grenoble Cedex 9, France
3 INRA Bioclimatologie, 84914 Avignon Cedex 9, France g ,
g
,
4 Laboratoire d'Ecologie Terrestre, UMR CNRS 5552, 31405 Toulouse Cedex 4, France
5 CESBIO, UMR CNRS 141, 31401 Toulouse Cedex 4, France Received: 9 September 1998 / Revised: 25 February 1999 / Accepted: 26 February 1999 Received: 9 September 1998 / Revised: 25 February 1999 / Accepted: 26 February 1999 carried out, e.g. HAPEX-MOBILHY (AndreÂ
et al.,
1986), FIFE (Sellers et al., 1992), EFEDA (Bolle et al.,
1993), and HAPEX-Sahel (Goutorbe et al., 1997). These
campaigns allowed successful calibration and validation
of several soil-vegetation-atmosphere transfer models
(hereafter denoted as SVATs), e.g. the ISBA (interaction
between soil, biosphere, and atmosphere) scheme devel-
oped by Noilhan and Planton (1989) suitable for
coupling with meteorological or climate models, or the
detailed SiSPAT scheme (simple soil-plant-atmosphere
transfer model) designed by Braud et al. (1995). The
development and implementation of SVATs in general
circulation models (GCM) or climate models clearly
demonstrated the positive impact of good representa-
tions of vegetation in atmospheric models, particularly
its ability to modify turbulent and radiative exchanges,
and to modulate the extraction of soil water. Abstract. The MUREX (monitoring the usable soil
reservoir experimentally) experiment was designed to
provide continuous time series of ®eld data over a long
period, in order to improve and validate the Soil-
vegetation-Atmosphere Transfer (SVAT) parameterisat-
ions employed in meteorological models. Intensive
measurements were performed for more than three
years over fallow farmland in southwestern France. MUREX: a land-surface ®eld experiment to study
the annual cycle of the energy and water budgets To
capture the main processes controlling land-atmosphere
exchanges, the local climate was fully characterised, and
surface water and energy ¯uxes, vegetation biomass, soil
moisture pro®les, surface soil moisture and surface and
soil temperature were monitored. Additional physiolog-
ical measurements were carried out during selected
periods to describe the biological control of the ¯uxes. The MUREX data of 1995, 1996, and 1997 are
presented. Four SVAT models are applied to the annual
cycle of 1995. In general, they succeed in simulating the
main features of the fallow functioning, although some
shortcomings are revealed. However, the limited duration of such campaigns (no
more than two months for most intensive ®eld cam-
paigns) prevented testing the model's ability to reproduce
the annual or inter-annual water cycle, for which longer
validation datasets are required. The motivation for a
long-term ®eld documentation of relevant processes was
con®rmed by the Project for Intercomparison of Land
Surface Parametrisation Schemes (PILPS). In PILPS
phase II (Chen et al., 1997), 23 schemes were run to
compare the various predicted components of the energy
balance with experimental data collected over a meadow
at Cabauw, the Netherlands. It was found that even for
this relatively simple type of surface, signi®cant dieren-
ces occurred between models and observations, and
models themselves. For example, the modelled annual
mean evaporation ranged between 32 W m)2 (about
400 mm y)1) and 56 Wm)2 (about 710 mm y)1), while
the measured value was 42 Wm)2 (530 mm y)1). Such
large errors of energy ¯ux estimates in atmospheric
models may have a powerful impact on predicted soil
moisture and surface runo, with a possible feedback on
cloud amount and precipitations. Koster and Milly
(1997) showed that the large scatter in PILPS ¯ux
calculations was related to the modelling of the interplay Key words. Hydrology (evapotranspiration; soil
moisture; water-energy interactions). Correspondence to: J.-C. Calvet
Me te o-France/CNRM, 42 Av. Coriolis, 31057
Toulouse Cedex 1, France
E-mail: jean-christophe.calvet@meteo.fr HAL Id: hal-00329127
https://hal.science/hal-00329127v1
Submitted on 18 Jun 2008 L’archive ouverte pluridisciplinaire HAL, est
destinée au dépôt et à la diffusion de documents
scientifiques de niveau recherche, publiés ou non,
émanant des établissements d’enseignement et de
recherche français ou étrangers, des laboratoires
publics ou privés. HAL is a multi-disciplinary open access
archive for the deposit and dissemination of sci-
entific research documents, whether they are pub-
lished or not. The documents may come from
teaching and research institutions in France or
abroad, or from public or private research centers. Distributed under a Creative Commons Attribution 4.0 International License Ann. Geophysicae 17, 838±854 (1999) Ó EGS ± Springer-Verlag 1999 1 Introduction The MUREX (monitoring the usable soil reservoir
experimentally) experiment (Bessemoulin et al., 1996;
Calvet et al., 1998a) was initiated in 1994. At that time,
a number of large-scale ®eld experiments focused on
land surface-atmosphere interactions had already been Correspondence to: J.-C. Calvet
Me te o-France/CNRM, 42 Av. Coriolis, 31057
Toulouse Cedex 1, France
E-mail: jean-christophe.calvet@meteo.fr 839 J.-C. Calvet et al.: MUREX: a land-surface ®eld experiment between transpiration and runo. However, it was not
possible to propose better parametrisations based on the
PILPS datasets because all the key SVAT variables (i.e. soil moisture, surface ¯uxes, and leaf area index, LAI)
were not measured at the same time during long enough
periods. (1) avoid any in¯uence of river water tables on the
water budget and (2) search for a natural canopy whose
characteristics do not change fundamentally from one
year to another. Since river water tables are often
close to the surface in valleys, the Toulouse region
oers a limited choice of sites satisfying this require-
ment. It was thus decided to select a fallow site over
one of the plateaux dominating the main watersheds of
the region. p
Based on this result, it was concluded that there was
real need for testing land-surface schemes against
measurements of all the components of the energy and
hydrological budgets, the seasonal variation of vegeta-
tion properties and soil moisture, and the variation of
the atmospheric forcing, during at least three annual
cycles. An additional motivation for undertaking such a
program was that some components of the hydrological
budget, such as runo or gravitational drainage, essen-
tial for understanding hydrologic systems, are only
accessible from long-term measurements (annual or
seasonal cycles). g
The MUREX fallow site (43°24¢N; 1°10¢E; altitude:
240 m) was an agricultural area abandoned in 1992. The
micrometeorological station (Fig. 1) was set up in June
1994. The ®eld was large enough (about 700 by 250 m)
to provide a good fetch. It was a rather ¯at parcel,
gently sloping down to the edge of a plateau dominating
the valley of the Touch river, 30 m below. The vegeta-
tion canopy of the MUREX site consisted of a dense
herbaceous agricultural fallow. The main plant species
were: Brachypodium sp and Potentilla reptans, especially,
and also Geranium rotundifolium, Erigeron canadensis,
and Rumex acetosa, as observed on day of year (DoY)
293 in 1995. 1 Introduction The canopy height ranged from 1 m when
vegetation was fully developed, to 0.1 m in winter. The
soil was a typical hydromorphic deep `boulbeÁne': the
mean texture of the 1 m surface soil layer was that of a
silt-loam (the sand and clay fractions were 14 and 28%,
respectively). However, strong vertical texture gradients
were observed: the proportion of clay increased from
17% at the surface to 40% at 1 m depth. On this type of
soil, at about 1 m, a local subsurface soil water
convergence might sometimes occur, caused by a tem-
porary perched water table over the clay bedrock. The resulting MUREX ®eld experiment presented is
a joint eort of Me te o-France/CNRM (Centre National
de Recherches Me te orologiques, Toulouse, France),
CESBIO (Centre d'Etudes Spatiales de la BiospheÁ re,
Toulouse, France) and LTHE (Laboratoire d'Etude des
Transferts en Hydrologie et Environnement, Grenoble,
France). The site is described and a summary of the
database is presented over three annual cycles (1995 to
1997). Finally,
dierent
modelling
approaches
are
applied to the data of 1995. 2 Main characteristics of the site Along with routine observations, more speci®c mea-
surements were made during selected periods: (1) direct
measurements of the soil moisture pro®le at the surface
(the top 5 cm) were performed during two 30-day
intensive observing periods in spring and autumn 1995
(Calvet et al., 1998a); (2) the integrated surface soil
moisture was routinely measured in 1997 by using
automatic capacitive probes; (3) the surface temperature
was derived from infrared radiometry from April to
November of each annual cycle; (4) the CO2 ¯uxes were
characterised from the summer 1997 to the spring of
1998; (5) the leaf stomatal conductance to water vapour
was measured on selected days throughout the studied
period, together with the leaf water potential; and (6) the
soil hydraulic properties (hydraulic and thermal con-
ductivities) were characterised. The latter measurements
were particularly useful to run the SiSPAT model (see
Sect. 6.3). Ra b b0 frc
1 ÿ rc 0:67
1670 qa0:08g rT 4
a
1 1 where r is the Stefan-Boltzmann constant, Ta and qa the
air temperature and speci®c humidity (respectively), at
screen-level, and rc is the cloud coverage
rc 2 0; 1. The rc parameter was measured at Francazal. The
regression coecients b and b0 were determined from the
available values of Ra at the MUREX site and from the
cloud coverage rc at Francazal. The regression param-
eters obtained (for 1995) are b 11:3 Wm)2 and
b0 0:9685, with a square correlation coecient
r2
of 73% and a standard error of 25 Wm)2. When cloud
coverage observations were not available, Eq. (1) was
applied
with
rc 0,
and
b 43:9 Wm)2
and
b0 1:0056. In this case, the value of r2 is 70% and
the standard error is 25 Wm)2. The same method was
applied to the 1996 and 1997 datasets, with similar
results. The 3-y measurements of the atmospheric
forcing
data
P; Rg; Ra; Ta; qa; U,
surface
¯uxes
(Rn; H; LE; G), and soil moisture, are summarised in
Figs. 2, 3, and 4. 3 Routine soil and atmospheric measurements The soil and atmospheric measurements of MUREX
were obtained using methods very similar to those of
HAPEX-MOBILHY (Andre et al., 1986): classical me-
teorological observations were combined with surface
energy and water budget measurements. 2 Main characteristics of the site The diculties in maintaining numerous automatic
measurements for a long period and in conducting
frequent investigations of the soil and vegetation at an
isolated site lead to the choice of a site close to CNRM,
30 km away from the city of Toulouse (France). Since
the aim of MUREX is to characterise the natural
interactions between surface ¯uxes, soil water content,
and
vegetation
functionning,
it
was
intended
to: The meteorological variables (precipitation, air tem-
perature and humidity, wind speed and direction) at the
site were monitored on a 30-min basis together with
surface temperature, solar radiation, and the surface
¯uxes: net radiation (Rn), sensible heat ¯ux (H), ground
heat ¯ux (G), and by dierence the latent heat ¯ux
LE = Rn)H)G. The other routine surface measure-
ments consisted of weekly pro®les of the deep soil Fig. 1. Photograph of the central part of the
MUREX fallow site (at the end of 1996) with
some of the instrumentation. From left to right:
the two-level SAMER system, the 10-m mast for
wind, screen-level temperature and humidity, and
IR-temperature, and the scaolding bearing the
radiation sensors Fig. 1. Photograph of the central part of the
MUREX fallow site (at the end of 1996) with
some of the instrumentation. From left to right:
the two-level SAMER system, the 10-m mast for
wind, screen-level temperature and humidity, and
IR-temperature, and the scaolding bearing the
radiation sensors J.-C. Calvet et al.: MUREX: a land-surface ®eld experiment 840 downwelling atmospheric thermal emission Ra), pres-
sure, and air temperature and humidity are missing, and
less than 1% of wind speed and precipitation data are
missing. In order to obtain a continuous atmospheric
forcing series, data from neighbouring automatic weath-
er stations (Poucharramet: 43°25¢N, 1°11¢E, altitude
204 m, and Ondes: 43°47¢N, 1°19¢E, altitude 108 m) and
of the Francazal airport station (43°32¢N, 1°22¢E,
altitude 164 m) were added to the database. Since the
downwelling atmospheric thermal emission (Ra) was not
measured in the supplementary weather stations, the 3%
missing data in the original dataset were completed by
the following formulation, adapted from Staley and
Jurica (1972): moisture content, a characterisation of the vegetation,
and 30-min recording of temperature pro®les in the soil. 3.1 Atmospheric measurements The equipment employed was similar to the SAMER
station described in Goutorbe (1991). Air temperature
and humidity were measured at screen-level (2 m). The
wind speed (U) and direction were measured by a
propeller anemometer at 10 m above the soil surface. The atmospheric pressure was measured and recorded
automatically. The station was also able to document
the surface energy balance: the net radiation was
measured together with the ground heat ¯ux, and the
sensible heat was calculated from two-level measure-
ments (1.5 m apart) of air temperature and wind speed. The accuracy of the two-level measurements (air tem-
perature and wind speed vertical gradients) was one
order of magnitude better than the original design: the
sensors were changed and thoroughly intercompared at
the same level, under very distinct atmospheric condi-
tions (i.e. dierent diurnal cycles, strong and low winds,
rain/no rain, etc.). However, other uncertainties (e.g. concerning the canopy height estimation) may aect the
obtained ¯uxes at short time scales, and the expected
accuracy was about 20% (Goutorbe, 1991). g
The main dierences between the three years are
related to changes in the rainfall regime: 1995 was
characterised
by
evenly
distributed
precipitation
(770 mm); 1996 was the wettest year (860 mm), and
1997 was a dry year (540 mm) marked by a spring and
an autumn drought (Fig. 2). The droughts of 1997 had a
signi®cant impact on the measured surface ¯uxes
(Fig. 4): the sensible heat ¯ux exceeded evapotranspira-
tion during the autumn 1997 (i.e. from September
onward), and the monthly evapotranspiration never
exceeded 300 MJ m)2, contrary to that which occurred
in 1995 and 1996. Note that the reduced maximum
monthly evapotranspiration in 1997 was also related to
lower values of incoming solar radiation (Fig. 2). 3.2 Deep soil moisture measurements Rainfall (P) was recorded automatically using a
tipping bucket rain gauge. Also, shortwave and total
(0.3±60 lm), upward and downward radiations were
measured. The MUREX station was very reliable for the
entire period studied. During 1995 for example less than
6% of surface ¯ux data are missing, 1 to 3% of the
radiation (the solar incident radiation Rg and the Deep soil moisture pro®les were obtained on a weekly
basis from neutron probe measurements. the measured
soil moisture pro®les corresponded to 0.10±0.15 m
intervals, from the surface to 1.35 m. The soil water
potential was estimated from tensiometric measure-
ments at 0.1 m intervals within the 0.3 m surface layer 841 J.-C. Calvet et al.: MUREX: a land-surface ®eld experiment Fig. 2. The measured precip-
itation and incoming radia-
tion, Rg and Ra, over the
MUREX fallow in 1995,
1996, and 1997. The plotted
precipitation and incoming
radiation are monthly sums of
the 30-min measured values Fig. 2. The measured precip-
itation and incoming radia-
tion, Rg and Ra, over the
MUREX fallow in 1995,
1996, and 1997. The plotted
precipitation and incoming
radiation are monthly sums of
the 30-min measured values and at 0.2 m intervals below, down to 1.3 m. The
measurements were performed at three positions within
the fallow site. They were selected after a preliminary
gravimetric water contents survey according to the
method proposed by Vachaud et al. (1985). In this
study, the average value is considered. Once properly
calibrated, the neutron probe technique allows accurate
measurements of the soil water content. The average
volumetric soil moisture measurements of the three years are displayed in Fig. 4. Again, the three annual
cycles contrast sharply: soil water extraction never
reached the 1 m depth in 1996, whereas water was
extracted over the whole pro®le (i.e. as deep as 1.35 m)
in 1995 from July to September. In 1997, water was
extracted below 1.35 m, from August onwards. It is
interesting to note that the precipitation occurring
during the autumn 1997 was not large enough to re®ll
the soil reservoir by the end of the year. Fig. 3. As in Fig. 2, except for
air temperature and humidity,
Ta and qa and wind speed. Air
temperature, humidity, and
wind speed are displayed as
monthly means with maximum
and minimum values indicated
with bars. 3.2 Deep soil moisture measurements Air temperature and
humidity are represented by
diamonds with solid lines, and
by boxes with solid thick lines,
respectively Fig. 3. As in Fig. 2, except for
air temperature and humidity,
Ta and qa and wind speed. Air
temperature, humidity, and
wind speed are displayed as
monthly means with maximum
and minimum values indicated
with bars. Air temperature and
humidity are represented by
diamonds with solid lines, and
by boxes with solid thick lines,
respectively Fig. 3. As in Fig. 2, except for
air temperature and humidity,
Ta and qa and wind speed. Air
temperature, humidity, and
wind speed are displayed as
monthly means with maximum
and minimum values indicated
with bars. Air temperature and
humidity are represented by
diamonds with solid lines, and
by boxes with solid thick lines,
respectively Fig. 3. As in Fig. 2, except for
air temperature and humidity, J.-C. Calvet et al.: MUREX: a land-surface ®eld experiment 842 Fig. 4. Monthly sums of the
measured ¯uxes (net radiation,
Rn; sensible heat ¯ux; H; latent
heat ¯ux LE; and ground heat
¯ux, G over the MUREX fal-
low in 1995, 1996, and 1997
(note that LE is estimated by
dierence), and a temporal dia-
gram of the soil volumetric
moisture within the 1.35 m soil
layer as measured with a neu-
tron probe over the MUREX
fallow Fig. 4. Monthly sums of the
measured ¯uxes (net radiation,
Rn; sensible heat ¯ux; H; latent
heat ¯ux LE; and ground heat
¯ux, G over the MUREX fal-
low in 1995, 1996, and 1997
(note that LE is estimated by
dierence), and a temporal dia-
gram of the soil volumetric
moisture within the 1.35 m soil
layer as measured with a neu-
tron probe over the MUREX
fallow Fig. 4. Monthly sums of the
measured ¯uxes (net radiation,
Rn; sensible heat ¯ux; H; latent
heat ¯ux LE; and ground heat
¯ux, G over the MUREX fal-
low in 1995, 1996, and 1997
(note that LE is estimated by
dierence), and a temporal dia-
gram of the soil volumetric
moisture within the 1.35 m soil
layer as measured with a neu-
tron probe over the MUREX
fallow Fig. 4. Monthly sums of the
measured ¯uxes (net radiation, Using the estimated soil moisture content change
together with the measured precipitation and cumulated
evaporation and cumulated evaporation, it is possible to
assess the water balance on a weekly basis. 3.3 Water balance Simple calculations based on tensiometric measure-
ments at the bottom of the pro®le indicate that
capillarity rises from below 1.35 m do not exceed
4 mm in 1995 (Calvet et al., 1998a). Therefore, most
of the water excess seems to be due to lateral movements
of water. In the case of the MUREX fallow, part of the
in¯ow may be supplied by upstream rain and irrigation
drained water (a large proportion of the uphill ®elds
consists of irrigated maize). According to the available measurements, the water
balance of the MUREX fallow is rather unrepresenta-
tive of the climatic or large-scale evaporation over
southwestern France. For example, the cumulated
estimated evapotranspiration is about 800 mm for
1995, whereas the value of the cumulated precipitation
for the same period is 770 mm. The evaporation excess
can be explained by a local subsurface soil water
convergence which is likely to occur in this kind of
terrain. The in situ measurements of bulk soil moisture,
precipitation and evapotranspiration can be employed
to estimate the weekly water excess Xs (Calvet et al.,
1998a): 3.2 Deep soil moisture measurements of the studied annual cycles represents a deep water gain
of about 200 mm. The in¯ow can be due to lateral
movements of water and (to a lesser extent) to capillarity
rises from soil layers deeper than 1.35 m. It produces a
signi®cant contribution to the local water balance,
especially during the drying periods. Fig. 4. Monthly sums of the
measured ¯uxes (net radiation,
Rn; sensible heat ¯ux; H; latent
heat ¯ux LE; and ground heat
¯ux, G over the MUREX fal-
low in 1995, 1996, and 1997
(note that LE is estimated by
dierence), and a temporal dia-
gram of the soil volumetric
moisture within the 1.35 m soil
layer as measured with a neu-
tron probe over the MUREX
fallow 4 Vegetation dynamics and the water
and carbon dioxide ¯uxes Fallow vegetation resulting from set-aside cultivated
®elds is currently widespread in western Europe. The
vegetation of the MUREX fallow consists of many C3
herbaceous plant species whose growing cycles overlap,
contributing to maintain a rather dense, evergreen
canopy. In order to characterise the fallow, the speci®c
contribution of each species was determined using a
simpli®ed point quadrats method (Daget and Poissonet,
1971), and the spatial heterogeneity of the vegetation
was investigated. Over a total number of 81 species
identi®ed on the site, the plants making up the canopy at
a given time represented 38 to 55 dierent species. However, the fallow was dominated by a few species
only: Brachypodium sp (45% of the plants), and 2 2 Xs DS RLE ÿ RP
2 Xs DS RLE ÿ RP layers (down to 0.70 m). This implies that a large part of
the soil water extraction occurring below 0.25 m (Fig. 4)
is due to capillarity rises from deeper soil layers, but
direct root extraction from below 0.25 m may occur. The deep root extraction could not be quanti®ed. Using the shortwave radiation measurements men-
tioned before, it was possible to determine the albedo
(a) of the canopy: a 0:20 0:04 in 1995 and 1996. In
1997,
a
slightly
higher
value
was
obtained:
a 0:21 0:04. Again, this higher value may be related
to a quicker wilting and yellowing of the leaves due to
the considerable water stress of 1997. In this section, speci®c biomass and ecophysiological
measurements performed with in the framework of
MUREX are presented. Green leaf area index (LAI) and
biomass measurements were performed by destructive
planimetric measurements by CESBIO and CNRM. Figure 6 shows the measured values of LAI together
with the dead and living above-ground biomass over the
three annual cycles. The dead vegetation residues tend to
form a rather dense vegetal mulch at the soil surface. In
1995, the fallow was mown on DoY 152, thus increasing
the mulch dead biomass. The rather large scatter of LAI
is due to the heterogeneous vegetation species distribu-
tion. However, clear interannual dierences emerge
from the series of observations displayed in Fig. 6. In
particular, the maximum LAI was lower in 1997 than in
1996 (about 2 and 3.5 m2 m)2, respectively), consistent
with the precipitation de®cit of 1997, compared with
1996. Water and carbon dioxide ¯uxes in the soil-plant-
atmosphere continuum widely depend on the vegetation. A speci®c study was conducted during the 1997 and
1998 growing seasons to determine the seasonal and
diurnal dynamics of the vegetation features as in¯uenced
by environmental conditions. Stomatal conductance and water potential were
measured on leaves of major species from the early
spring to the autumn of 1997, along with the biomass
and LAI dynamics of the ®eld vegetation. Xs DS RLE ÿ RP Xs DS RLE ÿ RP where DS is the change in the soil water storage (S)
between two consecutive measurements, from the sur-
face to a depth of 1.35 m, and RLE and RP are the
cumulated values of evapotranspiration and precipita-
tion over the period considered, respectively. The
cumulated positive values of Xs are presented in
Fig. 5, together with the cumulated precipitation. In
general, winter is a period of runo displaying negative
values of Xs. Conversely, the in¯ow episodes (positive
peak values of Xs) are observed from the beginning of
spring to the middle of the autumn. Their sum over any J.-C. Calvet et al.: MUREX: a land-surface ®eld experiment 843 Fig. 5. Cumulative precipita-
tion and positive values of the
water excess derived from the
total soil moisture content,
precipitation and evapotran-
spiration measurements over
the MUREX fallow in 1995,
1996, and 1997 Fig. 5. Cumulative precipita-
tion and positive values of the
water excess derived from the
total soil moisture content,
precipitation and evapotran-
spiration measurements over
the MUREX fallow in 1995,
1996, and 1997 Fig. 5. Cumulative precipita-
tion and positive values of the
water excess derived from the
total soil moisture content,
precipitation and evapotran-
spiration measurements over
the MUREX fallow in 1995,
1996, and 1997 Fig. 5. Cumulative precipita-
tion and positive values of the
water excess derived from the
total soil moisture content,
precipitation and evapotran-
spiration measurements over
the MUREX fallow in 1995,
1996, and 1997 Potentilla reptans (22%) were the main dominant
species, together with Erigeron canadensis, Epilobium
tetragonum, and Rumex acetosa. This kind of herba-
ceous canopy represents the ®rst stage of a plant
succession (once cultivation has stopped) leading to a
Quercus pubescens forest climax. Despite the wide area
covered by this type of fallow, they have not been
intensively investigated until now, especially concerning
their primary production, seasonal dynamics, and the
role they play in land-atmosphere exchanges. Potentilla reptans (22%) were the main dominant
species, together with Erigeron canadensis, Epilobium
tetragonum, and Rumex acetosa. This kind of herba-
ceous canopy represents the ®rst stage of a plant
succession (once cultivation has stopped) leading to a
Quercus pubescens forest climax. Despite the wide area
covered by this type of fallow, they have not been
intensively investigated until now, especially concerning
their primary production, seasonal dynamics, and the
role they play in land-atmosphere exchanges. Xs DS RLE ÿ RP The chamber
measurements enable estimation of the dierent com-
ponents of the net CO2 ¯ux: (1) the unscreened chamber
over the undisturbed canopy provides the net CO2-¯ux;
(2) by screening the chamber over the undisturbed
canopy, one obtain the total soil and leaf respiration
(photosynthesis is suppressed); and (3) by cutting the
vegetation before the observation, an estimate of soil
respiration is obtained (photosynthesis and leaf respira-
tion are suppressed). The ®rst term should correspond to
th
t
h
i
¯
d b
th
dd
l ti employed: chamber measurements were performed by
CESBIO (using a Ciras-1 PPSystems infrared gas
analyser), and CNRM used the eddy-correlation meth-
od (using a SOLENT ultrasonic anemometer, and a
LICOR LI-6262 infrared gas analyser). The chamber
measurements enable estimation of the dierent com-
ponents of the net CO2 ¯ux: (1) the unscreened chamber
over the undisturbed canopy provides the net CO2-¯ux;
(2) by screening the chamber over the undisturbed
canopy, one obtain the total soil and leaf respiration
(photosynthesis is suppressed); and (3) by cutting the
vegetation before the observation, an estimate of soil
respiration is obtained (photosynthesis and leaf respira-
tion are suppressed). The ®rst term should correspond to
the atmospheric ¯ux measured by the eddy-correlation
method. Figure 7 presents a series of bioclimatic
variables observed on DoY 247 in 1997. The atmo-
spheric net CO2-¯ux obtained by the two methods
(either chamber or eddy-correlation measurements) are
in good agreement. The canopy net assimilation An (i.e. the quantity of CO2 integrated into the vegetation
biomass through the photosynthesis process) may be
much higher than the measured atmospheric CO2-¯ux
presented in Fig. 7 because the plant may absorb the
CO2 emitted by the soil, also. An estimation of An is the
dierence between the atmospheric CO2-¯ux and soil
respiration: on DoY 247 in 1997 (Fig. 7), the obtained
net
assimilation
reaches
0.5 mg m)2 s)1
at
about
0900 LST and 1500 LST, whereas the atmospheric
CO2-¯ux does not exceed 0.3 mg m)2 s)1. The An and
evapotranspiration `plateau' observed between 0900 and
1500 LST (Fig. 7) may be related to stomatal closure
due to high temperatures or to a large saturation de®cit
of the air within the canopy. It is interesting to note that
the average behaviour of the canopy in terms of stomatal conductance may result from distinct diurnal
variations of this parameter from one plant species to
another (Fig. 7). Xs DS RLE ÿ RP The main
conclusions of this series of biophysical measurements
for the MUREX fallow are: (1) the plant species
exhibited low stomatal resistances, indicating their
potential ability to maintain high CO2 and water vapour
¯uxes; (2) some dierences between coexisting species
occurred, concerning diurnal variations of stomatal
conductance and water potential; (3) the diurnal vari-
ations of leaf stomatal resistance and water potential
diered greatly according to air saturation de®cits and
soil moisture conditions. As far as root biomass is concerned, an average value
of 160 g m)2 was obtained by CESBIO from measure-
ments performed on ®ve dierent dates in 1995. It must
be noted that most of the roots were found in the 0.25 m
surface soil layer. A few roots, representing a very small
part of the total root biomass, were observed in deeper Carbon dioxide ¯uxes were measured during the 1997
summer period, and in 1998. Two methods were J.-C. Calvet et al.: MUREX: a land-surface ®eld experiment 844 Fig. 6. The green leaf area
index and above-ground green
biomass (boxes) and necromass
(thick dashed line and triangles)
measured over the MUREX
fallow in 1995, 1996, and 1997. The discontinuity on day 152 is
due to the cutting of the vege-
tation Fig. 6. The green leaf area
index and above-ground green
biomass (boxes) and necromass
(thick dashed line and triangles)
measured over the MUREX
fallow in 1995, 1996, and 1997. The discontinuity on day 152 is
due to the cutting of the vege-
tation Fig. 6. The green leaf area
index and above-ground green
biomass (boxes) and necromass
(thick dashed line and triangles)
measured over the MUREX
fallow in 1995, 1996, and 1997. The discontinuity on day 152 is
due to the cutting of the vege-
tation Fig. 6. The green leaf area
index and above-ground green
biomass (boxes) and necromass
(thick dashed line and triangles)
measured over the MUREX
fallow in 1995, 1996, and 1997. The discontinuity on day 152 is
due to the cutting of the vege-
tation employed: chamber measurements were performed by
CESBIO (using a Ciras-1 PPSystems infrared gas
analyser), and CNRM used the eddy-correlation meth-
od (using a SOLENT ultrasonic anemometer, and a
LICOR LI-6262 infrared gas analyser). 5.1 Surface soil moisture measurements The surface temperatur
either infrared-derived (thick line) or measured in the soil (dashed li
here at 5-cm. The energy budget is given by the SAMER system
radiation Rn; heat ¯ux H: latent heat ¯ux LE; and ground heat
G). The air saturation de®cit (dashed line) is obtained from scre
level air temperature and humidity data, and the de®cit at the sur
(solid thick line) is computed from the infrared-derived temperatu
J.-C. Calvet et al.: MUREX: a land-surface ®eld experiment Fig. 7. Measurements characterising the bioclimatological function-
ing of the MUREX fallow on DoY 247 in 1997. The atmospheric
CO2 ¯ux and concentration are measured at 2.7 m above the soil
surface (solid lines); chamber measurements of the net CO2 ¯ux are
obtained over Potentilla- and Brachypodium-dominated spots (pluses
and diamonds, respectively); the negative CO2 ¯ux represented by
boxes corresponds to chamber measurements of soil respiration. The
CO2 ¯ux represented by the dashed line is an estimation of daytime net
assimilation given by the dierence between the atmospheric CO2 ¯ux and soil respiration. In situ measurements of leaf stomatal conduc-
tance are obtained over Potentilla- and Brachypodium-dominated
spots (pluses and diamonds, respectively). The surface temperature is
either infrared-derived (thick line) or measured in the soil (dashed line),
here at 5-cm. The energy budget is given by the SAMER system (net
radiation Rn; heat ¯ux H: latent heat ¯ux LE; and ground heat ¯ux
G). The air saturation de®cit (dashed line) is obtained from screen-
level air temperature and humidity data, and the de®cit at the surface
(solid thick line) is computed from the infrared-derived temperature The automatic method was implemented at the end
of 1996. Four capacitive probes (ML1 ThetaProbe,
Delta-T Devices 1996) were inserted into the soil surface
and kept vertical by using an adequate frame. The
output volumetric moisture represents an integrated
value over the top 6-cm at the soil surface. The
observations were recorded on a 15-min basis, and a
linear correction was applied to the raw data according
to manual measurements (oven-drying method). After
calibration, the rms dierence between the ThetaProbe
data
and
the
reference
measurements
was
about
0.01 m3 m)3. The average surface soil moisture is presented in
Fig. 8, together with the bulk soil moisture obtained
from the neutron-probe measurements over the 1.35 m
pro®le. 5.1 Surface soil moisture measurements Surface soil moisture was estimated either manually (by
weighing samples collected in the ®eld) or automatically
(in situ capacitive measurements). (
p
)
In 1995, soil moisture pro®les within the top 5-cm soil
layer were measured manually, twice a day, during two
30-day intensive observing periods (IOPs) in spring and
autumn: 1-from DoY 114 to 143 for the spring-IOP, 2-
from DoY 269 to 298 for the autumn-IOP. Each
measurement consisted of the gravimetric moisture of
36 (6 layers ´ 6 sites) soil samples, determined by a
direct oven-drying method. The soil moisture content
was measured: (1) from the surface to the 5-cm depth
with a resolution of 1 cm, and (2) for the 0.5-cm surface
layer. The volumetric moisture w (m3 m)3) is derived
from the gravimetric measurements h (kg of water per
kg of dry soil) by: w qb
qw
h
3 w qb
qw
h 3 where qb is the soil dry bulk density and qw the density
of liquid water (kg m)3). The value of qb did not change
signi®cantly from one IOP to the other: the measured
values over the surface 5-cm layer are 1452 87 and
1450 114 kg m)3 on DoY 123 and 286 of 1995,
respectively. J.-C. Calvet et al.: MUREX: a land-surface ®eld experiment 845 The automatic method was implemented at the end
of 1996. Four capacitive probes (ML1 ThetaProbe,
Delta-T Devices 1996) were inserted into the soil surface
and kept vertical by using an adequate frame. The
The average surface soil moisture is presented
Fig. 8, together with the bulk soil moisture obtain
from the neutron-probe measurements over the 1.35
pro®le. Interestingly, the measurements of 1997 sh
Fig. 7. Measurements characterising the bioclimatological function-
ing of the MUREX fallow on DoY 247 in 1997. The atmospheric
CO2 ¯ux and concentration are measured at 2.7 m above the soil
surface (solid lines); chamber measurements of the net CO2 ¯ux are
obtained over Potentilla- and Brachypodium-dominated spots (pluses
and diamonds, respectively); the negative CO2 ¯ux represented by
boxes corresponds to chamber measurements of soil respiration. The
CO2 ¯ux represented by the dashed line is an estimation of daytime net
assimilation given by the dierence between the atmospheric CO2 ¯ux
and soil respiration. In situ measurements of leaf stomatal cond
tance are obtained over Potentilla- and Brachypodium-domina
spots (pluses and diamonds, respectively). 5.1 Surface soil moisture measurements Interestingly, the measurements of 1997 show
that the surface soil moisture can be much lower than
the bulk soil moisture during long periods (e.g. Septem-
ber and October 1997). This is related to the vertical
gradients of texture and bulk density observed on the
site: the wilting-point soil moisture is lower close to the
surface than deeper in the soil, permitting lower values
of the soil water content. Direct soil evaporation is
expected to have a minor in¯uence on the decrease of J.-C. Calvet et al.: MUREX: a land-surface ®eld experiment 846 Fig. 8. Monthly averages of the
measured infrared-derived sur-
face temperature (thick solid
line), and of the soil temperature
at 1-cm (dashed line) and 50-cm
(solid line) below the soil surface,
and the measured soil water
content of the 1.35 m column
(diamonds) and surface soil wa-
ter content over the top 5±6 cm
(solid line), in 1995, 1996, and
1997 Fig. 8. Monthly averages of the
measured infrared-derived sur-
face temperature (thick solid
line), and of the soil temperature
at 1-cm (dashed line) and 50-cm
(solid line) below the soil surface,
and the measured soil water
content of the 1.35 m column
(diamonds) and surface soil wa-
ter content over the top 5±6 cm
(solid line), in 1995, 1996, and
1997 surface soil moisture because of the mulch screening
eect. surface soil moisture because of the mulch screening
eect. 6 Modelling approaches The main purpose of the MUREX experiment is to
improve the representation of the land-surface processes
in atmospheric models. Improved numerical experi-
ments and new analysis algorithms and parametrisations
based on these data are presented in other studies. In
particular, the ISBA scheme was employed in several
con®gurations to analyse the MUREX data: (1) the data
of 1995 were used to test an assimilation technique to
analyse the total soil water content from surface soil
moisture or surface temperature (Calvet et al., 1998a);
and (2) a new version of ISBA including CO2 assimi-
lation and plant growth (ISBA-Ags) was applied to the
data of 1995 and 1996 (Calvet et al., 1998b). Fig. 8. Monthly averages of the
measured infrared-derived sur-
face temperature (thick solid
line), and of the soil temperature
at 1-cm (dashed line) and 50-cm
(solid line) below the soil surface,
and the measured soil water
content of the 1.35 m column
(diamonds) and surface soil wa-
ter content over the top 5±6 cm
(solid line), in 1995, 1996, and
1997 5.2 Surface temperature and emissivity The SiSPAT model (Braud et al., 1995) belongs to
a more complex class of SVATs: the vegetation structure
is more detailed, and a multilayer description of the soil
is employed. Conversely, the ALiBi model (Olioso,
1992) was designed to analyse remote sensing data such
as infrared brightness temperature in order to diagnose
evapotranspiration and photosynthesis without moni-
toring the root-zone soil moisture. In ALiBi, the
characterisation of the physiological plant functioning
is improved. In this study, the four models are run over
1995 with the interpolated LAI curve obtained by Calvet
et al. (1998a) from the observed values. Also, the
photosynthesis modelled by ISBA-Ags and ALiBi are
compared. Calvet et al. (1998a): the Xs term is very signi®cant in the
case of the MUREX fallow, and the time series of Xs
must be prescribed in the evolution equation of w2 in
order to obtain a good agreement between simulated and
observed values of both surface evapotranspiration and
root-zone soil moisture. Furthermore, estimates of the
deep temperature are required to avoid severe drifts of
the cumulated model heat ¯ux in the soil. In this study,
the prescribed deep temperature is taken as the measured
temperature at 50 cm below the soil surface (Tÿ50). p
(
50)
The description of the surface ¯uxes Rn, H, and LE is
detailed in Noilhan and Planton (1989). The main
prescribed parameters of the surface involved in the ¯ux
calculation are: the surface albedo and emissivity (a and
es, respectively), the momentum and thermal roughness-
es (z0 and z0h, respectively) and the vegetation LAI and
minimal stomatal resistance (rsmin, expressed in units of
sm)1). This latter term depends on the plant species
making up the canopy, and in general must be calibrat-
ed. The surface albedo and emissivity are prescribed
from measurements and z0 is derived from the vegeta-
tion height estimates (h) through: z0 = 0.13 ´ h. Ta-
ble 1 displays the ISBA structure parameters either
obtained from the in situ measurements, derived from
the soil texture (wwilt, wsat), or calibrated (wfc, c, es, Cv,
rsmin, z0/z0h). The procedure to obtain calibrated values
is detailed in Calvet et al. (1998a). Note that in this
study, all the parameters were calibrated according to
Rn, H, and LE, whereas the cumulative G was also used
in Calvet et al. (1998a). 5.2 Surface temperature and emissivity In this study, six temperatures characterising the surface
were measured: the infrared temperature (TIR) is ob-
tained from an infrared radiometer, and the soil
temperature at 1, 5, 20, 50, and 120 cm below the soil
surface (Tÿ1; Tÿ5 and Tÿ20; Tÿ50, and Tÿ120, respective-
ly) were measured with platinum-resistance thermome-
ters. The surface temperature Ts was derived from TIR
and an eective emissivity (eeff ): Ts
T 4
IR=eeff
8
:
9
;
1=4
4 4 In this study, simulations performed by four models
over the 1995 annual cycle (ISBA, MAGRET, SiSPAT,
and ALiBi) are compared. All the models are driven by
measurements of incoming radiation, precipitation,
atmospheric pressure, air temperature and humidity,
and wind speed at a reference level. Also, vegetation
characteristics such as leaf area index and canopy height
must be prescribed. These parameters may change with
time. The ISBA (Noilhan and Planton, 1989) and
MAGRET (Lagouarde, 1991; Courault et al., 1996)
models have similar functions and input and output
variables: they both describe the water division between
the vegetation transpiration, the drainage, the surface
runo and the soil moisture increase or decrease, and
they both solve a single energy budget and compute the
surface temperature. This kind of models represents a
particularly robust approach to describe how land
surface functions over long periods, at low computing The value eeff 0:977 could be estimated from Eq. (4)
by assuming that Ts Tÿ1 when no vertical gradients of
temperature were observed at the soil surface (i.e. Tÿ1 Tÿ5) after sunset, between 2000 and 2100 Local
Standard Time (LST). The thermal emissivity of the surface (es) is a
parameter of the SVATs. The value of es can be
retrieved from Ts and the radiation measurements. Indeed, the measured upwelling thermal emission of
the surface (Rs) can be written as: Rs es rT 4
s
1 ÿ esRa
5 Rs es rT 4
s
1 ÿ esRa 5 The value es 0:97 minimises the rms dierence be-
tween the measured value of Rs, and the value given by
Eq. (5): the rms dierence is 10.6 Wm)2 and over 1995,
the mean bias is less than 2 Wm)2. 847 J.-C. Calvet et al.: MUREX: a land-surface ®eld experiment costs. 5.2 Surface temperature and emissivity Therefore, the Table 1 param-
eter values (especially Cv and z0/z0h) dier from those
given by Calvet et al. (1998a). The calibrated ratio z0/z0h
(Table 1) employed for ISBA (and MAGRET) reaches a
high value of 450, whereas a value of 10 is usually
employed for small vegetation canopies. However, many
authors have reported high values of z0/z0h (even higher
than 450) for a wide range of surfaces (see Verhoef et al.,
1997 for a review, and Trou¯eau et al., 1997). Verhoef
et al. (1997) suggest that z0/z0h = 10 is probably too
low for most natural surfaces. 6.2 MAGRET The working of MAGRET is detailed in Courault et al. (1996). It is very similar to ISBA. In particular, the input
atmospheric, radiation, and vegetation variables are the
same. Dierences arise from the way evapotranspira-
tion, and soil moisture are computed. Contrary to ISBA,
bare soil evaporation and vegetation transpiration are
not distinguished. Whereas they are calculated separate-
ly is ISBA, the total canopy evapotranspiration is
obtained using a bulk canopy resistance including
vegetation structure resistances, a resistance to soil
evaporation related to the surface dry soil layer, and the
plant stomatal resistance. This latter resistance is
calculated in the same way as in ISBA. Concerning soil
moisture, the two reservoir systems proposed by Jac-
quart and Choisnel (1995) is employed. Each reservoir
corresponds to a layer of wetted soil, the thickness of
which vary according to the computed loss (evapotran-
spiration) or gain (rainfall) of water. Since the MUREX
water excess cannot be included in a bulk soil reservoir
as in ISBA, the water excess was added to the
precipitation prescribed for MAGRET (Table 2). In order to better described the particular behaviour
of the MUREX fallow, the SiSPAT model was modi®ed
to account for the mulch eect (Gonzalez-Sosa et al.,
Submitted 1999). Indeed, Gonzalez-Sosa et al. (1999)
showed that, using SiSPAT, water and energy transfer
within the mulch layer must be explicitely described,
especially after the vegetation cutting in 1995, on DoY
152. Accounting for the mulch greatly improves the
SiSPAT simulations over MUREX. To summarise, the
mulched simulations present higher values of bulk soil
moisture and lower values of evapotranspiration. The
¯ux simulation after the vegetation cutting of DoY 152
is erroneous in non-mulched simulations. In particular
soil evaporation is greatly overestimated, which brings
about a rapid soil moisture depletion aecting the
quality of the ¯ux simulation afterwards. In the simu-
lations presented, SiSPAT was run with the same rsmin
than ISBA and MAGRET (Table 1). Displacement
height and roughness length for momentum were
calculated as a function of leaf area index and vegetation
height (Shuttleworth and Wallace, 1985). The z0/z0h
values are given by the formula of Brutsaert (1982, cited
by Verhoef et al., 1997). The rooting depth varies with
time from 0.5 to 0.7 m according to the observed soil
moisture pro®les. The maximum root density was set at
2 ´ 104 m m)3. 6.1 ISBA In this study, the most recent standard version of ISBA
(Noilhan and Mahfouf, 1996), implemented within the
Me te o-France global climate model ARPEGE by Ma-
hfouf et al. (1995) is employed. The ISBA scheme
simulates the surface ¯uxes (LE, H, G) and predicts the
evolution of the surface state variables using the
equations of the force-restore method of Deardor
(1977, 1978). Five variables (surface temperature Ts,
mean surface temperature T2, surface soil volumetric
moisture wg, total soil volumetric moisture w2, and the
canopy interception reservoir Wr) are obtained through
prognostic equations. It must be noted that ISBA does
not need a root distribution: w2 is the volumetric soil
moisture associated to a bulk layer of thickness d2
including the root zone. The surface soil moisture wg is
computed to estimate the evaporation from the soil
surface, whereas the transpired water is extracted from
w2. The surface water quantity from wg is included in w2. Initial values of soil variables are required. In this study,
the contribution of the water excess caused by a perched
aquifer or by deep capillarity rises (measured positive
values of Xs) is accounted for in the water budget as in Table 1. The soil and vegetation structure parameters over the MUREX fallow for ISBA and MAGRET
Symbol
De®nition
Value
ISBA and MAGRET:
d2
Soil root depth
1.35 m
CLAY
Clay fraction
28%
SAND
Sand fraction
14%
wwilt
Wilting point
0.20 m3 m)3
wfc
Field capacity
0.34 m3 m)3
a
Albedo
0.20
es
Emissivity
0.97
rsmin
Minimum stomatal resistance
50 sm)1
z0/z0h
Roughness length ratio
450
ISBA:
c
Deep heat transfer contribution
1
veg
Vegetation coverage
95%
Cv
Thermal coecient
2.0 ´ 10)5 Km2 J)1
MAGRET:
d
Extinction coecient for ground heat ¯ux
0.6 m2 m)2 Table 1. The soil and vegetation structure parameters over the MUREX fallow for ISBA and MAGRET J.-C. Calvet et al.: MUREX: a land-surface ®eld experiment 848 together with separate energy budgets of vegetation and
bare soil (Deardor, 1978). As a lower boundary
condition for the soil pro®le, the measured time
evolution of matric potential at 1.3 m was imposed. Since no deep measurement of soil temperature was
available in 1995, the average values of the 1996 and
1997 series (at 1.2 m) were used. 6.2 MAGRET A vegetation albedo of 0.15 was
employed, and the bare soil or mulch albedo was
calculated so that the total albedo equal 0.20. The other
vegetation parameters employed in SiSPAT are: the
critical leaf potential ()140 m), and the total plant
resistance (3.2 1012 s m)1). Another dierence concerns the calculation of the
ground heat ¯ux G and the eect of vegetation. In ISBA,
G is the residual of the energy budget. The obtained
value of G is employed in the pronostic equation for the
surface temperature, weighted by a thermal coecient
including a vegetation term Cv. In MAGRET, G is
computed from the temperature gradient at the surface
and an exponential attenuation term depending on LAI
times the extinction coecient d (Table 1). The MAG-
RET model computes the diurnal evolution of the
canopy albedo. In this study, the albedo module was
inactivated and the average value of 0.2 was imposed as
in ISBA. 6.3 SiSPAT Leaf stomatal conductances depend on incident PAR at
leaf surface, leaf water potential and air saturation
de®cit at leaf surface (Winkel and Rambal, 1990). Another important equation describes water transfers
from the soil to the leaves, where transpiration occurs. Like SiSPAT, it follows Van den Honert's formulation
and relates transpiration (LEt) to the variation of water
potential between the soil and the leaves (h and hv,
respectively) by means of an hydraulic conductance
(Gp), assumed constant for a given day: K
w Ksat
w=wsatc
7 7 where Ksat
is the saturated hydraulic conductivity
(m s)1) and c a scale factor. Field estimates of Ksat were
obtained from in®ltration tests performed using either
tension disk in®ltrometers (Vauclin and Chopart, 1992)
or Guelph in®ltrometers (Elrick and Reynolds, 1992),
for the 0±30 cm surface layer and the deepest layers,
respectively. The thermal conductivity k (expressed in units of
Wm)1 K)1) is given by the following equation, obtained
from in situ measurements by the line source method
(Laurent, 1989), completed with a laboratory charac-
terisation
of
the
volumetric
moisture
content
(w)
dependence: LEt Gp
h ÿ hm
9 9 LEt Gp
h ÿ hm A detailed description of water transfer, stomatal
conductance and photosynthesis calculation was pre-
sented by Olioso et al. (1996a). k 0:729 0:135
w=wsat
1:26 1 ÿ exp ÿ10:39
w=wsat1:84
h
i
8 y
(
)
In the case of the MUREX dataset, evapotranspira-
tion, canopy energy ¯uxes and photosynthesis were
estimated by inverting the ALiBi model daily, according
to the measured thermal brightness temperature (TIR) at
midday (174 days between DoY 125 and 317 in 1995):
each day, the simulated midday TIR was ®tted to the
measured value (0.25 K) by adjusting the hydraulic
conductance parameter Gp. The other necessary inputs
were: (1) either directly measured (LAI, canopy height,
air temperature, air humidity and wind speed at a
reference level above the canopy, solar and atmospheric
radiations, atmospheric pressure), (2) computed from
measured inputs (incident PAR, solar and PAR diuse
fraction and atmospheric radiation in the spectral range
of the thermal radiometer), and (3) estimated from the
values obtained for a soyabean crop in Avignon (Olioso
et al., 1996a). The latter parameters consisted of soil
thermal characteristics, leaf aerodynamic properties,
parameters for stomatal conductance and leaf photo-
synthesis, optical properties of leaves and soil and leaf
angle distribution. 6.3 SiSPAT Field estim
in®ltration tests perfor
®ltrometers (Vauclin an
trometers (Elrick and
ameters for the ®ve SiSPA
rated (*) using the SiSPAT
he van Genuchten (VG) and
(6) and (7), respectively): n
Porosity
wsat
(m3 m)3)
0.60
0.48
0.55
0.46
0.50
0.40
0.45
0.35
0.39
0.30
MUREX: a land-surface ® ated water content (m3 m)
he hydraulic conductivity c
Corey (1964):
aturated hydraulic condu
actor. Field estimates of Ksa
tion tests performed using
ters (Vauclin and Chopart
for the ®ve SiSPAT's horizons
using the SiSPAT model. They
Genuchten (VG) and Brooks and
(7), respectively): n and hg are sh
ity
wsat
(m3 m)3)
n (VG m
0.48
2.30
0.46
2.30
0.40
2.30
0.35
2.27
0.30
2.13
X: a land-surface ®eld experime Table 3. Soil parameters for the ®ve SiSPAT's horizons, either
measured or calibrated (*) using the SiSPAT model. They include
the coecients of the van Genuchten (VG) and Brooks and Correy
(BC) models (Eqs. (6) and (7), respectively): n and hg are shape and scale parameters of the VG model, and c is the shape parameter of
the BC model. The thermal conductivity k is given by Eq. (8) below
0.02 m. The saturated water content and hydraulic conductivity are
denoted by wsat and Ksat, respectively Horizon
(m)
Porosity
wsat
(m3 m)3)
n (VG model)
hg (VG model)
(m)
Ksat
(m s)1)
c (BC model)
k
(Wm)1 K)1)
0±0.005 m
0.60
0.48
2.30
)1.62
4.8 10)6
9.3
0.2 (*)
0.005±0.02 m
0.55
0.46
2.30
)1.62
9.8 10)6
9.3
0.6 (*)
0.02±0.1 m
0.50
0.40
2.30
)1.62
5.0 10)7 (*)
9.3
Eq. (8)
0.1±0.6 m
0.45
0.35
2.27
)1.70
7.5 10)8 (*)
10.6
Eq. (8)
0.6±1.35 m
0.39
0.30
2.13
)2.00
1.0 10)10 (*)
33.3
Eq. (8) w
wsat
1
h
hg
8
>
>
:
9
>
>
;
n
2
nÿ1
6 Like SiSPAT, the ALiBi model is a two-layered
energy balance model. Energy balance ¯uxes, as well as
photosynthesis, are calculated using detailed descrip-
tions of radiative transfer, turbulent transfer, water
transfer, stomatal conductance and leaf photosynthesis. 6 where wsat is the saturated water content (m3 m)3) and
hg (m) a scale factor. The hydraulic conductivity curve is
given by Brooks and Corey (1964): The vegetation conductance gs is given by the sum of
the leaf stomatal conductances over the whole canopy. 6.3 SiSPAT In particular, a maximum leaf
stomatal conductance of 15 mm s)1 was employed. The root zone water potential and the soil temperature
at a depth of 1 m were set to constant average values of
)0.5 MPa and 17 °C, respectively. Soil moisture in the 8 The dierent terms of the MUREX water budget for
1995, including soil moisture change and total runo,
are
simulated
by
ISBA,
MAGRET,
and
SiSPAT
(Table 2). Note that in the SiSPAT simulations the
water budget is not completely balanced (error of
15 mm) because of numerical cumulative inaccuracies
occurring during soil moisture saturation. The ALiBi
model presented below does not consider soil moisture
change. 6.3 SiSPAT As with ISBA and MAGRET, it was found that the
water excess Xs must be added to the water input in
order to simulate the water budget. In order to avoid the
water loss due to interception by the mulch and the
vegetation, the water excess was applied to the surface
soil layer (instead of directly increasing the precipita-
tion). The soil parameters employed in SiSPAT are
displayed in Table 3. Five horizons are distinguished,
from the surface to the total soil column depth (1.35 m). The employed suction curve, relating the volumetric
water content w to the matric potential h (expressed in
units of m) consists of the Van Genuchten (1980)
formula: In the SiSPAT model (Braud et al., 1995) the transfers
in the soil are described in more detail: the vertical
heterogeneity of the soil structure and texture is
accounted for, and a root distribution must be pre-
scribed. The heat and water exchanges are speci®ed
through Fourier and Richards' equations, and the soil
prognostic variables are vertical pro®les of temperature
and soil matric potential. Consistent with the accurate
description of the soil processes, the vegetation layer is
depicted more precisely than in ISBA and MAGRET:
the radiative transfer within the canopy is accounted for, Table 2. The MUREX 1995 water budget as calculated by ISBA, MAGRET, and SiSPAT
Model
Precipitation
(mm)
Water
excess (mm)
Evaporation
(mm)
Drainage
(mm)
Surface
runo (mm)
Soil moisure
change (mm)
Error (mm)
ISBA
773
184
856
95
0
6
0
MAGRET
957
±
884
125
0
)52
0
SiSPAT
773
184
858
1
101
)18
15 Table 2. The MUREX 1995 water budget as calculated by ISBA, MAGRET, and SiSPAT UREX 1995 water budget as calculated by ISBA, MAGRET, and SiSPAT er content (m3 m)3) a
ulic conductivity curve
964):
hydraulic conductiv
eld estimates of Ksat w
s performed using eith
uclin and Chopart, 199
ck and Reynolds, 199
and the deepest laye
ve SiSPAT's horizons, eit
SiSPAT model. They inclu
(VG) and Brooks and Cor
ively): n and hg are shape a
sat
m3 m)3)
n (VG model
48
2.30
46
2.30
40
2.30
35
2.27
30
2.13
surface ®eld experiment J.-C. Calvet et al.: MUREX: a land-surface ®eld experiment 849 g
9
>
>
;
n2
nÿ1
e saturated water conte
ctor. The hydraulic con
s and Corey (1964):
wsatc
the saturated hydrau
scale factor. 6.4 ALiBi The ALiBi model was proposed by Olioso (1992) to
relate evapotranspiration and photosynthesis to re¯ec-
tance and infrared brightness temperature measure-
ments. In order to derive evapotranspiration and
photosynthesis the ALiBi model has to be used in an
inverse mode (Olioso et al., 1996b): the modelled
thermal brightness temperature is ®tted to temperature
measurements by adjusting some input parameters
(usually related to water transfers through the plants). J.-C. Calvet et al.: MUREX: a land-surface ®eld experiment 850 ®rst 5 cm wg, which determines the soil evaporation in
ALiBi was either measured (DoY 125 to 143 and 270 to
298), or set to arbitrary values of 0.2 m3 m)3 in the
driest period (DoY 228 to 255) and 0.3 m3 m)3 else-
where. It was found that the value of wg has little
in¯uence on the ALiBi outputs in the case of MUREX. perfect simulations and values of E close to 0 corres-
pond to a basic representation of x by its constant
average value. Negative values of E show that the
model is not adequate. In general, the four models
perform realistic simulations of the surface ¯uxes. However, SiSPAT tends to underestimate evapotran-
spiration after the vegetation cutting (Fig. 10). The rms
error aecting the simulated G is rather large and the
skill score (not shown) displays negative values for all
the models, denoting an inconsistency of either the
measurements or the models. Such errors may be
explained by the in¯uence of the mulch: the explicit
simulations of the mulch eect by SiSPAT, in partic-
ular, shows that the presence of a mulch signi®cantly
modi®es the surface energy balance by decreasing soil
evaporation and soil temperature, and increasing soil
surface water content. Another problem caused by the
mulch is that in reality, the heat storage in the soil is
conditioned by the mulch whereas the measured G was
obtained from heat ¯ux plates buried at a 3-cm depth in
the soil. Although a correction was made to account for
the eect of the 3-cm soil top layer on G, the mulch
eect could not be properly characterised, and the soil
heat ¯ux at the surface of the mulch could not be
properly evaluated. It appears that the simulated G (at
the mulch surface) displays a much larger diurnal cycle
than the measured one. This aects the estimation of
the latent heat ¯ux, which was derived from the surface
energy budget. 7.1 Surface ¯uxes The values of the simulated instantaneous and cumu-
lated ¯uxes obtained with the structure parameters listed
in Tables 1 and 3 are compared with the observed values
in Table 4 and in Fig. 9, respectively. Figure 10 presents
the smoothed daily evapotranspiration over the 1995
annual cycle, either measured by the SAMER system or
modelled by ISBA, MAGRET, ALiBi, and SiSPAT. the
ALiBi simulations do not cover the entire annual cycle,
since they rely on TIR measurements, available in
summer and autumn 1995, only. The errors aecting
the simulated ¯uxes are listed in Table 4 in terms of rms
error, mean bias (simulated minus observed), and skill
score. The model skill score E is de®ned as: E 1 ÿ
P
i
xiOBS ÿ xiMOD
2
P
i
xiOBS ÿ xOBS
2
10 10 where x are half-hourly values of the ¯ux, either
measured or simulated (OBS and MOD subscripts,
respectively). A value of E of 100% corresponds to where x are half-hourly values of the ¯ux, either
measured or simulated (OBS and MOD subscripts,
respectively). A value of E of 100% corresponds to Table 4. Errors aecting the simulated half-hourly outputs of
ISBA, MAGRET, ALiBi, and SiSPAT in terms of rms error, mean
bias (simulated minus observed), and model skill score on:
evapotranspiration (LE), heat ¯ux (H), net radiation (Rn), soil heat ¯ux (G), surface temperature (Ts), and soil water storage
(w2 ´ d2). Note that ALiBi's simulation only covers some periods
and that ISBA's scores for the ALiBi's periods are shown in
parenthesis Table 4. 7 Model validation and comparison The model performances can be assessed by comparing
the model outputs with the measured values of surface
¯uxes, and for ISBA, MAGRET, and SiSPAT, the
surface temperature and the total soil moisture. 6.4 ALiBi On the other hand, the mean bias over G
is very small with SiSPAT (Table 4) and the cumulated
¯ux is simulated well (Fig. 9). The overestimation of Rn
with ISBA and MAGRET (Table 4) is consistent with
the overestimation of H and suggests that the modelled
surface temperature is underestimated (because more
energy is removed from the surface through the heat
¯ux in the model). The reverse conclusion can be drawn
from the results of ALiBi, which uses the observed bulk
temperature TIR to estimate the ¯uxes (both Rn and H
are underestimated), suggesting that the three models heat ¯ux (G), surface temperature (Ts), and soil water storage
(w2 ´ d2). Note that ALiBi's simulation only covers some periods
and that ISBA's scores for the ALiBi's periods are shown in
parenthesis 7.1 Surface ¯uxes 9)
is
well reproduced by ISBA, MAGRET, and SiSPAT,
with
a
rather
low
minimum
stomatal
resistance:
rsmin 50 s mÿ1. It must be noted that this value is
comparable with previous calibrations of ISBA over grasslands (Noilhan et al., 1993). It is interesting to
note that high evapotranspiration rates are also ob-
tained with ALiBi, based on surface temperature data
only, without describing the soil water budget and the
water excess. grasslands (Noilhan et al., 1993). It is interesting to
note that high evapotranspiration rates are also ob-
tained with ALiBi, based on surface temperature data
only, without describing the soil water budget and the
water excess. 7.1 Surface ¯uxes Errors aecting the simulated half-hourly outputs of
ISBA, MAGRET, ALiBi, and SiSPAT in terms of rms error, mean
bias (simulated minus observed), and model skill score on:
evapotranspiration (LE), heat ¯ux (H), net radiation (Rn), soil Model
LE
H
Rn
G
Ts
w2 ´ d2
ISBA
rms error
35 Wm)2
24 Wm)2
19 Wm)2
30 Wm)2
2.8 °C
17 mm
mean bias
4 Wm)2
11 Wm)2
13 Wm)2
)3 Wm)2
)2.2 °C
10 mm
skill score
91%
79%
99%
±
87%
90%
MAGRET
rms error
43 Wm)2
31 Wm)2
32 Wm)2
24 Wm)2
2.9 °C
21 mm
mean bias
6 Wm)2
8 Wm)2
10 Wm)2
)4 Wm)2
)1.6 °C
)1 mm
skill score
87%
66%
97%
±
87%
85%
ALiBi
rms error
33 Wm)2
31 Wm)2
23 Wm)2
31 Wm)2
±
±
(ISBA)
(38 Wm)2)
(24 Wm)2)
(20 Wm)2)
(33 Wm)2)
mean bias
9 Wm)2
)6 Wm)2
)6 Wm)2
)9 Wm)2
±
±
(6 Wm)2)
(10 Wm)2)
(14 Wm)2)
()2 Wm)2)
skill score
94%
73%
99%
±
±
±
(93%)
(84%)
(99%)
SiSPAT
rms error
41 Wm)2
23 Wm)2
25 Wm)2
33 Wm)2
3.9 °C
24 mm
mean bias
3 Wm)2
1 Wm)2
4 Wm)2
0 Wm)2
)0.4 °C
)15 mm
skill score
89%
79%
98%
±
75%
79% J.-C. Calvet et al.: MUREX: a land-surface ®eld experiment 851 have diculties in accounting for the mulch eect. As
far as the latent heat ¯ux is concerned, both instanta-
neous and cumulated ¯ux values are relatively well
simulated by the models (the model skill score is about
90% for all the models). After DoY 270 in 1995, all the
models
tend
to
overestimate
evapotranspiration
(Fig. 10). This may be caused, probably, by overesti-
mated values of LAI during this period (Fig. 6),
imposed on all the models. The high value of the
observed
cumulated
evapotranspiration
(Fig. 9)
is
well reproduced by ISBA, MAGRET, and SiSPAT,
with
a
rather
low
minimum
stomatal
resistance:
grasslands (Noilhan et al., 1993). It is interesting to
note that high evapotranspiration rates are also ob-
tained with ALiBi, based on surface temperature data
only, without describing the soil water budget and the
water excess. 7.2 Surface temperature
The errors on surface temperature are listed in Table 4. On average, the surface temperatures simulated by ISBA
and MAGRET tend to be lower than the measured Ts. Fig. 9. 7.1 Surface ¯uxes Comparison between
the simulated and measured
cumulated ¯uxes of water va-
pour (LE), heat (H), net radi-
ation (Rn), and heat storage in
the soil and the biomass (G)
over 1995. The thick solid line
represents the measured val-
ues. The simulations are rep-
resented by: ®ne solid line
(SiSPAT), and dashed lines
(®ne dashes for ISBA, and
thick dashes for MAGRET) Fig. 9. Comparison between
the simulated and measured
cumulated ¯uxes of water va-
pour (LE), heat (H), net radi-
ation (Rn), and heat storage in
the soil and the biomass (G)
over 1995. The thick solid line
represents the measured val-
ues. The simulations are rep-
resented by: ®ne solid line
(SiSPAT), and dashed lines
(®ne dashes for ISBA, and
thick dashes for MAGRET) Fig. 9. Comparison between
the simulated and measured
cumulated ¯uxes of water va-
pour (LE), heat (H), net radi-
ation (Rn), and heat storage in
the soil and the biomass (G)
over 1995. The thick solid line
represents the measured val-
ues. The simulations are rep-
resented by: ®ne solid line
(SiSPAT), and dashed lines
(®ne dashes for ISBA, and
thick dashes for MAGRET) have diculties in accounting for the mulch eect. As
far as the latent heat ¯ux is concerned, both instanta-
neous and cumulated ¯ux values are relatively well
simulated by the models (the model skill score is about
90% for all the models). After DoY 270 in 1995, all the
models
tend
to
overestimate
evapotranspiration
(Fig. 10). This may be caused, probably, by overesti-
mated values of LAI during this period (Fig. 6),
imposed on all the models. The high value of the
observed
cumulated
evapotranspiration
(Fig. 9)
is
well reproduced by ISBA, MAGRET, and SiSPAT,
with
a
rather
low
minimum
stomatal
resistance:
rsmin 50 s mÿ1. It must be noted that this value is
comparable with previous calibrations of ISBA over have diculties in accounting for the mulch eect. As
far as the latent heat ¯ux is concerned, both instanta-
neous and cumulated ¯ux values are relatively well
simulated by the models (the model skill score is about
90% for all the models). After DoY 270 in 1995, all the
models
tend
to
overestimate
evapotranspiration
(Fig. 10). This may be caused, probably, by overesti-
mated values of LAI during this period (Fig. 6),
imposed on all the models. The high value of the
observed
cumulated
evapotranspiration
(Fig. Fig. 9. Comparison between
the simulated and measured
cumulated ¯uxes of water va-
pour (LE), heat (H), net radi-
ation (Rn), and heat storage in
the soil and the biomass (G)
over 1995. The thick solid line
represents the measured val-
ues. The simulations are rep-
resented by: ®ne solid line
(SiSPAT), and dashed lines
(®ne dashes for ISBA, and
thick dashes for MAGRET) 7.2 Surface temperature The errors on surface temperature are listed in Table 4. On average, the surface temperatures simulated by ISBA
and MAGRET tend to be lower than the measured Ts. This is particularly true for ISBA, with a mean bias of
)2.2 °C. This is consistent with the bias between the Fig. 10. The daily evapotranspiration as measured and
modelled by ISBA, MAGRET, and ALiBi over 1995. The
annual cycle is smoothed using a 30-day moving average in
order to show the general trend Fig. 10. The daily evapotranspiration as measured and
modelled by ISBA, MAGRET, and ALiBi over 1995. The
annual cycle is smoothed using a 30-day moving average in
order to show the general trend Fig. 10. The daily evapotranspiration as measured and
modelled by ISBA, MAGRET, and ALiBi over 1995. The
annual cycle is smoothed using a 30-day moving average in
order to show the general trend Fig. 10. The daily evapotranspiration as measured and
modelled by ISBA, MAGRET, and ALiBi over 1995. The
annual cycle is smoothed using a 30-day moving average in
order to show the general trend Fig. 10. The daily evapotranspiration as measured and
modelled by ISBA, MAGRET, and ALiBi over 1995. The
annual cycle is smoothed using a 30-day moving average in
order to show the general trend J.-C. Calvet et al.: MUREX: a land-surface ®eld experiment 852 modelled and the observed Rn (Table 4). An explana-
tion is that ISBA and MAGRET, being single energy-
budget models, tend to simulate a surface temperature
which is representative of both vegetation and soil
surface. On the other hand, the mean bias on Ts given by
SiSPAT
is
low,
consistent,
again,
with
the
good
representation of the cumulated ¯uxes by this model
(Fig. 9). However, the SiSPAT skill score concerning Ts
is lower than the other models (75%, against 87% for
ISBA and MAGRET). This is due to a systematic
overestimation of daytime Ts and underestimation of
nighttime Ts by SiSPAT. impossible to account for texture gradients in simple
models like ISBA and MAGRET. In spite of these
diculties, ISBA seems to be the best model of soil
moisture for this dataset, with a skill score of 90%. One
factor that may explain the poor performance of
SiSPAT (skill score of 79%) is the large amount
of surface runo generated by the model, at the expense
of drainage (Table 2). 7.2 Surface temperature As far as surface soil moisture is concerned, a
comparison of the surface soil moisture wg simulated
by ISBA with the measured values was carried out by
Calvet et al. (1998a). It was shown that the ISBA wg
outputs compare well with the measurements corre-
sponding to a value integrated over the soil top 5-cm
layer. Using SiSPAT, the observed high values of wg are
better reproduced by explicitly modelling the transfers
within the mulch (Gonzalez-Sosa et al., 1999). 7.3 Soil moisture Figure 11 presents the values of the total soil water
content simulated by ISBA, MAGRET, and SiSPAT. The agreement is relatively good, with a rms error of 17
and 21 mm for ISBA and MAGRET, respectively, and
24 mm for SiSPAT (Table 4). As shown in Fig. 11, the
soil moisture simulated by ISBA may be higher than
®eld capacity in wintertime. Although this is physically
possible (soil moisture at saturation is higher than the
®eld capacity equilibrium value), it condradicts the
observations and may denote a problem in the para-
metrisation of in®ltration and drainage. MAGRET and
SiSPAT present lower soil moisture values during
autumn. This latter feature corresponds to the negative
soil water change obtained with these models in Table 2. The problem with the simulation of the root soil
moisture in ISBA (or MAGRET) is related to the
diculty of de®ning the root zone and the absence of a
description of the amount of water either directly
extracted by the roots or provided by the capillarity
rises (Boone et al., 1999). The soil water stress function
employed in the expression of stomatal resistance may
also dier from reality in some cases. Finally, it is 8 Summary and conclusions constrained by observations (LAI, and the radiative and
atmospheric forcing being the main factors). However,
estimating the surface temperature and the water
partitioning between vegetation transpiration, drainage,
surface runo, and soil moisture change is not easy, as
shown in Sect. 7. It is expected that the MUREX data
will be useful to test interactive vegetation SVAT
models, in which photosynthesis is computed and may
be employed to estimate the biomass production and
LAI (Calvet et al., 1998b). The MUREX continuous micrometeorological and soil
moisture observations of 1995, 1996, and 1997 are
presented. The MUREX experiment provided a unique
dataset comprising continuous atmospheric, vegetation
and soil measurements over more than three years, over
a fallow site. The observations showed that a dierent
amount of water was extracted from the soil reservoir
from one year to another, depending on the rainfall
regime. Most often than not, the soil water supply (from
either soil water storage or uphill water input) was large
enough to permit values of evapotranspiration higher
than the sensible heat ¯ux. In spite of the buer eect of
the soil, the growing cycle of the fallow studied showed a
clear dependence on the rainfall regime: the maximum
LAI varied between 2 and 3.5 according to the year
considered. Data analysis and model results show that
transfers of heat and water were in¯uenced by the
presence of a mulch in the vegetation layer. This led to
diculties especially in the estimation of the ground
heat ¯ux and consequently on the latent heat ¯ux,
derived as the residual of the energy budget. The
existence of lateral transfers associated with a perched
water table was also shown. However, the measurement
uncertainties make their exact quanti®cation dicult. (
)
Primarily, the aim of MUREX is to provide contin-
uous micrometeorological and soil moisture observa-
tions during several years, in order to assess the ability
of simple surface schemes employed in meteorology to
simulate properly the energy and water budgets of the
surface. Such a dataset could be useful to complete the
on-going intercomparison PILPS program (Chen et al.,
1997). In particular, the PILPS program has shown that
dierent parametrisations of soil water and runo can
cause a very large variability in the simulated root-zone
soil moisture. Also, the functional relationship between
the parametrisation of the root-zone soil moisture and
the evapotranspiration vary from one model to another
(Mahfouf et al., 1996). 8 Summary and conclusions It is likely that such a dataset will also be employed to
analyse the model shortcomings and to propose new
parametrisations. Moreover, the MUREX dataset may
be useful to test assimilation techniques. For example,
Calvet et al. (1998a), using the measurements of surface
soil moisture and surface temperature, have proposed
assimilation rules of these variables to retrieve the root-
zone soil water content. q
The data of 1995 were employed to assess the ability
of the surface schemes ISBA, MAGRET, SiSPAT, and
ALiBi to simulate properly the energy and water
budgets of a vegetated surface over a complete annual
cycle. The surface schemes describing soil moisture and
the water balance (ISBA, MAGRET, and SiSPAT) were
modi®ed to include the contribution of lateral transfers
to the water input. SiSPAT was also modi®ed to
explicitely model heat and water transfers within the
mulch. Although this allowed a signi®cant improvement
of the model results, further studies are needed to better
assess the in¯uence of mulch parameters on the results. Data from 1996 and 1997 will be used to re®ne the
mulch parameters and the occurrence of capillarity rises
and lateral in¯ows. Acknowledgements. This work was funded by the Conseil Re gio-
nal de Midi-Pyre ne es, the Programme National de Recherche en
Hydrologie, and by participants to the ®eld programme (Me te o-
France/CNRM, CESBIO, LTHE). Acknowledgements. This work was funded by the Conseil Re gio-
nal de Midi-Pyre ne es, the Programme National de Recherche en
Hydrologie, and by participants to the ®eld programme (Me te o-
France/CNRM, CESBIO, LTHE). Topical Editor J.-P. Duvel thanks Yongkang Xue and another
referee for their help in evaluating this paper. 7.4 Net assimilation of carbon dioxide Calvet et al.: MUREX: a land-surface ®eld experiment 7.4 Net assimilation of carbon dioxide As CO2 net assimilation by the canopy (An) was not
measured in 1995, only modelled annual cycles are
presented in Fig. 12. Values of An are provided by
ALiBi, and also by the new version of ISBA computing
photosynthesis (Calvet et al., 1998b): ISBA-Ags. The
curve displayed in Fig. 12 concerning ISBA-Ags was
obtained by imposing the same LAI values as for the
other models. Namely, ISBA-Ags was run in the
con®guration where photosynthesis and stomatal con-
ductance are calculated as in Jacobs et al. (1996),
without inferring plant growth. Both models give a
realistic magnitude of net assimilation (based on the
CO2-¯ux measurements of 1997 and 1998 presented in
Sect. 4) and the general trend is the same. However, the
An values given by ALiBi are 15% larger than the ISBA-
Ags ones, on average. Fig. 11. The measured (pluses) and simulated (solid and dashed lines)
soil water content of the 1.35 m soil column in 1995. The
measurements dier slightly from those presented in Calvet et al. (1998a) because the neutron probe output calibration was re®ned, and
all the measured sites are presented Fig. 11. The measured (pluses) and simulated (solid and dashed lines)
soil water content of the 1.35 m soil column in 1995. The
measurements dier slightly from those presented in Calvet et al. (1998a) because the neutron probe output calibration was re®ned, and
all the measured sites are presented
Fig. 12. The daily net CO2 assimilation as modelled by ISBA-Ags
(Calvet et al., 1998b) and ALiBi in 1995. The annual cycle is
smoothed using a 30-day moving average in order to show the general
trend Fig. 12. The daily net CO2 assimilation as modelled by ISBA-Ags
(Calvet et al., 1998b) and ALiBi in 1995. The annual cycle is
smoothed using a 30-day moving average in order to show the general
trend Fig. 11. The measured (pluses) and simulated (solid and dashed lines)
soil water content of the 1.35 m soil column in 1995. The
measurements dier slightly from those presented in Calvet et al. (1998a) because the neutron probe output calibration was re®ned, and
all the measured sites are presented Fig. 12. The daily net CO2 assimilation as modelled by ISBA-Ags
(Calvet et al., 1998b) and ALiBi in 1995. The annual cycle is
smoothed using a 30-day moving average in order to show the general
trend 853 J.-C. Acknowledgements. This work was funded by the Conseil Re gio-
nal de Midi-Pyre ne es, the Programme National de Recherche en
Hydrologie, and by participants to the ®eld programme (Me te o-
France/CNRM, CESBIO, LTHE). References Clim., 8(8), 2039±2057, 1995. Brutsaert,
W.,
Evaporation
into
the
atmosphere,
D. Reidel,
Dordrecht, 299 pp, 1982. Calvet, J.-C., J. Noilhan, and P. Bessemoulin, Retrieving the root-
zone soil moisture from surface soil moisture or temperature
estimates: A feasibility study based on ®eld measurements. J. Appl. Meteorol., 37(4), 371±386, 1998a. Mahfouf, J.-F., C. Ciret, A. Ducharne, P. Irannejad, J. Noilhan,
et al., Analysis of transpiration results from the RICE and
PILPS workshop. Global Planet. Change, 13, 73±88, 1996. Noilhan, J., and S. Planton, A simple parametrisation of land
surface processes for meteorological models, Mon. Weather
Rev., 117, 536±549, 1989. Calvet, J.-C., J. Noilhan, J.-L. Roujean, P. Bessemoulin, M. Cabelguenne, et al., An interactive vegetation SVAT model
tested against data from six contrasting sites, Agric. For. Meteorol., 92, 73±95, 1998b. Noilhan, J., and J.-F., Mahfouf, The ISBA land surface parametri-
sation scheme. Global Planet. Change, 13, 145±159, 1996. Chen, T. H., A. Henderson-Sellers, P. C. D. Milly, A. J. Pitman, A. C. M. Beljaars, et al., Cabauw experimental results from the
project for intercomparison of landsurface parametrisation
schemes (PILPS), J. Clim. 10, 1194±1215, 1997. Noilhan, J., J.-F. Mahfouf, A. Manzi, and S. Planton, Validation of
land-surface parametrisations: developments and experiments
at the French weather service, in Validation of models over
Europe (2), ECMWF seminar proceedings, Reading [7±11
September 1992] pp. 125±158, 1993. Courault, D., J.-P. Lagouarde, and B. Aloui, Evaporation for
maritime catchment combining a meteorological model with
vegetation information and airborne surface temperatures,
Agric. For. Meteorol., 82, 93±117, 1996. Olioso, A., Simulation des e changes d'e nergie et de masse d'un
couvert ve ge tal, dans le but de relier la transpiration et la
photosyntheÁ se aux mesures de re ¯ectance et de tempe rature de
surface. PhD Thesis, University of Montpellier, France, 1992. g
Daget, P., and J. Poissonet, Une me thode d'analyse phytologique
des prairies. CriteÁ res d'application. Ann. Agronom., 22, 5±41,
1971. Olioso, A., T. Carlson, and N. Brisson, Simulation of diurnal
transpiration and photosynthesis of a water stressed soybean
crop, Agric. For. Meteorol., 81, 41±59, 1996a. Deardor, J. W., A parametrisation of the ground surface moisture
content for use in atmosphere predictions models, J. Appl. Meteorol., 16, 1182±1185, 1977. Olioso, A., O. Taconet, and M. Ben Mehrez, Estimation of heat and
mass ¯uxes from IR brightness temperature, IEEE Trans. Geosci. Remote Sensing, 34, 1184±1190, 1996b. Deardor, J. References All the models are able to reproduce the general trend
of the measured variables. However, the surface tem-
perature is signi®cantly underestimated by ISBA and
MAGRET. The CO2 net assimilation simulated by
ALiBi is also signi®cantly dierent from the ISBA-Ags
results. It must be noted that any SVAT o-line
simulation requires prescribed boundary conditions. Together with the observed atmospheric forcing and
vegetation characteristics, such as LAI, the studied
models (apart from ALiBi) use estimates of water excess
Xs (Sect. 3.3) and deep soil temperature, obtained from
the ®eld measurements. Soil water excess, which results
from lateral transfers, cannot be simulated by 1D
SVATs and must be prescribed from observations. These ground boundary conditions are far from being
major components controlling the seasonal evolution
but need to be accounted for in order to simulate the
observed energy and water budgets. Of course, a part of
the model interseasonal to inter-annual variability is Andre , J.-C., J.-P. Goutorbe, and A. Perrier, HAPEX-MOBILHY:
a hydrologic atmospheric experiment for the study of water
budget and evaporation ¯ux at the climate scale, Bull. Am. Meteorol. Soc., 67(2), 138±144, 1986. Bessemoulin, P., J.-C. Calvet, G. Jaubert, J. Noilhan, D. Vignes, C. Tosca, I. Braud, R. Haverkamp, J.-L. Thony, and M. Vauclin,
MUREX : Un programme de suivi du cycle annuel des e changes
de masse et d'e nergie entre sol, ve ge tation, et atmospheÁ re. Premiers enseignements, in Atelier ExpeÂrimentation et Instru-
mentation, Me te o-France/CNRM, October 1996, Toulouse, pp. 289±294, 1996. Bolle, H. J., J.-C. Andre , J.-L. Arrue, H. K. Barth, P. Bessemoulin,
et al., EFEDA: European ®eld experiment in a deserti®cation-
threatened area, Ann. Geophysicae, 11, 173±189, 1993. Boone, A., J.-C. Calvet, and J. Noilhan, Inclusion of a third soil
layer in a land-surface scheme using the force-restore method,
J. Appl. Meteorol., in press, 1999. Braud, I., A. C. Dantas-Antonio, M. Vauclin, J.-L. Thony, and P. Ruelle, A simple soil-plant-atmosphere transfer model (Si-
SPAT): development and ®eld veri®cation, J. Hydrol., 166,
213±250, 1995. 854 J.-C. Calvet et al.: MUREX: a land-surface ®eld experiment Brooks, R. H., and A. T. Corey, Hydraulic properties of porous
media, Colo. State Univ. Fort Collins Hydrol. Pap., 3, 1964. Mahfouf, J.-F., A. O. Manzi, J. Noilhan, H. Giordani, and M. De que , The land surface scheme ISBA within the Me te o-France
climate model ARPEGE. Part I: implementation and prelimi-
nary results, J. References W., Ecient prediction of ground temperature and
moisture with inclusion of a layer of vegetation, J. Geophys. Res., 83, 1889±1903, 1978. Sellers, P., F. G. Hall, G. Asrar, D. E. Strebel, and R. E. Murphy,
An overview of the First International Satellite Land Surface
Climatology Project (ISLSCP) Field Experiment (FIFE), J. Geophys. Res., 97(D17), 18 345±18 371, 1992. Elrick, D. E., and W. D. Reynolds, Methods for analysing constant-
head well permeameter data, Soil Sci. Soc. Am. J., 56(1), 320±
323, 1992. Shuttleworth, W. J., and J. S. Wallace, Evaporation from space
crops ± an energy combination theory, Q. J. R. Meteorol. Soc.,
111, 839±855, 1985. Goutorbe, J.-P., A critical assessment of the Samer network
accuracy, In Eds T. J. Schmugge, and J. -C. Andre Land surface
evaporation. Measurement and parametrisation, Springer-Verlag,
Berlin Heidelberg, New York, pp. 171±182, 1991. Staley, D. O., and G. M. Jurica, Eective atmospheric emissivity
under clear skies, J. Appl. Meteorol., 11, 349±356, 1972. Goutorbe, J.-P., T. Lebel, A. J. Dolman, J. H. C. Gash, P. Kabat,
et al., An overview of HAPEX-Sahel: a study in climate and
deserti®cation, J. Hydrol., 188±189, 4±17, 1997. Trou¯eau, D., J. P. Lhomme, B. Monteny, and A. Vidal, Sensible
heat ¯ux and radiometric surface temperature over sparse
Sahelian vegetation. I. An experimental analysis of the kB)1
parameter, J. Hydrol., 188±189, 815±838, 1997. Jacobs, C. M. J., B. J. J. M. van den Hurk, and H. A. R. de Bruin,
Stomatal behaviour and photosynthetic rate of unstressed
grapevines in semi-arid conditions, Agric. For. Meteorol., 80,
111±134, 1996. Vachaud, G., A. Passerat de Silans, P. Balabanis, and M. Vauclin,
Temporal stability of spatially measured soil water probability
density function, Soil Sci. Soc. Am. J., 49(4), 822±828, 1985. Jacquart, C., and E. Choisnel, Un modeÁ le de bilan hydrique
simpli®e aÁ deux re servoirs utilisable en agrome te orologie, La
MeÂteÂorol., 8(9), 29±43, 1995. Van Genuchten, M. T., A closed-form equation for predicting the
hydraulic conductivity of unsaturated soils, Soil Sci. Soc. Am. J., 44, 892±899, 1980. Koster, R. D., and P. C. D. Milly, The interplay between
transpiration and runo formulations in land surface schemes
used with atmospheric models, J. Clim., 10(7), 1578±1591,
1997. Vauclin, M., and J.-L. Chopart, L'in®ltration multi-disques pour la
de termination in situ des caracte ristiques hydrodynamics de la
surface d'un sol gravillonaire de la Coà te d'Ivoire, Agron. Trop.,
46, 259±271, 1992. References Lagouarde, J.-P., Use of NOAA-AVHRR data combined with an
agrometeorological model for evaporation mapping, Int. J. Remote Sens., 12(9), 1853±1864, 1991. Verhoef, A., H. A. R. de Bruin, and B. J. J. M. Van den Hurk, Some
practical notes on the parameter kB)1 for sparse vegetation,
J. Appl. Meteorol., 36(5), 560±572, 1997. Winkel, T., and S. Rambal, Stomatal conductance of some
grapevines growing in the ®eld under a mediterranean environ-
ment, Agric. For. Meteorol., 51, 107±122, 1990. Laurent, J.-P., Evaluation des parameÁ tres thermiques d'un milieu
poreux: optimisation d'outils de mesure `in situ', Int. J. Heat
Mass Trans., 32(7), 1247±1259, 1989.
|
https://openalex.org/W4283642144
|
https://zenodo.org/record/6585368/files/V2I5-202.pdf
|
Zulu
| null |
O'ZBEKISTON RESPUBLIKASIDA FUQAROLIK
|
Zenodo (CERN European Organization for Nuclear Research)
| 2,022
|
cc-by
| 2,156
|
U U Ber
MAQOLA TARIXI
Qabul qilindi: 10- may 2022
Ma’qullandi: 14- may 2022
Chop etildi: 26- may 2022
KALIT SO’ZLAR
O’zbekiston
Respublikasining
fuqaroligiga qabul qilish va
uni
tiklash,
O’zbekiston
Respublikasining
fuqaroligidan
chiqish,O’zbekiston
Respublikasining
fuqaroligini
yo’qotganlik,
Konstitutsiyasiga rioya
etish majburiyatini. Ber
MAQOLA TARIXI
Qabul qilindi: 10- may 2022
Ma’qullandi: 14- may 2022
Chop etildi: 26- may 2022
KALIT SO’ZLAR
O’zbekiston
Respublikasining
fuqaroligiga qabul qilish va
uni
tiklash,
O’zbekiston
Respublikasining
fuqaroligidan
chiqish,O’zbekiston
Respublikasining
fuqaroligini
yo’qotganlik,
Konstitutsiyasiga rioya
etish majburiyatini. Ber
MAQOLA TARIXI
Qabul qilindi: 10- may 2022
Ma’qullandi: 14- may 2022
Chop etildi: 26- may 2022
KALIT SO’ZLAR
O’zbekiston
Respublikasining
fuqaroligiga qabul qilish va
uni
tiklash,
O’zbekiston
Respublikasining
fuqaroligidan
chiqish,O’zbekiston
Respublikasining
fuqaroligini
yo’qotganlik,
Konstitutsiyasiga rioya
etish majburiyatini. Ushbu
maqolada
Ozbekiston
Respublikasida
fuqarolikning tan olinishi, yo’qotilishi hamda ularning
tartibi haqida so’z yuritiladi. O’zbekiston Respublikasi
Konstitutsiyasining
VI
bobida
fuqarolik
haqida
tushuntirishlar va ma’lumotlar keltirilgan. Konstitutsiyada
O’zbekiston Respublikasining butun hududida yagona
fuqarolik o’rnatilishi, fuqarolikka ega bo’lish va uni yo’qotish
asoslari hamda tartibi qonun bilan belgilanishi, o’z
fuqarolarini davlat tashqarisida ham huquqiy himoya qilishi
va ularga homiylik qilishi ko’rsatishni kafolatlashi, chet el
fuqarolarining va fuqaroligi bo’lmagan shaxslarning huquq
normalariga muvofiq taminlashi haqida qoidalar ko’rsatilib
o’tilgan. KALIT SO ZLAR
O’zbekiston
Respublikasining
fuqaroligiga qabul qilish va
uni
tiklash,
O’zbekiston
Respublikasining
fuqaroligidan
chiqish,O’zbekiston
Respublikasining
fuqaroligini
yo’qotganlik,
Konstitutsiyasiga rioya
etish majburiyatini. O’zbekiston
Respublikasining
fuqaroligiga qabul qilish va
uni
tiklash,
O’zbekiston
Respublikasining
fuqaroligidan
chiqish,O’zbekiston
Respublikasining
fuqaroligini
yo’qotganlik,
Konstitutsiyasiga rioya
etish majburiyatini. Volume 2 Issue 5, May 2022 ISSN 2181-2853
Page 202
O’zbekiston Respublikasi fuqarolari jinsi,
irqi, millati, tili, dini, ijtimoiy kelib chiqishi,
e’tiqodi, shaxsiy va ijtimoiy mavqeyidan
qat’i
nazar,
qonun
oldida
tengdir. O’zbekiston Respublikasi o’z organlari va
mansabdor shaxslari orqali O’zbekiston
Respublikasi fuqarolari oldida mas’uldir,
O’zbekiston Respublikasining fuqarosi esa
davlat
oldida
mas’uldir. O’zbekiston
Respublikasi
fuqaroligiga qabul qilish,
O’zbekiston
Respublikasi
fuqaroligini
tiklash,
O’zbekiston
Respublikasi
fuqaroligidan chiqish yoki O’zbekiston
Respublikasi
fuqaroligini
yo’qotish
to’g’risidagi
qarorlar
O’zbekiston
Respublikasi
Prezidenti
tomonidan
farmonlar
tarzida
qabul
qilinadi
va
imzolangan kundan e’tiboran kuchga kiradi. 2005-yil 1-yanvarga qadar O’zbekiston
Respublikasida
doimiy
yashash
joyi
bo’yicha
ro’yxatga
olingan
hamda
muqaddam
chet
davlat
fuqaroligida
bo’lmagan shaxs xohish bildirgan taqdirda,
O’zbekiston Respublikasining fuqarosi deb
tan olinadi. O’zini
O’zbekiston
Respublikasining
fuqarosi deb tan olish to’g’risidagi ariza
bilan
murojaat
qilgan
kunga
qadar
O’zbekiston Respublikasida kamida o’n
besh yil doimiy yashab kelayotgan va Respublikasi
Prezidenti
tomonidan
farmonlar
tarzida
qabul
qilinadi
va
imzolangan kundan e’tiboran kuchga kiradi. U 2005-yil 1-yanvarga qadar O’zbekiston
Respublikasida
doimiy
yashash
joyi
bo’yicha
ro’yxatga
olingan
hamda
muqaddam
chet
davlat
fuqaroligida
bo’lmagan shaxs xohish bildirgan taqdirda,
O’zbekiston Respublikasining fuqarosi deb
tan olinadi. Respublikasi
Prezidenti
tomonidan
farmonlar
tarzida
qabul
qilinadi
va
imzolangan kundan e’tiboran kuchga kiradi. 2005-yil 1-yanvarga qadar O’zbekiston
Respublikasida
doimiy
yashash
joyi
bo’yicha
ro’yxatga
olingan
hamda
muqaddam
chet
davlat
fuqaroligida
bo’lmagan shaxs xohish bildirgan taqdirda,
O’zbekiston Respublikasining fuqarosi deb
tan olinadi. Respublikasi
Prezidenti
tomonidan
farmonlar
tarzida
qabul
qilinadi
va
imzolangan kundan e’tiboran kuchga kiradi. 2005-yil 1-yanvarga qadar O’zbekiston
Respublikasida
doimiy
yashash
joyi
bo’yicha
ro’yxatga
olingan
hamda
muqaddam
chet
davlat
fuqaroligida
bo’lmagan shaxs xohish bildirgan taqdirda,
O’zbekiston Respublikasining fuqarosi deb
tan olinadi. V l
2 I
5 M
2022
ISSN 2
O’zbekiston Respublikasi fuqarolari jinsi,
irqi, millati, tili, dini, ijtimoiy kelib chiqishi,
e’tiqodi, shaxsiy va ijtimoiy mavqeyidan
qat’i
nazar,
qonun
oldida
tengdir. O’zbekiston Respublikasi o’z organlari va
mansabdor shaxslari orqali O’zbekiston
Respublikasi fuqarolari oldida mas’uldir,
O’zbekiston Respublikasining fuqarosi esa
davlat
oldida
mas’uldir. O’zbekiston
Respublikasi
fuqaroligiga qabul qilish,
O’zbekiston
Respublikasi
fuqaroligini
tiklash,
O’zbekiston
Respublikasi
fuqaroligidan chiqish yoki O’zbekiston
Respublikasi
fuqaroligini
yo’qotish
to’g’risidagi
qarorlar
O’zbekiston O’zini
O’zbekiston
Respublikasining
fuqarosi deb tan olish to’g’risidagi ariza
bilan
murojaat
qilgan
kunga
qadar
O’zbekiston Respublikasida kamida o’n
besh yil doimiy yashab kelayotgan va ISSN 2181-2853 U EURASIAN JOURNAL OF LAW, FINANCE AND
APPLIED SCIENCES Innovative Academy Research Support Center
UIF = 8.3 | SJIF = 5.961
www.in-a ikkinchisining fuqaroligi o’zgarishiga sabab
bo’lmaydi. mazkur
davr mobaynida
chet
davlat
fuqaroligida bo’lmagan shaxs O’zbekiston
Respublikasining fuqarosi deb tan olinadi. Nikohning bekor qilinishi ushbu nikohda
tug’ilgan yoki nikoh bekor qilinguniga qadar
farzandlikka olingan bolaning fuqaroligi
o’zgarishiga sabab bo’lmaydi. O’zbekiston
Respublikasi o’z diplomatik vakolatxonalari
va
konsullik
muassasalari
orqali
O’zbekiston Respublikasi fuqarolari chet
davlatning
qonunchiligida,
O’zbekiston
Respublikasi va chet davlat ishtirokchi
bo’lgan xalqaro shartnomalarda o’ziga
berilgan
barcha
huquqlardan,
xalqaro
huquqning
umum
tan
olingan
normalaridan,
xalqaro
urf-odatlardan
to’laqonli foydalanish imkoniyatiga ega
bo’lishi uchun choralar ko’rishi shart. O’zbekiston
Respublikasining
konsullik
muassasalari,
shuningdek
O’zbekiston
Respublikasi
diplomatik
vakolatxonalarining konsullik bo’limlari
ham
O’zbekiston
Respublikasi
fuqarolarining huquqlarini himoya qilish va
ularning
buzilgan
huquqlarini
tiklash
bo’yicha choralar ko’radi. Ota-onasidan biri (yolg’iz otasi yoki onasi)
O’zbekiston Respublikasining fuqarosi deb
tan olingan, chet davlat fuqaroligini qabul
qilmagan va O’zbekiston Respublikasida
yashab kelayotgan bola xohish bildirgan
taqdirda,
O’zbekiston
Respublikasining
fuqarosi deb tan olinadi. O’zbekiston
Respublikasi
fuqarosining
pasporti, identifikatsiyalovchi kartasi yoki
O’zbekiston
Respublikasining
fuqaroligi
mavjud ekanligini ko’rsatuvchi boshqa
hujjat
O’zbekiston
Respublikasining
fuqaroligini tasdiqlovchi hujjatdir. Quyidagilar o’n olti yoshga to’lmagan
bolaning
O’zbekiston
Respublikasi
fuqaroligini tasdiqlovchi hujjatdir: Quyidagilar o’n olti yoshga to’lmagan
bolaning
O’zbekiston
Respublikasi
fuqaroligini tasdiqlovchi hujjatdir: tug’ilganlik to’g’risidagi guvohnoma yoki
ma’lumotnoma (bolaning tug’ilgan joyidan
qat’i nazar) — ota-onasidan birining
O’zbekiston Respublikasi fuqaroligi mavjud
ekanligini
ko’rsatuvchi
pasporti,
identifikatsiyalovchi kartasi yoki boshqa
hujjati ko’rsatilganda; O’zbekiston
Respublikasining
fuqaroligi
quyidagi hollarda olinadi: O’zbekiston
Respublikasining
fuqaroligi
quyidagi hollarda olinadi: tug’ilganlik to’g’risidagi guvohnoma yoki
ma’lumotnoma
(bola
O’zbekiston
Respublikasida
tug’ilgan
taqdirda)
—
fuqaroligi bo’lmagan shaxslar bo’lgan ota-
onaning (yolg’iz otaning yoki onaning)
O’zbekiston
Respublikasida
yashash
guvohnomasi ko’rsatilganda. -
tug’ilganlik bo’yicha; -
bola farzandlikka olinganda; -
O’zbekiston
Respublikasining
fuqaroligiga qabul qilish va uni tiklash
natijasida. -
O’zbekiston
Respublikasining
fuqaroligiga qabul qilish va uni tiklash
natijasida. O’zbekiston EURASIAN JOURNAL OF LAW, FINANCE AND
APPLIED SCIENCES
Innovative Academy Research Support Center
UIF = 8.3 | SJIF = 5.961
www.in-academy.uz Innovative Academy Research Support Center
UIF = 8.3 | SJIF = 5.961
www -
b)
uning
ota-onasidan
biri
O’zbekiston Respublikasi fuqarosi, boshqasi
esa fuqaroligi bo’lmagan shaxs bo’lsa yoki
bedarak yo’qolgan deb topilgan yoxud
uning turgan joyi noma’lum bo’lsa; -
b)
uning
ota-onasidan
biri
O’zbekiston Respublikasi fuqarosi, boshqasi
esa fuqaroligi bo’lmagan shaxs bo’lsa yoki
bedarak yo’qolgan deb topilgan yoxud
uning turgan joyi noma’lum bo’lsa; -
v)
uning
ota-onasidan
biri
O’zbekiston Respublikasi fuqarosi, boshqasi
esa chet davlat fuqarosi bo’lsa, tug’ilganlik
bo’yicha
O’zbekiston
Respublikasi
fuqaroligini oladi. b) tirikchilikning qonuniy manbaiga ega
bo’lsa; b) tirikchilikning qonuniy manbaiga ega
bo’lsa; v)
O’zbekiston
Respublikasining
Konstitutsiyasiga rioya etish majburiyatini
o’z zimmasiga olsa; O’zbekiston Respublikasining fuqaroligiga
qabul
qilish:
umumiy
tartibda,
soddalashtirilgan tartibda, alohida tartibda
amalga
oshiriladi. O’zbekiston
Respublikasining
fuqaroligiga
qabul
qilishning umumiy tartibi chet davlat
fuqarosiga
yoki
fuqaroligi
bo’lmagan
shaxsga nisbatan quyidagi hollarda amalga
oshiriladi, agar u: g) davlat tilini qonunchilikda belgilangan
tartibda
muloqot
qilish
uchun
zarur
darajada bilsa. g) davlat tilini qonunchilikda belgilangan
tartibda
muloqot
qilish
uchun
zarur
darajada bilsa. O’zbekiston Respublikasi Prezidenti milliy
manfaatlardan
kelib
chiqib,
chet
davlatlarning fuqarolariga va fuqaroligi
bo’lmagan
shaxslarga
“O’zbekiston
Respublikasining
fuqaroligi
to’g’risida”
qonunning 19 va 20-moddalari talablari
qo’llanilmagan
holda
O’zbekiston
Respublikasining
fuqaroligini
berishi
mumkin. O’zbekiston Respublikasi Prezidenti milliy
manfaatlardan
kelib
chiqib,
chet
davlatlarning fuqarolariga va fuqaroligi
bo’lmagan
shaxslarga
“O’zbekiston
Respublikasining
fuqaroligi
to’g’risida”
qonunning 19 va 20-moddalari talablari
qo’llanilmagan
holda
O’zbekiston
Respublikasining
fuqaroligini
berishi
mumkin. a) chet davlat fuqaroligidan chiqishni
rasmiylashtirgan bo’lsa; b) fuqaroligi bo’lmagan shaxsga yashash
guvohnomasi olingan kundan e’tiboran va
O’zbekiston Respublikasining fuqaroligiga
qabul qilish to’g’risidagi iltimosnoma bilan
murojaat qilgan kunga qadar O’zbekiston
Respublikasi hududida uzluksiz besh yil
davomida doimiy yashab kelayotgan bo’lsa;
v) tirikchilikning qonuniy manbaiga ega
bo’lsa; O’zbekiston
Respublikasining
fuqaroligi
quyidagilar oqibatida tugatiladi: O’zbekiston
Respublikasining
fuqaroligi
quyidagilar oqibatida tugatiladi: a)
O’zbekiston
Respublikasining
fuqaroligidan chiqish; b)
O’zbekiston
Respublikasining
fuqaroligini yo’qotganlik. O’zbekiston fuqaroligi
O’zbekiston
Respublikasining
xalqaro shartnomalarida va ushbu Qonunda
nazarda tutilgan boshqa asoslarga ko’ra
ham olinishi mumkin. O’zbekiston Respublikasi fuqarosining chet
davlat fuqarosi yoki fuqaroligi bo’lmagan
shaxs bilan nikohdan o’tishi, shuningdek
bunday nikohning bekor qilinishi er-
xotinning fuqaroligi o’zgarishiga sabab
bo’lmaydi. Bola o’zining tug’ilgan joyidan qat’i nazar,
agar tug’ilgan paytida: Er-xotindan
birining
O’zbekiston
Respublikasi
fuqaroligi
o’zgarishi -
a) ota-onasi (yolg’iz otasi yoki onasi)
O’zbekiston Respublikasi fuqarolari bo’lsa; Er-xotindan
birining
O’zbekiston
Respublikasi
fuqaroligi
o’zgarishi Er-xotindan
birining
O’zbekiston
Respublikasi
fuqaroligi
o’zgarishi -
a) ota-onasi (yolg’iz otasi yoki onasi)
O’zbekiston Respublikasi fuqarolari bo’lsa; ISSN 2181-2853 Volume 2 Issue 5, May 2022 Page 203 U EURASIAN JOURNAL OF LAW, FINANCE AND
APPLIED SCIENCES EURASIAN JOURNAL OF LAW, FINANCE AND
APPLIED SCIENCES EURASIAN JOURNAL OF LAW, FINANCE AND
APPLIED SCIENCES
Innovative Academy Research Support Center Innovative Academy Research Support Center
UIF = 8.3 | SJIF = 5.961
www.in-academy O’zbekiston Respublikasi Prezidentining
O’zbekiston
Respublikasining
fuqaroligi
masalalarini
hal
etishga
doir
konstitutsiyaviy
vakolatlarini
amalga
oshirish maqsadida besh yil muddatga
O’zbekiston
Respublikasi
Prezidenti
huzuridagi
Fuqarolik
masalalari
komissiyasi tuziladi. Komissiya
O’zbekiston
Respublikasi
Prezidenti
huzuridagi
maslahat-
ko’maklashuvchi
organdir. Komissiya
tarkibi O’zbekiston Respublikasi Prezidenti
tomonidan tasdiqlanadi. O’zbekiston Respublikasi Prezidentining
O’zbekiston
Respublikasining
fuqaroligi
masalalarini
hal
etishga
doir
konstitutsiyaviy
vakolatlarini
amalga
oshirish maqsadida besh yil muddatga
O’zbekiston
Respublikasi
Prezidenti
huzuridagi
Fuqarolik
masalalari
komissiyasi tuziladi. Komissiya
O’zbekiston
Respublikasi
Prezidenti
huzuridagi
maslahat-
ko’maklashuvchi
organdir. Komissiya
tarkibi O’zbekiston Respublikasi Prezidenti
tomonidan tasdiqlanadi. konsullik
okruglaridagi
mamlakatlarda
ozodlikdan
mahrum
etish
joylarida
saqlanayotgan O’zbekiston Respublikasi
fuqarolarining hisobini yuritadi, xorijga
doimiy
yashashga
chiqib
ketgan
O’zbekiston
Respublikasi
fuqarolarini
O’zbekiston
Respublikasi
Ichki
ishlar
vazirligining
har
chorakda
beradigan
so’rovnomalariga ko’ra doimiy konsullik
hisobi ro’yxatlari bo’yicha tekshiradi. Chet
davlat fuqaroligini ixtiyoriy ravishda olgan
O’zbekiston Respublikasining fuqarolarini
aniqlash uchun ichki ishlar organlari va
konsullik
muassasalari
manfaatdor
vazirliklar,
davlat
qo’mitalari
hamda
idoralarning
ma’lumotlarini
tekshiradi,
shuningdek
jismoniy
va
yuridik
shaxslarning
kelib
tushayotgan
murojaatlarini o’rganadi. gar o’n to’rt
yoshga
to’lmagan
bolaning
ota-onasi
(yolg’iz otasi yoki onasi) O’zbekiston
Respublikasi
fuqaroligini
olsa
yoki
O’zbekiston Respublikasining fuqaroligidan
chiqsa, ushbu bolaning fuqaroligi o’zgarishi
mumkin. Komissiya rais, rais o’rinbosari, kotib va
Komissiyaning boshqa a’zolaridan tarkib
topadi. Komissiyaning raisi, rais o’rinbosari va
boshqa a’zolari uning ishida jamoatchilik
asosida ishtirok etadi. Komissiya majlisida rais yo’q bo’lgan
taqdirda
uning
vazifalari
rais
o’rinbosarining
zimmasiga
yuklatiladi. O’zbekiston
Respublikasi
Prezidenti
huzuridagi
Fuqarolik
masalalari
komissiyasining reglamenti
O’zbekiston
Respublikasi
Prezidenti
tomonidan
tasdiqlanadi. Komissiya majlisida rais yo’q bo’lgan
taqdirda
uning
vazifalari
rais
o’rinbosarining
zimmasiga
yuklatiladi. O’zbekiston
Respublikasi
Prezidenti
huzuridagi
Fuqarolik
masalalari
komissiyasining reglamenti
O’zbekiston
Respublikasi
Prezidenti
tomonidan
tasdiqlanadi. Ota-ona (yolg’iz ota yoki ona) O’zbekiston
Respublikasi fuqaroligini o’zgartirganda
O’zbekiston Respublikasi fuqarosi bo’lgan
bolaning fuqaroligi, basharti bola fuqaroligi
bo’lmagan shaxs bo’lib qolmasa, o’zgarishi
mumkin. Ota-onalik huquqlaridan mahrum
etilgan
ota-onasining
fuqaroligi
o’zgarganda,
bolaning
fuqaroligi
o’zgarmaydi. Bolaning fuqaroligi o’zgargan
taqdirda, ota-onalik huquqlaridan mahrum
etilgan
ota-onasining
roziligi
talab
qilinmaydi. Ota-onasining
fuqaroligi
o’zgarganda yoki bola farzandlikka olingan
taqdirda o’n to’rt yoshdan o’n sakkiz
yoshgacha bo’lgan bolaning fuqaroligi faqat
uning yozma roziligi bilan o’zgartirilishi
mumkin. EURASIAN JOURNAL OF LAW, FINANCE AND
APPLIED SCIENCES
Innovative Academy Research Support Center
UIF = 8.3 | SJIF = 5.961
www.in-academy.uz Xorijda doimiy yashayotgan va yetti yil
mobaynida
doimiy
konsullik
hisobiga
turmagan
O’zbekiston
Respublikasining
fuqarolarini
aniqlash
uchun
konsullik
muassasalari:
konsullik
muassasasiga
shaxsan
murojaat
qilgan
shaxslarning
murojaatlarini
yoxud
konsullik
muassasalariga Tashqi ishlar vazirligi orqali
kelib tushgan ma’lumotlarni tekshiradi,
shaxsning
O’zbekiston
Respublikasi
fuqaroligiga mansubligi to’g’risidagi chet
davlatning so’rovnomalarini o’rganadi, o’z Xorijda doimiy yashayotgan va yetti yil
mobaynida
doimiy
konsullik
hisobiga
turmagan
O’zbekiston
Respublikasining
fuqarolarini
aniqlash
uchun
konsullik
muassasalari:
konsullik
muassasasiga
shaxsan
murojaat
qilgan
shaxslarning
murojaatlarini
yoxud
konsullik
muassasalariga Tashqi ishlar vazirligi orqali
kelib tushgan ma’lumotlarni tekshiradi,
shaxsning
O’zbekiston
Respublikasi
fuqaroligiga mansubligi to’g’risidagi chet
davlatning so’rovnomalarini o’rganadi, o’z g)
O’zbekiston
Respublikasining
Konstitutsiyasiga rioya etish majburiyatini
o’z zimmasiga olsa; g)
O’zbekiston
Respublikasining
Konstitutsiyasiga rioya etish majburiyatini
o’z zimmasiga olsa; d) davlat tilini muloqot qilish uchun zarur
darajada bilsa. Davlat tilini bilish darajasini
aniqlash tartibi O’zbekiston Respublikasi
Vazirlar Mahkamasi tomonidan belgilanadi. O’zbekiston Respublikasining fuqaroligiga
qabul qilishning soddalashtirilgan tartibi
chet davlat fuqarosi bo’lgan yoki fuqaroligi
bo’lmagan shaxs hisoblangan vatandoshga ISSN 2181-2853 Volume 2 Issue 5, May 2022 Page 204 U EURASIAN JOURNAL OF LAW, FINANCE AND
APPLIED SCIENCES
Innovative Academy Research Support Center O’zbekiston
Respublikasida
doimiy
yashovchi shaxslarga nisbatan fuqarolik
masalalariga oid qarorlarni ijro etish Ichki
ishlar vazirligi va uning tegishli bo’linmalari
zimmasiga, xorijda yashaydigan shaxslarga
nisbatan esa Tashqi ishlar vazirligi va
tegishli konsullik muassasalari zimmasiga
yuklatiladi.O’ziga
nisbatan
O’zbekiston
Respublikasining fuqaroligiga qabul qilish,
uni tiklash yoki undan chiqish bo’yicha
qaror qabul qilingan shaxs bu haqda
O’zbekiston Respublikasi Prezidentining
fuqarolik masalalariga oid qarori kuchga
kirgan kundan e’tiboran o’n besh kunlik ISSN 2181-2853 Volume 2 Issue 5, May 2022 Page 205 Page 205 U EURASIAN JOURNAL OF LAW, FINANCE AND
APPLIED SCIENCES
Innovative Academy Research Support Center
UIF = 8.3 | SJIF = 5.961
www.in-academy.uz EURASIAN JOURNAL OF LAW, FINANCE AND
APPLIED SCIENCES
Innovative Academy Research Support Center
UIF = 8.3 | SJIF = 5.961
www.in-academy.uz Innovative Academy Research Support Center
UIF = 8.3 | SJIF = 5.961
www.in-academ amal
qilmaslik
javobgarlikni
keltirib
chiqaradi. Ichki ishlar vazirligi, Tashqi
ishlar vazirligi hamda boshqa manfaatdor
vazirliklar, davlat qo’mitalari va idoralar
qonunning ijrosini, ijrochilarga muntazam
yetkazilishini hamda mohiyati va ahamiyati
aholi
o’rtasida
yanada
ko’proq
tushuntirilishi lozim deb o’ylayman. muddatda ichki ishlar organi yoki konsullik
muassasasi tomonidan xabardor qilinadi. Xulosa:
O’zbekiston
Respublikasida
fuqarolikka
ega
bo’lish
tartibi
va
fuqarolarning
huquqlardan
foydalanish
qonunda
belgilangan. O’zbekiston
Respublikasining
fuqaroligi
sohasida
yuzaga keladigan nizolar qonunchilikda
belgilangan tartibda hal etiladi. Bu qonunga
amal
qilmaslik
chiqaradi. Ichki
ishlar vazirligi ha
vazirliklar, davlat
qonunning ijrosin
yetkazilishini ham
aholi
o’rtasid
tushuntirilishi loz
Foydalanilgan adabiyotlar: muddatda ichki ishlar organi yoki konsullik
muassasasi tomonidan xabardor qilinadi. Xulosa:
O’zbekiston
Respublikasida
fuqarolikka
ega
bo’lish
tartibi
va
fuqarolarning
huquqlardan
foydalanish
qonunda
belgilangan. O’zbekiston
Respublikasining
fuqaroligi
sohasida
yuzaga keladigan nizolar qonunchilikda
belgilangan tartibda hal etiladi. Bu qonunga Foydalanilgan adabiyotlar: 1. http://lex.uz//docs/-20596 O’zbekiston Respublikasining Konstitutsiyasi
2. http://lex.uz//docs/-4761984 O’zbekiston Respublikasining fuqaroligi to’g’risida
3. O’zbekiston Respublikasining 2021-yil 14-iyundagi O’RQ-696-sonli Qonuni tahririda —
Qonunchilik ma’lumotlari milliy bazasi, 15.06.2021-y., 03/21/696/0553-son) Volume 2 Issue 5, May 2022 Page 206 ISSN 2181-2853
|
https://openalex.org/W3085014783
|
https://www.rrh.org.au/journal/download/pdf/6379/
|
English
| null |
COVID-19 leads to physically severe experiences for the rural elderly in Japan, during Obon
|
Rural and remote health
| 2,020
|
cc-by
| 787
|
FULL ARTICLE: ancestors sincerely and, during the Obon week, they get together
and hope for their ancestors’ fulfilled lives in heaven. ancestors sincerely and, during the Obon week, they get together
and hope for their ancestors’ fulfilled lives in heaven. ETHICS APPROVAL Unnan City Hospital clinical Ethics committee This work is licensed under a Creative Commons Attribution 4.0 International Licence Rural and Remote Health
rrh.org.au
James Cook University ISSN 1445-6354 Rural and Remote Health
rrh.org.au
James Cook University ISSN 1445-6354 Dear Editor The COVID-19 pandemic has not only been affecting people’s
health due to fear and symptoms of the infection, but has also had
negative social impacts on human interaction – especially on the
rural elderly of Japan preparing for the celebration of Obon. Obon,
a long holiday week in August, is unique to the Japanese people. It
is a special, traditional Japanese event related to Buddhism . It is
believed that during Obon the spirits of the ancestors come back
to each house and rest there. The Japanese go back to their
hometowns and celebrate Obon to ensure the happiness of their
ancestors in heaven. Notably, the rural elderly admire their
1 The COVID-19 pandemic has disrupted the celebration of Obon
this year. The rural elderly have suffered not only from the mental
stress of possible infections and the pain of not being able to see
their relatives and loved ones, but also from the physical stress of
the preparations of Obon. Working in the rural Japanese
community, one often listens to elderly patients’ worries regarding
COVID-19 and Obon. Rural family physicians should listen to the
accounts of the difficulties experienced by the elderly during Obon
from their perspective for various reasons. First, this pandemic has inhibited young people from returning to
their hometowns. At present, most COVID-19 cases are from urban
areas, and young people in these areas may risk infecting others,
even if they are asymptomatic . They may be anxious about
transmitting SARS-CoV-2 to their parents and grandparents, who
can become critically ill from the same. Rural areas isolate the
elderly and this pandemic has depressed them; the rural elderly
cannot meet their children and grandchildren , despite wishing to
meet them. 2
3 forced older people to engage in manual labor that exceeds their
physical abilities. Hence, the present condition has increased heat
strokes and musculoskeletal disorders among the rural elderly. During Obon in 2020, keeping in mind the pandemic, rural family
physicians should consider the pandemic’s negative impact in
terms of social isolation among the rural elderly. They should
consider the mental and physical stress among the elderly patients
resulting from manual labor, even though the patients themselves
may not provide appropriate information regarding their health
conditions . Although the rural elderly may not be able to avoid
labor related to Obon, rural family physicians’ continual listening
and advice can help alleviate their stress. Dear Editor 4 Second, the rural elderly had to prepare for Obon without
assistance from the young. During Obon, to welcome their
ancestors, people have to prepare not only food for the
celebration, but also clean their ancestors’ tombs and trim the
bushes surrounding their homes. Usually, the rural elderly get help
from the younger generations during Obon. However, the
pandemic has imposed a burden on the rural elderly. Although in
urban areas many companies can help provide manual labor to
prepare for Obon, rural areas do not have such services; this has Acknowledgements We would like to thank all the participants of this research. Ryuichi Ohta and Yoshinori Ryu, Community Care, Unnan City
Hospital, Japan This PDF has been produced for your convenience. Always refer to the live site https://www.rrh.org.au/journal/article/6379 for the
Version of Record. REFERENCES: solutions in the continuity of home care for rural older people: a
thematic analysis. Home Health Care Services Quarterly 2020;
39(2): 126-139. https://doi.org/10.1080/01621424.2020.1739185
PMid:32174235 1 Kingsley BNH. Obon, Festival of the Dead. Prairie Schooner 2017;
91(1): 81-83. https://doi.org/10.1353/psg.2017.0069
2 Nishiura H, Kobayashi T, Miyama T, Suzuki A, Jung SM, Hayashi K,
et al. Estimation of the asymptomatic ratio of novel coronavirus
infections (COVID-19). International Journal of Infectious Diseases
2020; 94: 154-155. https://doi.org/10.1016/j.ijid.2020.03.020
PMid:32179137 1 Kingsley BNH. Obon, Festival of the Dead. Prairie Schooner 2017;
91(1): 81-83. https://doi.org/10.1353/psg.2017.0069 4 Aoki T, Urushibara-Miyachi Y. A qualitative study of socially
isolated patients’ perceptions of primary care. Journal of General
and Family Medicine 2019; 20(5): 185-189. https://doi.org/10.1002
/jgf2.262 PMid:31516804 3 Ohta R, Ryu Y, Kitayuguchi J, Gomi T, Katsube T. Challenges and This PDF has been produced for your convenience. Always refer to the live site https://www.rrh.org.au/journal/article/6379 for the
Version of Record. This PDF has been produced for your convenience. Always refer to the live site https://www.rrh.org.au/journal/article/6379 for the
Ve sion of Reco d
|
https://openalex.org/W4255931800
|
https://www.researchsquare.com/article/rs-49199/v2.pdf?c=1631871154000
|
English
| null |
Are weight status and weight perception associated with academic performance among youth?
|
Research Square (Research Square)
| 2,020
|
cc-by
| 8,063
|
Are weight status and weight perception associated
with academic performance among youth?
Maram Livermore
Brock University Faculty of Applied Health Sciences
Markus J Duncan
Brock University Faculty of Applied Health Sciences
Scott T. Leatherdale
University of Waterloo Faculty of Applied Health Sciences
Karen Allison Patte
(
kpatte@brocku.ca
)
Brock University Faculty of Applied Health Sciences
https://orcid.org/0000-0002-5214-1943
Research article
Keywords: Obesity, overweight, academic achievement, weight perception, education, youth
Posted Date: September 23rd, 2020
DOI: https://doi.org/10.21203/rs.3.rs-49199/v2
License:
This work is licensed under a Creative Commons Attribution 4.0 International
License.
Read Full License
Version of Record: A version of this preprint was published on October 26th, 2020. See the published
version at https://doi.org/10.1186/s40337-020-00329-w. Are weight status and weight perception associated
with academic performance among youth? Maram Livermore
Brock University Faculty of Applied Health Sciences
Markus J Duncan
Brock University Faculty of Applied Health Sciences
Scott T. Leatherdale
University of Waterloo Faculty of Applied Health Sciences
Karen Allison Patte
(
kpatte@brocku.ca
)
Brock University Faculty of Applied Health Sciences
https://orcid.org/0000-0002-5214-1943
Research article
Keywords: Obesity, overweight, academic achievement, weight perception, education, youth
Posted Date: September 23rd, 2020
DOI: https://doi.org/10.21203/rs.3.rs-49199/v2
License:
This work is licensed under a Creative Commons Attribution 4.0 International
License. Read Full License
Version of Record: A version of this preprint was published on October 26th, 2020. See the published
version at https://doi.org/10.1186/s40337-020-00329-w. Research article Version of Record: A version of this preprint was published on October 26th, 2020. See the published
version at https://doi.org/10.1186/s40337-020-00329-w. Page 1/18 Page 1/18 Page 1/18 Abstract Background: Emerging evidence suggests perceptions of being overweight account for many of the
psychosocial consequences commonly associated with obesity. Previous research suggests an obesity
achievement gap, yet limited research has explored weight perception in association with academic
performance. Moreover, underweight perceptions have typically been excluded from research. The current
study examined how BMI classification and weight perception relate to academic performance in a large
cohort of youth. Methods: We used cross-sectional survey data from 61,866 grade 9-12 students attending the 122
Canadian schools that participated in Year 6 (2017/2018) of the COMPASS study. Mixed effect
regression models were used to examine associations between students’ BMI classification and weight
perceptions and their math and English/French course grades. All models were stratified by sex and
adjusted for sociodemographic covariates and school clustering. Results: For English/French grades, males and females with overweight or underweight perceptions were
less likely to achieve higher grades than their peers with perceptions of being at “about the right weight”,
controlling for BMI and covariates. For math grades, females with overweight perceptions, and all
students with underweight perceptions, were less likely to achieve higher grades than their peers with
“about the right weight” perceptions. All students with BMIs in the obesity range were less likely to report
grades of 60% or higher than their peers with “normal-weight” BMIs, controlling for weight perception and
covariates. Overweight BMIs were predictive of lower achievement in females for English/French grades,
and in males for math grades, relative to “normal-weight” BMIs. Results for students that did not respond
to the weight and weight perception items resembled those for obesity BMI and overweight/underweight
perceptions, respectively. Conclusions: Overall, this study demonstrates that an obesity achievement gap remains when controlling
for students’ perceptions of their weight, and that both underweight and overweight perceptions predict
lower academic performance, regardless of BMI classification. Results suggest barriers to academic
success exist among youth with larger body sizes, and those with perceptions of deviating from “about
the right weight”. Introduction About 35% of Canadian children and adolescents are at risk of having overweight or obesity [1]. Numerous studies have shown that childhood obesity is associated with various physical health
concerns [2]. Additionally, larger-bodied adolescents are at increased risk of adverse psychosocial
outcomes [3]. Strauss and Pollack reported that children and adolescents face many challenges but “few
problems in childhood have as significant an impact on emotional development as being overweight”
(p.747) [4]. In fact, children with overweight or obesity report lower quality of life scores than children
diagnosed with cancer [5]. Previous research also indicates the presence of an obesity achievement gap Page 2/18 for children and adolescents [6,7]. More specifically, some evidence suggests students with obesity have
poorer academic achievement, more absenteeism, higher dropout rates [6,8], and are less likely to pursue
and attain post-secondary education [9,10]. A recent meta-analysis by He et al. included 60 studies of weight status and academic performance and
found a pooled correlation (r=-.111) between higher Body Mass Index (BMI) and lower grades [11]. The
researchers concluded that the relationship between weight status and academic achievement was
moderated by geographical region, with lower grades more likely to be associated with high BMI in North
America than other cultures, possibly due to differing weight norms and given the stigma associated with
overweight and obesity. Emerging evidence suggests weight perception—individuals’ subjective appraisal
of their body weight—accounts for many of the psychosocial consequences commonly associated with
obesity [13-15]. That is, the perception of being overweight, rather than body weight itself, may account
for risks of lower self-concept and poor mental health. One of the potential factors linking obesity and the
perception of overweight to adverse outcomes is the experience of stigma and bias. An individual may
need to perceive themselves as overweight, in order to internalize the bias associated with
overweight/obesity. To date, the relationship between self-perceptions of weight and academic
achievement has been largely overlooked. To our knowledge, only one study has examined weight
perception as a predictor of academic achievement [28]. Among US adolescents (ages 14 to 17)
participating in the 2003 Youth Risk Behavior Study, Florin, Shultz, and Stettler found that perceived
overweight status was associated with lower grades, regardless of BMI classification, and that obesity
was no longer associated with grades when controlling for weight perception [28]. Introduction Replication is
necessary, particularly in more recent samples, given potential shifts in social weight norms (e.g., related
to obesity prevalence, body positive movements, and anti-obesity stigma and weight bias efforts). Furthermore, consideration of underweight perceptions in addition to overweight, and exploration of sex
or gender differences, are warranted. Related to thinness and muscularly sociocultural body ideals,
girls/women are more likely to report perceptions of underweight than boys/men, while boys/men tend to
be split between perceptions of underweight and overweight [16]. Regardless of body size, both
overweight and underweight perceptions appear detrimental to mental and physical health relative to
perceptions of being “about the right weight” [13-25]; however, the latter has received relatively limited
attention. Further examination of a potential weight-related achievement gap is critical to inform a learning
environment that will enable all youth to thrive. Deterrents to academic achievement in adolescence have
critical implications for future career opportunities and successful transitions to adulthood, with school
failure and dropout increasing the risk of later unemployment, poverty, lower quality life, criminality,
violence, and various health risk behaviors [6,7,29]. We sought to determine if larger-bodied students
report lower grades in secondary school than their peers with “normal-weight” BMIs. In addition, we
explored whether weight perceptions predicted grades, while controlling for weight status. It was
hypothesized that perceptions of being at “about the right weight” would provide a protective effect, and
any relationship between BMI classification and academic achievement would be reduced when
controlling for weight perception. Page 3/18 Design and Participants The COMPASS study is a prospective cohort study designed to collect longitudinal and hierarchical
health data from a large sample of grade 9 through 12 students (ages 14-19) enrolled in Canadian
secondary schools. Schools and school boards were purposely selected based on whether they permitted
active-information passive-consent parental permission protocols, which are critical for collecting robust
data among youth [30]. The COMPASS student questionnaire (Cq), a self-report paper-and-pencil survey,
is completed once annually by full school samples during one classroom period. All grade 9 through 12
students attending participating schools were eligible to participate and could decline at any time. A full
description of recruitment methods [31] and the COMPASS study are available in print [32] and online
(www.compass.uwaterloo.ca). The COMPASS study received ethics approval from the University of
Waterloo and Brock University Human Research Ethics Committee and all participating school boards. We used cross-sectional data from Year 6 (2017-2018 school year) of the COMPASS study, which
included 66,434 students at 122 secondary schools in Ontario (n=61), British Columbia (n=16), Alberta
(n=8) and Quebec (n=37). The overall student response rate in Year 6 was 81.85% of eligible students. Student non-participation primarily resulted from absences or scheduled study-periods during data
collection. Participants with missing outcome, sex, or covariate data were removed, leaving a final sample
of 61,866 adolescents. Methods Design and Participants Design and Participants Descriptive Statistics Descriptive statistics for all variables are described in Table 1. In this sample, 49.4% of participants
identified as male and 50.6% as female. About two-thirds (66.6%) of the sample identified as white and
one-third as non-white, mixed, or other race/ethnicity. In terms of weight status, 5.7%, 12.0%, 54.7% and
3.6% of the total sample had BMIs in the obesity, overweight, “normal-weight”, and underweight Descriptive statistics for all variables are described in Table 1. In this sample, 49.4% of participants
identified as male and 50.6% as female. About two-thirds (66.6%) of the sample identified as white and
one-third as non-white, mixed, or other race/ethnicity. In terms of weight status, 5.7%, 12.0%, 54.7% and
3.6% of the total sample had BMIs in the obesity, overweight, “normal-weight”, and underweight
categories, respectively, while 19.6% of students did not report their weight, and the remaining proportion
were missing sex, age, or height data to categorize BMI. Additionally, a total of 23.8% and 58.7% students
reported overweight and “about the right weight” perceptions respectively, while 1.4% did not respond to
the weight perception item. More males reported perceptions of “underweight” than females (21.2%
versus 11.1%), while more females reported perceptions of overweight (26.5% versus 20.6%) and “about
the right weight” (60.7% versus 56.2%) than males. Just over half of students (51.0-51.3%) reported math
and English/French grades above 60%. 3.6% of the total sample had BMIs in the obesity, overweight, normal weight , and underweight
categories, respectively, while 19.6% of students did not report their weight, and the remaining proportion
were missing sex, age, or height data to categorize BMI. Additionally, a total of 23.8% and 58.7% students
reported overweight and “about the right weight” perceptions respectively, while 1.4% did not respond to
the weight perception item. More males reported perceptions of “underweight” than females (21.2%
versus 11.1%), while more females reported perceptions of overweight (26.5% versus 20.6%) and “about
the right weight” (60.7% versus 56.2%) than males. Just over half of students (51.0-51.3%) reported math
and English/French grades above 60%. The concordance between weight perception and BMI category is presented in Figure 1. The weighted
Kappa was 0.392 among females and 0.370 among males. The majority of males (74.9%) and females
(81.3%) with BMIs in the obesity category reported perceptions of overweight. Measures Weight Status. Student weight status was defined by Body Mass Index (BMI; kg/m²) classification
determined based on student-reported height and weight [33], and the World Health Organization [34] age-
and sex-adjusted cut points (underweight, normal weight, overweight, obesity). A previous study found
the weight status measure to be reliable, valid, and valuable for use when objective methods are not
feasible [33]. Given the prevalence of missing BMI data, and as missing self-reported weight data may
not be missing at random [35], two separate categories were created for missing weight status based on
which variables were missing to determine BMI classification: missing BMI classification due to weight
not being reported, and missing BMI due to missing age, sex, or height. Weight Perception. Subjective perception of weight status was determined using the question, “How do
you describe your weight?” Response options included: “very underweight”, “slightly underweight”, “about
the right weight”, “slightly overweight” and “very overweight”. Responses were collapsed into three
categories: underweight, about right, and overweight. In addition, missing weight perception responses
were included as a fourth category. Covariates. Participant-reported race/ethnicity (categorized into white and, non-white minority,
multiethnic, or other) and school grade (9, 10, 11, 12, other [Secondary I- II in Quebec]) were entered into
the model as covariates. Also, student weekly spending/saving money (categorized into $1-$20, Page 4/18 $21-$100, >$100, don’t know) was included as an indicator of part-time employment and/or allowance,
as proxy for student-level SES in the absence of data on parental income or education data. $21-$100, >$100, don’t know) was included as an indicator of part-time employment and/or allowance,
as proxy for student-level SES in the absence of data on parental income or education data. Academic Performance. Academic performance was assessed using student-reported grades. Participants reported their approximate overall mark in their current or most recent math and English (in
Ontario, Alberta and BC schools) or French (in Quebec) courses. Grades were dichotomized as ≥ 60% or
<60% for both math and English/French grade models. ce. Academic performance was assessed using student-reported grades. Statistical Analysis R software [36] was used to conduct frequency descriptive statistics and mixed effect logistic regression
models. Separate models explored the association between BMI status and weight perception with math
grades and English/French course grades, stratified by sex and controlling for weekly spending money,
school grade, and race/ethnicity. Mixed models were used to account for school clustering by adding a
random intercept at the school level. No interaction effect between weight status and weight perception
was indicated when tested (results not reported). Descriptive Statistics Among those with
overweight BMIs, most females reported perceptions of overweight (62.2%); whereas more males reported
“about right” weight perceptions (54.5%) than overweight perceptions (42.4%). Most males (64.2%) and
females (61.5%) in the underweight BMI category reported underweight perceptions. In the “normal
weight” BMI category, more males reported underweight perceptions (27.5%) than females (12.2%), while
females were more likely to perceive their weight as “about right” (71.6%) or overweight (15.4%) than
males (65.4%; 5.9%). Page 5/18 Page 5/18 Page 5/18 Students with missing BMI data resembled those with “normal weight” BMIs, except a higher proportion
reported overweight perceptions. About half of students with missing BMI data reported “about right”
weight perceptions (51.1% males; 55.1% females), over a quarter reported overweight perceptions (25.1%
males; 30.8% females), and 9.7% of females and 19.0% of males reported underweight perceptions. Students with missing BMI data resembled those with “normal weight” BMIs, except a higher proportion
reported overweight perceptions. About half of students with missing BMI data reported “about right”
weight perceptions (51.1% males; 55.1% females), over a quarter reported overweight perceptions (25.1%
males; 30.8% females), and 9.7% of females and 19.0% of males reported underweight perceptions. Math Course Grades See Table 2 for model results testing BMI classification and weight perception as predictors of math
grades, after stratifying by sex and controlling for covariates. In males only, having over $100 a week
available for spending or saving and identifying as nonwhite, mixed, or other race/ethnicity were
associated with lower odds of reporting math grades over 60%, relative to males without any weekly
spending money and of white race/ethnicity, respectively. Students with BMIs in the obesity range were
less likely to report a math grade above 60% compared to those with BMIs considered “normal weight”. In
males only, overweight BMIs were associated with a lower likelihood of report a math grade above 60%
compared to those with “normal-weight” BMIs. No effect resulted for underweight BMI relative to “normal-
weight” BMI. In females only, weight perceptions of overweight predicted a lower likelihood of reporting
math grades above 60% when compared to perceptions of being “about the right weight”, controlling for
BMI classification and covariates. In both males and females, perceptions of underweight were
associated with lower likelihood of higher math grades than perceptions of being “about the right
weight”. Student with missing BMI data either due to not reporting weight or missing height, age or sex
data were more likely to have lower math grades when compared to those in the “normal-weight” BMI
category. Similarly, students with missing weight perception data were more likely to report lower math
grades when compared to those with “about the right weight” perceptions. Discussion The current study examined whether weight status and student perceptions of their weight were
associated with academic grades in selected secondary school courses (math and English/French) in a
large population study of Canadian youth. Results support the links between obesity, as determined by
BMI, and lower academic performance in both males and females. Overweight BMI classifications were
also associated with lower odds of high grades in English/French courses among females, and in math
classes among male students, relative to BMIs considered “normal-weight”. Similar to previous research
[28], females with overweight perceptions were less likely to achieve higher course grades in math, and
both males and females with overweight perceptions had lower odds of grades above 60% in their
English/French courses, relative to their peers with perceptions of being at “about the right weight”,
controlling for BMI classification and covariates. While no effect was found for underweight relative to
“normal-weight” BMI, females and males with perceptions of being underweight had lower odds of high
grades in math and English/French courses relative to students with “about the right weight” perceptions. Results for youth with missing BMI or weight perception data resembled those for obesity BMI
classifications and perceptions of overweight or underweight, respectively. Overall, this study
demonstrates that an obesity achievement gap remains when controlling for students’ perceptions of
their weight, and that weight perceptions—both underweight and overweight—predict lower academic
performance, regardless of BMI classification. Further research is needed to determine the mechanisms
underlying these relationships, in order to remove barriers to academic success among youth with larger
body sizes, and those with perceptions of deviating from “about the right weight”. To the best of our knowledge, only one previous study has examined weight perception as a predictor of
academic grades. The current study provides necessary replication and builds on existing literature, by
examining these relationships in a more recent and larger sample of youth. In a sample of approximately
11,000 US adolescents participating in the 2003 Youth Risk Behavior Study, perceptions of overweight
were found to be a stronger predictor of academic outcomes than BMI, and obesity was no longer a
significant predictor of academic performance when accounting for overweight perceptions [28]. Overweight perceptions have been linked to poor mental health, psychosocial distress, and low self-
esteem in adolescents [29,38-40], factors that have also been associated with lower academic
achievement [37]. English/French Course Grades Table 3 summarizes the resultant model testing BMI classification, weight perception, and covariates as
predictors of English/French grades by sex. Both males and females identifying as nonwhite, mixed, or
other race/ethnicity had lower odds of English/French grades above 60% than their counterparts
identifying as white. No differences in English/French grades resulted by spending money or grade. Both
male and female adolescents with BMIs in the obesity range, and females with overweight BMIs, were
less likely to report higher grades than their peers with BMIs considered “normal weight”. Both males and females with weight perceptions of overweight and underweight were less likely to report
higher grades (above 60%) in their English/French classes when compared to those with perceptions of
being at “about the right weight”. Students with missing BMI data due to either not reporting weight, or
because of missing height, age, or sex data, had significantly lower odds of reporting English/French
grades above 60%, when compared to those with “normal-weight” BMIs. Similarly, students with missing
weight perception data were less likely to have English/French grades above 60% when compared to
those with “about the right weight” perceptions. Page 6/18 Discussion In fact, researchers have reported perceiving oneself as overweight to be a stronger
predictor of behavioural issues and mental distress than actual weight status [13,27,40]. Similarly, the
current results indicate that perceptions of overweight significantly predicted poorer academic outcomes
independent of BMI classification, with the exception of math grades in males. However, unlike Florin et
al. [28], obesity BMI classifications remained predictive of lower grades when controlling for weight
perception. The current study suggests other factors appear to contribute to the obesity achievement gap, such as
parental education, mental health, or external weight bias. Students with obesity are more likely to
experience weight-based bullying within the school context [28,39]. Also, weight bias has been
documented in physical education teachers [9,43,44,54] and given its pervasiveness across the Page 7/18 Page 7/18 population [9,43], may contribute to educators’ perceptions of students’ academic abilities. Several
studies have indicated that the association between BMI and academic performance was no longer
significant when models adjusted for parental/familial characteristics [12,47]. For instance, Datar et al. concluded that overweight status is not a causal factor of lower academic performance, as weight-related
differences in test scores became insignificant when social and behavioural variables, such as SES and
parental time spent with the child were considered [47]. The authors cautioned that higher weight
students may be labeled as lower achievers, as weight is a more obvious marker than sociodemographic
characteristics. population [9,43], may contribute to educators’ perceptions
studies have indicated that the association between BMI an
significant when models adjusted for parental/familial cha
concluded that overweight status is not a causal factor of l
differences in test scores became insignificant when social
parental time spent with the child were considered [47]. The
students may be labeled as lower achievers, as weight is a
characteristics. Adolescents who perceive themselves as overweight may be at risk of internalizing weight stigma. Despite the high prevalence of obesity, weight stigma continues to be problematic. Stereotypes that
individuals living with obesity are lazy, unintelligent, or lack willpower contribute to stigma and
discrimination [43]. Internalized stigma, or self-stigma, occurs when individuals apply negative
stereotypes to themselves and believe that the stigma is deserved [44], leading to low self-esteem and
psychological distress [43]. Research indicates that bias toward individuals with overweight and obesity
persists in health care, employment, and home settings [9]. Education, however, has received less
research attention, particularly at the secondary school level. Discussion It is plausible that adolescents who perceive
themselves as overweight have lower self-concepts related to internalized weight stigma, which in turn,
contributes to poorer academic engagement and performance. That is, students who feel their weight is
“about right” may be more likely to succeed because they have not internalized negative stereotypes. Students with underweight perceptions also reported lower grades than those with “about right”
perceptions. Interestingly, the effect of underweight perceptions was more consistent than overweight
perceptions in predicting lower grades in both females and males, and across Math and English/French
grades. To our knowledge, no previous study has examined underweight perceptions in relation to
academic performance. While less studied, underweight perceptions have been associated with
depressive and anxiety symptoms in males [21,41,42] and suicidality and lower health-related quality of
life in all youth [25,27]. Based on this result, it is plausible that links between weight perception and
academic performance relate more to deviations from the social norm or sociocultural body ideals than
to weight stigma. Results are consistent with sociocultural body ideals of thinness for women and
muscularity for men, with more males reporting underweight perceptions than females. About one-fifth of
males and one-tenth of females reported underweight perceptions; yet only 1.5% and 2.1% of female and
males had BMIs classified as underweight, respectively. Another plausible explanation is that youth with
underweight perceptions were experiencing weight restrictions (e.g., due to food insecurity) and/or had
lost weight, while “normal-weight” by BMI, which in turn contributed to reduced ability to perform in
school. Results highlight the importance of including weight perceptions across the spectrum in future
research. Interestingly, students that did not report their body weight or weight perception tended to have the lowest
likelihood of achieving higher grades, comparable to the odds for obesity BMI classification. Far more
males and females with missing BMI data reported perceptions of overweight than their peers with Page 8/18 Page 8/18 “normal-weight” BMIs. Adolescents are less likely to report their weight as BMI increases and if they have
poor body image [35,45]. Missing self-reported weight status and perceptions may be influenced by an
awareness of societal norms of thinness/muscularity ideals and weight bias attitudes [45,46]. Discussion Hence, if
adolescents with larger body sizes did not report their weight due to concerns of judgment by others,
results lend support to the theory that negative perceptions regarding body weights outside of “about
right” contribute to lower academic performance. Future research should explore both internalized and externalized weight bias, and associated lower self-
concept and mental distress, as possible mechanisms explaining links between higher weight status, and
perceptions of overweight and underweight, with academic achievement. Upstream strategies targeting
the negative connotations of varying body sizes may prove valuable, to prevent the adverse psychosocial
outcomes associated with perceptions of being overweight or underweight. Enhanced efforts to prevent
weight-based bullying and promote weight acceptance are also advised. Previous research suggests
bullying victimization predicts changes from perceptions of being at “about the right weight” to
underweight and overweight perceptions among youth [55]. Limitations Several limitations require consideration. First, while the large sample supports generalizability, the
COMPASS study was not designed to be representative. Second, cross-sectional data was used to explore
relationships. Future longitudinal analysis of COMPASS data will assist in establishing temporality and
testing potential mechanistic contributors (e.g., mental health, self-concept, bullying victimization, school
connectedness). Third, the use of self-reported data carries risks of recall and social desirability bias. For
instance, lower achieving students may over report their achievement. However, a review of 37
independent samples found strong response validity of self-reported grades in high school students [48]. Similarly, weight status was based on student-reported height and weight, and as such, results likely
reflect greater concordance between weight perception and weight status than exists. However, time and
cost constraints preclude the feasibility of obtaining objective height and weight measures, not to
mention the potential harm of weight measurements in a school-based study. Also, a strong correlation
between measured and self-reported BMI has been shown in youth [33,49]. Fourth, the reference point
that youth used to answer the weight perception question is not entirely clear. That is, it is not known
whether respondents were comparing their weight to their ideal body, their peers, a medical standard, or
some other alternative. For instance, responses of ‘about the right weight’ may indicate weight
satisfaction rather than youths’ perception of how their weight compared to an external reference point. Lastly, as discussed, this study did not assess the contribution of self-esteem or mental health. Mental
health may confound the relationship between weight status or perception and academic achievement. Several studies have linked poor mental health with lower grades [37], and a recent intervention
promoting positive mental health and wellbeing significantly improved academic grades in participating
schools [50]. Future studies, using longitudinal designs, standardized test grades, and including mental
health, self-concept, and stigma measures should be considered. Several limitations require consideration. First, while the large sample supports generalizability, the
COMPASS study was not designed to be representative. Second, cross-sectional data was used to explore
relationships. Future longitudinal analysis of COMPASS data will assist in establishing temporality and
testing potential mechanistic contributors (e.g., mental health, self-concept, bullying victimization, school
connectedness). Third, the use of self-reported data carries risks of recall and social desirability bias. For
instance, lower achieving students may over report their achievement. However, a review of 37 connectedness). Conclusion Results support the existence of an achievement gap by both weight status and weight perception. This
study has widespread implications with over 2 million adolescents at risk of having overweight or obesity
in Canada [1], and over 40% of youth reporting weight perceptions other than “about right”. Research is
needed to further examine the mechanisms underlying these associations. Academic achievement sets a
lifelong trajectory of health and wellbeing. Lower academic achievement is linked to increased rates of
unemployment, poverty, criminality, and negative future health outcomes [6,7,29,51]. The obesity
achievement gap has been suggested as an early contributor to later SES disparities found by weight
status [52]. The present study contributes to a body of research that encourages the consideration of
both overweight and underweight perceptions and their potential impact on adolescent emotional and
physical health. Upstream strategies to prevent negative connotations associated with body sizes
divergent from “about right” and to promote weight acceptance merit consideration. Further exploration is
necessary to inform policies and interventions that foster a learning environment that will enable all
youth to thrive. Limitations Third, the use of self-reported data carries risks of recall and social desirability bias. For
instance, lower achieving students may over report their achievement. However, a review of 37
independent samples found strong response validity of self-reported grades in high school students [48]. Similarly, weight status was based on student-reported height and weight, and as such, results likely
reflect greater concordance between weight perception and weight status than exists. However, time and
cost constraints preclude the feasibility of obtaining objective height and weight measures, not to
mention the potential harm of weight measurements in a school-based study. Also, a strong correlation
between measured and self-reported BMI has been shown in youth [33,49]. Fourth, the reference point
that youth used to answer the weight perception question is not entirely clear. That is, it is not known
whether respondents were comparing their weight to their ideal body, their peers, a medical standard, or
some other alternative. For instance, responses of ‘about the right weight’ may indicate weight satisfaction rather than youths’ perception of how their weight compared to an external reference point. Lastly, as discussed, this study did not assess the contribution of self-esteem or mental health. Mental
health may confound the relationship between weight status or perception and academic achievement. Several studies have linked poor mental health with lower grades [37], and a recent intervention
promoting positive mental health and wellbeing significantly improved academic grades in participating
schools [50]. Future studies, using longitudinal designs, standardized test grades, and including mental
health, self-concept, and stigma measures should be considered. Page 9/18 Page 9/18 Page 9/18 Declarations Funding: The COMPASS study has been supported by a bridge grant from the CIHR Institute of Nutrition,
Metabolism and Diabetes (INMD) through the “Obesity – Interventions to Prevent or Treat” priority
funding awards (OOP-110788; awarded to SL), an operating grant from the CIHR Institute of Population
and Public Health (IPPH) (MOP-114875; awarded to SL), a CIHR project grant (PJT-148562; awarded to
SL), a CIHR bridge grant (PJT-149092; awarded to KP/SL), a CIHR project grant (PJT-159693; awarded to
KP), and by a research funding arrangement with Health Canada (#1617-HQ-000012; contract awarded to
SL). Acknowledgements: The authors would like to thank the schools, school boards, and students that have
participated in the COMPASS study, and all COMPASS staff and team members. Acknowledgements: The authors would like to thank the schools, school boards, and students that have
participated in the COMPASS study, and all COMPASS staff and team members. Competing interests: The authors declare that they have no competing interests/conflicts of interests. Ethics approval and consent to participate: The University of Waterloo Office of Research Ethics
(ORE#17264) and participating school boards approved all procedures. All students attending
participating schools were invited to participate using active-information passive-consent parental
permission protocols. Students could withdraw from the study at any time. Consent for publication: N/A Authors’ contributions: ML lead the writing of the manuscript. MJD conducted all analysis. STL is the PI
of the COMPASS Study. ML and KAP conceptualized the manuscript study. All authors (ML, MJD, STL,
KAP) contributed to manuscript revisions, results interpretation, and approved the final version. Page 10/18 Page 10/18 Availability of data and material: COMPASS study data is available upon request through completion and
approval of an online form: https://uwaterloo.ca/compass-system/information-researchers/data-usage-
application. The datasets used during the current study are available from the corresponding author on
reasonable request. Code availability: R software was used for all analysis. Further details available upon request. Code availability: R software was used for all analysis. Further details available upon request. ode availability: R software was used for all analysis. Further details availab References 1. Shentow-Bewsh R, Zuberi D. Reducing the prevalence of obesity in Canada: a call to action. Soc Work
Public Health. 2018;33(6):329-341. doi:10.1080/19371918.2018.1482252 2. Must A, Anderson SE. Effects of obesity on morbidity in children and adolescents. Nutr Clin Care. 2003;6(1):4-48. 3. Erickson SJ, Robinson TN, Haydel KF, Killen JD. Are overweight children unhappy? Body mass index,
depressive symptoms, and overweight concerns in elementary school children. Arch Pediatr Adolesc
Med 2000;154(9):931‐935. doi:10.1001/archpedi.154.9.931 3. Erickson SJ, Robinson TN, Haydel KF, Killen JD. Are overweight children unhappy? Body mass index,
depressive symptoms, and overweight concerns in elementary school children. Arch Pediatr Adolesc
Med 2000;154(9):931‐935. doi:10.1001/archpedi.154.9.931 4. Strauss RS, Pollack HA. Social marginalization of overweight children. Arch Pediatr Adolesc Med. 2003;157(8):746‐752. doi:10.1001/archpedi.157.8.746 4. Strauss RS, Pollack HA. Social marginalization of overweight children. Arch Pediatr Adolesc Med. 2003;157(8):746‐752. doi:10.1001/archpedi.157.8.746 5. Schwimmer JB, Burwinkle TM, Varni JW. Health-related quality of life of severely obese children and
adolescents. JAMA 2003;289(14):1813‐1819. doi:10.1001/jama.289.14.1813 5. Schwimmer JB, Burwinkle TM, Varni JW. Health-related quality of life of severely obese children and
adolescents. JAMA 2003;289(14):1813‐1819. doi:10.1001/jama.289.14.1813 6. Shore SM, Sachs ML, Lidicker JR, Brett SN, Wright AR, Libonati JR. Decreased Scholastic
Achievement in Overweight Middle School Students. Obesity 2008;16(7):1535‐1538. doi:10.1038/oby.2008.254 7. Benson R, Von Hippel PT, Lynch JL. Does more education cause lower BMI, or do lower-BMI
individuals become more educated? Evidence from the National Longitudinal Survey of Youth 1979. Soc Sci Med 2018;211:370‐377. doi:10.1016/j.socscimed.2017.03.042 8. Lanza HI, Huang DYC. Is obesity associated with school dropout? key developmental and ethnic
differences. J Sch Health. 2015;85(10):663-670. 8. Lanza HI, Huang DYC. Is obesity associated with school dropout? key developmental and ethnic
differences. J Sch Health. 2015;85(10):663-670. 9. Puhl RM, Heuer CA. The stigma of obesity: a review and update. Obesity 2009;17(5):941‐964. doi:10.1038/oby.2008.636 9. Puhl RM, Heuer CA. The stigma of obesity: a review and update. Obesity 2009;17(5):941‐964. doi:10.1038/oby.2008.636 10. Larsen JK, Kleinjan M, Engels R, Fisher JO, Hermans R. Higher weight, lower education: a longitudinal
association between adolescents’ body mass index and their subsequent educational achievement
level? J Sch Health. 2014;84(12):769-776. doi:10.1111/josh.12212 10. Larsen JK, Kleinjan M, Engels R, Fisher JO, Hermans R. Higher weight, lower education: a longitudinal
association between adolescents’ body mass index and their subsequent educational achievement
level? J Sch Health. 2014;84(12):769-776. doi:10.1111/josh.12212 11. He J, Chen X, Fan X, Cai Z, Huang F. Is there a relationship between body mass index and academic
achievement? A meta-analysis. Public Health. 2019;167:111-124. References doi:10.1016/j.puhe.2018.11.002 11. He J, Chen X, Fan X, Cai Z, Huang F. Is there a relationship between body mass index and academic
achievement? A meta-analysis. Public Health. 2019;167:111-124. doi:10.1016/j.puhe.2018.11.002 12. Yanfeng Li, Qi Dai, Jackson JC, Zhang J. Overweight is associated with decreased cognitive
functioning among school-age children and adolescents. Obesity. 2008;16(8):1809-1815. Page 11/18 13. Ter Bogt Tom FM, Van Dorsselaer SA, Monshouwer K, Verdurmen J, Engels R, Vollebergh W. Body
mass index and body weight perception as risk factors for internalizing and externalizing problem
behavior among adolescents. J Adolesc Health. 2006;39(1):27-34. 14. Roberts RE, Duong HT. Perceived weight, not obesity, increases risk for major depression among
adolescents. J Psychiatr Res. 2013;47(8):1110-1117. doi:10.1016/j.jpsychires.2013.03.019 15. Atlantis E, Ball K. Association between weight perception and psychological distress. Int J Obes. 2008;32(4):715-721. 16. Patte KA, Laxer RE, Qian W, Leatherdale ST. An analysis of weight perception and physical activity
and dietary behaviours among youth in the COMPASS study. SSM Popul Health 2016;2:841‐849. doi:10.1016/j.ssmph.2016.10.016 17. Neumark-Sztainer DR, Wall MM, Haines JI, Story MT, Sherwood NE, van den Berg PA. Shared risk and
protective factors for overweight and disordered eating in adolescents. Am J Prev Med. 2007;33(5):359‐369. doi:10.1016/j.amepre.2007.07.031 18. Khambalia A, Hardy LL, Bauman A. Accuracy of weight perception, life-style behaviours and
psychological distress among overweight and obese adolescents. J Paediatr Child Health
2012;48(3):220‐227. doi:10.1111/j.1440-1754.2011.02258.x 19. Osborn J, Naquin M, Gillan W, Bowers A. The impact of weight perception on the health behaviors of
college students. American Journal of Health Education 2016;47(5):287-298. doi:
10.1080/19325037.2016.1204966. 20. Gaylis JB, Levy SS, Hong MY. Relationships between body weight perception, body mass index,
physical activity, and food choices in Southern California male and female adolescents. International
Journal of Adolescence and Youth 2016;25;264-275, doi: 10.1080/02673843.2019.1614465 20. Gaylis JB, Levy SS, Hong MY. Relationships between body weight perception, body mass index,
physical activity, and food choices in Southern California male and female adolescents. International
Journal of Adolescence and Youth 2016;25;264-275, doi: 10.1080/02673843.2019.1614465 21. Schiefelbein EL, Mirchandani GG, George GC, Becker EA, Castrucci BC, Hoelscher DM. Association
between depressed mood and perceived weight in middle and high school age students: Texas 2004-
2005. Matern Child Health J 2012;16(1):169‐176. doi:10.1007/s10995-010-0733-1 21. Schiefelbein EL, Mirchandani GG, George GC, Becker EA, Castrucci BC, Hoelscher DM. Association
between depressed mood and perceived weight in middle and high school age students: Texas 2004-
2005. Matern Child Health J 2012;16(1):169‐176. doi:10.1007/s10995-010-0733-1 22. References Kim Y, Austin SB, Subramanian SV, Kawachi I. Body weight perception, disordered weight control
behaviors, and depressive symptoms among Korean adults: The Korea National Health and Nutrition
Examination Survey 2014. PLoS One 2014;3(6):e0198841. doi:10.1371/journal.pone.0198841 22. Kim Y, Austin SB, Subramanian SV, Kawachi I. Body weight perception, disordered weight control
behaviors, and depressive symptoms among Korean adults: The Korea National Health and Nutrition
Examination Survey 2014. PLoS One 2014;3(6):e0198841. doi:10.1371/journal.pone.0198841 23. Al Mamun A, Cramb S, McDermott BM, O'Callaghan M, Najman JM, Williams GM (2007). Adolescents' perceived weight associated with depression in young adulthood: a longitudinal study. Obesity 2007;15(12):3097‐3105. doi:10.1038/oby.2007.369 23. Al Mamun A, Cramb S, McDermott BM, O'Callaghan M, Najman JM, Williams GM (2007). Adolescents' perceived weight associated with depression in young adulthood: a longitudinal study. Obesity 2007;15(12):3097‐3105. doi:10.1038/oby.2007.369 24. Thurston IB, Sonneville KR, Milliren CE, Kamody RC, Gooding HC, Richmond TK. Cross-sectional and
prospective examination of weight misperception and depressive symptoms among youth with
overweight and obesity. Prev Sci 2007;18(2):152‐163. doi:10.1007/s11121-016-0714-8 24. Thurston IB, Sonneville KR, Milliren CE, Kamody RC, Gooding HC, Richmond TK. Cross-sectional and
prospective examination of weight misperception and depressive symptoms among youth with
overweight and obesity. Prev Sci 2007;18(2):152‐163. doi:10.1007/s11121-016-0714-8 25. Eaton DK, Lowry R, Brener ND, Galuska DA, Crosby AE. Associations of body mass index and
perceived weight with suicide ideation and suicide attempts among US high school students, Arch
Pediatr Adolesc Med. 2005;159(6):513‐519. doi:10.1001/archpedi.159.6.513 Page 12/18 Page 12/18 26. Farhat T, Iannotti RJ, Summersett-Ringgold F. Weight, weight perceptions, and health-related quality
of life among a national sample of us girls. J Dev Behav Pediatr. 2015;36(5):313‐323. doi:10.1097/DBP.0000000000000172 pp. 313–323, 2015. 27. Hayward J, Millar L, Petersen S, Swinburn B, Lewis AJ. When ignorance is bliss: weight perception,
body mass index and quality of life in adolescents. Int J Obes. 2014;38(10):1328‐1334. doi:10.1038/ijo.2014.78 28. Florin TA, Shults J, Stettler N. Perception of overweight is associated with poor academic
performance in US adolescents. J Sch Health. 2011;81(11):663-670. doi:10.1111/j.1746-
1561.2011.00642.x 29. McLaren L. Socioeconomic status and obesity. Epidemiol Rev. 2007;29(1):29-48. 30. Courser MW, Shamblen SR, Lavrakas PJ, Collins D, Ditterline P. The impact of active consent
procedures on nonresponse and nonresponse error in youth survey data: evidence from a new
experiment. Eval Rev. 2009;33(4):370-395. 31. Reel B, Bredin C, Leatherdale ST. COMPASS Year 5 and 6 School Recruitment and Retention:
Technical Report Series. 2018;5(1): Waterloo, Ontario: University of Waterloo. 32. Leatherdale ST, Brown KS, Carson V, et al. References The COMPASS study: a longitudinal hierarchical research
platform for evaluating natural experiments related to changes in school-level programs, policies and
built environment resources. BMC Public Health. 2014;14:331. doi:10.1186/1471-2458-14-331 33. Leatherdale ST, Laxer RE. Reliability and validity of the weight status and dietary intake measures in
the COMPASS questionnaire: are the self-reported measures of body mass index (BMI) and Canada’s
food guide servings robust? Int J Behav Nutr Phys Act. 2013.10(1):42. doi:10.1186/1479-5868-10-42 33. Leatherdale ST, Laxer RE. Reliability and validity of the weight status and dietary intake measures in
the COMPASS questionnaire: are the self-reported measures of body mass index (BMI) and Canada’s
food guide servings robust? Int J Behav Nutr Phys Act. 2013.10(1):42. doi:10.1186/1479-5868-10-42 34. Growth reference data for 5-19 years. https://www.who.int/growthref/en/. Accessed August 20,
2019. 35. Aceves-Martins M, Whitehead R, Inchley J, Giralt M, Currie C, Solà R. Self-reported weight and
predictors of missing responses in youth. Nutrition. 2018;53:54-58. doi:10.1016/j.nut.2018.01.003 35. Aceves-Martins M, Whitehead R, Inchley J, Giralt M, Currie C, Solà R. Self-reported weight and
predictors of missing responses in youth. Nutrition. 2018;53:54-58. doi:10.1016/j.nut.2018.01.003 36. R Core Team, 2019. R: A Language and Environment for Statistical Computing. http://www.R-
project.org/. 36. R Core Team, 2019. R: A Language and Environment for Statistical Computing. http://www.R-
project.org/. 37. Deighton J, Humphrey N, Belsky J, Boehnke J, Vostanis P, Patalay P. Longitudinal pathways between
mental health difficulties and academic performance during middle childhood and early
adolescence. Br J Dev Psychol. 2018;36(1):110-126. 37. Deighton J, Humphrey N, Belsky J, Boehnke J, Vostanis P, Patalay P. Longitudinal pathways between
mental health difficulties and academic performance during middle childhood and early
adolescence. Br J Dev Psychol. 2018;36(1):110-126. 38. Jansen W, van de Looij-Jansen PM, de Wilde EJ, Brug J. Feeling fat rather than being fat may be
associated with psychological well-being in young Dutch adolescents. J Adolesc Health. 2018;42(2):128-136. 38. Jansen W, van de Looij-Jansen PM, de Wilde EJ, Brug J. Feeling fat rather than being fat may be
associated with psychological well-being in young Dutch adolescents. J Adolesc Health. 2018;42(2):128-136. 39. Crosnoe R. Gender, obesity, and education. Sociol Educ 2007;80(3):241-260. 39. Crosnoe R. Gender, obesity, and education. Sociol Educ 2007;80(3):241-260. 40. Paxton SJ, Neumark-Sztainer D, Hannan PJ, Eisenberg ME. Body dissatisfaction prospectively
predicts depressive mood and low self-esteem in adolescent girls and boys. J Clin Child Adolesc
Psychol. 2006;35(4):539-549. Page 13/18 Page 13/18 41. Blashill AJ, Wilhelm S. References Body image distortions, weight, and depression in adolescent boys:
longitudinal trajectories into adulthood. Psychology of Men & Masculinity. 2014;15;4:445–451. doi:10.1037/a0034618 42. Isomaa R, Isomaa AL, Marttunen M, Kaltiala-Heino R, Björkqvist Longitudinal concomitants of
incorrect weight perception in female and male adolescents. Body Image. 2011;8(1):58–63. doi:10.1016/j.bodyim.2010.11.005 43. Pont SJ, Puhl R, Cook SR, Slusser W. Stigma experienced by children and adolescents with obesity. Pediatrics 2017;140(6):1-11. doi:10.1542/peds.2017-3034 44. Pearl RL, Puhl RM. Weight bias internalization and health: a systematic review. Obes Rev
2018;19(8):1141. doi:10.1111/obr.12701 45. Brener ND, Mcmanus T, Galuska DA, Lowry R, Wechsler H. Reliability and validity of self-reported
height and weight among high school students. J Adolesc Health 2003;2(4):281-287. doi:10.1016/S1054-139X(02)00708-5 46. Morrison TG, Kalin R, Morrison MA. Body-image evaluation and body-image investment among
adolescents: a test of sociocultural and social comparison theories. Adolescence. 2004; 39(155):571-
592. 47. Datar A, Sturm R, Magnabosco JL. Childhood overweight and academic performance: national study
of kindergartners and first graders. Obesity. 2004;12(1):58. 48. Kuncel NR, Credé M, Thomas LL. The validity of self-reported grade point averages, class ranks, and
test scores: a meta-analysis and review of the literature. Rev Educ Res. 2005;75(1):63. 49. Goodman E, Hinden BR, Khandelwal S. Accuracy of teen and parental reports of obesity and body
mass index. Pediatrics 2000;106(1):52-58. 50. Dix KL, Slee PT, Lawson MJ, Keeves JP. Implementation quality of whole-school mental health
promotion and students’ academic performance. Child Adolesc Ment Health. 2012;17(1):45-51. 51. Troiano RP, Flegal KM. Overweight children and adolescents: description, epidemiology, and
demographics. Pediatrics. 1998;101(3):497-504. 52. The heavy burden of obesity: the economics of prevention. 2019. https://doi.org/10.1787/67450d67-
en . Accessed October 18, 2019. 53. Rupp K, McCoy SM. Bullying perpetration and victimization among adolescents with overweight and
obesity in a nationally representative sample. Child Obes. 2019;15(5):323-330. doi:10.1089/chi.2018.0233 54. Peterson JL, Puhl RM, Luedicke J. An experimental assessment of physical educators’ expectations
and attitudes: the importance of student weight and gender. J Sch Health 2012;82(9):432-440. doi:10.1111/j.1746-1561.2012.00719.x 55. Patte KA, Qian W, Leatherdale ST. Predictors of one-year change in how youth perceive their weight. Journal of Obesity. 2020;7396948. doi:10.1155/2020/7396948 Page 14/18 Table 2. Mixed-effect model testing BMI category and weight perception as predictors of
Math grades ≥60% among youth Tables Table 1: Descriptive statistics for secondary school students in Year 6 (2017/2018) of the
COMPASS study (N=61,886) Table 1: Descriptive statistics for secondary school students in Year 6 (2017/2018) of the
COMPASS study (N=61,886)
Female
(n = 31,334)
Male
(n =30,552)
% (n)
% (n)
Grade
9
10
11
12
Othera
24.3 (7614)
24.5 (7677)
23.3 (7301)
14.9 (4669)
13.0 (4073)
24 (7332)
24.3 (7424)
23.1 (7058)
15.7 (4797)
12.9 (3941)
Race/ethnicity
White
Non-white, multiethnic, or other
66.7 (20889)
33.3 (10445)
66.6 (20355)
33.4 (10197)
BMI Classification
Underweight
“Normal Weight”
Overweight
Obesity
Missing weight
Missing age, sex, or height
1.5 (470)
56.8 (17798)
10.3 (3227)
4.1 (1285)
21.7 (6799)
5.6 (1755)
2.1 (642)
50.6 (15459)
13.3 (4063)
7.4 (2261)
19.4 (5927)
7.2 (2200)
Weight Perception
Underweight
“About the right weight”
Overweight
Missing
11.1 (3478)
60.7 (19019)
26.5 (8304)
1.7 (533)
21.2 (6477)
56.2 (17170)
20.6 (6294)
2.0 (611)
Weekly Spending Money
None
$1-$20
14.2 (4449)
25.9 (8116)
17.9 (5469)
24.7 (7546)
$21-$100
>$100
25.5 (7990)
16.9 (5295)
22.0 (6721)
20.4 (6233)
Don’t know
17.5 (5484)
15.0 (4583)
Math Grades
60-100%
0-59%
51.0 (15980)
49.0 (15354)
51.2 (15643)
48.8 (14909)
English/French Grades
60-100%
0-59%
51.3 (16074)
48.7 (15260)
51.1 (15612)
48.9 (14940)
a Secondary I-II in Quebec schools. BMI = body mass index. Table 1: Descriptive statistics for secondary school students in Year 6 (2017/2018) of the
COMPASS study (N=61,886) a Secondary I-II in Quebec schools. BMI = body mass index. a Secondary I-II in Quebec schools. BMI = body mass index. Table 2. Tables Mixed-effect model testing BMI category and weight perception as predictors of
Math grades ≥60% among youth Page 15/18 Females (N=31,334)
Males (N=30,552)
AOR
95% CI
AOR
95% CI
BMI Category (Reference: “Normal weight”)
Obesity
0.70**
0.59-0.82
0.62**
0.55-0.71
Overweight
0.90
0.80-1.02
0.87*
0.79-0.97
Underweight
0.93
0.70-1.25
0.85
0.68-1.07
Missing weight data
0.56**
0.51-0.61
0.62**
0.57-0.68
Missing sex, age, or height data
0.64**
0.55-0.74
0.65**
0.57-0.74
Weight Perception (Reference: “About the right weight”)
Overweight
0.82**
0.75-
0.89
0.91
0.83-
1.00
Underweight
0.77**
0.69-
0.86
0.90*
0.82-
0.97
Missing weight perception
Race/ethnicity (Reference: White)
Nonwhite, multiethnic, or other
Grade (Reference: 9)
10
11
12
Othera
Weekly Spending Money (Reference:
None)
$1-$20
$21-$100
>$100
Don’t know
0.64**
0.92
0.82**
0.96
0.91
1.10
0.94
0.94
0.94
1.05
0.50-
0.83
0.85-
1.00
0.74-
0.90
0.87-
1.07
0.81-
1.02
0.95-
1.26
0.84-
1.05
0.84-
1.05
0.83-
1.07
0.93-
1.19
0.59**
0.87**
0.80**
0.83**
0.88
1.14
0.93
0.89
0.86*
1.02
0.47-
0.73
0.81-
0.94
0.73-
0.88
0.75-
0.91
0.79-
0.99
0.99-
1.31
0.84-
1.02
0.80-
0.99
0.77-
0.95
0.91-
1.14 Note: Cross-sectional data from Year 6 (2017/2018) of the COMPASS study (N=61,886). The outcome reference
category is math grades <60%. Mixed models were used to account for school clustering by adding a random
intercept at the school level. * p < .01; ** p < .001 * p < .01; ** p < .001 a Secondary I-II in Quebec schools. Table 3. Mixed-effect model testing BMI category and weight perception as predictors of
English/French grades ≥60% among youth Page 16/18 Page 16/18 Females (N=31,334)
Males (N=30,552)
AOR
95% CI
AOR
95% CI
BMI Category (Reference: “Normal weight”)
Obesity
0.51**
0.41-0.64
0.63**
0.55-0.73
Overweight
0.71**
0.60-0.85
0.87
0.77-0.99
Underweight
1.04
0.68-1.59
0.93
0.71-1.21
Missing weight data
0.45**
0.40-0.50
0.54**
0.49-0.60
Missing sex, age, or height data
0.49**
0.40-0.59
0.58**
0.51-0.66
Weight Perception (Reference: “About the right weight”)
Overweight
0.76**
0.68-0.85
0.83**
0.75-0.92
Underweight
0.59**
0.51-0.69
0.87*
0.79-0.95
Missing weight perception
Race/ethnicity (Reference: White)
Nonwhite, multiethnic, or other
Grade (Reference: 9)
10
11
12
Othera
Weekly Spending Money (Reference: None)
$1-$20
$21-$100
>$100
Don’t know
0.50**
0.67**
0.96
0.87
1.07
1.19
0.88
0.95
0.88
0.99
0.36-0.68
0.60-0.75
0.83-1.10
0.75-1.00
0.90-1.27
0.98-1.44
0.75-1.03
0.80-1.11
0.74-1.06
0.84-1.18
0.66**
0.74**
0.94
0.94
1.06
1.12
1.08
1.05
0.93
1.09
0.51-0.84
0.68-0.80
0.85-1.05
0.84-1.05
0.93-1.21
0.97-1.29
0.97-1.21
0.93-1.18
0.83-1.05
0.96-1.23 Note: Cross-sectional data from Year 6 (2017/2018) of the COMPASS study (N=61,886). The outcome reference
category is English/French grades <60%. Mixed models were used to account for school clustering by adding a
random intercept at the school level.
* p < 01 ** p < 001 a Secondary I-II in Quebec schools. Figures Page 17/18 Figure 1
Weight perception and BMI category concordance among female and male secondary school students
participating in Year 6 (2017/2018) of the COMPASS study Figure 1 Weight perception and BMI category concordance among female and male secondary school students
participating in Year 6 (2017/2018) of the COMPASS study Weight perception and BMI category concordance among female and male secondary school students
participating in Year 6 (2017/2018) of the COMPASS study Page 18/18
|
https://openalex.org/W3004647783
|
https://link.springer.com/content/pdf/10.1007/s11665-020-04605-3.pdf
|
English
| null |
Microstructural Changes in Inconel 625 Alloy Fabricated by Laser-Based Powder Bed Fusion Process and Subjected to High-Temperature Annealing
|
Journal of materials engineering and performance
| 2,020
|
cc-by
| 5,070
|
Microstructural Changes in Inconel 625 Alloy Fabricated
by Laser-Based Powder Bed Fusion Process
and Subjected to High-Temperature Annealing
Kewin Gola, Beata Dubiel, and Izabela Kalemba-Rec Keywords
additive manufacturing, high-temperature annealing,
Inconel 625, laser powder bed fusion, microstructure Keywords
additive manufacturing, high-temperature annealing,
Inconel 625, laser powder bed fusion, microstructure 1. Introduction Inconel 625 is a nickel-based superalloy exhibiting high-
temperature strength, good corrosion resistance in several
aggressive environments and good weldability (Ref 1-3). Due
to its specific properties, it is widely used in different
applications, e.g., in automotive, aeronautics and chemical
industries. However,
like
other
nickel-based
superalloys,
Inconel 625 is difficult to machine. Therefore, additive
manufacturing (AM) of Inconel 625 is an attractive process
for fabrication of parts with complex geometry (Ref 4). One of
the mostly applied AM processes, which is used to make
Inconel 625 parts, is laser-based powder bed fusion (L-PBF), in
which the layers of powder are successively spread on the
platform and selectively fused by a laser beam scanning along
the predefined path. The process is repeating until the desired
shape is obtained. Due to the rapid melting followed by the
rapid cooling during the L-PBF processes, the thin layers of
melted powder undergo rapid dendritic solidification and thus,
in the post-built condition, Inconel 625 exhibits fine cellular–
dendritic microstructure (Ref 5). To reduce residual stresses
and thus avoid distortions before cutting built parts from the Therefore, the aim of the present study is characterization of
the microstructural changes that occur at micron-/submicron-
and nanoscale in the L-PBF Inconel 625 subjected to post-build
stress-relief annealing at a temperature of 980 C for 1 h, as
well as after subsequent annealing at a temperature of 600 C
for 5, 100 and 500 h. This article is an invited submission to JMEP selected from
presentations at The XXII Physical Metallurgy and Materials Science
Conference: Advanced Materials and Technologies (AMT 2019) held
June 9-12, 2019, in Bukowina Tatrzan´ska, Poland, and has been
expanded from the original presentation. JMEPEG (2020) 29:1528–1534
https://doi.org/10.1007/s11665-020-04605-3 JMEPEG (2020) 29:1528–1534
https://doi.org/10.1007/s11665-020-04605-3 The Author(s)
1059-9495/$19.00 Microstructural Changes in Inconel 625 Alloy Fabricated
by Laser-Based Powder Bed Fusion Process
and Subjected to High-Temperature Annealing
Kewin Gola, Beata Dubiel, and Izabela Kalemba-Rec g
p
Kewin Gola, Beata Dubiel, and Izabela Kalemba-Rec (Submitted August 30, 2019; in revised form January 14, 2020; published online February 3, 2020) The microstructure of the laser powder bed additively manufactured Inconel 625 in post-build stress-relief
annealed condition and subsequently annealed at a temperature of 600 C for 5, 100 and 500 h was
investigated by means of light microscopy as well as scanning and transmission electron microscopy. Stress-
relieved Inconel 625 exhibited fine cellular–dendritic microstructure characterized by the high dislocation
density. Selected area electron diffraction studies allowed to demonstrate that already after 5 h of annealing
at a temperature of 600 C, precipitation of the c¢¢ phase in the form of coherent nanoparticles occurred. With the prolongation of the duration of annealing at a temperature of 600 C from 5 to 500 h, the gradual
dissolution of intercellular areas and decrease in the dislocation density were accompanied by enhanced
precipitation of the c¢¢ phase nanoparticles. building platform, routinely stress-relief annealing is performed
at a temperature of 870 C for 1 h (Ref 5-7). Examination of
the microstructure of L-PBF-processed Inconel 625 in as-built
and stress-relief annealed condition was the subject of several
reports (Ref 5, 7-12). The microstructure after post-build
annealing at a temperature of 980 and 1150 C was also
examined (Ref 13). It was revealed that in the as-built
condition, the significant segregation of Ni and Cr to dendrite
cores as well as enrichment of the interdendritic regions in Nb
and Mo occurs (Ref 8, 9, 12, 13). Along the melt pool
boundaries, precipitates of the c¢¢ phase rich in Nb were
observed (Ref 10). Furthermore, in the stress-relieved condi-
tion, the presence of the c¢¢ phase, d phase as well as MC, M6C
and M23C6 carbides was reported (Ref 7, 11, 12). However, to
the knowledge of the authors, there is limited information about
the microstructure of L-PBF Inconel 625 subjected to the post-
build annealing at a temperature higher than the routinely
applied stress annealing. Moreover, there are only a few reports
on the evolution of the microstructure during the subsequent
exposure at a temperature of 600 C (Ref 9), which is the
working temperature of Inconel 625 in the chemical industry,
waste processing and energy systems applications (Ref 14-16). 2. Materials and Methods The specimens of Inconel 625 delivered by Bibus Menos
Sp. z o.o. (Poland) were fabricated by Direct Metal Laser
Sintering (DMLS), being the commercially available L-PBF
process of EOS GmbH (Germany). The chemical composition
of the alloy is as follows (wt.%): 63.9Ni-21.5Cr-1.3Fe-8.1Mo-
3.9Nb-0.5Si-0.2Al-0.2Ti-0.2Co-0.1C-0.1Mn. The
fabricated Kewin Gola, Beata Dubiel, and Izabela Kalemba-Rec, Faculty of
Metals Engineering and Industrial Computer Science, AGH University
of Science and Technology, al. Mickiewicza 30, 30-059 Krako´w,
Poland. Contact e-mail: kgola@agh.edu.pl. Journal of Materials Engineering and Performance 1528—Volume 29(3) March 2020 pools forming a so-called fish scale are visible. Figure 1(d)
displays the layers of deposited material observed in the
horizontal plane. The angle between the laser paths in the
successive layers is equal to 67, according to the scan strategy
applied by EOS in DMLS process. SEM observations revealed
the cellular–dendritic microstructure within the layers (Fig. 1c
and e), which is typical for L-PBF-processed alloys. The width
of cells was in the range of 300-400 nm. EDS microanalysis
revealed the preferred segregation of Nb and Mo to the cell
walls. The average concentration of Nb and Mo, determined
from measurements at 20 points, was equal to 1.2% Nb and
4.0% Mo inside the cells; meanwhile, at the cell walls, it
amounted to 5.8% Nb and 6.2% Mo. In both the vertical and
horizontal planes, besides the cellular forms, the elongated
columnar cells and fine dendrites with very short secondary
arms were present (Fig. 1c and e). Such elongated cells and
dendrites demonstrated epitaxial growth over melt pool bound-
aries. This shows that the re-melting of the underlying layers
during subsequent laser passes allowed for the restart of the
grain growth process in the next layers (Ref 18). In the
horizontal plane in the vicinity of melt pool boundaries, also the
coarser cellular regions occurred (Fig. 1e). Kruth et al. (Ref 19)
postulated that in L-PBF-processed alloys, such regions are not
overlapped with alternating layers and thus not re-melted by
laser, so in effect are less heat-affected, which leads to a coarse
microstructure. Similar findings have been reported in the laser-
assisted direct energy deposited Inconel 625 (Ref 20). 2. Materials and Methods TEM
microstructural analysis showed the high dislocation density
inside cells and splits of dislocations within the cell walls specimens in cuboidal shape 15 mm 9 15 mm 9 15 mm in
size were subjected to the post-build annealing at a temperature
of 980 C for 1 h and slowly cooled in an argon atmosphere. The purpose of this treatment was stress relieving and also
partial solution annealing. Subsequently, the specimens were
isothermally annealed at a temperature of 600 C for 5, 100 and
500 h and cooled in air. The microstructural investigation was
performed by means of light microscopy (LM), scanning and
transmission electron microscopy (SEM, TEM). The LM and
SEM analyses of polished sections were carried out using Axio
Imager M1m (ZEISS, Germany) and Nova NanoSEM 450
(FEI, USA) microscopes. Thin foils for TEM were prepared by
electropolishing. The TEM investigation was performed using
JEM-2010 ARP microscope (Jeol, Japan) equipped with INCA
EDS system (Oxford Instruments, UK). Selected area electron
diffraction (SAED) patterns were analyzed with the use of
JEMS v4.4230 software (Pierre Stadelmann, JEMS-SAS,
Switzerland). Dislocation density was estimated by the Ham
line-intercept method using TEM bright-field (BF) images
acquired in two-beam conditions (Ref 17). Journal of Materials Engineering and Performance 3. Results and Discussion The Ni content in the matrix was about
66.6%, while in the precipitates was in the range from 59.0 to
70.4%. In the case of Nb and Mo, in some particles, the
concentration of these elements was at a similar level as in the
matrix (e.g., precipitates no 1-3 in Table 1), while in most of
the analyzed particles, a significant enrichment in Nb and Mo
was observed. (Fig. 2a and b). The estimated dislocation density was equal to
3.65 9 1014 m2. This observation indicates that the annealing
at a temperature of 980 C for 1 h is not sufficient to relieve the
residual stresses in L-PBF Inconel 625. In the cell walls,
precipitates of the secondary phases were observed (Fig. 2c). Such particles were of globular or irregular shapes and up to
100 nm in size. Figure 2(d) shows the EDS spectrum of the
particle with globular morphology marked in Fig. 2(c) by an
arrow. Results of quantitative EDS microanalysis of the matrix
and exemplary precipitates are given in Table 1. The concen-
tration of Si, Cr and Fe in individual particles varied in a similar
range as in the matrix. The Ni content in the matrix was about
66.6%, while in the precipitates was in the range from 59.0 to
70.4%. In the case of Nb and Mo, in some particles, the
concentration of these elements was at a similar level as in the
matrix (e.g., precipitates no 1-3 in Table 1), while in most of
the analyzed particles, a significant enrichment in Nb and Mo
was observed. possible phases; thus, the undoubted phase identification was
not possible. In many previous studies, it was demonstrated that
Inconel 625 exhibits a complex precipitation behavior, which
varies depending on the production method, e.g., wrought,
powder metallurgy, welding, laser cladding or cold metal
transfer cladding (Ref 1, 2, 20-30). It has been shown that at a
temperature range of 550-900 C, precipitates of the c¢, c¢¢, d
and Ni2(Cr,Mo) Laves phases, as well as MC, M6C and M23C6
carbides, can be present in Inconel 625 fabricated using
different methods. The further studies are needed to determine
which exact phases are present in stress-relief annealed L-PBF
Inconel 625. Figure 3(a), (b), (c) and (d) shows the typical microstructure
of the L-PBF Inconel 625 specimen annealed at a temperature
of 600 C for 5 h observed on LM, SEM and TEM images. 3. Results and Discussion Figure 1(a) shows a composite of LM images corresponding
to horizontal (XY) and vertical (XZ and YZ) sections of the
stress-relief annealed specimen. The built direction is marked
by an arrow. In the vertical plane (Fig. 1b), the profiles of melt Fig. 1
Microstructure of the L-PBF Inconel 625 after stress-relief annealing at a temperature of 980 C for 1 h: (a) a composite of LM images
corresponding to vertical (XZ, YZ) and horizontal (XY) sections of the examined specimen, as well as LM and SEM images of the (b, c)
vertical and (d, e) horizontal sections Fig. 1
Microstructure of the L-PBF Inconel 625 after stress-relief annealing at a temperature of 980 C for 1 h: (a) a composite of LM images
corresponding to vertical (XZ, YZ) and horizontal (XY) sections of the examined specimen, as well as LM and SEM images of the (b, c)
vertical and (d, e) horizontal sections Volume 29(3) March 2020—1529 Fig. 2
TEM images of the (a) dislocations arrangements in several neighboring cells and (b) in the enlarged view in the individual cell, (c)
precipitates in the L-PBF Inconel 625 after stress-relief annealing at a temperature of 980 C for 1 h and (d) EDS spectrum of the particle with
globular morphology marked in (c) by arrow Fig. 2
TEM images of the (a) dislocations arrangements in several neighboring cells and (b) in the enlarged view in the individual cell, (c)
precipitates in the L-PBF Inconel 625 after stress-relief annealing at a temperature of 980 C for 1 h and (d) EDS spectrum of the particle with
globular morphology marked in (c) by arrow (Fig. 2a and b). The estimated dislocation density was equal to
3.65 9 1014 m2. This observation indicates that the annealing
at a temperature of 980 C for 1 h is not sufficient to relieve the
residual stresses in L-PBF Inconel 625. In the cell walls,
precipitates of the secondary phases were observed (Fig. 2c). Such particles were of globular or irregular shapes and up to
100 nm in size. Figure 2(d) shows the EDS spectrum of the
particle with globular morphology marked in Fig. 2(c) by an
arrow. Results of quantitative EDS microanalysis of the matrix
and exemplary precipitates are given in Table 1. The concen-
tration of Si, Cr and Fe in individual particles varied in a similar
range as in the matrix. Journal of Materials Engineering and Performance 3. Results and Discussion The
LM and SEM images did not reveal the pronounced changes in
the microstructure, as compared with the stress-relief annealed
condition. However, the dislocation density decreased to
2.77 9 1014 m2 (Fig. 3c). Besides the precipitates of globular
and irregular shapes located in the cell walls, the elongated
nanoparticles with oscillating contrast in TEM bright-field
images were observed. Examples of such precipitates are
marked by arrows in Fig. 3(d). SAED patterns allowed to
identify them as the precipitates of the c¢¢ phase. Figure 3(e) The trials of phase identification using SAED patterns
revealed that the examined particles may be the precipitates of
the c¢¢ phase (tetragonal, I4/mmm), d phase (orthorhombic,
Pmmm), P phase (orthorhombic, Pnma), Laves phase (hexag-
onal, P63mmc) or the M6C carbides (cubic, Fd3m). However,
the solutions of the diffraction patterns provided several 1530—Volume 29(3) March 2020 Table 1
Chemical composition of the matrix and precipitates in the stress-relieved Inconel 625 Elements concentration, at.% Element
Elements concentration, at.%
Matrix
Precipitates
1
2
3
1
2
3
4
5
6
7
8
9
10
Si
3.3
4.3
6.2
3.0
3.7
0.4
2.6
3.0
3.6
4.5
5.5
4.7
3.4
Cr
22.6
21.7
20.9
21.4
22.0
23.0
22.5
22.9
20.8
21.9
21.5
21.6
20.0
Fe
1.9
2.0
1.3
1.9
1.9
1.7
1.7
1.8
1.9
1.7
1.2
1.2
1.0
Ni
66.4
66.6
66.7
65.5
65.3
70.4
63.1
64.8
61.2
63.0
59.0
53.1
61.4
Nb
1.7
1.3
1.2
1.5
1.7
1.1
4.4
2.6
3.4
3.8
5.0
9.1
7.2
Mo
4.1
4.1
3.7
3.9
4.3
3.5
5.7
5.1
4.6
5.2
7.8
10.4
7.0 Fig. 3
Microstructure of the L-PBF Inconel 625 after stress-relief annealing at a temperature of 980 C for 1 h, followed by annealing at a
temperature of 600 C for 5 h: (a) LM image, (b) SEM image, as well as TEM images of (c) dislocations arrangements, (d) precipitates of the
c¢¢ phase (arrowed), (e) [100]c zone axis SAED pattern with the c¢¢ phase diffraction spots of the [100]c¢¢ and [010]c¢¢ zone axes Fig. Journal of Materials Engineering and Performance 3. Results and Discussion It should be pointed out that prolongation
of the annealing at a temperature of 600 C to 100 h did not
result in intensive precipitation of the d phase, as it was shown
in other studies even already after stress relieving at a
temperature of 870 C for 1 h (Ref 12). cooling rates in additive manufacturing are much higher, which
results in a very fine and highly segregated microstructure. Our
observations indicate that such refinement in L-PBF Inconel
625 contributes to precipitation of the c¢¢ phase at lower
temperature and after a shorter time than in heat-treated welds. Prolongation of the annealing time up to 100 h resulted in the
destabilization of the cellular microstructure manifested by
breaks in the continuity of the cell walls, as observed in LM
images (Fig. 4a) and more clearly in SEM images (Fig. 4b). Due to the intensification of recovery processes, dislocation
density inside the cells further decreased and was equal to
2.55 9 1014 m2 (Fig. 4c). Moreover, the c¢¢ phase particles
grew and precipitated in all three sets of the {100} planes of the
c phase matrix, what was demonstrated by the presence of the
three sets of diffraction spots form the [100], [010] and [001]
zone axes of the c¢¢ phase (Fig. 4d and e). Streaking of the c¢¢
diffraction spots in the principal cubic directions is related to
the plate-like shape of the particles precipitated on the {100} c
phase planes. EDS spectra acquired in the locations of the c¢¢
phase precipitates revealed the presence of the strong Ni, Cr,
Mo and Nb peaks (Fig. 4f). The c¢¢ phase nanoparticles were
below 100 nm in size, while the foil thickness was about p
(
)
After annealing for 500 h at a temperature of 600 C, the
progress in homogenizing of the cellular microstructure was
insignificant, as compared with the observations of the sample
annealed for 100 h (Fig. 5a and b). In turn, dislocation density
decreased to 1.68 9 1014 m2 (Fig. 5c). Pronounced intensi-
fication of the c¢¢ phase precipitation was also noticed. Densely
arranged precipitates were visible in TEM bright-field images
(Fig. 5d) as well as the dark-field images recorded in the [002]
spot of the c¢¢ phase (Fig. 5e and the corresponding SAED
pattern in Fig. 5f). 3. Results and Discussion 3
Microstructure of the L-PBF Inconel 625 after stress-relief annealing at a temperature of 980 C for 1 h, followed by annealing at a
temperature of 600 C for 5 h: (a) LM image, (b) SEM image, as well as TEM images of (c) dislocations arrangements, (d) precipitates of the
c¢¢ phase (arrowed), (e) [100]c zone axis SAED pattern with the c¢¢ phase diffraction spots of the [100]c¢¢ and [010]c¢¢ zone axes at a temperature of 870 C for 1 h (Ref 7). Our studies did not
confirm those findings for the specimen stress relieved at the
higher temperature of 980 C, whereas it was demonstrated that
precipitation of the c¢¢ phase occurs after subsequent annealing
at a temperature of 600 C for 5 h in the whole volume of the
L-PBF Inconel 625. Precipitation of the c¢¢ phase was also
observed in wrought Inconel 625 long-time service exposed
and subsequently annealed at the temperature of 650 C for 1 h
(Ref 2) as well as in TIG and GTA welds heat treated at 700 C
for 8 h (Ref 24) and 750 C for 4 h (Ref 25), respectively. Although welding and L-PBF processes seem similar, the shows the electron diffraction pattern of the matrix c phase
[100] zone axis with the lower intensity spots of the c¢¢ phase
[100] and [010] zone axes, which confirms that the precipitates
lie on the {100} planes of the c matrix. Body centered
tetragonal Ni3Nb-based c¢¢ phase exhibits ordered structure and
precipitates from the c solid solution in the form of coherent
precipitates (Ref 31). Precipitates of the c¢¢ phase were
observed in the thin foils prepared from the sections parallel
to the horizontal (XY) and vertical (XZ and YZ) sections of the
examined specimen. The evidence of the c¢¢ phase was reported
also in L-PBF Inconel 625 subjected to stress-relief annealing Volume 29(3) March 2020—1531 210 nm; thus, this result represents the sum of the c and c¢¢
phase compositions. It should be pointed out that prolongation
of the annealing at a temperature of 600 C to 100 h did not
result in intensive precipitation of the d phase, as it was shown
in other studies even already after stress relieving at a
temperature of 870 C for 1 h (Ref 12). 210 nm; thus, this result represents the sum of the c and c¢¢
phase compositions. Journal of Materials Engineering and Performance 3. Results and Discussion The density of the c¢¢ phase precipitates was
higher in the cell walls areas than in the cell interiors. This
observation can be correlated with the results of the EDS
microanalysis of the initial stress-relief condition showing the
high degree of Nb microsegregation to the cell walls, which
creates the more favorable conditions for precipitation of the c¢¢
phase. Fig. 4
Microstructure of the L-PBF Inconel 625 after stress-relief annealing at a temperature of 980 C for 1 h, followed by annealing at a
temperature of 600 C for 100 h: (a) LM image, (b) SEM image, as well as TEM images of (c) dislocations arrangements, (d) c¢¢ phase
precipitates marked by arrows, (e) [100]c zone axis SAED pattern with the c¢¢ phase diffraction spots of the [100]c¢¢, [010]c¢¢ and [001] c¢¢ zone
axes, (f) EDS spectrum from the area with the c¢¢ precipitate Fig. 4
Microstructure of the L-PBF Inconel 625 after stress-relief annealing at a temperature of 980 C for 1 h, followed by annealing at a
temperature of 600 C for 100 h: (a) LM image, (b) SEM image, as well as TEM images of (c) dislocations arrangements, (d) c¢¢ phase
precipitates marked by arrows, (e) [100]c zone axis SAED pattern with the c¢¢ phase diffraction spots of the [100]c¢¢, [010]c¢¢ and [001] c¢¢ zone
axes, (f) EDS spectrum from the area with the c¢¢ precipitate 1532—Volume 29(3) March 2020 Journal of Materials Engineering and Performance Fig. 5
Microstructure of the L-PBF Inconel 625 after stress-relief annealing at a temperature of 980 C for 1 h, followed by annealing at a
temperature of 600 C for 500 h: (a) LM image, (b) SEM image as well as TEM images of (c) of dislocations arrangements, (d, e) precipitates
of the c¢¢ phase in bright-field and dark-field in the [002]c¢¢ spot, (f) [100]c zone axis SAED pattern with the c¢¢ phase diffraction spots of the
[100]c¢¢, [010]c¢¢ and [001] c¢¢ zone axes; the [002]c¢¢ spot is highlighted in the bold circle Fig. 3. Results and Discussion 5
Microstructure of the L-PBF Inconel 625 after stress-relief annealing at a temperature of 980 C for 1 h, followed by annealing at a
temperature of 600 C for 500 h: (a) LM image, (b) SEM image as well as TEM images of (c) of dislocations arrangements, (d, e) precipitates
of the c¢¢ phase in bright-field and dark-field in the [002]c¢¢ spot, (f) [100]c zone axis SAED pattern with the c¢¢ phase diffraction spots of the
[100]c¢¢, [010]c¢¢ and [001] c¢¢ zone axes; the [002]c¢¢ spot is highlighted in the bold circle References 1. S. Floreen, G.E. Fuchs, and W.J. Yang, Metallurgy of alloy 625,
Superalloys 718, 625 and Various Derivatives, E.A. Loria, Ed., The
Minerals, Metals and Materials Society, Warrendale, 1994, p 13–37 4. Conclusions distribution and reproduction in any medium or format, as long as
you give appropriate credit to the original author(s) and the source,
provide a link to the Creative Commons licence, and indicate if
changes were made. The images or other third party material in this
article are included in the article’s Creative Commons licence,
unless indicated otherwise in a credit line to the material. If
material is not included in the article’s Creative Commons licence
and your intended use is not permitted by statutory regulation or
exceeds the permitted use, you will need to obtain permission
directly from the copyright holder. To view a copy of this licence,
visit http://creativecommons.org/licenses/by/4.0/. 1. The fine cellular–dendritic microstructure of L-PBF In-
conel 625 subjected to stress-relief annealing at a temper-
ature of 980 C for 1 h is characterized by high
dislocation density. y
2. Precipitates of the c¢¢ phase were found in L-PBF Inconel
625 annealed at a temperature of 600 C for 5, 100 and
500 h. The plate-shaped nanoparticles of the c¢¢ phase
precipitated on the {100} planes of the c phase matrix. 3. With the prolongation of annealing duration at a tempera-
ture of 600 C, the gradual dissolution of intercellular
areas and the decrease in the dislocation density was ob-
served. Acknowledgments 2. V. Shankar, K.B.S. Rao, and S.L. Mannan, Microstructure and
Mechanical Properties of Inconel 625 Superalloy, J. Nucl. Mater.,
2001, 228, p 222–232 This research was funded by the National Science Centre,
Poland, Grant No. 2017/27/B/ST8/02244. 3. L.E. Shoemaker, Alloys 625 and 725: Trends in Properties and
Applications, Superalloys 718, 625 and Various Derivatives, E.A. Loria, Ed., The Minerals, Metals and Materials Society, Warrendale,
2005, p 409–418 Journal of Materials Engineering and Performance This research was funded by the National Science Centre,
Poland, Grant No. 2017/27/B/ST8/02244. Open Access 4. J.A. Gonzalez, J. Mireles, S.W. Stafford, M.A. Perez, C.A. Terrazas,
and R.B. Wicker, Characterization of Inconel 625 Fabricated Using
Powder-Bed-Based Additive Manufacturing Technologies, J. Mater. Process. Technol., 2018, 264, p 200–210 This article is licensed under a Creative Commons Attribution 4.0
International License, which permits use, sharing, adaptation, Volume 29(3) March 2020—1533 Journal of Materials Engineering and Performance 19. J.P. Kruth, M. Badrossamay, E. Yasa, J. Deckers, L. Thijs, and J. Van
Humbeeck, Part and Material Properties in Selective Laser Melting of
Metals, in 16th International Symposium on Electromachining,
Shangai, China, 19–23 April 2010, Red Hook, NY (2014), pp. 3–14 5. E.A. Lass, M.R. Stoudt, M.E. Williams, M.B. Katz, L.E. Levine, T.Q. Phan, T.H. Gnaeupel-Herold, and D.S. Ng, Formation of the Ni3Nb d-
Phase in Stress-Relieved Inconel 625 Produced Via Laser Powder-Bed
Fusion Additive Manufacturing, Metall. Mater. Trans., 2017, 48, p
5547–5558 20. G.P. Dinda, A.K. Dasqupta, and J. Mazumder, Laser Aided Direct
Metal Deposition of Inconel 625 Superalloy: Microstructural Evolution
and Thermal Stability, Mater. Sci. Eng. A, 2009, 509, p 98–104 6. A. Kreitcberg, V. Brailovskia, and S. Turenneb, Effect of Heat
Treatment and Hot Isostatic Pressing on the Microstructure and
Mechanical Properties of Inconel 625 Alloy Processed by Laser
Powder Bed Fusion, Mater. Sci. Eng. A, 2017, 689, p 1–10 21. G.K. Dey, S. Albert, D. Srivastava, M. Sundararaman, and P. Mukhopadhyay, Microstructural Studies on Rapidly Solidified Inconel
625, Mater. Sci. Eng. A, 1989, 119, p 175–184 Powder Bed Fusion, Mater. Sci. Eng. A, 2017, 689, p 1–10 7. E.A. Lass, M.R. Stoudt, M.B. Katz, and M.E. Williams, Precipitation
and Dissolution of d and c¢¢ During Heat Treatment of a Laser Powder-
Bed Fusion Produced Ni-Based Superalloy, Scr. Mater., 2018, 154, p
83–86 22. L.M. Suave, J. Cornier, P. Villechaise, A. Soula, Z. Hervier, D. Bertheau, and J. Laigo, Microstructural Evolution during Thermal
Aging of Alloy 625: Impact of Temperature and Forming Process,
Metall. Mater. Trans. A Phys. Metall. Mater. Sci., 2014, 45, p 2963–
2982 8. F. Zhang, L.E. Levine, A.J. Allen, C.E. Campbell, E.A. Lass, S. Cheruvathur, M.R. Stoudt, M.E. Williams, and Y. Idell, Homogeniza-
tion Kinetics of a Nickel-Based Superalloy Produced by Powder Bed
Fusion Laser Sintering, Scr. Mater., 2017, 131, p 98–102 23. F.J. Rizzo and J. Radavich, Microstructural Characterization of PM
625-Type Materials, Superalloys 718, 625 and Various Derivatives,
E.A. Open Access Loria, Ed., The Minerals Metals and Materials Society, Warren-
dale, 1991, p 297–308 9. G. Marchese, M. Lorusso, S. Parizia, E. Bassini, J.-W. Lee, F. Calignano,
D. Manfredi, M. Terner, H.-U. Hong, D. Ugues, M. Lombardi, and S. Biamino, Influence of Heat Treatments on Microstructure Evolution and
Mechanical Properties of Inconel 625 Processed by Laser Powder Bed
Fusion, Mater. Sci. Eng. A, 2018, 729, p 64–75 24. F. Cortial, J.M. Corrieu, and C. Vernot-Loier, Influence of Heat
Treatments on Microstructure, Mechanical Properties, and Corrosion
Resistance of Weld Alloy 625, Superalloys 718, 625 and Various
Derivatives, E.A. Loria, Ed., The Minerals Metals and Materials
Society, Warrendale, 1994, p 859–870 10. K.N. Amato, J. Hernandez, L.E. Murr, E. Martinez, S.M. Gaytan, and
P.W. Shindo, Comparison of Microstructures and Properties for a Ni-
Base Superalloy (Alloy 625) Fabricated by Electron and Laser Beam
Melting, J. Mater. Sci. Res., 2012, 1, p 1–41 25. X. Xing, X. Di, and B. Wang, The Effect of Post-Weld Heat Treatment
Temperature on the Microstructure of Inconel 625 Deposited Metal, J. Alloys Compd., 2014, 593, p 110–116 11. T. Keller, G. Lindwall, S. Ghosh, L. Ma, B.M. Lane, F. Zhang, U.R. Kattner, E.A. Lass, J.C. Heigel, Y. Idell, M.E. Williams, A.J. Allen, J.E. Guyer, and L.E. Levine, Application of Finite Element, Phase-Field,
and CALPHAD-Based Methods to Additive Manufacturing of Ni-
Based Superalloy, Acta Mater., 2017, 139, p 244–253 26. T.E. Abioye, D.G. McCartney, and A.T. Clare, Laser Cladding of
Inconel 625 Wire for Corrosion Protection, J. Mater. Process. Technol.,
2015, 217, p 232–240 27. P. Petrzak, K. Kowalski, and M. Blicharski, Analysis of Phase
Transformations in Inconel 625 Alloy During Annealing, Acta Phys. Pol. A, 2016, 130, p 1041–1044 12. F. Zhang, L.E. Levine, A.J. Allen, M.R. Stoudt, G. Lindwall, E.A. Lass, M.E. Williams, Y. Idell, and C.E. Campbell, Effect of Heat
Treatment on the Microstructural Evolution of a Nickel-Based
Superalloy Additive-Manufactured by Laser Powder Bed Fusion, Acta
Mater., 2018, 152, p 200–214 28. B. Dubiel and J. Sieniawski, Precipitates in Additively Manufactured
Inconel 625 Superalloy, Materials, 2019, 12, p 1–11 29. M. Solecka, A. Kopia, A. Radziszewska, and B. Rutkowski,
Microstructure, Microsegregation and Nanohardness of CMT Clad
Layers of Ni-Base Alloy on 16Mo3 Steel, J. Alloys Compd., 2018, 751,
p 86–95 13. C. Li, R. White, X.Y. Fang, M. Weaver, and Y.B. Journal of Materials Engineering and Performance Open Access Gu, Microstructure
Evolution Characteristics of Inconel 625 Alloy from Selective Laser
Melting to Heat Treatment, Mater. Sci. Eng. A, 2017, 705, p 20–31 14. A. Kruse, A. Krupka, V. Schwarzkopf, C. Gamard, and T. Henningsen,
Influence of Proteins on Hydrothermal Gasification and Liquefaction of
Biomass. 2. Model Compounds, Ind. Eng. Chem. Res., 2007, 46, p 87–96 30. M. Rozmus-Go´rnikowska, Ł. Cieniek, M. Blicharski, and J. Kusin´ski,
Microstructure and Microsegregation of an Inconel 625 Weld Overlay
Produced on Steel Pipes by the Cold Metall Tranfer Technique, Arch. Metall. Mater., 2014, 59, p 1081–1084 15. S. Saka and D. Kusdiana, Biodiesel Fuel from Rapeseed Oil as
Prepared in Supercritical Methanol, Fuel, 2001, 80, p 225–231 31. B. Dubiel, A. Kruk, E. Ste˛pniowska, G. Cempura, D. Geiger, P. Formanek, J. Hernandez, P. Midgley, and A. Czyrska-Filemonowicz,
TEM, HRTEM, electron holography and electron tomography studies
of c¢ and c¢¢ nanoparticles in Inconel 718 superalloy, J. Microsc., 2009,
236, p 149–157 16. K. O¨ zdenkc¸i, C. De Blasio, G. Sarwar, K. Melin, J. Koskinen, and V. Alopaeus, Techo-Economic Feasibility of Supercritical Water Gasifi-
cation of Black Liquor, Energy, 2019, 189, p 116284 17. R.K. Ham, The Determination of Dislocation Densities in Thin Films,
Philos. Mag., 1961, 6, p 1183–1184 18. M. Calandri, S. Yin, B. Aldwell, F. Calignano, R. Lupoi, and D. Ugues,
Texture and Microstructural Features at Different Length Scales in
Inconel 718 Produced by Selective Laser Melting, Materials, 2019, 12,
p 1–32 Publisher’s Note
Springer Nature remains neutral with regard to
jurisdictional claims in published maps and institutional affilia-
tions. 1534—Volume 29(3) March 2020 Journal of Materials Engineering and Performance
|
https://openalex.org/W2022831310
|
https://zenodo.org/record/1938050/files/article.pdf
|
English
| null |
EARTHQUAKES RECORDED AT CHELTENHAM MAGNETIC OBSERVATORY JANUARY 24-31, 1906
|
Science
| 1,906
|
public-domain
| 1,565
|
EARTHQUAKES RECORDED AT OHELTENHAM MAU-
NET10 OBSERVATORY JANUARY 2431, 1906.1 1. January 24: 1. January 24:
North-South
East-West
Component. Component. Beginning
2 b. 04 m. 10 6. 2 h. 04 m. 28 s. Beginning principal portion 2
04
10
2
04
28
End principal portion
2
08
18
2
08
28
End
2
33
66
2
29
48
Maximum amplitude
2.0 mm. at
1.8 mm. at
2 h. 06 m. 32 6. 2 h. 06 m. 08 s. Average period of waves :
Maximum
11.0 s. 10.8 s. End
7.2
8.6
2. January 24:
Beginning
2 h. 42 m. 12 6. 2 h. 42 m. 58 s. Beginning principal portion 2
43
20
2
45
03
End principal portion
2
46
12
2
45
34
End
2
62
50
2
51
39
Maximum amplitude
1.5 mm. at
2.2 mm. at
2 h. 44 m. 14 s. 2 h. 44 m. 48 s. Average period of waves :
Maximum
10.5 6. 12.3 a. End
7.2
8.6
3. January 24:
Beginning
16 h. 68 m. 50 s. 16 h. 68 m. 20 6. Beginning principal portion 16
69
20
16
59
20
End principal portion
17
02
50
17
03
30
End
17
10
50
17
08
30
Maximum amplitude,
1.0 mm. at
0.6 mm. at
17 h. 01 m. 34 s. 17 h. 00 m. 26 s. Average period of waves :
Maximum
11.6 a. 9.2 a. Beg. prin. portion
9.1
--
End prin. portion
8.8
7.1 y
The process is an excretory one. By this
means, the rapidly accumulating concretions
of sodic urate are removed from the cells of
the mid-intestine, as are also; but incidentally,
certain unicellular parasites (Gregarinide). y
The process is an excretory one. By this
means, the rapidly accumulating concretions
of sodic urate are removed from the cells of
the mid-intestine, as are also; but incidentally,
certain unicellular parasites (Gregarinide). The nuclei lost by degeneration are replaced
by the mitotic division of the remaining nuclei
-this
occurring before the inner portion of
the epithelium is cast off. p
g
The nuclei lost by degeneration are replaced
by the mitotic division of the remaining nuclei
-this
occurring before the inner portion of
the epithelium is cast off. The peritrophic membrane, which always
envelopes a food-mass, is formed by the split-
ting of the intima, and is, therefore, a secre-
tion from the epithelium of the mid-intestine. XPECZAL ARTZCLEX.
EPITHELIAL
DEGENERATION,
REGENERATION
AND
SECRETION IN THE MID-INTESTINE OF
COLLEMBOLA. XPECZAL ARTZCLEX. EPITHELIAL
DEGENERATION,
REGENERATION
AND
SECRETION IN THE MID-INTESTINE OF
COLLEMBOLA. epithelium of the mid-intestine; (2) by special
clear cells in the middle region of the mid-
intestine; (3) by specialized cells in the pos-
terior region; these last give off proliferations
into the lumen, which become constricted off,
as free, rounded, cytoplasmic vesicles, which
break down in the alimentary canal and mingle
their contents with the food (much as in other
insects). INCoVemboEa, a degeneration of the inner
half of the epithelial wall of the mid-intestine
occurs in connection with each ecdysis. The
cells of the mid-intestine become confluent and
important changes of alveolation ensue; nearly
half the nuclei migrate toward the intima,
while the rest of the nuclei remain near the
basement membrane; a wall now forms be-
tween the two sets of nuclei, dividing the
epithelium into two concentric layers. The
inner of these two layers degenerates; the
cytoplasmic
reticulum
disintegrates;
the
nuclear membranes disappear and the chro-
matin granules become scattered, but remain
intact; niuch of the fluid substance is resorbed
into the remaining layer of cells. The dis-
organized epithelium, surrounded by a peri-
trophic membrane, is expelled through the
rectum shortly after the external moult. The novel r81e of the mid-intestine as an
organ of excretion is correlated with the ab-
sence of Malpighian tubes in CoElembokz. JUSTUS
W. FOLSO
MIRIAMU. WELLES MIRIAMU. WELLES ~ N T O M O L O ~ I ~ A L
LABORATO ~ N T O M O L O ~ I ~ A L
LABORATO UNIVERSITY ILLINOIS
OF UNIVERSITY ILLINOI
OF SCIENCE. SCIENCE. APBIL20, 1906.1 633 EARTHQUAKES RECORDED AT OHELTENHAM MAU-
NET10 OBSERVATORY JANUARY 2431, 1906.1 The wall that divides the originally single
laj-er of celb into two layers, splits into two
membranes, one of which surrounds the degen-
erating epithelium as a peritrophic mem-
brane, while the other forms the new intima
of the mid-intestine. The formation of new cells takes place
throughout the epithelium by mitosis; this
regeneration does not occur from local centers,
or 'crypts,' as it does in other insects; further-
more, no amitotic divisions are found at any
time. lComrnunicated by the superintendent of the
Coast and Geodetic Survey, Mr. 0. H. Tittmann. The observatory is situated at Cheltenham, Md.,
in latitude 38' 44'.0 N, and longitude 76" 50'.5
west of Greenwich. The. times recorded are Secretion is performed (1) by the general SCIENCE. [N. S. VOL. XXIIZ. NO. 590. [N. S. VOL. XXIIZ. NO. 590. 4. January 25:
Beginning
35 h. 47 m. 00 s. 15 h. 46 m. 45 a. Beginning principal portion 15
47
15
15
49
14
End principal portion
15
48
15
16
50
45
End
15
54
00
15
53
24
Maximum amplitude
1.0 mm. at
1.5 mm. at
15 h. 47 m. 38 s. 15 h. 49 m. 44 s. Average period of waves :
Beginning
3.2 s. 3.3 s. Beg. prin. portion
3.4
8.9
Muximum
3.2
8.9
5. January 27:
Beginning
511.19m.00~. 5 h. 19 m. 17 s. Beginning principal portion
5
21
58
5
21
52
End principal portion
5
27
08
5
29
02
End
5
43
00
5
51
27
Maximum amplitude
1.4 mm. at
2.2 mm. at
5 h. 26 m. 58 s. 5 h. 27 m. 02 s. Average period of waves:
Beginning
Principal portion
Maximum
End
6. January 31:
Beginning
10 h. 43 m. 44 s. 10 h. 43 m. 33 s. Beginning principal portion I0
50
23
10
49
43
End principal portion
11
1G
23
11
17
06
End
14
08
00
14
18
00
Maximum amplitude
77 mm. at
66 mm. at
11 h. 02 m. 43s. 11 h.07m.4G s. Average period of waves:
Beginning
3.6 s. 2.8 s. Principal portion
25.2
18.0
End
18.5
18.0 been completed and the results are submitted
in this article. The methods followed were
the ones officially adopted by the Association
of Agricultural Chemists, and the results are
expressed in percentages of the moisture-free
sample. EARTHQUAKES RECORDED AT OHELTENHAM MAU-
NET10 OBSERVATORY JANUARY 2431, 1906.1 It is to be regretted that the carbonic
acid could not be determined, but the lack of
inaterial made it impossible: Insoluble matter ................ Sol. Si. 0, ...................... K,O ........................... Na,O .......................... CaO ........................... lMgO ........................... MnO ........................... Pe,O,AI,O, (mostly Al) ........... P,O, ........................... SO, ............................ Water and organic matter.. ...... Total nitrogen .................. Insoluble matter ................ Sol. Si. 0, ...................... K,O ........................... Na,O .......................... CaO ........................... lMgO ........................... MnO ........................... Pe,O,AI,O, (mostly Al) ........... P,O, ........................... SO, ............................ Water and organic matter.. ...... Total nitrogen .................. Some facts in connection with this analysis
are of peculiar interest. I n the first place the
per cent. of soluble inatter is large and the
greater part of it is Al. This gives some
insight into the origin of the silt. Another
noteworthy feature is the large amount of
potash and its ratio to the soda. While in
eastern soils it is usual for the potash to be in
excess of the soda, the proportion seems larger
than customary. In western soils the soda is
generally in excess of the potash, and this
mould indicate that the silt analyzed originally
came from a semi-arid or humid region. How-
ever, considering the analysis as a whole, there
would be no question about pronouncing this
silt to be an excellent soil. A11 this plant
food is boing removed from the land and
carried either to the sea or to the mouth of
the river. For the sake of argument let us
assume that the above analysis represents the
average composition of the silt carried by the
Mississippi during the entire year. This is
doubtless not quite true, but will serve as a
basis for some calculations. One estimate of
the total amount of silt carried by the Missis-
sippi during a year places the figures at 443,-
'750,000 tons. Assuming this to be true, the
following table gives in tons the amount of
various substances removed in this silt dur-
ing the year : Multiplying ratio of both pointers, 10. Period of north-south component pendulum
about 25 seconds, of east-west component pen-
dulum about 20 seconds. W. F. WALLIS W. F. WALLIS,
Observer in charge. W. F. WALLIS,
Observer in charge. AEALITSIS OF 3fISSISSIPPI RIVER SILT. ON October 13, 1905, there appeared in
SCIENCE
a coniplete analysis of the water of
the Mississippi River, and toward the close
of it the author made the statement that the
silt from the water sample had been saved and
would be subjected to a plant-food analysis
at a later date. Such an analysis has now seventy-fifth meridian mean civil time, counting
the hours continuously from midnight to mid-
night. 'The maximum amplitudes of this earthquake
were too large to be recorded. The pendulums
struck the brushes on both sides. The ampli-
tudes given were measured on the trace and are
probably much too small.
|
https://openalex.org/W2020769038
|
https://epub.ub.uni-muenchen.de/34445/1/10.1038_cddis.2015.24.pdf
|
English
| null |
Selective inhibition of HDAC8 decreases neuroblastoma growth in vitro and in vivo and enhances retinoic acid-mediated differentiation
|
Cell death and disease
| 2,015
|
cc-by
| 15,211
|
Selective inhibition of HDAC8 decreases
neuroblastoma growth in vitro and in vivo and enhances
retinoic acid-mediated differentiation I Rettig1, E Koeneke1, F Trippel2, WC Mueller3, J Burhenne4, A Kopp-Schneider5, J Fabian1, A Schober6, U Fernekorn6, A von Deimling7,
HE Deubzer1,8,9, T Milde1,8, O Witt1,8 and I Oehme*,1 I Rettig1, E Koeneke1, F Trippel2, WC Mueller3, J Burhenne4, A Kopp-Schneider5, J Fabian1, A Schober6, U Fernekorn6, A von Deimling7,
HE Deubzer1,8,9, T Milde1,8, O Witt1,8 and I Oehme*,1 For differentiation-defective malignancies, compounds that modulate transcription, such as retinoic acid and histone deacetylase
(HDAC) inhibitors, are of particular interest. HDAC inhibitors are currently under investigation for the treatment of a broad spectrum
of cancer diseases. However, one clinical drawback is class-specific toxicity of unselective inhibitors, limiting their full anticancer
potential. Selective targeting of individual HDAC isozymes in defined tumor entities may therefore be an attractive alternative
treatment approach. We have previously identified HDAC family member 8 (HDAC8) as a novel target in childhood neuroblastoma. Using small-molecule inhibitors, we now demonstrate that selective inhibition of HDAC8 exhibits antineuroblastoma activity
without toxicity in two xenograft mouse models of MYCN oncogene-amplified neuroblastoma. In contrast, the unselective HDAC
inhibitor vorinostat was more toxic in the same models. HDAC8-selective inhibition induced cell cycle arrest and differentiation
in vitro and in vivo. Upon combination with retinoic acid, differentiation was significantly enhanced, as demonstrated by elongated
neurofilament-positive neurites and upregulation of NTRK1. Additionally, MYCN oncogene expression was downregulated in vitro
and tumor cell growth was markedly reduced in vivo. Mechanistic studies suggest that cAMP-response element-binding protein
(CREB) links HDAC8- and retinoic acid-mediated gene transcription. In conclusion, HDAC-selective targeting can be effective in
tumors exhibiting HDAC isozyme-dependent tumor growth in vivo and can be combined with differentiation-inducing agents. Cell Death and Disease (2015) 6, e1657; doi:10.1038/cddis.2015.24; published online 19 February 2015 The development of neuroblastoma, the most common
extracranial solid tumor of childhood, is hypothesized to be
related to maturation defects of neural crest-derived precursor
cells of the peripheral sympathetic nervous system. The long-
term overall survival probability of high-risk neuroblastoma
patients is o50%. In addition, chemotherapy-treated patients
struggle with therapy-related immediate and long-term toxi-
cities (reviewed in Brodeur1). Received 29.10.14; revised 29.12.14; accepted 12.1.15; Edited by M Agostini
Abbreviations: 13-cis RA, 13-cis retinoic acid; Ac-H4, acetylated histone 4; Ac-SMC3, acetylated structural maintenance of chromosomes 3 protein; Ac-tubulin,
acetylated tubulin; AMC, Academic Medical Center; AP, alkaline phosphatase; ATRA, all-trans retinoic acid; Cpd2, compound 2 (1-naphthohydroxamic acid); CREB, cAMP-
response element-binding protein; GEO, Gene Expression Omnibus; GGT, gamma-glutamyltransferase; GPT, glutamic pyruvic transaminase; HDAC, histone deacetylase;
HDAC8, HDAC family member 8; H&E, hematoxylin and eosin; INSS, International Neuroblastoma Staging System; MTD, maximum tolerable dose; NEF, neurofilament;
NTRK1, neurotrophic tyrosine kinase receptor type 1; OT, glutamic oxaloacetic transaminase; PBMC, peripheral blood mononuclear cells; P-H3, phosphorylated histone 3;
SAHA, suberoylanilide hydroxamic acid; TrkA, tropomyosin receptor kinase A; TSA, trichostatin A Citation: Cell Death and Disease (2015) 6, e1657; doi:10.1038/cddis.2015.24
& 2015 Macmillan Publishers Limited
All rights reserved 2041-4889/15
www.nature.com/cddis Citation: Cell Death and Disease (2015) 6, e1657; doi:10.1038/cddis.2015.24
& 2015 Macmillan Publishers Limited
All rights reserved 2041-4889/15
www.nature.com/cddis OPEN OPEN 1Clinical Cooperation Unit Pediatric Oncology, German Cancer Research Center (DKFZ), Heidelberg, Germany; 2Department of Pediatric Surgery, Dr. von Hauner
Children's Hospital, Ludwig-Maximilians-University Munich, Munich, Germany; 3Department of Neuropathology, Institute for Pathology, University of Leipzig, Leipzig,
Germany; 4Department of Clinical Pharmacology and Pharmacoepidemiology, University Hospital Heidelberg, Heidelberg, Germany; 5Department of Biostatistics, German
Cancer Research Center (DKFZ), Heidelberg, Germany; 6Ilmenau University of Technology, Institute for Micro- and Nanotechnologies, Ilmenau, Germany; 7Department of
Neuropathology, Institute of Pathology, Ruprecht-Karls-University Heidelberg, and Clinical Cooperation Unit Neuropathology, German Cancer Research Center (DKFZ),
Heidelberg, Germany and 8Department of Pediatric Oncology, Hematology and Immunology, University Hospital Heidelberg, Heidelberg, Germany
*Corresponding author: I Oehme, Clinical Cooperation Unit Pediatric Oncology (G340), German Cancer Research Center (DKFZ), Im Neuenheimer Feld 280, Heidelberg
69120, Germany. Tel: +49 6221 423388; Fax: +49 6221 423277; E-mail: i.oehme@dkfz.de
9Current adress: Department of Pediatric Hematology, Oncology and Bone Marrow Transplantation, Charité - University Medicine Berlin, CVK, CC17, Augustenburger Platz 1,
13353 Berlin ediatric Hematology, Oncology and Bone Marrow Transplantation, Charité - University Medicine Berlin, CVK, CC17, Augustenburger Platz 1 Selective inhibition of HDAC8 decreases
neuroblastoma growth in vitro and in vivo and enhances
retinoic acid-mediated differentiation Thus, more neuroblastoma-
optimized therapy approaches aiming at oncogenic molecular
targets, for example, affecting neuroblastoma cell maturation
and proliferation, are believed to improve therapeutic efficacy,
reduce toxicity and avoid long-term side effects.2 Small
molecules, which influence gene transcription, for example,
histone deacetylase (HDAC) inhibitors, are known to induce
maturation of differentiation-defective tumor cells, such as
neuroblastoma cells (reviewed in Witt et al.3). The human family of HDACs is grouped into four classes
based on their homology to yeast HDACs. Three of the four
classes (class I, II and IV) have a zinc-dependent catalytic
mechanism and are the so-called classical HDACs. HDAC
family member 8 (HDAC8) together with HDAC1, 2 and 3
compose class I.4 HDACs remove acetyl groups from lysine
residues of numerous nuclear and cytosolic proteins, affecting
gene transcription as well as many cellular
pathways
(Choudhary et al.5 and Glozak et al.6 and reviewed in Marks
et al.7). Small-molecule inhibitors of HDAC enzymatic activity
bind to the highly conserved catalytic domain and hence
unselectively inhibit the activity of all zinc-dependent HDACs. Vorinostat (SAHA: suberoylanilide hydroxamic acid) was the
first HDAC inhibitor to be approved for clinical use by the FDA
for the treatment of refractory cutaneous T-cell lymphoma.8 In
several clinical trials, vorinostat was active against leukemia HDAC8 impairs cell death and differentiation
I Rettig et al 2 2 and lymphoma, but rather modestly effective against solid
tumors.9 As HDACs are involved in numerous essential
cellular
processes,
the
clinical
use
of
broad-spectrum HDAC inhibitors is associated with dose-limiting side effects
such as thrombocytopenia, fatigue, nausea, diarrhea and
anorexia (reviewed in Lane and Chabner9 and Witt et al.10). Cell Death and Disease Selective inhibition of HDAC8 decreases
neuroblastoma growth in vitro and in vivo and enhances
retinoic acid-mediated differentiation All neuroblastoma cell lines used were HDAC8
positive,
independent of MYCN
oncogene amplification
status. In contrast, infant fibroblasts displayed much lower
HDAC8 protein levels (Supplementary Figure S1C). To
confirm HDAC8 as a potential therapeutic target in vivo, we
examined the influence of HDAC8 knockdown on tumor
growth in a neuroblastoma xenograft mouse model. For this
purpose, we used BE(2)-C, a well-established cell line
representing
a
relatively
chemotherapy-resistant,
highly
aggressive, MYCN-amplified and p53 mutant neuroblastoma
cell model. Cells transiently transfected with either negative
control or HDAC8 targeting siRNA were xenografted sub-
cutaneously into athymic mice. The specificity of the used
siRNA was previously defined.15 HDAC8 depletion delayed
tumor growth with a significant difference in tumor weight at
day 9 after tumor transplantation (Figure 1a). HDAC8 was
clearly depleted at the beginning. However, owing to the
transient nature of the transfection, the expression of HDAC8
increased again over time (Supplementary Figure S1D). This
result supported our hypothesis to use HDAC8 as a
neuroblastoma drug target in vivo. Because similar toxicity profiles have been independently
observed in many clinical trials involving HDAC inhibitors, it is
believed that these are drug class-specific side effects, which
limit their full anticancer potential. Thus, there is an ongoing
debate in the field of HDAC inhibitor development as to whether
selective inhibitors targeting only a single HDAC family member
that drives tumor growth and survival would result in a larger
therapeutic window and greater efficacy compared with broad-
spectrum inhibitors currently in clinical use.11 So far, in vivo
proofof concept demonstrating that superior antitumoral activity
using an HDAC-isozyme-selective inhibitor can be achieved is
lacking. In this regard, HDAC family member 8 appears to be an
attractive target, as crystal structure analysis revealed a unique
second metal binding site in close proximity to the main catalytic
domain,12 which distinguishes this HDAC isozyme from the
other classical deacetylases. This finding led to the design of
linkerless hydroxamic acid-based inhibitors13,14 that fit in this
secondary pocket and display a much higher selectivity for
HDAC8 over the other classical HDACs. We have previously
shown that the expression of HDAC8 correlates with advanced
tumor stage and poor outcome in neuroblastoma.15 Here, we
verify small-molecule inhibition of HDAC8 as a novel therapeu-
tic strategy alone and as a potent enhancer of retinoic acid-
mediated differentiation and propose the cAMP-response
element-binding protein (CREB) as a link between HDAC8
and retinoic acid-mediated gene transcription. Selective inhibition of HDAC8 decreases
neuroblastoma growth in vitro and in vivo and enhances
retinoic acid-mediated differentiation cell death
proliferation
differentiation
(1-Naphthohydroxamic Acid)
cell culture
in vivo
half-life 15 min
MTD 50mg/kg/d
40 µM (single)
PCI-48000
PCI-48012
half-life 1 h
MTD 40mg/kg/d
cell culture
in vivo
4 µM (single)
2 µM (combination)
IC50: 34µM
IC50: 0.02µM
20 µM (combination)
Compound2 (Cpd2)
PCI-34051
0
20
40
60
80
100
120
10
30
50
70
90
110
130
BE
neuroblastoma
UW
medulloblastoma
FB
untransformed
NB
MB
UT
solvent
Cpd2
PCI-34051
PCI-48000
+ - - - + - - - + - - - + - - - + - - -
- + - - - + - - - + - - - + - - - + - -
- - + - - - + - - - + - - - + - - - + -
- - - + - - - + - - - + - - - + - - - +
+ - - - + - - - + - - - + - - - + - - -
- + - - - + - - - + - -
- + - - - + - -
- - + - - - + - - - + -
- - + - - - + -
- - - + - - - + - - - + - - - + - - - +
+
-
-
-
-
+
-
-
-
-
+
-
-
-
-
+
* * *
*
* *
*
* *
* * *
*
*
*
*
*
*
*
*
kDa
Ac-H4
-actin
Ac-tubulin
11
55
42
Ac-SMC3
-actin
42
150
-actin
42
Ac-H4
-actin
Ac-tubulin
NC siH8
kDa
11
55
42
NC siH8
kDa
150
Ac-SMC3
-actin
42
NC
siRNA
HDAC8
siRNA
0
75
50
25
100
125
***
blank
solvent
PCI-48000
Cpd2
TSA
101
100
102
103
104
Ac-K (log FL-2)
IMR-32
BE(2)-C
Ke
IM
SY
SK
DA
ON
As
Cpd2
PCI-
34051
PCI-
48000
sease NC
siRNA
HDAC8
siRNA
0
75
50
25
100
125
*** cell death
proliferation
differentiation
(1-Naphthohydroxamic Acid)
cell culture
in vivo
half-life 15 min
MTD 50mg/kg/d
40 µM (single)
PCI-48000
PCI-48012
half-life 1 h
MTD 40mg/kg/d
cell culture
in vivo
4 µM (single)
2 µM (combination)
IC50: 34µM
IC50: 0.02µM
20 µM (combination)
Compound2 (Cpd2)
PCI-34051
kDa
Ac-H4
-actin
Ac-tubulin
11
55
42
Ac-SMC3
-actin
42
150
-actin
42
Ac-H4
-actin
Ac-tubulin
NC siH8
kDa
11
55
42
NC siH8
kDa
150
Ac-SMC3
-actin
42
NC
siRNA
HDAC8
siRNA
0
75
50
25
100
125
***
blank
solvent
PCI-48000
Cpd2
TSA
101
100
102
103
104
Ac-K (log FL-2)
IMR-32
BE(2)-C
Cpd2
PCI-
34051
PCI-
48000 (1-Naphthohydroxamic Acid)
cell culture
in vivo
half-life 15 min
MTD 50mg/kg/d
40 µM (single)
IC50: 34µM
20 µM (combination)
Compound2 (Cpd2) PCI-48000
PCI-48012
half-life 1 h
cell culture
in vivo
4 µM (single)
IC50: 0.02µM
PCI-34051 PCI-34051 (1-Naphthohydroxamic Acid)
Compound2 (Cpd2) PCI-48000 cell death
proliferation
differentiation
kDa
Ac-H4
-actin
Ac-tubulin
11
55
42
Ac-SMC3
-actin
42
150
-actin
42
Ac-H4
-actin
Ac-tubulin
NC siH8
kDa
11
55
42
NC siH8
kDa
150
Ac-SMC3
-actin
42
blank
solvent
PCI-48000
Cpd2
TSA
101
100
102
103
104
Ac-K (log FL-2)
IMR-32
BE(2)-C
Cpd2
PCI-
34051
PCI-
48000 0
20
40
60
80
100
120
10
30
50
70
90
110
130
BE
neuroblastoma
UW
medulloblastoma
FB
untransformed
NB
MB
UT
solvent
Cpd2
PCI-34051
PCI-48000
+ - - - + - - - + - - - + - - - + - - -
- + - - - + - - - + - - - + - - - + - -
- - + - - - + - - - + - - - + - - - + -
- - - + - - - + - - - + - - - + - - - +
+ - - - + - - - + - - - + - - - + - - -
- + - - - + - - - + - -
- + - - - + - -
- - + - - - + - - - + -
- - + - - - + -
- - - + - - - + - - - + - - - + - - - +
+
-
-
-
-
+
-
-
-
-
+
-
-
-
-
+
* * *
*
* *
*
* *
* * *
*
*
*
*
*
*
*
*
Ke
IM
SY
SK
DA
ON
As Cell Death and Disease HDAC8 impairs cell death and differentiation
I Rettig et al HDAC8 impairs cell death and differentiation
I Rettig et al 3 3 HDAC8 protein expression in a diversity of neuroblastoma
cell lines. Figure 1
Characterization of HDAC8 inhibitors in vitro. (a) In vivo validation of HDAC8 as a drug target. Tumor weight of explanted BE(2)-C neuroblastoma xenografts after
siRNA-mediated transient knockdown of HDAC8 or negative control (NC) in NMRI Foxn1 nude mice (cohorts of 20 animals each). Cells were transplanted 48 h after knockdown.
Data are represented as mean tumor weight ± S.E.M. Mann Whitney was used for statistics (***Po0.001). (b) Summary of the HDAC8 inhibitors used in the study. The chemical
structures of linkerless hydroxamic acid-based inhibitors selective for HDAC8, Cpd2 and PCI-34051 are shown. PCI-48000 and PCI-48012 are more stable variants of PCI-34051.
(c) Flow cytometric analysis of intracellular whole lysine acetylation levels of IMR-32 cells. Cells were treated with HDAC8 inhibitors Cpd2 (40 μM) and PCI-48000 (4 μM) for 24 h.
TSA (trichostatin A, 150 nM) was used as positive control. Solvent: DMSO control. Blank: secondary antibody only. (d) Left panel: western blot analysis of histone 4, α-tubulin and
SMC3 acetylation levels in BE(2)-C cells treated with increasing concentrations of Cpd2 (0, 40, 100 and 200 μM), PCI-34051 (0, 4, 30 and 100 μM) and PCI-48000 (0, 4, 20 and
100 μM). Actin served as a loading control. Right panel: Western blot analysis of histone 4 acetylation, α-tubulin acetylation and SMC3 acetylation in HDAC8-depleted (siH8)
versus negative control siRNA (NC)-transfected BE(2)-C cells. (e) Determination of total cell numbers 6 days after treatment of neuroblastoma cell lines (MYCN amplified:
BE = BE(2)-C, IM = IMR-32 and Ke = Kelly; non-amplified: SY = SH-SY5Yand SK = SK-N-AS), medulloblastoma cell lines (UW = UW-288-2, DA = DAOYand ON = ONS76)
and proliferative infant fibroblasts (FB), as well as astrocytes (As) with HDAC8 inhibitors Cpd2 (40 μM), PCI-34051 (4 μM) and PCI-48000 (4 μM). Cell count was normalized to
solvent-treated cells. T-test was used for statistics. *Po0.0001, bars represent mean values and error bars represent S.E.M. Selective inhibition of HDAC8 decreases
neuroblastoma growth in vitro and in vivo and enhances
retinoic acid-mediated differentiation HDAC8-selective inhibitors display antineuroblastoma
activity in vitro. To inhibit specifically HDAC8 activity in
neuroblastoma,
we
decided
to
use
and
compare
two
structurally divergent HDAC8 inhibitors, which are both
reported to be selective for HDAC8: Compound 2 (Cpd2; 1-
naphthohydroxamic
acid13)
and
PCI-34051.14
For
PCI-
-34051, two more stable variants exist: PCI-48000 and
PCI-48012. A comparison of the compounds and concentra-
tions applied is shown in Figure 1b. We confirmed HDAC8
selectivity of Cpd2 and PCI-34051 using in vitro cell-free
biochemical
assays
of
all
classical
HDACs
1–11
(Supplementary Table 1). This determined HDAC6 and
HDAC1 as potential off-target HDACs at concentrations
above 75 μM for Cpd2 and above 30 μM for PCI-34051. Thus, acetylation levels of HDAC6 or HDAC1 substrates can
be used as an indicator for unselective HDAC family inhibition
for the two HDAC8-selective compounds. We therefore
established working concentrations for the selective use of
the inhibitors on neuroblastoma cells at 40 μM for Cpd2 and
4 μM for PCI-34051 and PCI-48000. We confirmed the
HDAC8-selective
inhibition
under
these
conditions:
(i)
increase in whole lysine acetylation levels (Figure 1c and
Supplementary Table 2), (ii) increase in acetylation level of Results HDAC8 as a neuroblastoma drug target in vivo. We have
recently demonstrated that knockdown and inhibition of
HDAC8 in neuroblastoma cell cultures induced cell cycle
arrest and differentiation.15 The aim of the current study was
to evaluate HDAC8 as a potential drug target in vivo. Before
this, we investigated HDAC8 expression in human tumor
samples. We reanalyzed publically available expression data
from the AMC (Academic Medical Center) cohort (GEO
accession no. GSE16476) of neuroblastoma patients using
the web-based R2 microarray database (http://r2.amc.nl).16
The expression of HDAC8 significantly differed between the
tumor stages. HDAC8 expression levels significantly corre-
lated with INSS (International Neuroblastoma Staging Sys-
tem) stage 4 and poor overall survival (Supplementary
Figures 1A and B). These results confirm our previously
published observation with human tumor samples from the
German neuroblastoma trial.15 We additionally tested the Cell Death and Disease HDAC8 impairs cell death and differentiation
I Rettig et al the HDAC8 substrate core cohesion complex protein SMC317
(Figure 1d), (iii) absence of HDAC6 substrate acetylation
(tubulin;18 Figure 1d and Supplementary Figure S2) and (iv)
absence of HDAC1-3 substrate acetylation (histone 4;19
Figure 1d and Supplementary Figure S2). Treatment with the
pan-HDAC inhibitor TSA (trichostatin A), which was used as a
control for unselective HDAC inhibition, resulted in a much
stronger induction of global lysine acetylation (Figure 1c and
Supplementary Table 2). This is in line with the higher amount
of acetylated substrates as a consequence of inhibition of
multiple HDACs. As predicted, increasing concentrations of
Cpd2 and PCI-34051, far above their IC50 values, displayed
HDAC6 off-target activity and induced tubulin acetylation
(Figure 1d). with a plasma peak concentration of 8.24 μM after adminis-
tration of a single dose of 100 mg/kg, suggesting that the
MTD of 40 mg/kg results in therapeutic plasma levels above
the HDAC8 in vitro IC50, but still in the range of selectivity. As
a pharmacodynamic marker for HDAC8 inhibitory action
in vivo, we analyzed lysine acetylation levels in peripheral
blood mononuclear cells (PBMC) after intraperitoneal admin-
istration of Cpd2 and PCI-48012 at MTD concentrations
(Supplementary Table 2). HDAC8 selectivity was further
controlled by the absence of tubulin acetylation and absence
of histone 4 acetylation in PBMC (Figure 2c). Vorinostat was
used as a control for unselective HDAC inhibition. Results The MTD
of vorinostat in mice (150 mg/kg per day) is well established
and was taken from the published literature.20 p
Next, we determined the in vivo efficacy of HDAC8-selective
inhibition
in
neuroblastoma
xenograft
mouse
models. Because the pharmacokinetic studies revealed a relatively
short half-life for Cpd2, we subsequently used PCI-48012 for
treatment efficacy studies. One week after tumor cell injection,
mice established a palpable tumor mass and the animals were
then randomly assigned to treatment and control cohorts. Treatment with PCI-48012 significantly delayed tumor growth
compared with solvent-treated control group. Moreover,
HDAC8-selective inhibition delayed neuroblastoma tumor
growth more efficiently (BE(2)-C model; Figure 2d) or equally
efficiently (IMR-32 model; Figure 2e) compared with vorinostat
treatment. Whereas vorinostat treatment induced clear signs
of toxicity, such as diarrhea and weight loss at the applied
dose, HDAC8-selective inhibitor treatment was well tolerated
(Figure 2f). The treatment of a panel of MYCN-amplified as well as
MYCN-single-copy neuroblastoma cell lines with HDAC8
inhibitors for 6 days significantly decreased cell numbers. Of
note, medulloblastoma cell lines were less susceptible
compared with neuroblastoma cells. Untransformed cells
(infant fibroblasts and astrocytes) hardly responded to the
treatment (Figure 1e). So far, these results point towards
HDAC8 as a promising and selective target for neuroblastoma
treatment. HDAC8-selective inhibitors display antineuroblastoma
activity in vivo. To determine the in vivo efficacy of
HDAC8-selective inhibitors in neuroblastoma, we started to
characterize the inhibitors for in vivo use. We first determined
toxicity profiles, maximum tolerable doses (MTDs) and
plasma levels of the two HDAC8-selective inhibitors Cpd2
and PCI-48012 for application in mice in vivo. We chose the
latter compound as a more in vivo stable variant of PCI-48000
with improved pharmacokinetic properties (Figure 1b). Both
inhibitors were intraperitoneally injected into NMRI Foxn1
nude mice in increasing doses from 40 up to 400 mg/kg
per day for 2 × 5 days. Each dose was tested in a cohort of
three animals. Resulting MTDs were validated within cohorts
of eight animals for each inhibitor. Dose-limiting toxicities
(DLTs) were determined by monitoring body weight, where a
reduction of ≥20% from starting body weight indicated
toxicity. Additional experiments performed to check for toxicity
included clinical chemistry and hematological blood para-
meters, as well as histological examination of a panel of
organs for toxicity using hematoxylin and eosin (H&E) stain
after killing the animals. As a reference, the blood of 12
untreated NMRI Foxn1 nude mice was analyzed. Results A mixed linear model with fixed slope for each group and random intercept for each mouse was used for statistical analysis (SAS ProcM
= 0.0216; **P = 0.0015; ***Po0.0001). (e) Growth curve of IMR-32 neuroblastoma xenografts in nude mice treated intraperitoneally with either solvent, HDAC8 inh
-48012 (40 mg/kg per day) or unselective HDAC inhibitor vorinostat (150 mg/kg per day). Treatment started 7 days after implantation of tumors (black arrow). Results In these
studies, we determined the MTD for HDAC8 inhibitors Cpd2
at 50 mg/kg per day and for PCI-48012 at 40 mg/kg per day. At these concentrations, neither body weight (Figure 2a) nor
blood parameters were critically changed according to clinical
toxicity criteria (Table 1). DLTs of Cpd2 included weight loss
and signs of liver toxicity, as evidenced by elevated plasma
liver enzymes and detection of necrotic areas on histological
liver
examination. DLTs
of
PCI-48012
included
severe
inflammation of the intestine at organ examination, but no
noticeable changes in blood parameters. Pharmacokinetic
studies after intraperitoneal administration of the inhibitors
identified the half-life of Cpd2 to be ~ 15 min, with a plasma
peak concentration of ~ 30 μM, matching the in vitro HDAC8
IC50 of 34 μM (Figure 2b). PCI-48012 has a half-life of ~ 1 h, HDAC8-selective
inhibition
promotes
differentiation,
delays cell proliferation and induces cell death in vitro
and in vivo. We have previously demonstrated that HDAC8
depletion of neuroblastoma cells via siRNA transfections
resulted in an upregulation of markers for differentiation (NEF
and NTRK1) as well as G0/G1 cell cycle arrest with p21WAF1/
CIP1 upregulation. In line with our previous data, all three
HDAC8 inhibitors induced morphological changes pointing
towards neuronal differentiation, which was associated with
an increase in neurofilament (Figure 3a) and TrkA protein
expression (Figure 3b). In addition, all HDAC8 inhibitors used
delayed neuroblastoma population growth (Figure 3c) and
induced the cell-cycle inhibitor p21WAF1/CIP1 (Figure 3d). Furthermore, differentiation and proliferation disruption was
followed by cell death (Figure 3e). Corresponding to these in vitro data, quantitative immuno-
histochemical analysis of the dissected tumor material from
the HDAC8-inhibitor-treated mice revealed an increase in
neurofilament-positive tumors (7/12 = 58.3%) compared with
control groups (1/12 = 8.3%), and a significantly lower amount
of cells positive for phosphorylated histone H3, a marker for
mitosis (Figure 4). Consistent with the cell culture results, induction of
differentiation was also accompanied by induction of cell
death in vivo. The amount of tumor cells positive for active
caspase-3
was
significantly
increased. Taken
together,
selective
inhibition
of
HDAC8
enhanced
differentiation,
reduced tumor growth and induced cell death in vitro and
in vivo. Results Cell Death and Disease HDAC8 impairs cell death and differentiation
I Rettig et al 5 5 start of
treatment
0
500
1000
1500
2000
days after
tumor transplantation
0
10
15
5
2500
20
solvent
vorinostat
PCI-48012
0 10 20 30 40 50 60 70 80 90100
time [min]
100
1000
10000
100000
t1/2 = 15.3min
0
10
20
30
-30
-20
-10
0
40
50
100
200
PCI-
48012
(mg/kg/d)
0
1
2
3
4
7
8
9
10
11
time [d]
0
10
20
30
-30
-20
-10
0
50
100
200
300
400
Cpd2
(mg/kg/d)
0
1
2
3
4
7
8
9
10
11
time [d]
PCI-
48012
Cpd2
0h
2h
0h
2h
0h
2h
vorinostat
Ac-H4
-actin
Ac-tubulin
kDa
11
55
42
solvent
vorinostat
PCI-48012
-10
0
10
20
days of treatment
10
15
5
20
-20
0
0
days after
tumor transplantation
0
20
30
10
100
200
300
400
500
600
start of
treatment
solvent
vorinostat
PCI-48012
*
***
**
***
***
*
(IMR-32)
(BE(2)-C)
gure 2
Characterization of HDAC8 inhibitors in vivo. (a) NMRI Foxn1 nude mice were treated with different concentrations of HDAC8 inhibitors Cpd2 and PCI-48012
T is defined by loss of body weight of ≥20%. Solvent control = 0 mg/kg per day. Data points represent mean values and error bars represent S.E.M. (b) Determination o
of HDAC8 inhibitor Cpd2 in the blood plasma of NMRI Foxn1 nude mice. Cpd2 was injected intraperitoneally at a dose of 100 mg/kg and whole blood was collected at se
e points (0, 10, 20, 30, 45, 60, 75 and 90 min) in cohorts of six animals, respectively. Resulting concentrations were defined by HPLC analysis from blood sera. (c) Wester
alysis of histone 4 and α-tubulin acetylation levels in PBMC cells of untreated mice (0 h) and respective animals upon treatment with HDAC8 inhibitors PCI-48012 (40 m
d Cpd2 (50 mg/kg) 2 h after intraperitoneal injection. Broad-spectrum HDAC inhibitor vorinostat (100 mg/kg) was used as a positive control. Actin served as a loading co
Growth curve of BE(2)-C neuroblastoma xenografts in nude mice treated intraperitoneally with either solvent, HDAC8 inhibitor PCI-48012 (40 mg/kg per day) or unsel
AC inhibitor vorinostat (150 mg/kg per d) in cohorts of 12 animals each. Treatment started 7 days after implantation of tumors (black arrow). Data are represented as
mor volume ± S.E.M. Results Dat 5 0 10 20 30 40 50 60 70 80 90100
100
1000
10000
100000
t1/2 = 15.3min
0
10
20
30
-30
-20
-10
0
40
50
100
200
PCI-
48012
(mg/kg/d)
0
1
2
3
4
7
8
9
10
11
time [d]
0
10
20
30
-30
-20
-10
0
50
100
200
300
400
Cpd2
(mg/kg/d)
0
1
2
3
4
7
8
9
10
11
time [d]
PCI-
48012
Cpd2
0h
2h
0h
2h
0h
2h
vorinostat
Ac-H4
-actin
Ac-tubulin
kDa
11
55
42 0
10
20
30
-30
-20
-10
0
40
50
100
200
PCI-
48012
(mg/kg/d)
0
1
2
3
4
7
8
9
10
11
time [d]
0
10
20
30
-30
-20
-10
0
50
100
200
300
400
Cpd2
(mg/kg/d)
0
1
2
3
4
7
8
9
10
11
time [d] 0
10
20
30
-30
-20
-10
0
50
100
200
300
400
Cpd2
(mg/kg/d)
0
1
2
3
4
7
8
9
10
11
time [d] 0
10
20
30
-30
-20
-10
0
40
50
100
200
PCI-
48012
(mg/kg/d)
0
1
2
3
4
7
8
9
10
11
time [d] 0 10 20 30 40 50 60 70 80 90100
100
1000
10000
100000
t1/2 = 15.3min
0
-30
-20
-10
0
1
2
3
4
7
8
9
10
11
time [d]
0
-30
-20
-10
0
1
2
3
4
7
8
9
10
11
time [d]
PCI-
48012
Cpd2
0h
2h
0h
2h
0h
2h
vorinostat
Ac-H4
-actin
Ac-tubulin
kDa
11
55
42 0 10 20 30 40 50 60 70 80 90100
time [min]
100
1000
10000
100000
t1/2 = 15.3min 0
PCI-
48012
Cpd2
0h
2h
0h
2h
0h
2h
vorinostat
Ac-H4
-actin
Ac-tubulin
kDa
11
55
42 1 time [min] solvent
vorinostat
PCI-48012
-10
0
10
20
days of treatment
10
15
5
20
-20
0
0
days after
tumor transplantation
0
20
30
10
100
200
300
400
500
600
start of
treatment
solvent
vorinostat
PCI-48012
***
***
*
(IMR-32) start of
treatment
0
500
1000
1500
2000
days after
tumor transplantation
0
10
15
5
2500
20
solvent
vorinostat
PCI-48012
time [min]
solvent
vorinostat
PCI-48012
-10
0
10
20
days of treatment
10
15
5
20
-20
0
0
days after
tumor transplantation
0
20
30
10
100
200
300
400
500
600
start of
treatment
solvent
vorinostat
PCI-48012
*
***
**
***
***
*
(IMR-32)
(BE(2)-C) solvent
vorinostat
PCI-48012
-10
0
10
20
days of treatment
10
15
5
20
-20
0 start of
treatment
0
500
1000
1500
2000
days after
tumor transplantation
0
10
15
5
2500
20
solvent
vorinostat
PCI-48012
[
]
*
***
**
(BE(2)-C) days after
tumor transplantation
(BE(2)-C) Figure 2
Characterization of HDAC8 inhibitors in vivo. Results (a) NMRI Foxn1 nude mice were treated with different concentrations of HDAC8 inhibitors Cpd2 and PCI-48012. The
DLT is defined by loss of body weight of ≥20%. Solvent control = 0 mg/kg per day. Data points represent mean values and error bars represent S.E.M. (b) Determination of half-
life of HDAC8 inhibitor Cpd2 in the blood plasma of NMRI Foxn1 nude mice. Cpd2 was injected intraperitoneally at a dose of 100 mg/kg and whole blood was collected at several
time points (0, 10, 20, 30, 45, 60, 75 and 90 min) in cohorts of six animals, respectively. Resulting concentrations were defined by HPLC analysis from blood sera. (c) Western blot
analysis of histone 4 and α-tubulin acetylation levels in PBMC cells of untreated mice (0 h) and respective animals upon treatment with HDAC8 inhibitors PCI-48012 (40 mg/kg)
and Cpd2 (50 mg/kg) 2 h after intraperitoneal injection. Broad-spectrum HDAC inhibitor vorinostat (100 mg/kg) was used as a positive control. Actin served as a loading control. (d) Growth curve of BE(2)-C neuroblastoma xenografts in nude mice treated intraperitoneally with either solvent, HDAC8 inhibitor PCI-48012 (40 mg/kg per day) or unselective
HDAC inhibitor vorinostat (150 mg/kg per d) in cohorts of 12 animals each. Treatment started 7 days after implantation of tumors (black arrow). Data are represented as mean
tumor volume ± S.E.M. A mixed linear model with fixed slope for each group and random intercept for each mouse was used for statistical analysis (SAS ProcMixed)
(*P = 0.0216; **P = 0.0015; ***Po0.0001). (e) Growth curve of IMR-32 neuroblastoma xenografts in nude mice treated intraperitoneally with either solvent, HDAC8 inhibitor
PCI-48012 (40 mg/kg per day) or unselective HDAC inhibitor vorinostat (150 mg/kg per day). Treatment started 7 days after implantation of tumors (black arrow). Data are
represented as mean tumor volume ± S.E.M. A mixed linear model with fixed slope for each group and random intercept for each mouse was used for statistical analysis (SAS
ProcMixed) (*P = 0.0349; ***Po0.0001). (f) Changes in body weight of nude mice in a long-term treatment of 3 × 5 days with solvent, HDAC8 inhibitor PCI-48012 (40 mg/kg
per day) and broad-spectrum HDAC inhibitor vorinostat (150 mg/kg per day). Data are represented as mean tumor volume ± S.E.M. Results The combined treatment increased the amount of
neurite outgrowths as well as neurite length in neuroblastoma
cell lines (Figures 5a–c and Supplementary Figure S3A) and
increased the expression of neurofilament (Figure 5a and
Supplementary Figure S3A), as well as NTRK1 (Figure 5d). Additionally, the combined treatment decreased cell numbers
(Supplementary Figures S3B–D) and the ability to form
colonies in longer-term assays (Figure 5e and Supplementary
Figure S3E). One already well-known result of retinoic acid
treatment in neuroblastoma is the reduction of MYCN
oncogene expression in amplified tumors.21 Interestingly,
the cotreatment with HDAC8-selective inhibitors even further
reduced MYCN protein levels in BE(2)-C and IMR-32 cells
(Figure 5f and Supplementary Figure S3F). caused severe swelling of the abdomen. Treatment of BE(2)-C
xenografted NMRI nude mice for 2 × 5 days with HDAC8
inhibitor, or 13-cis RA revealed that the treatment with HDAC8
inhibitor alone was more potent to reduce tumor growth than
13-cis RA treatment alone. The combination of both agents
was more potent than either treatment alone and resulted in
synergistic
effects,
which
was
determined
using
log-
transformed tumor volumes with a linear mixed model
(Figure 6a and Supplementary Figure S4B). The in vivo
combination of PCI-48012 and 13-cis RA decreased tumor
proliferation and increased tumor cell death in comparison
with either single treatment alone (Figure 6b). In summary,
HDAC8-selective
inhibition
enhanced
antineuroblastoma
activity of retinoic acid treatment in vitro and in vivo. Forced HDAC8 overexpression counteracts retinoic acid-
mediated differentiation. Next, we used stably HDAC8-
overexpressing BE(2)-C cells for longer-term colony assays
(18 days), and treated the cells with ATRA. During this time
period, untreated cells (HDAC8 and empty vector control)
grew to a high density and began to detach, and thus the
differences were not visible anymore. However, reseeding of
detached
cells
floating
in
the
supernatant
into
new
plates revealed a substantial difference between HDAC8-
overexpressing and empty vector control cells (Figure 7a,
labeled
as
‘supernatant’). In
addition,
ATRA
efficiently
reduced the ability of empty vector control cells to form
colonies and significantly more colonies were formed by
ATRA-treated HDAC8-overexpressing cells (Figure 7a), sug-
gesting a direct mechanistic link between HDAC8 and retinoic
acid signaling. With this in mind, we focused on CREB,
because CREB is involved in both retinoic acid23 and HDAC8
signaling (Figure 7b). Results Cell Death and Disease HDAC8 impairs cell death and differentiation
I Rettig et al 6 6 Table 1 Mean indicated values of blood parameters from HDAC8-inhibitor-treated NMRI Foxn1 nude mice (± S.D.) at MTD concentrations in comparison with normal
range
Solvent (n = 8)
Cpd2 (n = 8)
PCI-48012 (n = 8)
Normal range (n = 12)
Sodium
133.1 ±8.9
136.8 ± 5.3
138.9 ± 5.3
123.4–135.7 mmol/l
Potassium
5.8 ± 0.9
6.9 ± 2.2
6.4 ± 2.2
3.8–6.4 mmol/l
Calcium
2.3 ± 0.1
2.4 ± 0.4
2.4 ± 0.2
2.1–2.4 mmol/l
Chloride
101.9 ± 3.9
99.9 ± 2.2
102.0 ± 3.7
97.3–109.7 mmol/l
Creatinine
0.2 ± .04
0.2 ± 0.04
0.2 ± 0.02
0.1–0.3 mmol/l
Phosphate
2.6 ± 0.5
1.4 ± 1.0
1.9 ± 1.2
2.0–2.9 U/l
Urea
26.3 ± 4.2
23.5 ± 6.4
30.0 ± 7.2
17.4–38.3 mg/dl
Uric acid
0.8 ± 0.8
0.9 ± 1.0
0.7 ± 0.3
0.4–3.2 mg/dl
Glucose
203.9 ± 28.0
187.8 ± 39.1
174.0 ± 18.6
176.8–238.6 mg/dl
GOT/AST
110.0 ± 5.2
100.3 ± 46.9
167.5 ± 93.0
68.3–308.7 mg/dl
GPT/ALT
66.4 ± 30.1
66.5 ± 10.9
127.3 ± 73.3
29.8–143.8 mg/dl
AP
98.6 ± 26.5
123.8 ± 35.8
81.8 ± 33.7
62.5–117.6 U/l
GGT
−3.0 ± 1.8
−2.0 ± 1.0
0.5 ± 0.5
−1.9 to 1.0 U/l
Bilirubin total
o0.2
o0.2
o0.2
o0.2 U/l
Leukocytes
3.0 ± 1.0
4.4 ± 1.8
4.5 ± 1.3
1.8–6.1/nl
Erythrocytes
7.6 ± 0.3
7.6 ± 0.3
7.0 ± 0.9
5.5–8.2/pl
Thrombocytes
641.0 ± 61.4
682.0 ± 151.1
690.9 ± 253.1
242.5–806.8/nl
Hemoglobin
12.6 ± 0.7
12.4 ± 0.7
11.6 ± 1.6
9.4–12.7 g/dl
Hematocrit
0.4 ± 0.02
0.4 ± 0.01
0.4 ±0.04
0.3–0.4 l/l
Abbreviations: AP, alkaline phosphatase; GGT, gamma-glutamyltransferase (gammaGT); GOT, glutamic oxaloacetic transaminase; GPT, glutamic pyruvic
transaminase Abbreviations: AP, alkaline phosphatase; GGT, gamma-glutamyltransferase (gammaGT); GOT, glutamic oxaloacetic transaminase; GPT, glutamic pyruvic
transaminase Enhancement of differentiation with combined treatment
of neuroblastoma cells with HDAC8 inhibitor and retinoic
acid in vitro and in vivo. Because we observed induction of
differentiation
following
selective
HDAC8
inhibition,
we
wondered whether it is possible to further enhance the
differentiation phenotype via combination with retinoic acid, a
differentiating agent currently in use for neuroblastoma
therapy. In these experiments, we used the HDAC8 inhibitors
in lower concentrations (Cpd2 20 μM and PCI-48000 2 μM),
which were still sufficient to inhibit HDAC8 (Supplementary
Table 2). Results HDAC8 has been described to be
involved
in
the
phosphatase-mediated
inactivation
of
CREB.24 Indeed, stably HDAC8-overexpressing BE(2)-C
cells
displayed
lower
levels
of
phosphorylated
CREB
(Figure 7c). In addition, expression of tyrosine hydroxylase, In
addition,
we
combined
HDAC8-selective
inhibitor
PCI-48012 with retinoic acid in vivo. For these experiments,
the clinically approved drug 13-cis retinoic acid (13-cis RA)
was used instead of ATRA because of its prolonged in vivo
half-life in comparison with ATRA.22 We first determined a
tolerable combination regimen of HDAC8 inhibitor plus retinoic
acid. HDAC8 inhibitor PCI-48012 (40 mg/kg per day) was
injected intraperitoenally in combination with 13-cis RA at
different concentrations into NMRI nude mice. Toxicity of the
combination was assessed by monitoring body weight
(Supplementary Figure S4A). These experiments revealed a
tolerable dose for 13-cis RA of 10 mg/kg per day, as
intraperitoneal administration in a dose of 20 mg/kg per day Cell Death and Disease Cpd2
PCI-34051 PCI-48000
+
+
-
-
-
+
0
20
40
60
80
0
24
48
72
96
144
time [h]
60
40
20
0
80
TrkA
-actin
kDa
140
42
Cpd2
PCI-34051 PCI-48000
+
+
-
-
-
+
p21
-actin
kDa
42
21
solvent
PCI-34051
Cpd2
-
+
-
-
-
-
-
+
-
-
-
-
-
+
-
PCI-48000
-
-
-
-
+
**
**
*
0.00
0.25
0.50
0.75
1.00
1.25
1.50
0
24
48
72
96 120 144
time [h]
0.0
1.0
2.0
3.0
4.0
0
24
48
72
96 120 144
time [h]
0.0
0.50
0.75
1.00
1.25
0
24
48
72
96 120 144
time [h]
1.50
0.25
solvent
PCI-34051
Cpd2
untreated
solvent
PCI-34051
Cpd2
untreated
solvent
PCI-34051
Cpd2
untreated
PCI-48000
solvent
PCI-34051
Cpd2
120
***
***
***
***
***
***
**
**
*
IMR-32
BE(2)-C
Kelly
Kelly
BE(2)-C
IMR-32
IMR-32
gure 3
Phenotype characterization of HDAC8 inhibitors in vitro. (a) Flow cytometric quantitative analysis of neurofilament protein expression. BE(2)-C cells were treate
HDAC8 inhibitors or solvent for 6 days. Blank (–/–/–/–): secondary antibody only. (b) Western blot analysis of TrkA protein levels. Kelly cells were treated with HDAC
bitors or solvent for 72 h. Actin served as a loading control. (c) Growth curves of BE(2)-C, IMR-32 and Kelly cells treated with HDAC8 inhibitors Cpd2 (40 μM) and PCI-3405
M). (d) Western blot analysis of p21WAF1/CIP1 protein levels. Results (b) Western blot analysis of TrkA protein levels. Kelly cells were treated with HDAC8
inhibitors or solvent for 72 h. Actin served as a loading control. (c) Growth curves of BE(2)-C, IMR-32 and Kelly cells treated with HDAC8 inhibitors Cpd2 (40 μM) and PCI-34051
(4 μM). (d) Western blot analysis of p21WAF1/CIP1 protein levels. IMR-32 cells were treated with HDAC8 inhibitors (+) Cpd2 (40 μM), PCI-34051 (4 μM) and PCI-48000 (4 μM) or
solvent (−) for 72 h. Actin served as a loading control. (e) Time-resolved determination of dead cells via automated cell counting and trypan blue staining (% trypan blue positive)
in IMR-32 cells after treatment with HDAC8 inhibitors; data points represent mean values and error bars represent S.E.M. which has been described to be regulated by CREB via direct
promoter activation (Armstrong et al.25 and references
therein), was repressed in stably HDAC8-overexpressing
BE(2)-C cells (Figure 7d). Thus, we hypothesize that HDAC8
and ATRA signaling partly converge at the level of CREB and
the combined treatment of neuroblastoma cells with HDAC8
inhibitors and retinoic acid enhances differentiation. Results IMR-32 cells were treated with HDAC8 inhibitors (+) Cpd2 (40 μM), PCI-34051 (4 μM) and PCI-48000 (4 μM)
vent (−) for 72 h. Actin served as a loading control. (e) Time-resolved determination of dead cells via automated cell counting and trypan blue staining (% trypan blue positiv
MR-32 cells after treatment with HDAC8 inhibitors; data points represent mean values and error bars represent S.E.M. HDAC8 impairs cell death and differentiation
I Rettig et al HDAC8 impairs cell death and differentiation
I Rettig et al Cpd2
PCI-34051 PCI-48000
+
+
-
-
-
+
60
40
20
0
80
TrkA
-actin
kDa
140
42
solvent
PCI-34051
Cpd2
-
+
-
-
-
-
-
+
-
-
-
-
-
+
-
PCI-48000
-
-
-
-
+
**
**
*
0.00
0.25
0.50
0.75
1.00
1.25
1.50
0
24
48
72
96 120 144
time [h]
0.0
1.0
2.0
3.0
4.0
0
24
48
72
96 120 144
time [h]
0.0
0.50
0.75
1.00
1.25
0
24
48
72
96 120 144
time [h]
1.50
0.25
solvent
PCI-34051
Cpd2
untreated
solvent
PCI-34051
Cpd2
untreated
solvent
PCI-34051
Cpd2
untreated
***
***
***
***
***
***
IMR-32
BE(2)-C
Kelly
Kelly
BE(2)-C 60
40
20
0
80
solvent
PCI-34051
Cpd2
-
+
-
-
-
-
-
+
-
-
-
-
-
+
-
PCI-48000
-
-
-
-
+
**
**
*
BE(2)-C Cpd2
PCI-34051 PCI-48000
+
+
-
-
-
+
TrkA
-actin
kDa
140
42
Kelly TrkA 0.0
0.50
0.75
1.00
1.25
0
24
48
72
96 120 144
time [h]
1.50
0.25
solvent
PCI-34051
Cpd2
untreated
***
***
Kelly 0.00
0.25
0.50
0.75
1.00
1.25
1.50
0
24
48
72
96 120 144
time [h]
0.0
1.0
2.0
3.0
4.0
0
24
48
72
96 120 144
time [h]
solvent
PCI-34051
Cpd2
untreated
solvent
PCI-34051
Cpd2
untreated
***
***
***
***
IMR-32
BE(2)-C time [h] time [h] time [h] Cpd2
PCI-34051 PCI-48000
+
+
-
-
-
+
p21
-actin
kDa
42
21
IMR-32 0
20
40
60
80
0
24
48
72
96
144
time [h]
PCI-48000
solvent
PCI-34051
Cpd2
120
**
**
*
IMR-32 time [h] time [h] Figure 3
Phenotype characterization of HDAC8 inhibitors in vitro. (a) Flow cytometric quantitative analysis of neurofilament protein expression. BE(2)-C cells were treated
with HDAC8 inhibitors or solvent for 6 days. Blank (–/–/–/–): secondary antibody only. Discussion Shown are representative pictures from tumors of HDAC8-inhibitor-treated animals (PCI-48012 40 mg/kg
per day) and solvent-treated animals. H&E-stained sections, magnification ×400. Neurofilament (brownish color); bar chart reflects the quantitative analysis of neurofilament-
positive tumors (percentage). The quantification was carried out blinded to the treatment. Phosphorylated histone H3-positive cells (brownish color); bar chart reflects the
quantitative analysis of phospho-histone H3-positive cell numbers; t-test was used for statistics; bars represent mean values of 60 pictures (12 tumors per treatment and 5
pictures per slide) and error bars represent S.D. Active caspase-3 (brownish color); bar chart reflects the quantitative analysis of caspase-3-positive cell numbers; t-test was used
for statistics; bars represent mean values, error bars represent S.D. Scale bars = 100 μm. Different regions are shown for the different stainings Figure 5
Combined treatment of neuroblastoma cells with HDAC8 inhibitors and retinoic acid in vitro. (a) Representative pictures showing crystal violet-stained BE(2)-C cells
6 days after treatment with HDAC8 inhibitors Cpd2 (20 μM) or PCI-34051 (2 μM) and ATRA (10 μM) (scale bar = 200 μm). Immunofluorescent pictures show neurofilament (red)
and DAPI (blue) staining of BE(2)-C cells 6 days after treatment with HDAC8 inhibitors Cpd2 (20 μM) or PCI-34051 (2 μM) and ATRA (10 μM) (scale bar = 100 μm). (b) Bar chart
reflects the quantitative analysis of the number of neurites per cell of BE(2)-C cells. The quantification was carried out blinded to the treatment. T-test was used for statistics
(*Po0.05, **Po0.01 and ***Po0.001). Bars represent mean values and error bars represent S.D. (c) Bar chart reflects the quantitative analysis of the number of neurites per
cell of IMR-32 cells. T-test was used for statistics (*Po0.05, **Po0.01 and ***Po0.001). Bars represent mean values and error bars represent S.D. (d) Bar diagram displays
NTRK1 mRNA expression 6 days after treatment of BE(2)-C cells with HDAC8 inhibitors Cpd2 (20 μM) and PCI-48000 (1 μM) in combination with ATRA (10 μM). (e) Longer-term
colony assay of BE(2)-C cells treated once with HDAC8 inhibitors Cpd2 (C; 20 μM) and PCI-34051 (P; 2 μM), and ATRA (A; 10 μM) for at least 10 days. Results were quantified
and are displayed in the bar diagram. Bars represent mean values and error bars represent S.E.M. (***Po0.001). Discussion The treatment outcome of high-risk neuroblastoma tumors still
needs to be improved. Ongoing studies focus on the
identification of novel therapeutic approaches aiming at
oncogenic molecular targets to improve therapy efficacy for
stage 4 patients.26,27 The use of HDAC inhibitors is emerging Cell Death and Disease HDAC8 impairs cell death and differentiation
I Rettig et al 8 8 as an effective treatment strategy for cancer therapy.28,29
Several HDAC inhibitors are being tested in phase I–III clinical
trials and show significant responses in leukemias and
lymphoma. Most HDAC inhibitors block the activity of multiple
HDAC isozymes involved in numerous biological processes,
as demonstrated by the loss of function studies in mice. For example, the knockout of class I HDACs 1 and 3 manifests in
early embryonic lethality in mice and depletion of HDAC2
results in lethal cardiac hypertrophy.30–32 Simultaneous
inhibition of the activity of these HDACs yields a high potential
for toxicities that result in dose-limiting side effects, restricting
the full anticancer potential of HDAC inhibitors. Hence, ***
PCI-48012
solvent
5
10
15
0
20
25
***
10
20
30
40
0
20
40
60
80
0
100
PCI-48012
+
-
PCI-48012
+
-
PCI-48012
+
- ***
PCI-48012
solvent
5
10
15
0
20
25
***
10
20
30
40
0
20
40
60
80
0
100
PCI-48012
+
-
PCI-48012
+
-
PCI-48012
+
-
Figure 4
Phenotype characterization of HDAC8 inhibitors in vivo. Shown are representative pictures from tumors of HDAC8-inhibitor-treated animals (PCI-48012 40 mg/kg
per day) and solvent-treated animals. H&E-stained sections, magnification ×400. Neurofilament (brownish color); bar chart reflects the quantitative analysis of neurofilament-
positive tumors (percentage). The quantification was carried out blinded to the treatment. Phosphorylated histone H3-positive cells (brownish color); bar chart reflects the
quantitative analysis of phospho-histone H3-positive cell numbers; t-test was used for statistics; bars represent mean values of 60 pictures (12 tumors per treatment and 5
pictures per slide) and error bars represent S.D. Active caspase-3 (brownish color); bar chart reflects the quantitative analysis of caspase-3-positive cell numbers; t-test was used
for statistics; bars represent mean values, error bars represent S.D. Scale bars = 100 μm. Different regions are shown for the different stainings 5
10
15
0
20
25
***
PCI-48012
+
- ***
10
20
30
40
0
PCI-48012
+
- Figure 4
Phenotype characterization of HDAC8 inhibitors in vivo. Discussion (f) Western blot analysis of MYCN protein levels in BE(2)-C
cells 6 days after treatment with HDAC8 inhibitors Cpd2 (20 μM) or PCI-34051 (2 μM) and ATRA (10 μM) Cell Death and Disease HDAC8 impairs cell death and differentiation
I Rettig et al 9 9 9 study demonstrates superiority of selective HDAC isozyme
targeting versus pan-HDAC inhibition in terms of toxicity and
efficacy in a tumor model that is dependent on HDAC8. inhibition of one single HDAC isozyme could be more effective
and less toxic than the unspecific inhibition of several HDAC
family members by creating a larger therapeutic window. This ily members by creating a larger therapeutic window. This
efficacy in a tumor model that is dependent on HDAC8. Discussion ATRA
PCI
Cpd2
0.0
0.2
0.4
0.6
0.8
***
***
**
**
0.0
0.6
0.8
1.0
1.2
1.4
***
***
*
***
solvent
Cpd2
PCI34051
ATRA
PCI + ATRA
Cpd2 + ATRA
-
+
-
+
-
+
-
-
+
+
-
-
-
-
-
-
+
+
0.4
0.2
ATRA
PCI
Cpd2
-
+
-
+
-
+
-
-
+
+
-
-
-
-
-
-
+
+
ATRA
PCI
Cpd2
-
+
-
-
-
+
-
-
-
+
-
+
-
-
-
-
+
-
+
-
+
MYCN
-actin
kDa
42
55
1.0
0.2
1.1
0.8
0.5 0.03 0.03
ratio
ATRA
PCI
Cpd2
-
+
-
+
-
+
-
-
+
+
-
-
-
-
-
-
+
+
0.0
2.5
5.0
7.5
10.0
12.5
15.0
Solvent (S)
C A
ATRA (A)
Cpd2 (C)
PCI-34051
(P)
P A
***
***
0.0
0.5
50
100
1.0
S C P A CA PA
IMR-32
BE(2)-C
BE(2)-C
BE(2)-C
BE(2)-C
BE(2)-C solvent
Cpd2
PCI34051
ATRA
PCI + ATRA
Cpd2 + ATRA
BE(2)-C PCI34051 ATRA ATRA
PCI
Cpd2
0.0
0.2
0.4
0.6
0.8
***
***
**
**
0.0
0.6
0.8
1.0
1.2
1.4
***
***
*
***
-
+
-
+
-
+
-
-
+
+
-
-
-
-
-
-
+
+
0.4
0.2
ATRA
PCI
Cpd2
-
+
-
+
-
+
-
-
+
+
-
-
-
-
-
-
+
+
ATRA
PCI
Cpd2
-
+
-
-
-
+
-
-
-
+
-
+
-
-
-
-
+
-
+
-
+
MYCN
-actin
kDa
42
55
1.0
0.2
1.1
0.8
0.5 0.03 0.03
ratio
ATRA
PCI
Cpd2
-
+
-
+
-
+
-
-
+
+
-
-
-
-
-
-
+
+
0.0
2.5
5.0
7.5
10.0
12.5
15.0
Solvent (S)
C A
ATRA (A)
Cpd2 (C)
PCI-34051
(P)
P A
***
***
0.0
0.5
50
100
1.0
S C P A CA PA
IMR-32
BE(2)-C
BE(2)-C
BE(2)-C
BE(2)-C ATRA
PCI
Cpd2
0.0
0.2
0.4
0.6
0.8
***
***
**
**
0.0
0.6
0.8
1.0
1.2
1.4
***
***
*
***
-
+
-
+
-
+
-
-
+
+
-
-
-
-
-
-
+
+
0.4
0.2
ATRA
PCI
Cpd2
-
+
-
+
-
+
-
-
+
+
-
-
-
-
-
-
+
+
ATRA
PCI
Cpd2
-
+
-
+
-
+
-
-
+
+
-
-
-
-
-
-
+
+
0.0
2.5
5.0
7.5
10.0
12.5
15.0
IMR-32
BE(2)-C
BE(2)-C ATRA
PCI
Cpd2
0.0
0.2
0.4
0.6
0.8
***
***
**
**
0.0
0.6
0.8
1.0
1.2
1.4
***
***
*
***
-
+
-
+
-
+
-
-
+
+
-
-
-
-
-
-
+
+
0.4
0.2
ATRA
PCI
Cpd2
-
+
-
+
-
+
-
-
+
+
-
-
-
-
-
-
+
+
ATRA
PCI
Cpd2
-
+
-
-
-
+
-
-
-
+
-
+
-
-
-
-
+
-
+
-
+
MYCN
-actin
kDa
42
55
1.0
0.2
1.1
0.8
0.5 0.03 0.03
ratio
ATRA
PCI
Cpd2
-
+
-
+
-
+
-
-
+
+
-
-
-
-
-
-
+
+
0.0
2.5
5.0
7.5
10.0
12.5
15.0
Solvent (S)
C A
ATRA (A)
Cpd2 (C)
PCI-34051
(P)
P A
***
***
0.0
0.5
50
100
1.0
S C P A CA PA
IMR-32
BE(2)-C
BE(2)-C
BE(2)-C
BE(2)-C ATRA
PCI
Cpd2
0.0
0.2
0.4
0.6
0.8
***
***
**
**
-
+
-
+
-
+
-
-
+
+
-
-
-
-
-
-
+
+
BE(2)-C 0.0
0.6
0.8
1.0
1.2
1.4
***
***
*
***
0.4
0.2
ATRA
PCI
Cpd2
-
+
-
+
-
+
-
-
+
+
-
-
-
-
-
-
+
+
IMR-32 ATRA
PCI
Cpd2
-
+
-
+
-
+
-
-
+
+
-
-
-
-
-
-
+
+
0.0
2.5
5.0
7.5
10.0
12.5
15.0
BE(2)-C Solvent (S)
C A
ATRA (A)
Cpd2 (C)
PCI-34051
(P)
P A Solvent (S)
C A
ATRA (A)
Cpd2 (C)
PCI-34051
(P)
P A Cell Death and Disease HDAC8 impairs cell death and differentiation
I Rettig et al 10 0
200
400
600
800
1000
1200
1400
days after tumor transplantation
tumor volume [mm3]
tumor volume log[ml]
0
10
15
5
start of
treatment
*
***
***
4.5
5.0
5.5
6.0
6.5
days of treatment
0
4
6
2
8
10
solvent
13-cis RA
PCI-48012
PCI-48012 + 13-cis RA
13-cis RA
PCI-48012
PCI-48012 +
13-cis RA
solvent
additive effect
*
***
*** 0
200
400
600
800
1000
1200
1400
days after tumor transplantation
tumor volume [mm3]
0
10
15
5
start of
treatment
*
solvent
13-cis RA
PCI-48012
PCI-48012 + 13-cis RA 0
200
400
600
800
1000
1200
1400
days after tumor transplantation
tumor volume [mm3]
tumor volume log[ml]
0
10
15
5
start of
treatment
PCI-48012
solvent
13-cis RA
PCI-48012 +
13-cis RA
0
10
20
30
40
RA
PCI-48012
*
**
-
-
-
+
+
+
+
-
RA
PCI-48012
-
+
-
-
+
+
-
+
0
1
2
3
4
5
6
7
8
**
*
*
***
***
4.5
5.0
5.5
6.0
6.5
days of treatment
0
4
6
2
8
10
solvent
13-cis RA
PCI-48012
PCI-48012 + 13-cis RA
13-cis RA
PCI-48012
PCI-48012 +
13-cis RA
solvent
additive effect
*
***
***
cleaved caspase-3
phospho-histone H3
H&E
% P-H3 positive cells /image
cleaved caspase-3
positive cells /image
gure 6
Combined treatment with HDAC8 inhibitors and retinoic acid in vivo. Discussion (a) Left panel: Growth curves of BE(2)-C neuroblastoma xenografts in NMRI Foxn1
ated i.p. with HDAC8 inhibitor PCI-48012 (40 mg/kg/d), 13-cis retinoic acid (10 mg/kg/d) or combination of both compounds. Cohorts consisted of 12 animals each
rted 7 days after implantation of tumors (black arrow). Data are presented as mean tumor volume ± S.E.M.. Right panel: For assessment of in vivo effects, tumor vol
-transformed. A linear mixed model was used with fixed effect for slope in each group and random intercept for each mouse. Statistical analysis of additive or synerg
s performed using SAS Proc Mixed by evaluation of the interaction effect. The right panel shows the predicted change of log tumor volumes over time, displayed
es with group-specific slope. The red line shows the hypothetical change of log tumor volume if the combinational treatment was additive. Synergistic effects pro
ow the red line. *Po0.05, ***Po0.001. (b) Immunohistological analysis of tumor material. Shown are representative pictures from tumors of animals treated
AC8-inhibitor, 13-cis retinoic acid or the combination of both compounds. Hematoxylin & eosin (H&E) stained sections, magnification 400x. Phosphorylated histone H
s were counted and quantified as a percent of cells per image; t-test was used for statistics (*Po0.05, **Po0.01). Active caspase-3 (brownish color); bar chart
antitative analysis of caspase-3 positive cell numbers; t-test was used for statistics; bars represent mean values, error bars represent S.D. Scale bars = 100 μm
ions are shown for the different stainings tumor volume log[ml]
*
4.5
5.0
5.5
6.0
6.5
days of treatment
0
4
6
2
8
10
13-cis RA
PCI-48012
PCI-48012 +
13-cis RA
solvent
additive effect
*
***
*** tumor volume log[ml] days of treatment PCI-48012
solvent
13-cis RA
PCI-48012 +
13-cis RA
0
10
20
30
40
RA
PCI-48012
-
-
RA
PCI-48012
0
1
2
3
4
5
6
7
8
days of treatment
cleaved caspase-3
phospho-histone H3
H&E
% P-H3 positive cells /image
cleaved caspase-3
positive cells /image RA
PCI-48012
-
+
-
-
+
+
-
+
0
1
2
3
4
5
6
7
8
**
*
% P-H3 positive cells /image 0
10
20
30
40
RA
PCI-48012
*
**
-
-
-
+
+
+
+
-
cleaved caspase-3
positive cells /image Figure 6
Combined treatment with HDAC8 inhibitors and retinoic acid in vivo. (a) Left panel: Growth curves of BE(2)-C neuroblastoma xenografts in NMRI Foxn1 nude mice
treated i.p. Discussion with HDAC8 inhibitor PCI-48012 (40 mg/kg/d), 13-cis retinoic acid (10 mg/kg/d) or combination of both compounds. Cohorts consisted of 12 animals each. Treatment
started 7 days after implantation of tumors (black arrow). Data are presented as mean tumor volume ± S.E.M.. Right panel: For assessment of in vivo effects, tumor volumes were
log-transformed. A linear mixed model was used with fixed effect for slope in each group and random intercept for each mouse. Statistical analysis of additive or synergistic effects
was performed using SAS Proc Mixed by evaluation of the interaction effect. The right panel shows the predicted change of log tumor volumes over time, displayed as straight
lines with group-specific slope. The red line shows the hypothetical change of log tumor volume if the combinational treatment was additive. Synergistic effects produce a line
below the red line. *Po0.05, ***Po0.001. (b) Immunohistological analysis of tumor material. Shown are representative pictures from tumors of animals treated with either
HDAC8-inhibitor, 13-cis retinoic acid or the combination of both compounds. Hematoxylin & eosin (H&E) stained sections, magnification 400x. Phosphorylated histone H3-positive
cells were counted and quantified as a percent of cells per image; t-test was used for statistics (*Po0.05, **Po0.01). Active caspase-3 (brownish color); bar chart reflects the
quantitative analysis of caspase-3 positive cell numbers; t-test was used for statistics; bars represent mean values, error bars represent S.D. Scale bars = 100 μm. Different
regions are shown for the different stainings Figure 6
Combined treatment with HDAC8 inhibitors and retinoic acid in vivo. (a) Left panel: Growth curves of BE(2)-C neuroblastoma xenografts in NMRI Foxn1 nude mice
treated i.p. with HDAC8 inhibitor PCI-48012 (40 mg/kg/d), 13-cis retinoic acid (10 mg/kg/d) or combination of both compounds. Cohorts consisted of 12 animals each. Treatment
started 7 days after implantation of tumors (black arrow). Data are presented as mean tumor volume ± S.E.M.. Right panel: For assessment of in vivo effects, tumor volumes were
log-transformed. A linear mixed model was used with fixed effect for slope in each group and random intercept for each mouse. Statistical analysis of additive or synergistic effects
was performed using SAS Proc Mixed by evaluation of the interaction effect. The right panel shows the predicted change of log tumor volumes over time, displayed as straight
lines with group-specific slope. The red line shows the hypothetical change of log tumor volume if the combinational treatment was additive. Discussion Synergistic effects produce a line
below the red line. *Po0.05, ***Po0.001. (b) Immunohistological analysis of tumor material. Shown are representative pictures from tumors of animals treated with either
HDAC8-inhibitor, 13-cis retinoic acid or the combination of both compounds. Hematoxylin & eosin (H&E) stained sections, magnification 400x. Phosphorylated histone H3-positive
cells were counted and quantified as a percent of cells per image; t-test was used for statistics (*Po0.05, **Po0.01). Active caspase-3 (brownish color); bar chart reflects the
quantitative analysis of caspase-3 positive cell numbers; t-test was used for statistics; bars represent mean values, error bars represent S.D. Scale bars = 100 μm. Different
regions are shown for the different stainings We chose HDAC8-selective targeting in neuroblastoma for
this direct comparison since in our previous studies we
observed a particular oncogenic function of HDAC8 in
neuroblastoma among all HDAC family members investigated:
advanced-stage and metastasized neuroblastoma tumors
express high levels of HDAC8.15 Here, we demonstrate the
antineuroblastoma efficacy of an HDAC8 inhibitor in vivo at
concentrations avoiding unspecific and therapy-limiting side
effects. Vorinostat has been reported to inhibit primarily
HDACs 1, 2, 3 and 6,33 which are the ubiquitously expressed HDAC family members that are hypothesized to be respon-
sible for unwanted side effects.10 In contrast, HDAC8
expression has been found to be rather tissue-specific.34
Thus, in terms of unspecific side effects, the targeting of one
single enzyme seems to be superior to pan-HDAC inhibition
when applied in an appropriate tumor entity that displays
oncogenic dependency on that particular HDAC family
member. Nevertheless, besides HDAC8, other HDAC family
members also control tumor-suppressive functions in neuro-
blastoma. For example, HDAC1 in the sensitization of Cell Death and Disease HDAC8 impairs cell death and differentiation
I Rettig et al 11 -
+
+
+
+
+
-
-
-
-
-
-
BE(2)-C
supernatant
ATRA
HDAC8
+
+
-
+
HDAC8
GAPDH
ratio
1.0
4.1
44
kDa
37
1
10
100
1000
ATRA
HDAC8
NC vector
SN
RA
cell death &
differentiation
HDAC8i
proliferation
survival
CREB
HDAC8
P-CREB
GAPDH
P-CREB
CREB
43
kDa
43
HDAC8 -
+
ratio
0.7
0.5
1.0
1.0
0.9
1.2
1.0
1.2
ratio
37
+
-
BE(2)-C
1
2
3
ATRA
HDAC8
-
-
0
4
+
+
-
+
-
+
*
Figure 7
HDAC8 overexpression counteracts ATRA treatment. (a) Longer-term colony assay (18 days) of BE(2)-C cells stably overexpressing HDAC8, treated with ATRA
(10 μM), where indicated. Discussion Supernatant containing detached floating cells from densely grown untreated cells was transferred into a new 6-well plate and colonies were stained
10 days after replating. Results were quantified and are displayed in the bar diagram. Bars represent mean values and error bars represent S.E.M. (***Po0.001). SN,
supernatant cells. (b) Hypothetical model: HDAC8 and ATRA signaling might converge at the level of CREB. Combined treatment of neuroblastoma cells with HDAC8 inhibitors
and retinoic acid might enhance CREB-mediated differentiation. (c) Western blot analysis of BE(2)-C cells stably overexpressing HDAC8, where indicated, for P-CREB and CREB
levels. (d) Bar diagram displays TH mRNA expression 72 h after treatment of BE(2)-C cells stably overexpressing HDAC8, where indicated, with ATRA (10 μM) (*Po0.05). -
+
+
+
+
+
-
-
-
-
-
-
BE(2)-C
supernatant
ATRA
HDAC8
+
+
-
+
HDAC8
GAPDH
ratio
1.0
4.1
44
kDa
37
1
10
100
1000
ATRA
HDAC8
NC vector
SN BE(2)-C 1
2
3
ATRA
HDAC8
-
-
0
4
+
+
-
+
-
+
* GAPDH
P-CREB
CREB
43
kDa
43
HDAC8 -
+
ratio
0.7
0.5
1.0
1.0
0.9
1.2
1.0
1.2
ratio
37
+
-
BE(2)-C RA
cell death &
differentiation
HDAC8i
proliferation
survival
CREB
HDAC8
P-CREB Figure 7
HDAC8 overexpression counteracts ATRA treatment. (a) Longer-term colony assay (18 days) of BE(2)-C cells stably overexpressing HDAC8, treated with ATRA
(10 μM), where indicated. Supernatant containing detached floating cells from densely grown untreated cells was transferred into a new 6-well plate and colonies were stained
10 days after replating. Results were quantified and are displayed in the bar diagram. Bars represent mean values and error bars represent S.E.M. (***Po0.001). SN,
supernatant cells. (b) Hypothetical model: HDAC8 and ATRA signaling might converge at the level of CREB. Combined treatment of neuroblastoma cells with HDAC8 inhibitors
and retinoic acid might enhance CREB-mediated differentiation. (c) Western blot analysis of BE(2)-C cells stably overexpressing HDAC8, where indicated, for P-CREB and CREB
levels. (d) Bar diagram displays TH mRNA expression 72 h after treatment of BE(2)-C cells stably overexpressing HDAC8, where indicated, with ATRA (10 μM) (*Po0.05). multidrug-resistant neuroblastoma cell lines to cytotoxic
agents,35 HDAC2 in repressing miR-183-mediated tumor
suppression36 and HDAC3 in negatively regulating tumor
suppressor GRHL1.37 Recently, we have discovered a novel
function for HDAC10 in promoting autophagy-mediated cell
survival. Animal studies
M Plasma
samples (50 μl) were analyzed for Cpd2 concentration after liquid/liquid extraction with
tert.-butylmethylether using HPLC coupled to tandem mass spectrometry (LC/MS/MS;
TSP 2000, TSQ 7000; Thermo Fisher Scientific, Waltham, MA, USA). The extracts
were gradient chromatographed on a Hydro-RP column (4 μM, 150 × 2.1 mm2;
Phenomenex, Torrance, CA, USA) at 40 °C using 5 mM ammonium acetate and
acetonitrile at 0.6 ml/min. The eluent was directly infused into the electrospray ion
source (negative ions) of the MS/MS system and the mass transitions of Cpd2 (m/z
186.1 →142.0 and m/z 186 →58.0) and vorinostat (internal standard; m/z
268.1 →97.0) were monitored in the multiple reaction monitoring mode. Cpd2 and
vorinostat eluted within 6 min and the respective peak area ratios were used for the
quantification of Cpd2 using linear regression in the calibrated range from 100 to
20 000 ng/ml. The limit of quantification was 100 ng/ml and the correlation coefficient
was 40.99. Accuracy and precision control using quality control samples revealed
deviation consistently below 15%, which is in accordance with the FDA guideline,
Guidance for Industry: Bioanalytical Method Validation. All animal studies were
approved by the German Cancer Research Center (DKFZ) institutional animal care
and use committee and the Regional Administrative Council Karlsruhe, Germany. All
experiments were in accordance with the relevant regulatory standards. Taken together, our data demonstrate effectiveness of an
HDAC-selective inhibitor in a preclinical model of neuroblastoma. From a clinical perspective, combination of HDAC8 inhibition with
retinoic acid treatment might be a promising strategy in the
maintenance treatment of high-risk neuroblastoma. Materials and Methods
Cell culture and transfections. All cell lines were grown under standard
conditions as described previously.38 Human neuroblastoma cell lines BE(2)-C
(ECACC, Salisbury, UK), IMR-32 (DSMZ, Braunschweig, Germany), Kelly (DSMZ),
SH-SY5Y (DSMZ), SK-N-AS and SH-EP (both generously provided by the
laboratory of M Schwab), and medulloblastoma cell lines UW-228-2, DAOY and
ONS7647 were grown under standard conditions in DMEM with L-glutamine, 4.5 g/l
glucose (Lonza, Basel, Switzerland) and 1% non-essential amino acids (NEAA)
(Invitrogen, Darmstadt, Germany), or RPMI1640 with L-glutamine (Lonza) and 1%
NEAA, or EMEM with L-glutamine. All media were supplemented with 10% fetal
bovine serum (FBS) (Sigma, Munich, Germany). Non-transformed, proliferatively
active primary human skin fibroblasts from an infant donor were a friendly gift from
Petra Boukamp, German Cancer Research Center (DKFZ), Heidelberg, Germany. Fibroblasts were maintained in DMEM/HAM´s F12 (Invitrogen), supplemented with
10% FBS and 1% NEAA. Animal studies
M Animal studies
Mouse xenograft studies with HDAC8 knockdown: BE(2)-C cells
transiently transfected with siRNA against HDAC8 or negative control were
resuspended in Matrigel (BD, Franklin Lakes, NJ, USA) and 20 U/ml heparin 48 h
after transfection. Next, 1.3 × 106 viable cells were implanted in 100 μl Matrigel into
the subcutaneous tissue of right flank of 5- to 6-week-old female athymic nude mice
(NMRI Foxn1; Harlan Laboratories Inc., Nijmegen, The Netherlands) in cohorts of
20 animals each. Cell numbers were established via trypan blue exclusion. Tumor
dimensions were measured and tumor volume (mm3) was calculated as
(L × W2) × π/6 (L indicating length, W indicating width (all in mm)). At days 3, 6
and 9 after tumor cell injection, three randomly selected mice from each group were
killed, and tumors were explanted. Specimens used for isolation of total RNA or
protein were shock frozen in liquid nitrogen immediately after removal and stored at
−80 °C. A part of each tumor sample was formalin fixed and embedded in paraffin
for immunohistochemical analysis. The transfection efficiency of siRNA in BE(2)-C
cells was 90% as determined by fluorescently labeled siRNA siGLO Lamin A/C
(Dharmacon, Lafayette, CO, USA). Preparation of PBMC cells: Whole blood (100 μl) was collected from the
femoral vein of untreated athymic NMRI nude mice (HsdCpb:NMRI-Foxn1nu; Harlan
Laboratories Inc.), and respective animals were killed 2 h after intraperitoneal
injection of substances. For the analysis of whole acetylation levels, mice were killed
at half-life time of the inhibitors. Whole blood with heparin was added to RBC lysis
buffer (diluted 1 : 10) (BioLegend, San Diego, CA, USA) at a ratio of 1 : 4. Next, red
blood cell lysis was performed according to the manufacturer's instructions. PBMCs
were lysed in SDS lysis buffer for western blot or processed for FACS analysis. Determination of inhibitor plasma half-life: To determine the half-life of
HDAC8 inhibitor Cpd2 in vivo, the substance was injected intraperitoneally at a dose
of 100 mg/kg into athymic NMRI nude mice (Harlan). Whole blood (100 μl) was
collected from the femoral vein of animals at time points 0, 5, 10, 15, 20, 30, 45, 60,
75 and 90 min after injection (six animals per time point). Plasma was obtained by
centrifugation at 10 000 r.c.f. for 10 min at 4 °C and stored at −80 °C. Discussion However, the effects on CREB phosphorylation were
rather modest and further experiments are required to unravel
completely the link between retinoic acid and HDAC8 inhibitor
combination treatment-induced cell death and differentiation
mechanism in neuroblastoma. control (pCEP4/hygro-FLAG). Transfected cells were selected with hygromycin
(400 μg/ml) for 3 weeks. A mixed population of hygromycin-resistant cells was used
for experimental analysis. on-target validation in tumor tissue. Quite recently, the
cohesion complex protein SMC317 and the tumor suppressor
ARID1A44 have been described to be deacetylated in an
HDAC8-dependent manner. Additionally, HDAC8 has been
linked to phosphatase PP1-mediated CREB inactivation.24
Activation of CREB during ATRA treatment has been
described before45 and phosphorylation of CREB has been
postulated
to
mediate
retinoic
acid-induced
caspase-8
expression in neuroblastoma cells.23 Caspase-8 is frequently
inactivated by epigenetic silencing in many tumors, including
neuroblastomas, and the expression was induced by the use
of epigenetically acting compounds in combination with
interferon-γ.46 As RA treatment has been described to activate
CREB and HDAC8 expression inactivates CREB, we hypothe-
size that the combined treatment with RA and HDAC8
inhibitors might synergistically act on CREB-mediated cell
differentiation. Indeed, our data show that a combination of
both HDAC8 inhibitor and 13-cis retinoic acid, a currently
applied drug in neuroblastoma treatment protocols, strongly
enhanced differentiation in cell culture and decreased tumor
growth in vivo. Mechanistically, HDAC8 counteracts retinoic
acid-mediated CREB signaling and inhibition of clonogenic
growth and combined treatment with HDAC8 inhibitor and
ATRA enhance CREB phosphorylation. This suggests that the
molecular action of both compounds converge at the level of
CREB. However, the effects on CREB phosphorylation were
rather modest and further experiments are required to unravel
completely the link between retinoic acid and HDAC8 inhibitor
combination treatment-induced cell death and differentiation
mechanism in neuroblastoma. Discussion HDAC10 depletion in neuroblastoma cells interrupted
autophagic flux and induced accumulation of autophago-
somes, lysosomes and a substrate of the autophagic
degradation pathway, p62/SQSTM1. Disrupted autophagy
was associated with sensitization to cytotoxic drug treatment
in a panel of highly malignant neuroblastoma cell lines.38 application of HDAC8-selective inhibitors. So far, HDAC8
has been reported to have specific function in the development
of cranial neural crest cells of mice, as knockout of HDAC8
leads to skull instability and perinatal death.39 Recently,
human HDAC8 variants have been described, which are
associated with X-linked intellectual disability disorders. All
males that carried the variant in HDAC8 showed microcephaly
and distinct focal deformation of the skull.39 HDAC8 itself has
been described to be regulated by SOX4,40 a transcription
factor of the SOX family, which is required for B-lymphocyte
development,41 and also during development of the sympa-
thetic nervous system.42,43 Our experiments also focused on the understanding of the
molecular function of HDAC8 in neuroblastoma to find
rationally
targeted
treatment
combinations
for
clinical y
The identification of HDAC8-specific substrates is still of
major interest and would be the best tool for HDAC8 inhibitor Cell Death and Disease HDAC8 impairs cell death and differentiation
I Rettig et al 12 on-target validation in tumor tissue. Quite recently, the
cohesion complex protein SMC317 and the tumor suppressor
ARID1A44 have been described to be deacetylated in an
HDAC8-dependent manner. Additionally, HDAC8 has been
linked to phosphatase PP1-mediated CREB inactivation.24
Activation of CREB during ATRA treatment has been
described before45 and phosphorylation of CREB has been
postulated
to
mediate
retinoic
acid-induced
caspase-8
expression in neuroblastoma cells.23 Caspase-8 is frequently
inactivated by epigenetic silencing in many tumors, including
neuroblastomas, and the expression was induced by the use
of epigenetically acting compounds in combination with
interferon-γ.46 As RA treatment has been described to activate
CREB and HDAC8 expression inactivates CREB, we hypothe-
size that the combined treatment with RA and HDAC8
inhibitors might synergistically act on CREB-mediated cell
differentiation. Indeed, our data show that a combination of
both HDAC8 inhibitor and 13-cis retinoic acid, a currently
applied drug in neuroblastoma treatment protocols, strongly
enhanced differentiation in cell culture and decreased tumor
growth in vivo. Mechanistically, HDAC8 counteracts retinoic
acid-mediated CREB signaling and inhibition of clonogenic
growth and combined treatment with HDAC8 inhibitor and
ATRA enhance CREB phosphorylation. This suggests that the
molecular action of both compounds converge at the level of
CREB. Animal studies
M 13-Cis retinoic acid (Sigma) was dissolved in 90% DMSO, 10% PBS and
given by intraperitoneal injection at a concentration of 10 mg/kg per day for
2 × 5 days. Tumor volume (mm3) was calculated as described above (L × W × H) ×
π/6, where L indicates length, W indicates width and H indicates height (all in mm)). At explantation, tumor material was processed as described above. Intracellular protein quantification via FACS. Cells were collected,
washed with PBS and fixed with 2% paraformaldehyde for 15 min at room
temperature (RT). Cells were permeabilized by incubation with 50 μg/ml digitonin for
6 min at RT. Cells were incubated with primary antibody overnight at 4 °C. After
staining with Cy3-conjugated anti-rabbit antibody, cells were analyzed by FACS
using CellQuest software (BD Biosciences, San Jose, CA, USA). The following
primary antibodies were used: anti-acetylated lysine (polyclonal; Cell Signaling) and
anti-neurofilament-M (clone EP2460; Epitomics, Burlingame, CA, USA). Crystal violet staining. Cells were fixed with 2% paraformaldehyde for
15 min at RT. After washing with PBS, sterile filtered crystal violet (0.005% in H2O)
was added for 1 h at RT. Cells were washed three times with H2O. Morphometrical
studies were performed on the photomicroscope CKX41 (Olympus, Hamburg,
Germany) and analyzed with cellB software (Olympus). All animal studies were approved by the German Cancer Research Center (DKFZ)
institutional animal care and use committee and the Regional Administrative Council
Karlsruhe, Germany. All experiments were in accordance with the relevant regulatory
standards. Immunohistochemistry. Immunohistochemical stainings were performed
semiautomatically on 5-μM-thick sections of formalin-fixed, paraffin-embedded
specimens using VENTANA Benchmark XT (Strasbourg, France). The following
antibodies were
applied
according
to
the
manufacturer's
protocols:
anti-
neurofilament protein (Clone2F11; DakoCytomation, Hamburg, Germany; diluted
1 : 200), anti-phospho-histone H3 (Ser10) (polyclonal; Biocare Medical, Concord,
CA, USA; diluted 1 : 100), anti-active caspase-3 (polyclonal; Abcam, Cambridge,
UK; diluted 1 : 100). Antigen retrieval was performed before primary antibody
exposure. Primary antibody binding sites were visualized using the Ultra-View
Universal DAB Detection Kit (Ventana/Roche, Mannheim, Germany). Nuclei were
automatically counterstained by hematoxylin and blueing reagent (Ventana). From
each specimen, five representative digital images of neurofilament, phosphohistone
H3- and caspase-3-stained sections were captured at × 200 magnification (high
power field, HPF) using a photomicroscope (CKX41; Olympus). Cell imaging
software cellB (Olympus) was used for the acquisition of microscopic images. The
quantification was performed by operators who were blinded to the treatment group
represented in the images. Animal studies
M Positive cells were counted with the counting tool of the
cellB software (Olympus). Hematoxylin and eosin (H&E) staining was analyzed at
× 400 magnification (HPF) using photomicroscope Axioskop 2 (Zeiss, Oberkochen,
Germany) equipped with AxioCam ICc1 (Zeiss) and AxioVision Software Release
4.6 (Zeiss). Reagents. HDAC8-selective inhibitors Cpd213 (stock concentration 250 mM),
PCI-34051 (Figure 2a)14 (stock 20 mM; Pharmacyclics Inc., Sunnyvale, CA, USA),
PCI-48000 (stock 20 mM; Pharmacyclics) and PCI-48012 (Pharmacyclics), as well
as unselective HDAC inhibitors vorinostat (SAHA: suberoylanilide hydroxamic acid)
(stock 100 mM; Selleck Chemicals, Houston, TX, USA) and TSA (stock 1 mM;
Merck, Darmstadt, Germany), were dissolved in DMSO. Cpd2, PCI-34051, and
PCI-48000 were used in in vitro experiments. For in vivo experiments, PCI-48012
was applied. PCI-48000 and PCI-48012 are chemical modifications of PCI-34051
with similar HDAC8 selectivity profiles and similar IC50 values, but longer plasma
half-life in the case of PCI-48012. All-trans retinoic acid (stock 10 mM; Sigma) was
dissolved in ethanol. In vitro HDAC activity assay. The inhibitory activity of HDAC8-selective
inhibitors towards all classical 1–11 HDACs was investigated by Reaction Biology
Corporation (Malvern, PA, USA). The catalytic domains of human HDACs were
expressed by baculovirus expression system in Sf9 cells. Enzymes were stored in
50 mM Tris-HCl (pH8.0), 138 mM NaCl, 20 mM glutathione and 10% glycerol, and
were stable for 46 months at −80 °C. The purity was ascertained by SDS-PAGE. Peptide substrate was conjugated with AMC. Reaction buffer was 25 mM Tris-Cl,
pH8.0, 137 mM NaCl, 2.7 mM KCl, 1 mM MgCl2 and 0.1 mg/ml BSA. The HDAC
reaction was performed at 30 °C for 2 h before adding the developer reagent. The
free AMC was detected with excitation of 360 nm and emission 460 nm at kinetic
mode for 90 min. IC50 values were calculated with GraphPad Prism version 3.0a
(GraphPad Software Inc., San Diego, CA, USA). Statistical analysis. For in vitro experiments, either a two-sided t-test or an
one-way analysis of variance was carried out (SigmaPlot 12.0). Means are
represented in bar charts, and error bars represent the S.D. of at least three
independent experiments. P-values o0.05 were considered significant. For
assessment of in vivo effects, tumor volumes were log-transformed. A linear mixed
model was used with fixed slope for each group and random intercept for each
mouse. Statistical analysis of additive or synergistic effects was performed using
SAS Proc Mixed (Version 9.2; SAS Institute, Cary, NC, USA). Animal studies
M Neuroblastoma cell lines were last genotyped in April
2012 (DSMZ). Medulloblastoma cell lines were last genotyped in October 2012
(DKFZ). All cell lines were routinely tested for mycoplasma contamination. Human
infant astrocytes (obtained from S Pfister, German Cancer Research Center
(DKFZ)) were grown as described previously.38 Determination of MTD: HDAC8-selective inhibitors were intraperitoneally
injected into athymic NMRI nude mice (Harlan) in increasing doses from 40 to
400 mg/kg per day. DLTs were monitored by observation of general condition and
determination of body weight. We used several criteria for objectively measuring
DLT. These included a decrease in body weight of ≥20% compared with baseline
body weight, blood analysis of standard parameters for clinical chemistry and
hematology (central laboratory of University Hospital Heidelberg, Heidelberg,
Germany) and organ histopathology (formalin-fixed, paraffin-embedded organs), by
H&E stain for the detection of necrosis (Prof. Dr. AD Gruber, Institute for Veterinary
Pathology, Berlin, Germany). These measurements assured lack of toxicities at
MTD concentrations of inhibitors. Transient transfections were performed as described previously.15 The following
siRNAs were used: HDAC8 (siRNA1 (ID 120597, exons 1 and 2; Ambion
(Huntingdon, UK) Ltd)) and corresponding negative control siRNA (Silencer Negative
Control 1; Ambion). Generation of stable cell lines: Human BE(2)-C cell lines stably expressing
HDAC8 were established by transfection using Effectene (Qiagen, Hilden, Germany)
with pCEP4/hygro-FLAG HDAC8 cDNA or hygromycin-resistant empty vector Cell Death and Disease HDAC8 impairs cell death and differentiation
I Rettig et al 13 (clone AC-15; Sigma), anti-P-CREB (Ser133; Cell Signaling, Danvers, MA, USA),
anti-CREB (48H2; Cell Signaling) and anti-GAPDH (clone 6C5; Merck). Mouse xenograft studies with HDAC8 inhibitors: Neuroblastoma cells
were resuspended in Matrigel and 20 U/ml heparin. Next, 2 × 106 (BE(2)-C) or
4 × 106 (IMR-32) viable cells were implanted in 100 μl Matrigel into the
subcutaneous tissue of right flank of 5- to 6-week old female athymic nude mice
(Harlan). Mice were randomly assigned to groups of 12 individuals bearing similarly
sized tumors. HDAC8-selective inhibitor PCI-48012 at MTD (40 mg/kg per day) was
dissolved in 90% DMSO (HybriMax; Sigma), 10% PBS and given by intraperitoneal
injection for 2 × 5 days. HDAC inhibitor vorinostat was dissolved in 100% DMSO and
given by intraperitoneal injection at a concentration of 150 mg/kg per day for
2 × 5 days. Animal studies
M Tumor weights were
analyzed with the nonparametric Mann–Whitney test (GraphPad Prism version
3.0a). Error bars represent the S.E.M. P-values o0.05 were considered significant. Cell counting, cell viability, cell death and colony assay. Cells were
collected, pooled with corresponding supernatant, centrifuged and resuspended in
1.5 ml cell culture media. Cell count as well as cell viability was measured by
automated trypan blue staining with Vi-Cell XR Cell Viability Analyzer from Beckman
Coulter (Krefeld, Germany). Caspase-3-like protease activity assay was performed
as described previously.15 Colony assay. In six-well plates, 500 cells were seeded and treated as
indicated. Viable colonies were stained after a minimum of 10 days with crystal
violet. For quantification, the mean intensity of each well of the 8-bit binary picture
was measured with ImageJ software (U. S. National Institutes of Health, Bethesda,
MD, USA; http://imagej.nih.gov/ij/). Web-based gene expression analysis: R2 (R2: microarray analysis and
visualization platform; http://r2.amc.nl) was used to investigate HDAC8 expression in
a publically available cohort of primary neuroblastoma patients (Academic Medical
Center (AMC) cohort; Gene Expression Omnibus (GEO) database accession no. GSE16476). The following probeset was used to detect HDAC8 expression:
223345_at. Patient characteristics were published previously.16 Real-time, reverse-transcription polymerase chain reaction. Real-
time PCR was performed as described previously.15 Data were normalized against
housekeeping genes SDHA and HPRT 48 and set in relation to negative control. Conflict of Interest The authors declare no conflict of interest. The authors declare no conflict of interest. Western blot analysis. Western blot analysis was performed as described
previously.15 For detection of phosphoproteins, PhosStop (Roche, Penzberg,
Germany) was added to the lysis buffer. The following antibodies were used for
detection: anti-HDAC8 (H-145) (polyclonal; Santa Cruz, Santa Cruz, CA, USA), anti-
acetyl tubulin (clone 6-11B-1; Sigma), anti-acetyl-histone H4 (polyclonal; Upstate,
Lake Placid, NY, USA), anti-acetyl-SMC3 (provided by Prof. K Shirahige, University
of Tokyo, Tokyo, Japan), anti p21waf1/cip1 (clone CP74; Merck), anti-Trk (C-14)
(polyclonal; Santa Cruz), anti-neurofilament-M (polyclonal; Merck), anti-β-actin Acknowledgements. We thank A Bittmann and R Straub for excellent technical
assistance, and P Boukamp (Genetics of Skin Carcinogenesis, German Cancer
Research Center (DKFZ), Heidelberg, Germany) for kindly providing us with primary
human skin fibroblasts from an infant donor. We gratefully thank Sriram
Balasubramanian and Pharmacyclics Inc. (Sunnyvale, CA, USA) for the HDAC8-
selective inhibitors PCI-34051, PCI-48000 and PCI-48012 and Scott M Ulrich Cell Death and Disease Cell Death and Disease HDAC8 impairs cell death and differentiation
I Rettig et al 14 14 (Department of Chemistry, Ithaca College, Ithaca, NY, USA) for the HDAC8-selective
inhibitor Cpd2. We thank Prof. Katsuhiko Shirahige for kindly providing us with the Ac-
SMC3 antibody. This work was supported by the German Federal Ministry of
Economy and Technology through ZIM (to IO, UF and AS), and by grants from the
Wilhelm Sander Foundation (Grant no 2009.008.1) (to IO), the BMBF through
NGFNplus (to HED and OW), the University of Heidelberg through the FRONTIER
program (to HED) and a B Braun foundation grant (to TM). 26. Matthay KK, George RE, Yu AL. Promising therapeutic targets in neuroblastoma. Clin
Cancer Res 2012; 18: 2740–2753. 27. Pugh TJ, Morozova O, Attiyeh EF, Asgharzadeh S, Wei JS, Auclair D et al. The genetic
landscape of high-risk neuroblastoma. Nat Genet 2013; 45: 279–284. 28. Minucci S, Pelicci PG. Histone deacetylase inhibitors and the promise of epigenetic
(and more) treatments for cancer. Nat Rev Cancer 2006; 6: 38–51. (
)
29. Yoo CB, Jones PA. Epigenetic therapy of cancer: past, present and future. Nat Rev Drug
Discov 2006; 5: 37–50. 30. Lagger G, O'Carroll D, Rembold M, Khier H, Tischler J, Weitzer G et al. Essential function of
histone deacetylase 1 in proliferation control and CDK inhibitor repression. EMBO J 2002;
21: 2672–2681. 1. Brodeur GM. Neuroblastoma: biological insights into a clinical enigma. Nat Rev Cancer
2003; 3: 203–216. 31. The authors declare no conflict of interest. Fabian J, Lodrini M, Oehme I, Schier MC, Thole TM, Hielscher T et al. GRHL1 acts as tumor
suppressor in neuroblastoma and is negatively regulated by MYCN and HDAC3. Cancer Res
2014; 74: 2604–2616. 8. Kelly WK, O'Connor OA, Krug LM, Chiao JH, Heaney M, Curley T et al. Phase I study of an
oral histone deacetylase inhibitor, suberoylanilide hydroxamic acid, in patients with
advanced cancer. J Clin Oncol 2005; 23: 3923–3931. 8. Kelly WK, O'Connor OA, Krug LM, Chiao JH, Heaney M, Curley T et 38. Oehme I, Linke JP, Böck BC, Milde T, Lodrini M, Hartenstein B et al. Histone deacetylase 10
promotes autophagy-mediated cell survival. Proc Natl Acad Sci USA 2013; 110:
E2592–2601. 9. Lane AA, Chabner BA. Histone deacetylase inhibitors in cancer therapy. J Clin Oncol 2009;
27: 5459–5468. 10. Witt O, Deubzer HE, Milde T, Oehme I. HDAC family: What are the cancer relevant targets? Cancer Lett 2009; 277: 8–21. 39. Haberland M, Mokalled MH, Montgomery RL, Olson EN. Epigenetic control of skull
morphogenesis by histone deacetylase 8. Genes Dev 2009; 23: 1625–1630. 11. Balasubramanian S, Verner E, Buggy JJ. Isoform-specific histone deacetylase inhibitors: the
next step? Cancer Lett 2009; 280: 211–221. p
g
y
y
40. Higuchi T, Nakayama T, Arao T, Nishio K, Yoshie O. SOX4 is a direct target gene of FRA-2 40. Higuchi T, Nakayama T, Arao T, Nishio K, Yoshie O. SOX4 is a direct target gene of FRA-2
and induces expression of HDAC8 in adult T-cell leukemia/lymphoma. Blood 2013; 121:
3640–3649. g
,
y
,
,
,
g
g
and induces expression of HDAC8 in adult T-cell leukemia/lymphoma. Blood 2013; 121:
3640–3649. 12. Somoza JR, Skene RJ, Katz BA, Mol C, Ho JD, Jennings AJ et al. Structural snapshots of
human HDAC8 provide insights into the class I histone deacetylases. Structure 2004; 12:
1325–1334. 41. Schilham MW, Oosterwegel MA, Moerer P, Ya J, de Boer PA, van de Wetering M et al. Defects in cardiac outflow tract formation and pro-B-lymphocyte expansion in mice lacking
Sox-4. Nature 1996; 380: 711–714. 13. Krennhrubec K, Marshall BL, Hedglin M, Verdin E, Ulrich SM. Design and evaluation of
'Linkerless' hydroxamic acids as selective HDAC8 inhibitors. Bioorg Med Chem Lett 2007;
17: 2874–2878. 42. Potzner MR, Tsarovina K, Binder E, Penzo-Méndez A, Lefebvre V, Rohrer H et al. Sequential
requirement of Sox4 and Sox11 during development of the sympathetic nervous system. The authors declare no conflict of interest. Montgomery RL, Potthoff MJ, Haberland M, Qi X, Matsuzaki S, Humphries KM et al. Maintenance of cardiac energy metabolism by histone deacetylase 3 in mice. J Clin Invest
2008; 118: 3588–3597. 2. Van Roy N, De Preter K, Hoebeeck J, Van Maerken T, Pattyn F, Mestdagh P et al. The emerging molecular pathogenesis of neuroblastoma: implications for improved risk
assessment and targeted therapy. Genome Med 2009; 1: 74. 32. Trivedi CM, Luo Y, Yin Z, Zhang M, Zhu W, Wang T et al. Hdac2 regulates the cardiac
hypertrophic response by modulating Gsk3 beta activity. Nat Med 2007; 13: 324–331. 3. Witt O, Deubzer HE, Lodrini M, Milde T, Oehme I. Targeting histone deacetylases in
neuroblastoma. Curr Pharm Des 2009; 15: 436–447. 33. Bradner JE, West N, Grachan ML, Greenberg EF, Haggarty SJ, Warnow T et al. Chemical
phylogenetics of histone deacetylases. Nat Chem Biol 2010; 6: 238–243. 4. Buggy JJ, Sideris ML, Mak P, Lorimer DD, McIntosh B, Clark JM. Cloning and
characterization of a novel human histone deacetylase, HDAC8. Biochem J 2000; 350(Part 1):
199–205. 34. Waltregny D, De Leval L, Glénisson W, Ly Tran S, North BJ, Bellahcène A et al. Expression
of histone deacetylase 8, a class I histone deacetylase, is restricted to cells showing smooth
muscle differentiation in normal human tissues. Am J Pathol 2004; 165: 553–564. 5. Choudhary C, Kumar C, Gnad F, Nielsen ML, Rehman M, Walther TC et al. Lysine
acetylation targets protein complexes and co-regulates major cellular functions. Science
2009; 325: 834–840. 35. Keshelava N, Davicioni E, Wan Z, Ji L, Sposto R, Triche TJ et al. Histone deacetylase 1 gene
expression and sensitization of multidrug-resistant neuroblastoma cell lines to cytotoxic
agents by depsipeptide. J Natl Cancer Inst 2007; 99: 1107–1119. 6. Glozak MA, Sengupta N, Zhang X, Seto E. Acetylation and deacetylation of non-histone
proteins. Gene 2005; 363: 15–23. 36. Lodrini M, Oehme I, Schroeder C, Milde T, Schier MC, Kopp-Schneider A et al. MYCN and
HDAC2 cooperate to repress miR-183 signaling in neuroblastoma. Nucleic Acids Res 2013;
41: 6018–6033. p
7. Marks P, Rifkind RA, Richon VM, Breslow R, Miller T, Kelly WK. Hist 7. Marks P, Rifkind RA, Richon VM, Breslow R, Miller T, Kelly WK. Histone deacetylases and
cancer: causes and therapies. Nat Rev Cancer 2001; 1: 194–202. cancer: causes and therapies. Nat Rev Cancer 2001; 1: 194–202. 37. 1. Brodeur GM. Neuroblastoma: biological insights into a clinical enigma. Nat Rev Cancer
2003; 3: 203–216. The authors declare no conflict of interest. Development 2010; 137: 775–784. 14. Balasubramanian S, Ramos J, Luo W, Sirisawad M, Verner E, Buggy JJ. A novel histone
deacetylase 8 (HDAC8)-specific inhibitor PCI-34051 induces apoptosis in T-cell lymphomas. Leukemia 2008; 22: 1026–1034. 43. Bergsland M, Werme M, Malewicz M, Perlmann T, Muhr J. The establishment of neuronal
properties is controlled by Sox4 and Sox11. Genes Dev 2006; 20: 3475–3486. 15. Oehme I, Deubzer HE, Wegener D, Pickert D, Linke JP, Hero B et al. Histone deacetylase 8
in neuroblastoma tumorigenesis. Clin Cancer Res 2009; 15: 91–99. 44. Olson DE, Udeshi ND, Wolfson NA, Pitcairn CA, Sullivan ED, Jaffe JD et al. An unbiased
approach to identify endogenous substrates of "histone" deacetylase 8. ACS Chem Biol
2014; 17: 2210–6. 16. Molenaar JJ, Koster J, Ebus ME, van Sluis P, Westerhout EM, de Preter K et al. Copy
number defects of G1-cell cycle genes in neuroblastoma are frequent and correlate with high
expression of E2F target genes and a poor prognosis. Genes Chromosomes Cancer 2012;
51: 10–19. 45. Chu PW, Cheung WM, Kwong YL. Differential effects of 9-cis, 13-cis and all-trans retinoic
acids on the neuronal differentiation of human neuroblastoma cells. Neuroreport 2003; 14:
1935–1939. 17. Deardorff MA, Bando M, Nakato R, Watrin E, Itoh T, Minamino M et al. HDAC8 mutations in
Cornelia de Lange syndrome affect the cohesin acetylation cycle. Nature 2012; 489:
313–317. 46. Fulda S, Debatin KM. 5-Aza-2'-deoxycytidine and IFN-gamma cooperate to sensitize for
TRAIL-induced apoptosis by upregulating caspase-8. Oncogene 2006; 25: 5125–5133. 47. Milde T, Oehme I, Korshunov A, Kopp-Schneider A, Remke M, Northcott P et al. HDAC5 and
HDAC9 in medulloblastoma: novel markers for risk stratification and role in tumor cell growth. Clin Cancer Res 2010; 16: 3240–3252. 18. Hubbert C, Guardiola A, Shao R, Kawaguchi Y, Ito A, Nixon A et al. HDAC6 is a microtubule-
associated deacetylase. Nature 2002; 417: 455–458. 19. Ng HH, Bird A. Histone deacetylases: silencers for hire. Trends Biochem Sci 2000; 25:
121–126. 48. Fischer M, Skowron M, Berthold F. Reliable transcript quantification by real-time reverse
transcriptase-polymerase chain reaction in primary neuroblastoma using normalization to
averaged expression levels of the control genes HPRT1 and SDHA. J Mol Diagn 7: 89–96. 48. Fischer M, Skowron M, Berthold F. Reliable transcript quantification by real-time reverse
transcriptase-polymerase chain reaction in primary neuroblastoma using normalization to
averaged expression levels of the control genes HPRT1 and SDHA. J Mol Diagn 7: 89–96. 48. Supplementary Information accompanies this paper on Cell Death and Disease website (http://www.nature.com/cddis) The authors declare no conflict of interest. Fischer M, Skowron M, Berthold F. Reliable transcript quantification by real-time reverse
transcriptase-polymerase chain reaction in primary neuroblastoma using normalization to
averaged expression levels of the control genes HPRT1 and SDHA. J Mol Diagn 7: 89–96. 20. Lindemann RK, Newbold A, Whitecross KF, Cluse LA, Frew AJ, Ellis L et al. Analysis of the
apoptotic and therapeutic activities of histone deacetylase inhibitors by using a mouse model
of B cell lymphoma. Proc Natl Acad Sci USA 2007; 104: 8071–8076. 21. Thiele CJ, Reynolds CP, Israel MA. Decreased expression of N-myc precedes retinoic
acid-induced morphological differentiation of human neuroblastoma. Nature 1985; 313:
404–406. Cell Death and Disease is an open-access journal
published by Nature Publishing Group. This work is
licensed under a Creative Commons Attribution 4.0 International
Licence. The images or other third party material in this article are
included in the article’s Creative Commons licence, unless indicated
otherwise in the credit line; if the material is not included under the
Creative Commons licence, users will need to obtain permission from
the licence holder to reproduce the material. To view a copy of this
licence, visit http://creativecommons.org/licenses/by/4.0 22. Armstrong JL, Redfern CP, Veal GJ. 13-Cis retinoic acid and isomerisation in paediatric
oncology – is changing shape the key to success? Biochem Pharmacol 69: 1299–1306. 23. Jiang M, Zhu K, Grenet J, Lahti JM. Retinoic acid induces caspase-8 transcription via
phospho-CREB and increases apoptotic responses to death stimuli in neuroblastoma cells. Biochim Biophys Acta 2008; 1783: 1055–1067. 24. Gao J, Siddoway B, Huang Q, Xia H. Inactivation of CREB mediated gene transcription by
HDAC8 bound protein phosphatase. Biochem Biophys Res Commun 2009; 379: 1–5. 25. Ghee M, Baker H, Miller JC, Ziff EB. AP-1, CREB and CBP transcription factors differentially
regulate the tyrosine hydroxylase gene. Brain Res Mol Brain Res 1998; 55: 101–114. Supplementary Information accompanies this paper on Cell Death and Disease website (http://www.nature.com/cddis) Supplementary Information accompanies this paper on Cell Death and Disease website (ht Supplementary Information accompanies this paper on Cell Death and Disease website (http://www.nature.com/cddis) Cell Death and Disease
|
https://openalex.org/W4205555724
|
https://bg.copernicus.org/articles/19/2079/2022/bg-19-2079-2022.pdf
|
English
| null |
Comment on bg-2021-331
| null | 2,022
|
cc-by
| 14,501
|
Organic matter characteristics of a rapidly eroding permafrost cliff
in NE Siberia (Lena Delta, Laptev Sea region) Jongejans (loeka.jongejans@awi.de) and Jens Strauss (jens.strauss@awi.de) Correspondence: Loeka L. Jongejans (loeka.jongejans@awi.de) and Jens Strauss (jens.strauss@awi.de) Received: 10 December 2021 – Discussion started: 13 December 2021 Received: 10 December 2021 – Discussion started: 13 December 2021 Received: 10 December 2021 – Discussion started: 13 December 2021
Revised: 15 March 2022 – Accepted: 15 March 2022 – Published: 14 April 2022 Revised: 15 March 2022 – Accepted: 15 March 2022 – Published: 14 April 2022 Revised: 15 March 2022 – Accepted: 15 March 2022 – Published: 14 April 2022 lar geochemical and carbon isotopic analyses of Late Pleis-
tocene Yedoma permafrost and Holocene cover deposits, dis-
continuously spanning the last ∼52 kyr. We showed that the
ancient permafrost exposed in the Sobo-Sise cliff has a high
organic carbon content (mean of about 5 wt %). The oldest
sediments stem from Marine Isotope Stage (MIS) 3 intersta-
dial deposits (dated to 52 to 28 cal ka BP) and are overlaid by
last glacial MIS 2 (dated to 28 to 15 cal ka BP) and Holocene
MIS 1 (dated to 7–0 cal ka BP) deposits. The relatively high
average chain length (ACL) index of n-alkanes along the
cliff profile indicates a predominant contribution of vascu-
lar plants to the OM composition. The elevated ratio of iso-
and anteiso-branched fatty acids (FAs) relative to mid- and
long-chain (C ≥20) n-FAs in the interstadial MIS 3 and the
interglacial MIS 1 deposits suggests stronger microbial activ-
ity and consequently higher input of bacterial biomass during Abstract. Organic carbon (OC) stored in Arctic permafrost
represents one of Earth’s largest and most vulnerable ter-
restrial carbon pools. Amplified climate warming across the
Arctic results in widespread permafrost thaw. Permafrost de-
posits exposed at river cliffs and coasts are particularly sus-
ceptible to thawing processes. Accelerating erosion of terres-
trial permafrost along shorelines leads to increased transfer
of organic matter (OM) to nearshore waters. However, the
amount of terrestrial permafrost carbon and nitrogen as well
as the OM quality in these deposits is still poorly quantified. We define the OM quality as the intrinsic potential for further
transformation, decomposition and mineralisation. Here, we
characterise the sources and the quality of OM supplied to
the Lena River at a rapidly eroding permafrost river shore-
line cliff in the eastern part of the delta (Sobo-Sise Island). Our multi-proxy approach captures bulk elemental, molecu- Research article Research article Biogeosciences, 19, 2079–2094, 2022
https://doi.org/10.5194/bg-19-2079-2022
© Author(s) 2022. This work is distributed under
the Creative Commons Attribution 4.0 License. Organic matter characteristics of a rapidly eroding permafrost cliff
in NE Siberia (Lena Delta, Laptev Sea region) Tuerena8,
Lutz Schirrmeister1, Sebastian Wetterich1,c, Alexander Kizyakov9, Cornelia Karger3, and Jens Strauss1 1Permafrost Research Section, Alfred Wegener Institute, Helmholtz Centre for Polar and Marine Research,
14473 Potsdam, Germany Permafrost Research Section, Alfred Wegener Institute, Helmholtz Centre for Polar and Marine Res
4473 Potsdam, Germany y
2Institute of Geosciences, University of Potsdam, 14476 Potsdam, Germany 2Institute of Geosciences, University of Potsdam, 14476 Potsdam, Germany 3Organic Geochemistry Section, Helmholtz Centre Potsdam GFZ German Research Centre for Geosciences,
14473 Potsdam, Germany 3Organic Geochemistry Section, Helmholtz Centre Potsdam GFZ German Research Centre for Geosciences,
14473 Potsdam, Germany 4Marine Geochemistry Section, Alfred Wegener Institute, Helmholtz Centre for Polar and Marine Research,
27570 Bremerhaven, Germany
5 4Marine Geochemistry Section, Alfred Wegener Institute, Helmholtz Centre for Polar and Marine Research,
27570 Bremerhaven, Germany 5Faculty of Geosciences, University of Bremen, 28359 Bremen, Germany y
y
y
6Department of Geography and Environmental Sciences, Northumbria University, Newcastle-upon-Tyne, NE1 8ST, UK
7Institute of Carbon Cycles, Helmholtz Centre Hereon, 21502 Geesthacht, Germany 6Department of Geography and Environmental Sciences, Northumbria University, Newcastle-upon-Tyne, NE1 8ST, UK
7I
i
f C b
C
l
H l h l
C
H
21502 G
h
h G 6Department of Geography and Environmental Sciences, Northumbria University, Newcastle-upon-Tyne, NE1 8ST, UK
7Institute of Carbon Cycles, Helmholtz Centre Hereon, 21502 Geesthacht, Germany 8Scottish Association for Marine Science, Oban, PA37 1QA, UK
9 8Scottish Association for Marine Science, Oban, PA37 1QA, UK
9 8Scottish Association for Marine Science, Oban, PA37 1QA, UK hology and Glaciology Department, Faculty of Geography, Lomonosov Moscow State University,
119991, Russia 9Cryolithology and Glaciology Department, Faculty of Geography, Lomonosov Moscow State University,
Moscow, 119991, Russia anow at: Department of Environmental Science and Analytical Chemistry, Stockholm University, Stockholm, 11418, Sweden
bnow at: Department of Human Geography, Stockholm University, Stockholm, 11418, Sweden
cnow at: Institute of Geography, Technische Universität Dresden, 01069 Dresden, Germany anow at: Department of Environmental Science and Analytical Chemistry, Stockholm University,
b anow at: Department of Environmental Science and Analytical Chemistry, Stockholm University, Stockholm, 11418, Sweden
bnow at: Department of Human Geography, Stockholm University, Stockholm, 11418, Sweden
cnow at: Institute of Geography, Technische Universität Dresden, 01069 Dresden, Germany p
y
y,
y,
,
,
bnow at: Department of Human Geography, Stockholm University, Stockholm, 11418, Sweden
cnow at: Institute of Geography Technische Universität Dresden 01069 Dresden Germany bnow at: Department of Human Geography, Stockholm University, Stockholm, 11418, Sweden ow at: Institute of Geography, Technische Universität Dresden, 01069 Dresden, Germany Correspondence: Loeka L. Organic matter characteristics of a rapidly eroding permafrost cliff
in NE Siberia (Lena Delta, Laptev Sea region) Charlotte Haugk1,2,a, Loeka L. Jongejans1,2, Kai Mangelsdorf3, Matthias Fuchs1, Olga Ogneva1,4,5, Juri Palmtag6,b,
Gesine Mollenhauer4,5, Paul J. Mann6, P. Paul Overduin1, Guido Grosse1,2, Tina Sanders7, Robyn E. 1
Introduction Climate warming puts permafrost, especially ice-rich per-
mafrost, in the terrestrial Arctic at risk of thawing
(e.g. Strauss et al., 2021b). Permafrost, by definition, is
ground that stays below 0 ◦C for 2 or more consecutive
years. Terrestrial permafrost ecosystems are affected by on-
going climate warming with consequences for geomorpho-
logical, hydrological and biogeochemical processes from a
local to regional scale (IPCC, 2019). Almost twice as much
carbon is stored in the permafrost region than what is cur-
rently contained in the atmosphere (Hugelius et al., 2014;
Mishra et al., 2021), making permafrost carbon dynamics a
globally relevant issue (Grosse et al., 2011; Schuur et al.,
2008; Strauss et al., 2021a; Turetsky et al., 2020). Total esti-
mated soil organic carbon (SOC) storage for the permafrost
region is ∼1100–1600 Gt, of which 181 ± 54 Gt is attributed
to deep permafrost (below 3 m depth) of the Yedoma region
(Hugelius et al., 2014; Strauss et al., 2021b, 2013). In our study, we measure molecular biomarkers (n-
alkanes, n-fatty acids) and use established biomarker prox-
ies and indices such as the average chain length (ACL) of
n-alkanes, the carbon preference index (CPI) and the higher
plant fatty acid (HPFA) index to test whether they mirror the
OM degree of decomposition and reflect the OM quality in
ancient permafrost deposits. Additionally, analyses of the to-
tal organic carbon content (TOC), the stable carbon isotope
ratios (δ13C of TOC), the total nitrogen (TN) content and
TOC/TN (here referred to as the C/N ratio) are applied to
our sample set. Hierarchical clustering is used to identify the
stratigraphical units along the sample profile based on the
major changes in OM composition. Warming throughout the Arctic prolongs the season for
permafrost thaw and open ice-free water bodies, resulting
in increasing erosion of ice- and carbon-rich permafrost
sediments exposed at coasts (Günther et al., 2013; Jones
et al., 2020). Very ice-rich permafrost deposits (i.e. 50–
90 vol % ice) such as the Late Pleistocene Yedoma Ice Com-
plex (Schirrmeister et al., 2013; Strauss et al., 2017) are par-
ticularly vulnerable to rapid thermo-denudation and thermo-
erosion processes along river shores (Costard et al., 2014;
Kanevskiy et al., 2016; Stettner et al., 2018; Fuchs et al.,
2020). Vonk et al. (2013b) showed that Yedoma ice-wedge
meltwater can increase the decomposition of organic mat-
ter (OM) due to co-metabolising effects. C. Haugk et al.: Eroding permafrost cliff carbon characteristics C. Haugk et al.: Eroding permafrost cliff carbon characteristics 2080 these climatically warmer periods. The overall high carbon
preference index (CPI) and higher plant fatty acid (HPFA)
values as well as high C/N ratios point to a good quality of
the preserved OM and thus to a high potential of the OM for
decomposition upon thaw. A decrease in HPFA values down-
wards along the profile probably indicates stronger OM de-
composition in the oldest (MIS 3) deposits of the cliff. The
characterisation of OM from eroding permafrost leads to a
better assessment of the greenhouse gas potential of the OC
released into river and nearshore waters in the future. to the nearshore zone and onto the Arctic Shelf (Mann et
al., 2022; Sanders et al., 2022). The Arctic river discharge
has increased significantly in recent decades, transporting
organic-rich waters to the nearshore area (Holmes et al.,
2012, 2021). Increased riverbank erosion of Arctic rivers fol-
lowing warming during the last few decades constitutes an
important mechanism of carbon export from land to water
(Zhang et al., 2017, 2021; Fuchs et al., 2020). The studies
cited here stress the need to better understand the interactions
between thawing permafrost and river and nearshore waters. The study of fossil biomolecules and other OM charac-
teristics provides insights into the composition and level
of OM decomposition and hence can greatly improve es-
timates of the greenhouse gas potential of thaw-mobilised
OM from permafrost deposits (Andersson and Meyers, 2012;
Sánchez-García et al., 2014). A few studies have previously
focused on molecular biomarkers in north-eastern Siberian
permafrost deposits (e.g. Zech et al., 2010; Höfle et al.,
2013; Strauss et al., 2015; Stapel et al., 2016; Jongejans et
al., 2020). In general, the abundance and distribution of n-
alkanes, which are long-chained, single-bond hydrocarbons,
are used for OM characterisation where the chain length of
n-alkanes indicates OM sources. 1
Introduction Another poten-
tial impact of the decomposition of terrestrial OM and dis-
charge with Arctic river water is the change in biochemical
properties that may increase ocean acidification and anthro-
pogenic carbon dioxide uptake from the atmosphere (Semile-
tov et al., 2016). Furthermore, Semiletov et al. (2016) es-
timated that 57 % of the terrestrial organic carbon in the
East Siberian Shelf originates from ancient Pleistocene-age
permafrost C, such as Yedoma deposits adjacent to river or
coastal zones. Extensive river networks like the Lena River,
especially in their delta zones, carry large nutrient OM loads j
g
p
Thus, the OM characteristics of permafrost deposits,
rapidly eroding at a cliff site in the eastern Lena Delta, are
analysed for the first time for biomarkers. The set of frozen
samples was obtained along a 25 m vertical cliff profile with
a relatively high sampling density of about 1 m covering all
exposed cryostratigraphic units. In this study, we aim (1) to
characterise the OM composition of ancient permafrost that
accumulated under different climate conditions; (2) to assess
the degree of decomposition that the OM already experi-
enced; and (3) to hypothesise, based on the decomposition
legacy, the potential of future decomposability and microbial
decomposition of the permafrost OM. C. Haugk et al.: Eroding permafrost cliff carbon characteristics C. Haugk et al.: Eroding permafrost cliff carbon characteristics 2081 Figure 1. Overview of the Sobo-Sise Yedoma cliff. (a) Location of the Sobo-Sise Yedoma cliff in the Lena Delta in north-eastern Siberia
(Landsat 5 mosaic (band combination 5, 4, 3) including scenes from 2009 and 2010; Landsat 5 image courtesy of the US Geological Survey);
(b) picture of the Sobo-Sise Yedoma cliff from the east to west; (c) cross-section of the cliff profile (adapted from Wetterich et al., 2020a)
indicating the three vertically sampled sections: SOB18-01, SOB18-03 and SOB18-06. Figure 1. Overview of the Sobo-Sise Yedoma cliff. (a) Location of the Sobo-Sise Yedoma cliff in the Lena Delta in north-eastern Siberia
(Landsat 5 mosaic (band combination 5, 4, 3) including scenes from 2009 and 2010; Landsat 5 image courtesy of the US Geological Survey);
(b) picture of the Sobo-Sise Yedoma cliff from the east to west; (c) cross-section of the cliff profile (adapted from Wetterich et al., 2020a)
indicating the three vertically sampled sections: SOB18-01, SOB18-03 and SOB18-06. Sea, with a mean annual discharge of 525 km3 yr−1 (Holmes
et al., 2021). It also transports summer “heat” from the south
to the north (Yang et al., 2005). The study area on Sobo-Sise
Island (Fig. 1a–b) is located in the continuous permafrost
zone. The island stretches between the Sardakhskaya and
Bykovskaya main channels in the eastern part of the delta. In addition to the modern floodplain, there are three geomor-
phic units in the delta (Grigoriev, 1993; Schwamborn et al.,
2002). While the first unit consists of Holocene floodplains
and could occur in the whole delta area, the second unit con-
sists of Late Pleistocene and Holocene fluvial deposits that
are mostly located in the north-western part of the delta and
are cut off from the current delta dynamics (Schirrmeister
et al., 2011b). The third geomorphological unit consists of
erosional remnants of a Late Pleistocene accumulation plain
with ice-rich Yedoma Ice Complex deposits and is present mainly in the west, south and east of the delta (Schwamborn
et al., 2002; Wetterich et al., 2008; Morgenstern et al., 2011). According to a landform classification for Sobo-Sise Island,
43 % of the land surface is occupied by Yedoma uplands and
Yedoma slopes and 43 % is thermokarst basins with the re-
maining 14 % being thermokarst lakes (Fuchs et al., 2018). 2
Study area The Lena River forms the largest delta in the Arctic, cover-
ing an area of 29×103 km2 (Schneider et al., 2009), and dis-
charges the second-highest freshwater load into the Laptev Biogeosciences, 19, 2079–2094, 2022 https://doi.org/10.5194/bg-19-2079-2022 3.2
Sedimentological organic matter parameters Prior to bulk geochemical analyses, all samples were freeze-
dried (Sublimator, ZIRBUS technology), ground and ho-
mogenised (Fritsch PULVERISETTE 5 planetary mill; 8 min
at 360 rotations per minute). Total elemental carbon (TC) and
total nitrogen (TN) content of sediment samples in weight
percentage (wt %) were measured with a carbon–nitrogen–
sulfur analyser (vario EL III, Elementar) with a detection
limit of 0.1 wt % for carbon and nitrogen. Samples below
this detection limit were set to 0.05 wt % so that the statis-
tics could be calculated. Total organic carbon (TOC) con-
tent in weight percentage (wt %) was measured with a TOC
analyser (vario MAX C, Elementar; analytical accuracy of
0.1 wt %). The TOC to TN (C/N) ratio has been used as
a rough first indicator of the degree of OM decomposition
with decreasing values indicating proceeding decomposition
(Palmtag et al., 2015). The stable carbon isotope ratio (δ13C)
of TOC reflects both the initial contribution from different
plant species and plant components and OM decomposition
processes (Gundelwein et al., 2007). Samples for δ13C anal-
yses were treated with hydrochloric acid (20 mL, 1.3 M) to
remove carbonates, heated on a hotplate (97.7 ◦C for 3 h) and
subsequently washed with distilled water. The samples were
filtered (Whatman Grade GF/B, nominal particle retention
of 1.0 µm), after which the residue was dried and ground. All δ13C samples were measured using a DELTA V Advan-
tage isotope ratio mass spectrometer (MS) equipped with a
Flash 2000 analyser (Thermo Fisher Scientific; analytical ac-
curacy of 0.15 ‰), using helium as a carrier gas. The δ13C
(13C/12C) value is reported in per mille (‰) compared to the
standard ratio Vienna Pee Dee Belemnite (VPDB). The Sobo-Sise Yedoma cliff has an average height of 22 m
with a maximum height of 27.7 m above the river water level
(m a.r.l.) (Fuchs et al., 2020) and is affected by fluvio-thermal
erosion. The current average shoreline retreat rate, which was
calculated using satellite data, is 15.7 m yr−1 (2015–2018),
which is remarkably high (Fuchs et al., 2020). C. Haugk et al.: Eroding permafrost cliff carbon characteristics et al. (2011b) attributed parts of the third geomorphological
unit in the Lena Delta in the western and southern parts of
the delta to remnants of a Yedoma accumulation plain. This
formed during the Late Pleistocene when the Lena River had
its delta farther north. Radiocarbon ages corroborated that
Yedoma deposits on Sobo-Sise Island accumulated during
the Late Pleistocene between about 52 and 15 cal ka BP. Sub-
stantial hiatuses were found at about 36–29 cal ka BP and at
20 to 17 cal ka BP, which may be related to fluvial erosion
and/or changed discharge patterns of the Lena River (Wet-
terich et al., 2020a). Middle to Late Holocene ages from
6.36 to 2.5 cal kyr BP were found in the uppermost cover
deposits of the cliff, which is also in agreement with other
cover deposits found on top of Yedoma such as on the nearby
Bykovsky Peninsula (Schirrmeister et al., 2002; Grosse et al.,
2007). (∼20 × 10 × 10 cm) from the cliff wall. Samples were col-
lected after cleaning and scraping off the outermost unfrozen
and frozen parts of the cliff wall in order to collect frozen,
uncontaminated samples. Then, the samples were lifted up-
wards, cleaned and subsampled for biomarker analysis. In to-
tal, we collected 61 sediment samples, of which 28 were se-
lected for biomarker analysis at about 1 m intervals covering
the entire exposed section. The samples were stored frozen
in pre-combusted glass jars, apart from 9 samples (SOB18-
06-09 to SOB18-06-34) which were initially stored in plastic
whirl packs before being transferred in a frozen state to glass
containers after transport to the laboratories. 3.2
Sedimentological organic matter parameters In compar-
ison, retreat rates were lower for other Yedoma cliffs such
as on Kurungnakh Island in the central Lena Delta (4.1–
6.9 m yr−1; Stettner et al., 2018) and at the Itkillik exposure
in Alaska (11 m yr−1; Kanevskiy et al., 2016) but even higher
for Muostahk Island (29.4 m yr−1; Günther et al., 2013) and
Cape Mamontov Klyk (21 m yr−1; Günther et al., 2015). The
Sobo-Sise Yedoma cliff (72◦32′ N, 128◦17′ E; Fig. 1c) ex-
tends over 1660 m in length and faces north towards the Sar-
dakhskaya Channel. Here, the water discharge amounts to
about 8000 m3 s−1 during the summer-low period (Fedorova
et al., 2015) and the Lena River is ice-covered for about
8 months per year between October and May. The river ice
thickness reaches up to 2 m. Water depth at the beginning of
the Sardakhskaya Channel (close to Stolp and Sardakh is-
lands) can reach up to 22 m (Fedorova et al., 2015) and is
approximately 11 m in front of the Sobo-Sise Yedoma cliff,
allowing for water flow underneath the river ice cover during
the winter months (Fuchs et al., 2020). C. Haugk et al.: Eroding permafrost cliff carbon characteristics The terrain is affected by thermokarst processes (Nitze and
Grosse, 2016) and surface thaw subsidence (Chen et al.,
2018). The distinct surface morphology of Sobo-Sise Island in-
cludes Yedoma uplands intersected by thermo-erosional val-
leys and thermokarst basins. Syngenetic permafrost forma-
tion in polygonal tundra landscapes over long periods in the
Late Pleistocene formed thick deposits with large ice wedges
that are exposed at the cliff (Schirrmeister et al., 2011b, 2020;
Strauss et al., 2015; Jongejans et al., 2018). Schirrmeister https://doi.org/10.5194/bg-19-2079-2022 Biogeosciences, 19, 2079–2094, 2022 2082 3.4.2
Carbon preference index The CPI (carbon preference index) was originally introduced
by Bray and Evans (1961) as the ratio of odd- to even-
numbered n-alkanes and indicates the level of OM transfor-
mation, which decreases with progressing maturation. OM
decomposition leading to lower CPI values is a measure of
thermal alteration referring to rocks or oils on a geologi-
cal timescale. However, this ratio, as well as the very simi-
lar odd-over-even predominance (OEP) ratio, was previously
used in Quaternary permafrost deposits as an indicator for
OM decomposition (Zech et al., 2009; Strauss et al., 2015;
Struck et al., 2020; Jongejans et al., 2020). Based on these
studies, we refer to values over 5 as less degraded OM of
high quality. Equation (2) describes the CPI and was modi-
fied after Marzi et al. (1993) using C23−33 as a chain length
interval. 3.4.1
Average chain length The n-alkane average chain length (ACL) is the weighted
average number of carbon atoms used for determining OM
sources. Long-chain odd-numbered n-alkanes (C ≥21) are
essential constituents that serve as biomarkers for higher ter-
restrial plants (Eglinton and Hamilton, 1967; Eglinton and
Eglinton, 2008; Schäfer et al., 2016), whereas shorter chain
lengths indicate bryophyte, bacterial or algal origin (Cran-
well, 1984; Rieley et al., 1991; Kuhn et al., 2010). A change
in the ACL can suggest a change in the terrestrial source
biota. We used the equation (Eq. 1) first described by Poyn-
ter (1989) but with a chain interval from C23 to C33 following
Strauss et al. (2015) and Jongejans et al. (2018): ACL =
Pi · Ci
PCi
,
(1) ACL =
Pi · Ci
PCi
, (1) where C denotes concentration and i the carbon number. 3.3.1
Extraction and fraction separation Following freeze-drying and grinding, biomarker subsam-
ples were transferred into glass jars. Extraction and separa-
tion were conducted according to Schulte et al. (2000) and
Strauss et al. (2015). Samples were processed in two batches,
each containing 14 samples. We weighed between 8 and 11 g
in extraction cell bodies fit for the accelerated solvent extrac-
tor (ASE 200 Dionex). Dichloromethane / methanol (volume
ratio of 99 : 1) was used as a solvent mixture for OM extrac-
tion. Each sample was held in a static phase (5 min heating
phase, 20 min at 75 ◦C and 5 MPa). Dissolved compounds
were then further concentrated at ∼42 ◦C using a closed-
cell concentrator (TurboVap 500 Zymark), and the remain-
ing solvent was evaporated under N2. Afterwards, internal
standards were added: 5α-androstane for the aliphatic frac-
tion, ethylpyrene for the aromatic fraction, 5α-androstan-17-
one for the NSO (nitrogen, sulfur and/or oxygen) neutral
polar fraction and erucic acid for the NSO fatty-acid frac-
tion (80 µL each from respective 100 µg mL−1 standard so-
lutions). Subsequently, an asphaltene precipitation was per-
formed to remove compounds with higher molecular com-
plexity (asphaltenes) by dissolving the extracts in a small
amount of dichloromethane and adding a 40-fold excess of n-
hexane. Precipitated asphaltenes were removed by filtration
through a sodium-sulfate-filled funnel. Subsequently, the n-
hexane-soluble portion was separated by medium-pressure
liquid chromatography (MPLC) (Radke et al., 1980) into
three fractions of different polarities: aliphatic hydrocarbons,
aromatic hydrocarbons and polar hetero-compounds (NSO
compounds). Finally, the NSO fractions of 13 samples were
split into an acid and neutral polar (alcohol) fraction using
a KOH-impregnated column. While the n-fatty-acid potas-
sium salts were attached to the silica gel, the neutral po-
lar compounds were eluted with dichloromethane. After re-
mobilising the n-fatty acids by protonation of their salts
with formic acid, the n-fatty-acid fraction was obtained with
dichloromethane. C. Haugk et al.: Eroding permafrost cliff carbon characteristics stant flow of 1 mL min−1. After injection, the compounds
of interest were separated on an SGE BPX5 fused-silica
capillary column (50 m length, 0.22 mm i.d., 0.25 µm film
thickness) using the following temperature conditions: ini-
tial temperature of 50 ◦C (1 min isothermal), heating rate
of 3 ◦C min−1 to 310 ◦C, held isothermally for 30 min. The
MS operated in the electron impact mode at 70 eV. Full-scan
mass spectra were recorded from m/z 50–600 at a scan rate
of 2.5 scans s−1. Using the software Xcalibur (Thermo Fisher
Scientific), peaks in the GC-MS run were quantified using
the internal standards for n-alkanes and n-fatty acids. All
biomarker concentrations are expressed in micrograms per
gram of dry sediment (µg g−1 sed.) and per gram of TOC
(µg g−1 TOC). ysed the n-alkane distributions of all 28 samples and selected
13 samples for the analysis of n-fatty acids. The selection of
the n-fatty acids was made to cover the entire profile contin-
uously (approximately every 2 m). ysed the n-alkane distributions of all 28 samples and selected
13 samples for the analysis of n-fatty acids. The selection of
the n-fatty acids was made to cover the entire profile contin-
uously (approximately every 2 m). 3.3
Lipid biomarker analyses The Sobo-Sise Yedoma cliff was sampled in three overlap-
ping vertical sediment profiles (Fig. 1b) covering the en-
tire exposed permafrost section (profile SOB18-01, 24.1 to
15.7 m a.r.l.; profile SOB18-03, 18.2 to 10.2 m a.r.l.; pro-
file SOB18-06, 13.4 to 0.9 m a.r.l.). Each profile was cry-
olithologically described (see Wetterich et al., 2018, 2020a),
and samples were collected at 0.5 m intervals by rappelling
down on a rope from the top of the cliff. We used an
axe and hammer to extract defined cubes of frozen ground Lipid biomarkers provide information on a molecular level
about the source of OM, the environmental conditions during
deposition and the degree of decomposition. In this study, we
focused on n-alkanes in the aliphatic OM fraction and n-fatty
acids in the polar hetero-compound fraction. Changes in their
relative abundance can provide an indication of the degree of
decomposition (Kim et al., 2005) as outlined below. We anal- Biogeosciences, 19, 2079–2094, 2022 https://doi.org/10.5194/bg-19-2079-2022 2083 C. Haugk et al.: Eroding permafrost cliff carbon characteristics https://doi.org/10.5194/bg-19-2079-2022 3.5
Data analysis In order to identify the stratification along the cliff based
on the OM characteristics of the permafrost sediments, the
data set was clustered using a constrained agglomerative hi-
erarchical clustering of a distance matrix (chclust of the ri-
oja package, in R version 4.0.4) (Juggins, 2019). We ap-
plied the non-parametric Kruskal–Wallis (> two groups)
test for statistical analyses of the data to compare all ma-
jor parameters (TOC, C/N, n-alkanes, ACL, CPI, short
and long n-fatty acids, HPFA index, and (iso- + anteiso-
branched) / (mid- and long-chain n-fatty acids)) between the
identified clusters. In the Results section, we report the p val-
ues; the correlation coefficients are reported in the Supple-
ment (Table S1). For each sample, the absolute n-fatty-acid (FA) concentra-
tion was measured and the most abundant homologue’s chain
length was identified. In addition, we looked at the share of
iso- and anteiso-branched FAs, which are indicators for mi-
crobial biomass (Rilfors et al., 1978; Stapel et al., 2016). Fur-
thermore, we calculated the higher plant fatty acid (HPFA)
index, which is the relative amount of the long-chain n-
fatty acids to long-chain n-alkanes in the sediments. The
HPFA index was introduced by Strauss et al. (2015) follow-
ing the principles of the HPA index of Poynter (1989), only
with using fatty acids instead of wax alcohols (Eq. 3). The
HPFA index reflects the degree of preservation of OM due to
the higher lability of n-fatty acids in relation to n-alkanes
(Canuel and Martens, 1996). The preferential decomposi-
tion of fatty acids is due to their functional group leading
to a chemical polarisation within the molecule forming an
attack point for geochemical or microbiological decompo-
sition and/or decarboxylation (Killops and Killops, 2013). Therefore, a decrease in the HPFA index indicates increased
OM decomposition. We use this index for internal compar-
ison where higher values (above the mean) indicate a com-
paratively higher-quality OM. C. Haugk et al.: Eroding permafrost cliff carbon characteristics C. Haugk et al.: Eroding permafrost cliff carbon characteristics 2084 2084
C. Haugk et al.: Eroding permafrost cliff carbon characteristics Figure 2. Biogeochemical parameters of the Sobo-Sise Yedoma cliff: total organic carbon (TOC) content, total nitrogen (TN) content,
carbon-to-nitrogen (C/N) ratio, bulk stable carbon isotope ratios (δ13C), radiocarbon ages and modelled age in calibrated kiloyears before
present (cal kyr BP). Data points are displayed over cliff height from the cliff top at 25 m above river level (a.r.l.) to cliff bottom at 0 m a.r.l. The three sections of SOB18 are plotted separately for each parameter (black, dark grey and light grey circles). Units I, II (grey rectangle)
and III correspond to Marine Isotope Stage (MIS) 1 to 3, respectively. The radiocarbon ages were published in Wetterich et al. (2020a). 3.3.2
GC-MS measurements and compound
quantification n-Alkanes and n-fatty acids were analysed using gas chro-
matography coupled with a mass spectrometer (GC-MS; GC n-Alkanes and n-fatty acids were analysed using gas chro-
matography coupled with a mass spectrometer (GC-MS; GC
– Trace GC Ultra and MS – DSQ, both Thermo Fisher Sci-
entific). Prior to the analyses, n-fatty acids were methylated
with diazomethane. The GC was equipped with a cold in-
jection system operating in the splitless mode. The injector
temperature was programmed from 50 to 300 ◦C at a rate
of 10 ◦C s−1. Helium was used as carrier gas with a con- CPI23−33 =
Podd C23−31
+
Podd C25−33
2
Peven C22−32
(2) (2) Biogeosciences, 19, 2079–2094, 2022 https://doi.org/10.5194/bg-19-2079-2022 2084
C. Haugk et al.: Eroding permafrost cliff carbon characteristics
Figure 2. Biogeochemical parameters of the Sobo-Sise Yedoma cliff: total organic carbon (TOC) content, total nitrogen (TN) content,
carbon-to-nitrogen (C/N) ratio, bulk stable carbon isotope ratios (δ13C), radiocarbon ages and modelled age in calibrated kiloyears before
present (cal kyr BP). Data points are displayed over cliff height from the cliff top at 25 m above river level (a.r.l.) to cliff bottom at 0 m a.r.l. The three sections of SOB18 are plotted separately for each parameter (black, dark grey and light grey circles). Units I, II (grey rectangle)
and III correspond to Marine Isotope Stage (MIS) 1 to 3, respectively. The radiocarbon ages were published in Wetterich et al. (2020a). 4.2.2
n-Fatty acids TOC content was highest in the topmost sample and
ranged from < 0.1 wt % (below the detection limit, sample
SOB18-01-18 at 15.7 m a.r.l.) to 25.5 wt % (SOB18-01-01 at
24.1 m a.r.l.). Both the minimum and the maximum TOC val-
ues were found within profile SOB18-01 (Fig. 2). The av-
erage TOC content was 4.9 wt % (standard deviation (SD)
4.7, n = 28) and values decreased from the cliff top down-
wards with two values higher than 10 wt % at 24.1 m a.r.l. (SOB18-01-01; 25.5 wt %) and at 16.2 m a.r.l. (SOB18-03-
05; 11.3 wt %). TN content had an average of 0.3 wt % (SD
0.2), and the highest value was also found at 24.1 m a.r.l. (SOB18-01-01; 0.8 wt %) and the lowest at 15.7 m a.r.l. (SOB18-01-18; < 0.1 wt %). C/N ratios ranged from 7.2 to
30.5 (Fig. 2) and displayed, except for the uppermost sam-
ple, only little variability over the cliff profile with a mean
of 13.2 (SD 4.2). The δ13C values ranged from −25.2 ‰
(SOB18-01-12 at 18.7 m a.r.l.) to −29.4 ‰ (SOB18-03-05
at 16.2 m a.r.l.) and had an average of −27.2 ‰ (SD 1.1). There was a significant negative correlation between δ13C
and C/N values (r −0.59 (Pearson correlation), p < 0.01). Radiocarbon ages of selected plant remains were determined
on a MICADAS system. These data and a Bayesian age–
depth model were published by Wetterich et al. (2020a). The
laboratory methods were described in detail by Mollenhauer
et al. (2021). The radiocarbon ages ranged from 2.50 (at
23.7 m a.r.l.) to 51.88 cal kyr BP (at 0.9 m a.r.l.). We found n-FAs with carbon numbers between C8 and C32. The n-FAs showed a strong even-over-odd carbon number
predominance. Furthermore, hydroxy FAs (C6 to C8), iso-
branched FAs (C10 to C19), anteiso-branched FAs (C11, C12,
C13, C15 and C17), monounsaturated FAs (C16 to C20 and
C24), unsaturated iso- and anteiso-branched FAs (C17), cy-
clopropyl FAs (C17 and C19), di- and triunsaturated FAs
(C18), and phytanic acid were detected (Fig. S2, Table S2). Concentration of long-chain n-FAs (C24 to C32) ranged
from 290 µg g−1 TOC at 24.1 m a.r.l. to 2346 µg g−1 TOC
at 1.4 m a.r.l. (mean 1041 µg g−1 TOC, SD 655). The most
abundant long-chain n-FA was n-C24 for all samples, ex-
cept at 16.7 m a.r.l. (n-C26) (Figs. 4.2.1
n-Alkanes n-Alkanes were detected in the range between n-C14 and n-
C35 and showed a strong odd-over-even carbon number pre-
dominance. The relative n-alkane concentration increased in
the lower part of the cliff, closer to the river level. Relative
n-alkane concentrations ranged from 1 to 172 µg g−1 TOC
(mean 30, SD 42) for the short-chain (C14 to C20) n-alkanes
and from 119 to 3214 µg g−1 TOC (mean 1068, SD 886) for
the long-chain (C21 to C33) n-alkanes (Fig. 3). The absolute
concentrations were also higher in the lower part of the cliff
for the short-chain (mean 1 µg g−1 sed., SD 1) and long-chain
n-alkanes (mean 41 µg g−1 sed., SD 31). The main dominat-
ing n-alkane chain length was n-C27 in the lower part of the
cliff and alternated between n-C27 and n-C29 in the upper
part (Fig. S1). Four samples were dominated by the n-C31
n-alkane. 4.2.2
n-Fatty acids 2 and S2). The mid-
chain n-FA (C21 to C23) concentration ranged from 121
to 1250 µg g−1 TOC (mean 463 µg g−1 TOC, SD 314) and
was highest at 22.7 m a.r.l. The short-chain n-FA concentra-
tion (C8 to C20) ranged from 120 to 968 µg g−1 TOC (mean
560 µg g−1 TOC, SD 212) and was highest in the bottom
sample at 1.4 m a.r.l. Among the short-chain FAs, the n-
C16 dominated all samples (Fig. S2). The iso- and anteiso-
branched FAs were more abundant in the bottom section of
the cliff and lowest in the middle section, and the ratio of
iso- and anteiso-branched saturated fatty acids (C11, C13, C15
and C17) to mid- and long-chain (C ≥21) n-FAs ranged from
0.03 to 0.32 (mean 0.13, SD 0.09; Figs. 3 and S2). The HPFA
index had a mean value of 0.63 (SD 0.11, n = 13), a mini-
mum of 0.45 (SOB18-06-17 at 5.5 m a.r.l.) and a maximum
of 0.86 (SOB18-01-04 at 22.7 m a.r.l.) close to the cliff top. Overall, HPFA values below 16 m a.r.l. were slightly lower
than in the upper section. 4
Results The uppermost sediments of the cliff consisted of Holocene-
age sediments (from 24.1 to 22.5 m a.r.l.; upper part of
SOB18-01) on top of Late Pleistocene Yedoma sediments
from 22.2 m a.r.l. down to the cliff base at the river wa-
ter level. A detailed cryostratigraphic description is given in
Wetterich et al. (2020a). HPFA =
Pn −fattyacids C24C26C28
Pn −fattyacids C24C26C28 + Pn −alkanes C27C29C31
(3) (3) Biogeosciences, 19, 2079–2094, 2022 https://doi.org/10.5194/bg-19-2079-2022 2085 4.3
Clustering We identified three main sub-groups using agglomerative hi-
erarchical clustering (Fig. 4a): unit I from 24.1 to 22.7 m a.r.l. (n = 3), unit II from 21.7 to 16.7 m a.r.l. (n = 7) and unit III
from 16.2 to 1.4 m a.r.l. (n = 18). Our clustering matched
the three cryostratigraphic units as defined by Wetterich et
al. (2020a), which further corresponded to MIS 1, MIS 2 and
MIS 3 from the top to the bottom. The TOC content (Fig. 4b)
and C/N ratio were significantly the highest in unit I and low-
est in unit II (p < 0.01 and p < 0.05, respectively; Fig. 2 and
Table S1). The short-chain (C14 to C20) and long-chain (C21
to C33) n-alkane concentration, expressed in µg g−1 TOC,
was higher in unit III, but the differences were only signif-
icant for the short-chain n-alkanes (p < 0.01) (Fig. 4c). The
short-chain (C8 to C20), mid-chain (C21 to C23) and long-
chain (C24 to C32) n-fatty-acid concentrations expressed in
µg g−1 TOC did not differ significantly between the units. The ACL and CPI values were similar for each unit (Fig. 3). The HPFA index was significantly different between the units
(p < 0.05) with the highest values in unit I and lowest values
in unit III (Fig. 4d). The share of iso- and anteiso-branched The n-alkane-based ACL showed variations between 27.1
(SOB18-06-19 at 4.5 m a.r.l.) and 29.0 (SOB18-01-06 at
21.7 m a.r.l.) with a mean of 28.0 across the cliff (SD 0.50;
Fig. 3). The CPI of n-alkanes (n-C23 to n-C33) ranged from
5.76 (SOB18-06-15; 6.5 m a.r.l.) to 16.29 (SOB18-06-05;
11.5 m a.r.l.) with a mean value of 9.89 (SD 2.79). Below
7 m a.r.l., the CPI significantly decreased. https://doi.org/10.5194/bg-19-2079-2022 Biogeosciences, 19, 2079–2094, 2022 2086
C. Haugk et al.: Eroding permafrost cliff carbon characteristics
Figure 3. Biomarker parameters of the Sobo-Sise Yedoma cliff. The first four columns show n-alkane parameters: short-chain (C14–C20),
and long-chain (C21–C33) n-alkane concentrations (both in µg g−1 TOC), the n-alkane average chain length (ACL23−33), and n-alkane
carbon preference index (CPI23−33). The last five columns show n-fatty-acid parameters: short-chain (C8–C20), mid-chain (C21–C23) and
long-chain (C24–C32) n-fatty-acid concentrations (in µg g−1 TOC); the ratio of iso- and anteiso-branched saturated fatty acids (C11, C13,
C15 and C17) to mid- and long-chain (C ≥21) n-fatty acids; and the higher plant fatty acid (HPFA) index. 4.3
Clustering Data points are displayed over
cliff height from the cliff top at 25 m above river level (a.r.l.) to cliff bottom at 0 m a.r.l. The three sections of SOB18 are plotted separately
for each parameter (black, dark grey and light grey circles). Units I, II (grey rectangle) and III correspond to Marine Isotope Stage (MIS) 1
to 3. C. Haugk et al.: Eroding permafrost cliff carbon characteristics 2086 2086
C. Haugk et al.: Eroding permafrost cliff carbon characteristics Figure 3. Biomarker parameters of the Sobo-Sise Yedoma cliff. The first four columns show n-alkane parameters: short-chain (C14–C20),
and long-chain (C21–C33) n-alkane concentrations (both in µg g−1 TOC), the n-alkane average chain length (ACL23−33), and n-alkane
carbon preference index (CPI23−33). The last five columns show n-fatty-acid parameters: short-chain (C8–C20), mid-chain (C21–C23) and
long-chain (C24–C32) n-fatty-acid concentrations (in µg g−1 TOC); the ratio of iso- and anteiso-branched saturated fatty acids (C11, C13,
C15 and C17) to mid- and long-chain (C ≥21) n-fatty acids; and the higher plant fatty acid (HPFA) index. Data points are displayed over
cliff height from the cliff top at 25 m above river level (a.r.l.) to cliff bottom at 0 m a.r.l. The three sections of SOB18 are plotted separately
for each parameter (black, dark grey and light grey circles). Units I, II (grey rectangle) and III correspond to Marine Isotope Stage (MIS) 1
to 3. Cooler and drier summer conditions as well as unstable
(draining and rewetting phases) aquatic conditions were re-
constructed for MIS 2 (Wetterich et al., 2021). In our study,
the elevated ratio of iso- and anteiso-branched FAs relative
to mid- and long-chain (C ≥21) n-FAs in unit I (MIS 1)
and III (MIS 3) compared to unit II (MIS 2; Figs. 4e and
S2) suggests stronger microbial activity during the warmer
MIS 3 and MIS 1 periods (Rilfors et al., 1978; Stapel et al.,
2016) and points to a higher input of bacterial biomass dur-
ing that time. Additionally, we found a significant abundance
of short-chain FAs, especially n-C16 in all samples (Fig. S2). However, these FAs are common not only in bacterial but
also in eukaryotic microorganisms (Gunstone et al., 2007)
and thus represent a mixing signal. Therefore, we focused
here on iso- and anteiso-branched FAs as they are more spe-
cific biomarkers for bacterial biomass (Kaneda, 1991). 4.3
Clustering FAs compared to mid- and long-chain n-FAs was highest in
unit III and lowest in unit II (Fig. 4e), but the differences
were not significant. C. Haugk et al.: Eroding permafrost cliff carbon characteristics 2087 Figure 4. Statistical separation of the Sobo-Sise Yedoma cliff pro-
file and selected carbon parameters of the separated cliff units. (a) Clustering of samples with y axis representing cliff height from
the cliff top at 25 m above river level (a.r.l.) to cliff bottom at
0 m a.r.l. Unit I corresponds to Marine Isotope Stage (MIS) 1, unit II
to MIS 2 and unit III to MIS 3. Resulting box plots allow better vi-
sualisation of the OM distribution along the Sobo-Sise Yedoma cliff
profile: (b) total organic carbon (TOC) content (in wt %), (c) short-
chain n-alkane concentration (in µg g−1 TOC), (d) higher plant fatty
acid (HPFA) index, and (e) ratio of iso- and anteiso-branched sat-
urated fatty acids (C11, C13, C15 and C17) to mid- and long-chain
(C ≥21) n-fatty acids. posits contained rather less decomposed twigs and grass re-
mains as well as single peaty lenses (15–20 cm in diam-
eter) and peat layers (10–20 up to 130 cm thick). A sim-
ilar occurrence of single-twig remains (2–4 mm in diame-
ter), dark brown spots, finely dispersed organic remains and
peaty lenses (5–25 cm in diameter) was found in MIS 2 de-
posits, while MIS 1 deposits contained many more peaty
components, i.e. numerous peaty lenses (2–25 cm in diam-
eter), which was reflected in higher TOC values compared to
MIS 3 and MIS 2 deposits (Wetterich et al., 2020a). At 16.2 m a.r.l., we found a peak in TOC (11.3 wt %,
SOB18-03-05) and a simultaneously depleted δ13C value
(−29.4 ‰). High TOC and low δ13C values have been found
to be indicative for peat accumulation and low decompo-
sition under wetter conditions in a more anaerobic regime
(Wetterich et al., 2009; Schirrmeister et al., 2011a; Strauss
et al., 2012). These peat layers can form by moss accumu-
lation which is hardly decomposed and/or incorporated soon
upon accumulation. From a biomarker perspective, this sam-
ple was not much different regarding the biomolecular com-
position, indicating a similar organic biomass (Fig. S2). The
higher relative abundance of iso- and anteiso-FAs relative
to mid- and long-chain FAs in this sample may point to in-
creased microbial activity and thus microbial decomposition
during the time of deposition in this OM-rich sample. C. Haugk et al.: Eroding permafrost cliff carbon characteristics Con-
sidering palaeoenvironmental studies from Kurungnakh Is-
land in the central Lena Delta, palaeosol formation was inten-
sified by relatively warm and wet summers during the climate
optimum of the interstadial MIS 3 between 40 and 32 ka BP
(Wetterich et al., 2008, 2014). Therefore, it is very likely that
this layer is a buried palaeosol layer containing peaty mate-
rial. Figure 4. Statistical separation of the Sobo-Sise Yedoma cliff pro- Figure 4. Statistical separation of the Sobo-Sise Yedoma cliff pro-
file and selected carbon parameters of the separated cliff units. (a) Clustering of samples with y axis representing cliff height from
the cliff top at 25 m above river level (a.r.l.) to cliff bottom at
0 m a.r.l. Unit I corresponds to Marine Isotope Stage (MIS) 1, unit II
to MIS 2 and unit III to MIS 3. Resulting box plots allow better vi-
sualisation of the OM distribution along the Sobo-Sise Yedoma cliff
profile: (b) total organic carbon (TOC) content (in wt %), (c) short-
chain n-alkane concentration (in µg g−1 TOC), (d) higher plant fatty
acid (HPFA) index, and (e) ratio of iso- and anteiso-branched sat-
urated fatty acids (C11, C13, C15 and C17) to mid- and long-chain
(C ≥21) n-fatty acids. Figure 4. Statistical separation of the Sobo-Sise Yedoma cliff pro-
file and selected carbon parameters of the separated cliff units. (a) Clustering of samples with y axis representing cliff height from
the cliff top at 25 m above river level (a.r.l.) to cliff bottom at
0 m a.r.l. Unit I corresponds to Marine Isotope Stage (MIS) 1, unit II
to MIS 2 and unit III to MIS 3. Resulting box plots allow better vi-
sualisation of the OM distribution along the Sobo-Sise Yedoma cliff
profile: (b) total organic carbon (TOC) content (in wt %), (c) short-
chain n-alkane concentration (in µg g−1 TOC), (d) higher plant fatty
acid (HPFA) index, and (e) ratio of iso- and anteiso-branched sat-
urated fatty acids (C11, C13, C15 and C17) to mid- and long-chain
(C ≥21) n-fatty acids. 5.1.1
Organic matter source We found that the n-alkane distributions were dominated
by the long-chain n-alkanes (C ≥21) and that short-chain
n-alkanes only played a marginal role (Fig. S1). The most
abundant n-alkane homologues in the entire data set were n-
C27, n-C29 and n-C31, which indicates that the OM stemmed
from higher land plants (Eglinton and Hamilton, 1967). This
is confirmed by the dominance of long-chain n-FAs (C24–
C32) with a strong even-over-odd carbon number predom-
inance (Fig. S2). The relatively high ACL across the cliff
(Fig. 3) indicates a predominant contribution of vascular
plants, which corroborates the pollen record presented by
Wetterich et al. (2021). Their results indicated the presence
of tundra–steppe vegetation during MIS 3–2, while MIS 1
pollen spectra of the uppermost three samples indicated a
shift from tundra–steppe to shrub–tundra vegetation. Occa-
sional warmer-than-today summers were recorded during the
early MIS 3 as well as the presence of low-centre polygons
with favourable (stable) aquatic conditions during MIS 3. The source and nature of the OM preserved in permafrost
influence both its quantity and its quality (Jongejans et
al., 2018). TOC, TN, C/N and δ13C variations result from
changes in biomass productivity and/or decomposition, from
different OM sources, from changes in depositional condi-
tions influencing OM preservation, and from different char-
acteristics of Cryosol formation. Generally, enriched δ13C
and low TOC and C/N values, as we found in unit II and
III (Fig. 2), are typical of Yedoma deposits that formed dur-
ing cold stages (Schirrmeister, 2012). However, climate vari-
ations during the last ice age were differentiated into warmer
interstadials (e.g. MIS 3) and colder stadial periods (e.g. MIS 2) which climatically triggered changes in vegetation
and Cryosol formation. At the Sobo-Sise Yedoma cliff, the Biogeosciences, 19, 2079–2094, 2022 https://doi.org/10.5194/bg-19-2079-2022 5.2
Implications Nitzbon et al. (2020) found that terrestrial permafrost-locked
OC will be significantly thaw-affected by 2100, and it could
even be up to 3-fold (12-fold) more under warming sce-
nario RCP4.5 (RCP8.5) compared to previous estimates if in-
cluding thermokarst-inducing processes. Deep OM as char-
acterised in our study can be released by deep disturbance
processes such as thermokarst development, thermal erosion
or riverbank erosion. Our findings show that freshly thawed
and high-quality OM was frozen in the cliff sediments and
is now being mobilised rapidly as is shown by the high an-
nual erosion rates of 15.7 m yr−1 as reported by Fuchs et
al. (2020) (2015–2018, in the long term 9 m yr−1). Further-
more, we suggest that the very ice-rich cliff wall sections are
not exposed to aerobic conditions for very long time peri-
ods before being eroded into the Lena River. Thus, aerobic
microbial decomposition of the OM at the cliff front is pre-
sumably playing only a minor role. Additionally, cliff ero-
sion is mainly driven by thermo-erosion and niche forma-
tion at the base of the ice-rich Yedoma cliff, resulting in
block failure instead of slow gradual cliff retreat (Fuchs et
al., 2020). Accordingly, some of the OM in the cliff may not
even become exposed to the air and thaw at all before be-
ing eroded into the river. Fuchs et al. (2020) showed an av-
erage loss of 5.2 × 106 kg OC yr−1 and 0.4 × 106 kg N yr−1
(2015–2018). For the OC flux sourced from permafrost and
peat deposits (and in particular from erosive locations like
our study site on Sobo-Sise Island), Wild et al. (2019) esti-
mated 0.9 × 108 kg C yr−1. Previous studies showed that mineral-associated OM can
make up a substantial fraction of the OM in permafrost soils,
which protects the OM from decomposition (Dutta et al.,
2006; Mueller et al., 2015). We found that the biomarker
concentrations are negatively correlated with the mean grain
size published by Wetterich et al. (2020b) (Fig. S3). Espe-
cially the negative correlation between the grain size and the
short- (p < 0.01), mid- and long-chain n-FA concentrations
(p < 0.05) are significant. This suggests that, even though
the n-FAs are generally vulnerable with respect to decom-
position, the n-FAs might to some extent be protected from
OM decomposition upon mobilisation and transport. C. Haugk et al.: Eroding permafrost cliff carbon characteristics A higher decomposition for
the lower cliff part is also supported by the highest ratio of the
iso- and anteiso-branched FAs vs. mid- and long-chain FAs
in the MIS 3 deposits (Fig. 4e), indicating a higher relative
amount of microbial biomass and suggesting higher micro-
bial activity during this warmer interval. the control of these processes on the finally preserved OM
biomarker signatures. 5.1.2
Organic matter quality OM from different vegetation types was incorporated from
the active layer into the permafrost during and after dif-
ferent phases of decomposition. The biogeochemical and
biomarker proxies outlined in the previous section mainly
describe the sources and composition of permafrost OM. In
addition, biomarker ratios provide information on the decom-
position level of the OM and, with that, the potential “decom-
posability” (quality) of the respective permafrost OM upon
thaw. The OM assessment of this study via agglomerative
clustering found an overall high OM quality of the Sobo-
Sise Yedoma deposits with high CPI values (mean 9.89)
and higher C/N ratios (mean 13.24) compared to the other
Yedoma deposits such as those on the Buor Khaya Penin-
sula (central Laptev Sea) with mean C/N values of about 10
(Strauss et al., 2015). The elevated C/N value in the top sam-
ple of the Sobo-Sise record likely results from the influence
of modern plants rooted in the active layer. For the rest of the
profile, the C/N ratios were rather uniform. The high CPI in
our study is comparable to other Yedoma sites as reported
by Strauss et al. (2015) for the Buor Khaya Peninsula (mean TOC values were higher during MIS 1 and 3 compared to the
last glacial (MIS 2) deposits, suggesting higher OM accumu-
lation, which was presumably triggered by higher biomass
production. The TOC values from the MIS 3 and MIS 2 sedi-
ments of the Sobo-Sise Yedoma cliff (< 0.1 wt %–11.3 wt %,
mean 4.0 wt %) were significantly higher (p < 0.01) than
those of other Siberian Yedoma sites (< 0.1 wt %–27 wt %,
mean 3.0 wt %; 17 study sites, 719 samples) but very simi-
lar to data from Kurungnakh Island (mean 3.8 wt %) which
is located about 70 km west-south-west of Sobo-Sise Island
in the central Lena Delta (Strauss et al., 2012, 2020). Likely,
the high TOC values in the Sobo-Sise record are a result of
past wetter conditions leading to the formation of peat layers. Comparably to the Kurungnakh Island Yedoma record, the
Sobo-Sise Yedoma cliff is characterised by silty sediments
with multiple layers enriched in peat pointing to palaeosol
formation during permafrost aggradation. The MIS 3 de- https://doi.org/10.5194/bg-19-2079-2022 Biogeosciences, 19, 2079–2094, 2022 C. Haugk et al.: Eroding permafrost cliff carbon characteristics 2088 11.6) and Jongejans et al. (2018) for the Baldwin Peninsula
(western Alaska; 12.2). At the Sobo-Sise Yedoma cliff, the
CPI values scattered around a mean of 9.89 and decreased in
the lowermost 7 m of the cliff profile. This could probably in-
dicate a higher level of OM decomposition for the lower cliff
part but can also be influenced by the vegetation type and
species prevailing during the early MIS 3 with stagnant wa-
ter and partly warmer-than-today summer climate conditions
(Wetterich et al., 2021). The HPFA values (0.45–0.86, me-
dian 0.61) are a bit higher compared to Yedoma deposits in-
vestigated by Strauss et al. (2015) on the Buor Khaya Penin-
sula (0.15–0.69, median 0.54). Overall, the HPFA signifi-
cantly decreased downwards (Fig. 4d; p < 0.05), especially
below unit II, which suggests that the OM is further decom-
posed downwards. This fits the assumption that there was
more time for OM decomposition for the older lower cliff
parts of the palaeo-active layer. A higher decomposition for
the lower cliff part is also supported by the highest ratio of the
iso- and anteiso-branched FAs vs. mid- and long-chain FAs
in the MIS 3 deposits (Fig. 4e), indicating a higher relative
amount of microbial biomass and suggesting higher micro-
bial activity during this warmer interval. 11.6) and Jongejans et al. (2018) for the Baldwin Peninsula
(western Alaska; 12.2). At the Sobo-Sise Yedoma cliff, the
CPI values scattered around a mean of 9.89 and decreased in
the lowermost 7 m of the cliff profile. This could probably in-
dicate a higher level of OM decomposition for the lower cliff
part but can also be influenced by the vegetation type and
species prevailing during the early MIS 3 with stagnant wa-
ter and partly warmer-than-today summer climate conditions
(Wetterich et al., 2021). The HPFA values (0.45–0.86, me-
dian 0.61) are a bit higher compared to Yedoma deposits in-
vestigated by Strauss et al. (2015) on the Buor Khaya Penin-
sula (0.15–0.69, median 0.54). Overall, the HPFA signifi-
cantly decreased downwards (Fig. 4d; p < 0.05), especially
below unit II, which suggests that the OM is further decom-
posed downwards. This fits the assumption that there was
more time for OM decomposition for the older lower cliff
parts of the palaeo-active layer. C. Haugk et al.: Eroding permafrost cliff carbon characteristics the aquatic system of the Lena Delta and nearshore Laptev
Sea and Arctic Ocean areas, which will affect local but likely
also regional biogeochemical cycles in the marine realm
(Grotheer et al., 2020; Tanski et al., 2021; Mann et al.,
2022) and the shelf seas. Once mobilised and transported
into inland waters, permafrost-derived OC can be rapidly
used by aquatic microorganisms, increasing OM decompo-
sition in riverine and coastal Arctic waters (Vonk et al.,
2012, 2013a, 2015; Drake et al., 2015; Mann et al., 2015). Vonk et al. (2012) studied the organic matter exported from
the fast-eroding Yedoma cliff of Muostakh Island, which is
located about 120 km south-east of the Sobo-Sise Yedoma
cliff. They found that decomposition of Yedoma OM prior
to delivery to the ocean was substantial. In their study of
OM mobilisation by retrogressive thaw slumps in Canada,
Bröder et al. (2021) found that the majority of the exported
OC was derived from permafrost deposits. However, they
also found that Pleistocene permafrost deposits mostly con-
tained less labile, slowly cycling permafrost OC. Similarly,
Bröder et al. (2019) showed that more than half of the car-
bon transported and deposited on the shelf sea floor likely
resists decomposition on a centennial scale while the rest
decays relatively slowly. Furthermore, OM mobilised from
Pleistocene or Holocene permafrost by rapid thermokarst and
mass-wasting processes contribute different shares of partic-
ulate and dissolved organic matter (Kokelj et al., 2021; Shakil
et al., 2020), which has implications for decomposition and
transport. In addition, Karlsson et al. (2011) hypothesised
that Yedoma OC, associated with mineral-rich matter from
coastal erosion, is ballasted and thus quickly settles to the
bottom. Increasing thermokarst and mass-wasting processes,
as well as river and coastal erosion, will continue to mobilise
both labile and recalcitrant OM from Pleistocene permafrost
deposits, and it is still largely unknown what short- and long-
term effects this will have on the release of greenhouse gases
and alteration of biochemical processes in nearshore waters. position for the bottom 7 m (CPI) or even for the bottom 15 m
(HPFA) of the cliff profile of MIS 3 age, and less OM accu-
mulation during MIS 2 in contrast to the warmer MIS 3 and
1 sequences is assumed. C. Haugk et al.: Eroding permafrost cliff carbon characteristics At the Sobo-Sise Yedoma cliff, rep-
resenting an example of rapidly eroding permafrost shore-
lines in the Lena Delta, OM with a high decomposition po-
tential is being mobilised from almost all sections of the cliff
profile. This material is suggested to rapidly enter the flu-
vial and probably also the offshore aquatic ecosystem. Thus,
OM mobilisation at the Sobo-Sise Yedoma cliff and simi-
larly eroding permafrost sites bear the potential to impact the
carbon dynamics, the biogeochemistry, and the riverine and
nearshore marine ecosystems. Data availability. The
data
presented
in
this
study
are
freely
available
in
the
PANGAEA
data
repository
(https://doi.org/10.1594/PANGAEA.935672,
Haugk
et
al.,
2022). Cryolithological
and
geochronological
data
from
the
Sobo-Sise
Yedoma
cliff
are
available
in
PANGAEA
(https://doi.org/10.1594/PANGAEA.919470,
Wetterich
et
al.,
2020b). Supplement. The supplement related to this article is available on-
line at: https://doi.org/10.5194/bg-19-2079-2022-supplement. Author contributions. CH and JS designed this study and drafted a
first version of the manuscript. CH carried out the lipid biomarker
analyses and interpretation, with help from LLJ, CK and KM. SW,
LS and AK conducted the sampling and field studies. CH, JS and
LLJ led the manuscript writing. All co-authors contributed to the
manuscript writing process. Competing interests. The contact author has declared that neither
they nor their co-authors have any competing interests. 5.2
Implications On top
of the preferred decomposition of short-chain n-alkanes over
their long counterpart (Elias et al., 2007), the stronger neg-
ative correlation (even though not significant) between the
grain size and the long-chain n-alkanes compared to their
short-chain counterpart could suggest that the latter might be
more vulnerable to decomposition or might reflect the dif-
ferent sources of these biomolecules. While long-chain n-
alkanes are derived from higher land plants and enter the soil
by deposition, short-chain n-alkanes might contain a signifi-
cant portion of microbial biomass, whose abundance depends
on the availability of appropriate substrates. g
y
By using a biomarker approach (e.g. n-alcohols, n-fatty
acids, n-alkanes) on sub-aquatic sediments, van Dongen et
al. (2008) found a lower degree of decomposition of the old
terrestrial OM released by the great eastern Arctic rivers, in-
cluding the Lena River, compared to the western Eurasian
Arctic. Thus, they predicted greater remineralisation rates
and release of carbon dioxide and methane. Our biomarker
findings of terrestrial permafrost fit well into this scenario. Winterfeld et al. (2015) studied the lignin phenol composi-
tion of the Lena River, Lena Delta and Laptev Sea nearshore
zones and proposed that OM decomposition is consider-
able after permafrost thawing on land and during transport
and sedimentation in the water. The present study on the
OM origin and the annual OC erosion rates at the Sobo-
Sise Yedoma cliff complements ongoing research on mobil-
isation of permafrost-locked carbon from Late Pleistocene
Yedoma deposits, while thermal erosion is a widespread and
climate-sensitive phenomenon in the Yedoma domain, cov-
ering nearly 5 × 105 km2 in Siberia and Alaska (Strauss et
al., 2021b). This indicates the high potential of thermal ero-
sion for mobilisation and release upon thaw of not only
large amounts of carbon but also well-preserved OM into As outlined above, three stages of permafrost aggradation
on Sobo-Sise Island linked to climatic variability were iden-
tified according to Wetterich et al. (2020a). OM preservation
during these stages is strongly impacted by the duration of
freezing and thawing periods, the associated presence and
absence of oxygen in the soil, the related level of microbial
activity, and/or physical protection of the OM by the inor-
ganic matrices (e.g. Fe complexation) (Freeman et al., 2001;
Hedges and Keil, 1995; Lützow et al., 2006). As these factors
are all closely interlinked, it is almost impossible to decipher Biogeosciences, 19, 2079–2094, 2022 https://doi.org/10.5194/bg-19-2079-2022 2089 https://doi.org/10.5194/bg-19-2079-2022 References Fuchs, M., Nitze, I., Strauss, J., Gunther, F., Wetterich, S.,
Kizyakov, A., Fritz, M., Opel, T., Grigoriev, M. N., Maksimov, G. T., and Grosse, G.: Rapid Fluvio-Thermal Erosion of a Yedoma
Permafrost Cliff in the Lena River Delta, Front. Earth Sci., 8,
336, https://doi.org/10.3389/feart.2020.00336, 2020. Andersson, R. A. and Meyers, P. A.: Effect of climate change on de-
livery and degradation of lipid biomarkers in a Holocene peat se-
quence in the Eastern European Russian Arctic, Org. Geochem.,
53, 63–72, https://doi.org/10.1016/j.orggeochem.2012.05.002,
2012. Fuchs, M., Ogneva, O., Sanders, T., Schneider, W., Polyakov, V.,
Becker, O. O., Bolshiyanov, D., Mollenhauer, G., and Strauss, J.:
CACOON Sea – water sampling along the Sardakhskaya chan-
nel and near shore of the Laptev Sea (chapter 3.26) in: Reports on
Polar and Marine Research – Russian-German Cooperation: Ex-
peditions to Siberia in 2019, edited by: Fuchs, M., Bolshiyanov,
D., Grigoriev, M. N., Morgenstern, A., Pestryakova, L. A., Tsi-
bizov, L., and Dill, A., Alfred Wegener Institute, Bremerhaven,
https://doi.org/10.48433/BzPM_0749_2021, 2021. Bray, E. E. and Evans, E. D.: Distribution of n-paraffins as a clue
to recognition of source beds, Geochim. Cosmochim. Ac., 22,
2–15, https://doi.org/10.1016/0016-7037(61)90069-2, 1961. Bröder, L., Andersson, A., Tesi, T., Semiletov, I., and Gustafs-
son, Ö.: Quantifying Degradative Loss of Terrigenous Or-
ganic Carbon in Surface Sediments Across the Laptev and
East
Siberian
Sea,
Glob. Biogeochem. Cy.,
33,
85–99,
https://doi.org/10.1029/2018GB005967, 2019. Grigoriev, M. N.: Cryomorphogenesis in the Lena Delta, Permafrost
Institute Press, Yakutsk, 1993. Canuel, E. A. and Martens, C. S.: Reactivity of recently de-
posited organic matter: Degradation of lipid compounds near the
sediment-water interface, Geochim. Cosmochim. Ac., 60, 1793–
1806, https://doi.org/10.1016/0016-7037(96)00045-2, 1996. Grosse, G., Schirrmeister, L., Siegert, C., Kunitsky, V. V.,
Slagoda, E. A., Andreev, A. A., and Dereviagyn, A. Y.:
Geological
and
geomorphological
evolution
of
a
sedi-
mentary
periglacial
landscape
in
Northeast
Siberia
dur-
ing
the
Late
Quaternary,
Geomorphology,
86,
25–51,
https://doi.org/10.1016/j.geomorph.2006.08.005, 2007. Chen, J., Günther, F., Grosse, G., Liu, L., and Lin, H.: Sentinel-1
InSAR Measurements of Elevation Changes over Yedoma Up-
lands on Sobo-Sise Island, Lena Delta, Remote Sens., 10, 1152,
https://doi.org/10.3390/rs10071152, 2018. Grosse, G., Harden, J., Turetsky, M., McGuire, A. D., Camill,
P., Tarnocai, C., Frolking, S., Schuur, E. A. G., Jorgenson, T.,
Marchenko, S., Romanovsky, V., Wickland, K. P., French, N.,
Waldrop, M., Bourgeau-Chavez, L., and Striegl, R. G.: Vul-
nerability of high-latitude soil organic carbon in North Amer-
ica to disturbance, J. Geophys. Res.-Biogeo., 116, G00K06,
https://doi.org/10.1029/2010jg001507, 2011. 6
Conclusions Disclaimer. Publisher’s note: Copernicus Publications remains
neutral with regard to jurisdictional claims in published maps and
institutional affiliations. Sedimentological and biogeochemical analyses showed that
the sediments exposed at the Sobo-Sise Yedoma cliff con-
tain a high TOC content (mean 5 wt %) and well-preserved
OM (C/N mean 13.2, mean CPI 9.89) in comparison to
other Yedoma permafrost sites. Our study corroborated the
palaeoenvironmental data from the Sobo-Sise Yedoma cliff
from previous research which suggested that Yedoma forma-
tion during the interstadial MIS 3 and the accumulation of
the topmost Holocene deposits (MIS 1) were associated with
more microbial activity than during the stadial MIS 2. In
addition, our findings suggest that mainly high-quality OM
has been freeze-locked perennially into permafrost during
the Late Pleistocene to Holocene. Although the OM qual-
ity seems to be overall fairly high (TOC, C/N and CPI),
biomarker parameters indicate a higher level of OM decom- Acknowledgements. This
study
was
carried
out
within
the
CACOON project and the joint Russian–German expeditions
Lena 2018 and Sobo-Sise 2018 supported by the Samoylov Re-
search Station. We thank Michael Fritz (AWI Potsdam) and Aleksey
Aksenov (Arctic and Antarctic Research Institute, St Petersburg) for
supporting the permafrost sampling on Sobo-Sise in August 2018. We thank Justin Lindemann (AWI Potsdam) and Anke Sobotta
(German Research Centre for Geoscience) as well as Hanno Meyer
and Mikaela Weiner (AWI ISOLAB Facility) for support with lab-
oratory analysis. Biogeosciences, 19, 2079–2094, 2022 https://doi.org/10.5194/bg-19-2079-2022 C. Haugk et al.: Eroding permafrost cliff carbon characteristics 2090 Financial support. This research has been supported by the Bun-
desministerium für Bildung und Forschung (grant no. 03F0806A),
UK Research and Innovation (grant no. NE/R012806/1) and the
Deutsche Forschungsgemeinschaft (grant no. WE4390/7-1). Eglinton, T. I. and Eglinton, G.: Molecular proxies for paleocli-
matology, 275, 1–16, https://doi.org/10.1016/j.epsl.2008.07.012,
2008. Fedorova, I., Chetverova, A., Bolshiyanov, D., Makarov, A., Boike,
J., Heim, B., Morgenstern, A., Overduin, P. P., Wegner, C.,
Kashina, V., Eulenburg, A., Dobrotina, E., and Sidorina, I.:
Lena Delta hydrology and geochemistry: long-term hydrological
data and recent field observations, Biogeosciences, 12, 345–363,
https://doi.org/10.5194/bg-12-345-2015, 2015. The article processing charges for this open-access
publication were covered by the Alfred Wegener Institute,
Helmholtz Centre for Polar and Marine Research (AWI). Freeman, C., Evans, C. D., Monteith, D. T., Reynolds, B., and Fen-
ner, N.: Export of organic carbon from peat soils, Nature, 412,
785–785, https://doi.org/10.1038/35090628, 2001. Review statement. This paper was edited by Sebastian Naeher and
reviewed by two anonymous referees. Review statement. This paper was edited by Sebastian Naeher and
reviewed by two anonymous referees. Fuchs, M., Grosse, G., Strauss, J., Günther, F., Grigoriev, M., Max-
imov, G. M., and Hugelius, G.: Carbon and nitrogen pools in
thermokarst-affected permafrost landscapes in Arctic Siberia,
Biogeosciences, 15, 953–971, https://doi.org/10.5194/bg-15-
953-2018, 2018. C. Haugk et al.: Eroding permafrost cliff carbon characteristics W., and John, S.: Or-
ganic matter composition and stabilization in a polygonal tun-
dra soil of the Lena Delta, Biogeosciences, 10, 3145–3158,
https://doi.org/10.5194/bg-10-3145-2013, 2013. Holmes, R. M., McClelland, J. W., Peterson, B. J., Tank, S. E.,
Bulygina, E., Eglinton, T. I., Gordeev, V. V., Gurtovaya, T. Y.,
Raymond, P. A., Repeta, D. J., Staples, R., Striegl, R. G., Zhuli-
dov, A. V., and Zimov, S. A.: Seasonal and Annual Fluxes of
Nutrients and Organic Matter from Large Rivers to the Arc-
tic Ocean and Surrounding Seas, Estuar. Coasts, 35, 369–382,
https://doi.org/10.1007/s12237-011-9386-6, 2012. Kokelj, S. V., Kokoszka, J., van der Sluijs, J., Rudy, A. C. A.,
Tunnicliffe, J., Shakil, S., Tank, S. E., and Zolkos, S.: Thaw-
driven mass wasting couples slopes with downstream systems,
and effects propagate through Arctic drainage networks, The
Cryosphere, 15, 3059–3081, https://doi.org/10.5194/tc-15-3059-
2021, 2021. Kuhn, T. K., Krull, E. S., Bowater, A., Grice, K., and Gleixner,
G.: The occurrence of short chain n-alkanes with an even
over odd predominance in higher plants and soils, 41, 88–95,
doi:10.1016/j.orggeochem.2009.08.003, 2010. Holmes,
R. M.,
Shiklomanov,
A. I.,
Suslova,
A.,
Tre-
tiakov,
M.,
McClelland,
J. W.,
Scott,
L.,
Spencer,
R. G. M.,
and
Tank,
S. E.:
River
Discharge:
https:
//arctic.noaa.gov/Report-Card/Report-Card-2021/ArtMID/
8022/ArticleID/953/River-Discharge (last access: 14 March
2022), 2021. Lützow, M. v., Kögel-Knabner, I., Ekschmitt, K., Matzner, E.,
Guggenberger, G., Marschner, B., and Flessa, H.: Stabiliza-
tion of organic matter in temperate soils: mechanisms and
their relevance under different soil conditions – a review, Eu-
rop. J. Soil Sci., 57, 426–445, https://doi.org/10.1111/j.1365-
2389.2006.00809.x, 2006. Hugelius, G., Strauss, J., Zubrzycki, S., Harden, J. W., Schuur, E. A. G., Ping, C.-L., Schirrmeister, L., Grosse, G., Michaelson, G. J., Koven, C. D., O’Donnell, J. A., Elberling, B., Mishra, U.,
Camill, P., Yu, Z., Palmtag, J., and Kuhry, P.: Estimated stocks
of circumpolar permafrost carbon with quantified uncertainty
ranges and identified data gaps, Biogeosciences, 11, 6573–6593,
https://doi.org/10.5194/bg-11-6573-2014, 2014. Mann, P. J., Eglinton, T. I., McIntyre, C. P., Zimov, N.,
Davydova, A., Vonk, J. E., Holmes, R. M., and Spencer,
R. G.: Utilization of ancient permafrost carbon in head-
waters of Arctic fluvial networks, Nat. Commun., 6, 7856,
https://doi.org/10.1038/ncomms8856, 2015. IPCC: Special Report on the Ocean and Cryosphere in a Chang-
ing Climate, Intergov, Panel on Climate Change, Monaco,
https://doi.org/10.1017/9781009157964, 2019. Mann, P. C. Haugk et al.: Eroding permafrost cliff carbon characteristics 2091 Peninsula, Northeastern Siberia, Front. Environ. Sci., 8, 118,
https://doi.org/10.3389/fenvs.2020.00118, 2020. Peninsula, Northeastern Siberia, Front. Environ. Sci., 8, 118,
https://doi.org/10.3389/fenvs.2020.00118, 2020. F. D., Harwood, J. L., and Dijkstra, A. J., CRC press (Tylor Fran-
cis Group), 51–156, 2007. F. D., Harwood, J. L., and Dijkstra, A. J., CRC press (Tylor Fran-
cis Group), 51–156, 2007. Juggins, S.: Package: rioja: Analysis of Quaternary Science Data,
2019. Günther, F., Overduin, P. P., Sandakov, A. V., Grosse, G., and Grig-
oriev, M. N.: Short- and long-term thermo-erosion of ice-rich
permafrost coasts in the Laptev Sea region, Biogeosciences, 10,
4297–4318, https://doi.org/10.5194/bg-10-4297-2013, 2013. Kaneda, T.: Iso- and anteiso-fatty acids in bacteria: biosynthesis,
function, and taxonomic significance, Microbiol. Rev., 55, 288–
302, https://doi.org/10.1128/mr.55.2.288-302.1991, 1991. Günther, F., Overduin, P. P., Yakshina, I. A., Opel, T., Baranskaya,
A. V., and Grigoriev, M. N.: Observing Muostakh disappear: per-
mafrost thaw subsidence and erosion of a ground-ice-rich island
in response to arctic summer warming and sea ice reduction, The
Cryosphere, 9, 151–178, https://doi.org/10.5194/tc-9-151-2015,
2015. Kanevskiy, M., Shur, Y., Strauss, J., Jorgenson, T., Fortier,
D., Stephani, E., and Vasiliev, A.: Patterns and rates of
riverbank
erosion
involving
ice-rich
permafrost
(yedoma)
in
northern
Alaska,
Geomorphology,
253,
370–384,
https://doi.org/10.1016/j.geomorph.2015.10.023, 2016. Karlsson, E. S., Charkin, A., Dudarev, O., Semiletov, I., Vonk,
J. E., Sánchez-García, L., Andersson, A., and Gustafsson, Ö.:
Carbon isotopes and lipid biomarker investigation of sources,
transport and degradation of terrestrial organic matter in the
Buor-Khaya Bay, SE Laptev Sea, Biogeosciences, 8, 1865–1879,
https://doi.org/10.5194/bg-8-1865-2011, 2011. Haugk, C., Jongejans, L. L., Mangelsdorf, K., Fuchs, M., Ogneva,
O., Palmtag, J., Mollenhauer, G., Mann, P. J., Overduin, P. P., Grosse, G., Sanders, T., Tuerena, R. E., Schirrmeister,
L., Wetterich, S., Kizyakov, A., Karger, C., and Strauss, J.:
Organic matter characteristics using lipid biomarker analysis
of a rapidly eroding permafrost cliff, PANGAEA [data set],
https://doi.org/10.1594/PANGAEA.935672, 2022. Killops, S. D. and Killops, V. J.: Introduction to organic geochem-
istry, John Wiley & Sons, New York, 393 pp., 2013. Hedges, J. I. and Keil, R. G.: Sedimentary organic matter preserva-
tion: an assessment and speculative synthesis, Mar. Chem., 49,
81–115, https://doi.org/10.1016/0304-4203(95)00008-F, 1995. Kim, S., Stanford, L., Rodgers, R., Marshall, A., Walters,
C., Qian, K., Wenger, L., and Mankiewicz, P.: Microbial
Alteration of the Acidic and Neutral Polar NSO Com-
pounds Revealed by Fourier Transform Ion Cyclotron Res-
onance Mass Spectrometry, Org. Geochem., 36, 1117–1134,
https://doi.org/10.1016/j.orggeochem.2005.03.010, 2005. Höfle, S., Rethemeyer, J., Mueller, C. References Costard, F., Gautier, E., Fedorov, A., Konstantinov, P., and
Dupeyrat, L.: An Assessment of the Erosion Potential of
the Fluvial Thermal Process during Ice Breakups of the
Lena River (Siberia), Permafrost Perigl. Proc., 25, 162–171,
https://doi.org/10.1002/ppp.1812, 2014. Cranwell,
P. A.:
Lipid
geochemistry
of
sediments
from
Upton
Broad,
a
small
productive
lake,
7,
25–37,
https://doi.org/10.1016/0146-6380(84)90134-7, 1984. Grotheer, H., Meyer, V., Riedel, T., Pfalz, G., Mathieu, L., Hefter,
J., Gentz, T., Lantuit, H., Mollenhauer, G., and Fritz, M.: Burial
and Origin of Permafrost-Derived Carbon in the Nearshore Zone
of the Southern Canadian Beaufort Sea, Geophys. Res. Lett., 47,
e2019GL085897, https://doi.org/10.1029/2019GL085897, 2020. Drake, T. W., Wickland, K. P., Spencer, R. G., McKnight, D. M., and Striegl, R. G.: Ancient low-molecular-weight or-
ganic acids in permafrost fuel rapid carbon dioxide produc-
tion upon thaw, P. Natl. Acad. Sci. USA, 112, 13946–13951,
https://doi.org/10.1073/pnas.1511705112, 2015. Gundelwein, A., Muller-Lupp, T., Sommerkorn, M., Haupt, E. T. K., Pfeiffer, E. M., and Wiechmann, H.: Carbon in tun-
dra soils in the Lake Labaz region of arctic Siberia, Eu-
rop. J. Soil Sci., 58, 1164–1174, https://doi.org/10.1111/j.1365-
2389.2007.00908.x, 2007. Dutta,
K.,
Schuur,
E. A. G.,
Neff,
J. C.,
and
Zimov,
S. A.: Potential carbon release from permafrost soils of
Northeastern Siberia, Glob. Change Biol., 12, 2336–2351,
https://doi.org/10.1111/j.1365-2486.2006.01259.x, 2006. Gunstone, F. D. and Harwood, J. L.: Occurrence and characterisa-
tion of oils and fats, in: The lipid handbook, edited by: Gunstone, Eglinton, G. and Hamilton, R. J.: Leaf Epicuticular Waxes, Science,
156, 1322–1335, https://doi.org/10.1126/science.156.3780.1322,
1967. https://doi.org/10.5194/bg-19-2079-2022 Biogeosciences, 19, 2079–2094, 2022 C. Haugk et al.: Eroding permafrost cliff carbon characteristics P., Strauss,
J., Laboor, S., and Boike, J.: Fast response of cold ice-rich
permafrost in northeast Siberia to a warming climate, Nat. Commun., 11, 2201, https://doi.org/10.1038/s41467-020-15725-
8, 2020. Schirrmeister, L.: Late Glacial to Holocene landscape dynamics of
Arctic coastal lowlands in NE Siberia, Quat. Int., 279, p. 434,
https://doi.org/10.1016/j.quaint.2012.08.1413, 2012. Schirrmeister, L., Froese, D. G., Tumskoy, V., Grosse, G., and Wet-
terich, S.: Yedoma: Late Pleistocene ice-rich syngenetic per-
mafrost of Beringia, in: Encyclopedia of Quaternary Sciences,
2 Edn., edited by: Elias, S. A., Elsevier, Amsterdam, 542–552,
2013. Nitze,
I. and
Grosse,
G.:
Detection
of
landscape
dynam-
ics in the Arctic Lena Delta with temporally dense Land-
sat time-series stacks, Remote Sens. Environ., 181, 27–41,
https://doi.org/10.1016/j.rse.2016.03.038, 2016. Palmtag, J., Hugelius, G., Lashchinskiy, N., Tamstorf, M. P.,
Richter, A., Elberling, B., and Kuhry, P.: Storage, Landscape Dis-
tribution, and Burial History of Soil Organic Matter in Contrast-
ing Areas of Continuous Permafrost, Arct. Antarct. Alp. Res., 47,
71-88, https://doi.org/10.1657/AAAR0014-027, 2015. Schirrmeister, L., Dietze, E., Matthes, H., Grosse, G., Strauss, J.,
Laboor, S., Ulrich, M., Kienast, F., and Wetterich, S.: The gene-
sis of Yedoma Ice Complex permafrost – grain-size endmember
modeling analysis from Siberia and Alaska, Quat. Sci. J., 69, 33–
53, https://doi.org/10.5194/egqsj-69-33-2020, 2020. Schneider, J., Grosse, G., and Wagner, D.: Land cover classifi-
cation of tundra environments in the Arctic Lena Delta based
on Landsat 7 ETM+ data and its application for upscaling
of methane emissions, Remote Sens. Environ., 113, 380–391,
https://doi.org/10.1016/j.rse.2008.10.013, 2009. Poynter, J. G.: Molecular stratigraphy: The recognition of palaeo-
climatic signals in organic geochemical data, PhD, School of
Chemistry, University of Bristol, Bristol, 324 pp., 1989. Radke, M., Willsch, H., and Welte, D. H.: Preparative hy-
drocarbon group type determination by automated medium
pressure liquid chromatography, Anal. Chem., 52, 406–411,
https://doi.org/10.1021/ac50053a009, 1980. Schulte, S., Mangelsdorf, K., and Rullkötter, J.: Organic mat-
ter preservation on the Pakistan continental margin as revealed
by biomarker geochemistry, Org. Geochem., 31, 1005–1022,
https://doi.org/10.1016/S0146-6380(00)00108-X, 2000. Rieley, G., Collier, R., Jones, D., and Eglinton, G.: The biogeo-
chemistry of Ellesmere Lake, UK – I: source correlation of leaf
wax inputs to the sedimentary lipid record, Org. Geochem., 17,
901–912, https://doi.org/10.1016/0146-6380(91)90031-E, 1991. Schuur, E. A. G., Bockheim, J., Canadell, J. G., Euskirchen, E.,
Field, C. B., Goryachkin, S. V., Hagemann, S., Kuhry, P., Lafleur,
P. C. Haugk et al.: Eroding permafrost cliff carbon characteristics 2092 nitrogen fluxes of the Lena River Delta, Ambio, 51, 423–438,
https://doi.org/10.1007/s13280-021-01665-0, 2022. nitrogen fluxes of the Lena River Delta, Ambio, 51, 423–438,
https://doi.org/10.1007/s13280-021-01665-0, 2022. D., Fuchs, M., Palmtag, J., Kuhry, P., Treat, C. C., Zubrzycki,
S., Hoffman, F. M., Elberling, B., Camill, P., Veremeeva, A.,
and Orr, A.: Spatial heterogeneity and environmental predictors
of permafrost region soil organic carbon stocks, Sci. Adv., 7,
eaaz5236, https://doi.org/10.1126/sciadv.aaz5236, 2021. D., Fuchs, M., Palmtag, J., Kuhry, P., Treat, C. C., Zubrzycki,
S., Hoffman, F. M., Elberling, B., Camill, P., Veremeeva, A.,
and Orr, A.: Spatial heterogeneity and environmental predictors
of permafrost region soil organic carbon stocks, Sci. Adv., 7,
eaaz5236, https://doi.org/10.1126/sciadv.aaz5236, 2021. Schäfer, I. K., Lanny, V., Franke, J., Eglinton, T. I., Zech,
M., Vysloužilová, B., and Zech, R.: Leaf waxes in litter
and topsoils along a European transect, SOIL, 2, 551–564,
https://doi.org/10.5194/soil-2-551-2016, 2016. Mollenhauer, G., Grotheer, H., Torben, G., Bonk, E., and Hefter,
J.: Standard operation procedures and performance of the MI-
CADAS radiocarbon laboratory at Alfred Wegener Institute
(AWI), Germany, Nuclear Instruments and Methods in Physics
Research Section B: Beam Interactions with Materials and
Atoms, 496, 45–51, https://doi.org/10.1016/j.nimb.2021.03.016,
2021. Schirrmeister, L., Siegert, C., Kuznetsova, T., Kuzmina, S., An-
dreev, A., Kienast, F., Meyer, H., and Bobrov, A.: Paleoenvi-
ronmental and paleoclimatic records from permafrost deposits
in the Arctic region of Northern Siberia, Quat. Int., 89, 97–118,
https://doi.org/10.1016/S1040-6182(01)00083-0, 2002. Schirrmeister, L., Grosse, G., Wetterich, S., Overduin, P. P.,
Strauss, J., Schuur, E. A. G., and Hubberten, H.-W.: Fossil or-
ganic matter characteristics in permafrost deposits of the north-
east Siberian Arctic, J. Geophys. Res.-Biogeo., 116, G00M02,
https://doi.org/10.1029/2011jg001647, 2011a. Morgenstern, A., Grosse, G., Gunther, F., Fedorova, I., and
Schirrmeister, L.: Spatial analyses of thermokarst lakes and
basins in Yedoma landscapes of the Lena Delta, Cryosphere, 5,
849–867, https://doi.org/10.5194/tc-5-849-2011, 2011. Mueller,
C. W.,
Rethemeyer,
J.,
Kao-Kniffin,
J.,
Löpp-
mann, S., Hinkel, K. M., and G Bockheim, J.: Large
amounts of labile organic carbon in permafrost soils of
northern
Alaska,
Global
Change
Biol.,
21,
2804–2817,
https://doi.org/10.1111/gcb.12876, 2015. Schirrmeister, L., Kunitsky, V., Grosse, G., Wetterich, S., Meyer,
H., Schwamborn, G., Babiy, O., Derevyagin, A., and Siegert,
C.: Sedimentary characteristics and origin of the Late Pleis-
tocene Ice Complex on north-east Siberian Arctic coastal
lowlands and islands – A review, Quat. Int., 241, 3–25,
https://doi.org/10.1016/j.quaint.2010.04.004, 2011b. Nitzbon, J., Westermann, S., Langer, M., Martin, L. C. C. Haugk et al.: Eroding permafrost cliff carbon characteristics J., Strauss, J., Palmtag, J., Dowdy, K., Ogneva, O., Fuchs,
M., Bedington, M., Torres, R., Polimene, L., Overduin, P., Mol-
lenhauer, G., Grosse, G., Rachold, V., Sobczak, W. V., Spencer,
R. G. M., and Juhls, B.: Degrading permafrost river catchments
and their impact on Arctic Ocean nearshore processes, Ambio,
51, 439-455, https://doi.org/10.1007/s13280-021-01666-z, 2022. Jongejans, L. L., Strauss, J., Lenz, J., Peterse, F., Mangels-
dorf, K., Fuchs, M., and Grosse, G.: Organic matter char-
acteristics in yedoma and thermokarst deposits on Bald-
win Peninsula, west Alaska, Biogeosciences, 15, 6033–6048,
https://doi.org/10.5194/bg-15-6033-2018, 2018. Marzi, R., Torkelson, B. E., and Olson, R. K.: A Revised
Carbon Preference Index, Org. Geochem., 20, 1303–1306,
https://doi.org/10.1016/0146-6380(93)90016-5, 1993. p
g
g
Jongejans, L. L., Mangelsdorf, K., Schirrmeister, L., Grigoriev,
M. N., Maksimov, G. M., Biskaborn, B. K., Grosse, G.,
and Strauss, J.: n-Alkane Characteristics of Thawed Per-
mafrost Deposits Below a Thermokarst Lake on Bykovsky Mishra, U., Hugelius, G., Shelef, E., Yang, Y., Strauss, J., Lupachev,
A., Harden, J. W., Jastrow, J. D., Ping, C. L., Riley, W. J., Schuur,
E. A. G., Matamala, R., Siewert, M., Nave, L. E., Koven, C. https://doi.org/10.5194/bg-19-2079-2022 Biogeosciences, 19, 2079–2094, 2022 C. Haugk et al.: Eroding permafrost cliff carbon characteristics C. Haugk et al.: Eroding permafrost cliff carbon characteristics Strauss, J., Schirrmeister, L., Grosse, G., Wetterich, S., Ul-
rich,
M.,
Herzschuh,
U.,
and
Hubberten,
H.-W.:
The
deep permafrost carbon pool of the Yedoma region in
Siberia and Alaska, Geophys. Res. Lett., 40, 6165–6170,
https://doi.org/10.1002/2013GL058088, 2013. Vonk, J. E., Sánchez-García, L., Dongen, B. E. van, Alling, V.,
Kosmach, D., Charkin, A., Semiletov, I. P., Dudarev, O. V.,
Shakhova, N., Roos, P., Eglinton, T. I., Andersson, A., and
Gustafsson, Ö.: Activation of old carbon by erosion of coastal
and subsea permafrost in Arctic Siberia, Nature, 489, 137–140,
https://doi.org/10.1038/nature11392, 2012. Strauss, J., Schirrmeister, L., Mangelsdorf, K., Eichhorn, L., Wet-
terich, S., and Herzschuh, U.: Organic-matter quality of deep
permafrost carbon – a study from Arctic Siberia, Biogeosciences,
12, 2227–2245, https://doi.org/10.5194/bg-12-2227-2015, 2015. Vonk, J. E., Mann, P. J., Davydov, S., Davydova, A., Spencer, R. G. M., Schade, J., Sobczak, W. V., Zimov, N., Zimov, S., Bulygina,
E., Eglinton, T. I., and Holmes, R. M.: High biolability of ancient
permafrost carbon upon thaw, Geophys. Res. Lett., 40, 2689–
2693, https://doi.org/10.1002/grl.50348, 2013a. Strauss, J., Schirrmeister, L., Grosse, G., Fortier, D., Hugelius, G.,
Knoblauch, C., Romanovsky, V., Schädel, C., Schneider von
Deimling, T., Schuur, E. A. G., Shmelev, D., Ulrich, M., and
Veremeeva, A.: Deep Yedoma permafrost: A synthesis of depo-
sitional characteristics and carbon vulnerability, Earth-Sci. Rev.,
172,
75–86,
https://doi.org/10.1016/j.earscirev.2017.07.007,
2017. Vonk, J. E., Mann, P. J., Dowdy, K. L., Davydova, A., Davydov, S. P., Zimov, N., Spencer, R. G. M., Bulygina, E. B., Eglinton, T. I.,
and Holmes, R. M.: Dissolved organic carbon loss from Yedoma
permafrost amplified by ice wedge thaw, Environ. Res. Lett., 8,
035023, https://doi.org/10.1088/1748-9326/8/3/035023, 2013b. Strauss, J., Laboor, S., Schirrmeister, L., Grosse, G., Fortier, D.,
Hugelius, G., Knoblauch, C., Romanovsky, V. E., Schädel,
C., Schneider von Deimling, T., Schuur, E. A. G., Shmelev,
D., Ulrich, M., and Veremeeva, A.: Geochemical, lithological,
and geochronological characteristics of sediment samples from
Yedoma and thermokarst deposits in Siberia and Alaska 1998–
2016, PANGAEA, https://doi.org/10.1594/PANGAEA.919062,
2020. Vonk, J. E., Tank, S. E., Bowden, W. B., Laurion, I., Vincent, W. F., Alekseychik, P., Amyot, M., Billet, M. F., Canário, J., Cory,
R. M., Deshpande, B. N., Helbig, M., Jammet, M., Karlsson,
J., Larouche, J., MacMillan, G., Rautio, M., Walter Anthony, K. M., and Wickland, K. P.: Reviews and syntheses: Effects of per-
mafrost thaw on Arctic aquatic ecosystems, Biogeosciences, 12,
7129–7167, https://doi.org/10.5194/bg-12-7129-2015, 2015. Strauss, J., Abbott, B., Hugelius, G., Schuur, E. A. C. Haugk et al.: Eroding permafrost cliff carbon characteristics 2093 ter Anthony, K., Zimov, N., and Grosse, G.: Circum-Arctic Map
of the Yedoma Permafrost Domain, Front. Earth Sci., 758360,
https://doi.org/10.3389/feart.2021.758360, 2021b. Shakil, S., Tank, S. E., Kokelj, S. V., Vonk, J. E., and Zolkos,
S.: Particulate dominance of organic carbon mobilization from
thaw slumps on the Peel Plateau, NT: Quantification and im-
plications for stream systems and permafrost carbon release,
Environ. Res. Lett., 15, 114019, https://doi.org/10.1088/1748-
9326/abac36, 2020. Struck, J., Bliedtner, M., Strobel, P., Schumacher, J., Bazarradnaa,
E., and Zech, R.: Leaf wax n-alkane patterns and compound-
specific δ13C of plants and topsoils from semi-arid and arid Mon-
golia, Biogeosciences, 17, 567–580, https://doi.org/10.5194/bg-
17-567-2020, 2020. Stapel, J. G., Schirrmeister, L., Overduin, P. P., Wetterich,
S., Strauss, J., Horsfield, B., and Mangelsdorf, K.: Micro-
bial lipid signatures and substrate potential of organic mat-
ter in permafrost deposits – implications for future greenhouse
gas production, J. Geophys. Res.-Biogeo., 121, 2652–2666,
https://doi.org/10.1002/2016JG003483, 2016. Tanski, G., Bröder, L., Wagner, D., Knoblauch, C., Lantuit, H.,
Beer, C., Sachs, T., Fritz, M., Tesi, T., Koch, B. P., Haghipour,
N., Eglinton, T. I., Strauss, J., and Vonk, J. E.: Permafrost
Carbon and CO2 Pathways Differ at Contrasting Coastal Ero-
sion Sites in the Canadian Arctic, Front. Earth Sci., 9, 630493,
https://doi.org/10.3389/feart.2021.630493, 2021. Stettner, S., Beamish, A. L., Bartsch, A., Heim, B., Grosse, G.,
Roth, A., and Lantuit, H.: Monitoring Inter- and Intra-Seasonal
Dynamics of Rapidly Degrading Ice-Rich Permafrost Riverbanks
in the Lena Delta with TerraSAR-X Time Series, Remote Sens.,
10, 51, https://doi.org/10.3390/rs10010051, 2018. Turetsky, M. R., Abbott, B. W., Jones, M. C., Anthony, K. W., Ole-
feldt, D., Schuur, E. A. G., Grosse, G., Kuhry, P., Hugelius, G.,
Koven, C., Lawrence, D. M., Gibson, C., Sannel, A. B. K., and
McGuire, A. D.: Carbon release through abrupt permafrost thaw,
Nat. Geosci., 13, 138–143, https://doi.org/10.1038/s41561-019-
0526-0, 2020. Strauss, J., Schirrmeister, L., Wetterich, S., Borchers, A., and
Davydov, S. P.: Grain-size properties and organic-carbon stock
of Yedoma Ice Complex permafrost from the Kolyma low-
land, northeastern Siberia, Glob. Biogeochem. Cy., 26, GB3003,
https://doi.org/10.1029/2011GB004104, 2012. van Dongen, B. E., Semiletov, I., Weijers, J. W. H., and Gustafs-
son, O. R.: Contrasting lipid biomarker composition of terres-
trial organic matter exported from across the Eurasian Arctic by
the five great Russian Arctic rivers, Global Biogeochem. Cy., 22,
GB1011, https://doi.org/10.1029/2007gb002974, 2008. C. Haugk et al.: Eroding permafrost cliff carbon characteristics M., and Lee, H.: Vulnerability of permafrost carbon to climate
change: Implications for the global carbon cycle, BioScience, 58,
701–714, https://doi.org/10.1641/B580807, 2008. Rilfors, L., Wieslander, A., and Ståhl, S.: Lipid and protein com-
position of membranes of Bacillus megaterium variants in the
temperature range 5 to 70◦C, J. Bacteriol., 135, 1043–1052,
https://doi.org/10.1128/jb.135.3.1043-1052.1978, 1978. Schwamborn, G., Rachold, V., and Grigoriev, M. N.: Late Quater-
nary sedimentation history of the Lena Delta, Quat. Int., 89, 119–
134, https://doi.org/10.1016/S1040-6182(01)00084-2, 2002. Schwamborn, G., Rachold, V., and Grigoriev, M. N.: Late Quater-
nary sedimentation history of the Lena Delta, Quat. Int., 89, 119–
134, https://doi.org/10.1016/S1040-6182(01)00084-2, 2002. Semiletov, I., Pipko, I., Gustafsson, Ö., Anderson, L. G., Sergienko,
V., Pugach, S., Dudarev, O., Charkin, A., Gukov, A., Bröder,
L., Andersson, A., Spivak, E., and Shakhova, N.: Acidifica-
tion of East Siberian Arctic Shelf waters through addition of
freshwater and terrestrial carbon, Nat. Geosci., 9, 361–365,
https://doi.org/10.1038/ngeo2695, 2016. Sánchez-García, L., Vonk, J. E., Charkin, A. N., Kosmach, D.,
Dudarev, O. V., Semiletov, I. P., and Gustafsson, Ö.: Char-
acterisation of three regimes of collapsing Arctic Ice Com-
plex deposits on the SE Laptev Sea coast using biomarkers
and dual carbon isotopes, Permafrost Periglac., 25, 172–183,
https://doi.org/10.1002/ppp.1815, 2014. Se
etov, .,
p o, ., Gusta sso , Ö.,
de so ,
. G., Se g e
o,
V., Pugach, S., Dudarev, O., Charkin, A., Gukov, A., Bröder,
L., Andersson, A., Spivak, E., and Shakhova, N.: Acidifica-
tion of East Siberian Arctic Shelf waters through addition of
freshwater and terrestrial carbon, Nat. Geosci., 9, 361–365,
https://doi.org/10.1038/ngeo2695, 2016. Sanders, T., Fiencke, C., Fuchs, M., Haugk, C., Juhls, B., Mol-
lenhauer, G., Ogneva, O., Overduin, P., Palmtag, J., Povazhniy,
V., Strauss, J., Tuerena, R., Zell, N., and Dähnke, K.: Seasonal https://doi.org/10.5194/bg-19-2079-2022 https://doi.org/10.5194/bg-19-2079-2022 Biogeosciences, 19, 2079–2094, 2022 C. Haugk et al.: Eroding permafrost cliff carbon characteristics 2094 (NE Siberia, Russia), Palaeogeogr. Palaeocl., 279, 73–95,
https://doi.org/10.1016/j.palaeo.2009.05.002, 2009. (NE Siberia, Russia), Palaeogeogr. Palaeocl., 279, 73–95,
https://doi.org/10.1016/j.palaeo.2009.05.002, 2009. Winterfeld, M., Goñi, M. A., Just, J., Hefter, J., and Mollenhauer,
G.: Characterization of particulate organic matter in the Lena
River delta and adjacent nearshore zone, NE Siberia – Part 2:
Lignin-derived phenol compositions, Biogeosciences, 12, 2261–
2283, https://doi.org/10.5194/bg-12-2261-2015, 2015. Wetterich, S., Tumskoy, V., Rudaya, N., Andreev, A. A.,
Opel,
T.,
Meyer,
H.,
Schirrmeister,
L.,
and
Hüls,
M.:
Ice
Complex
formation
in
Arctic
East
Siberia
dur-
ing
the
MIS3
Interstadial,
Quat. Sci. Rev.,
84,
39–55,
https://doi.org/10.1016/j.quascirev.2013.11.009, 2014. Yang, D., Liu, B., and Ye, B.: Stream temperature changes over
Lena River Basin in Siberia, Geophys. Res. Lett., 32, L05401,
https://doi.org/10.1029/2004GL021568, 2005. Wetterich, S., Kizyakov, A., Fritz, M., Aksenov, A., Schirrmeis-
ter, L., and Opel, T.: Expedition Report: Permafrost research
on Sobo-Sise Island (Lena Delta), Reports on Polar and Marine
Research, 102–113, https://doi.org/10.2312/BzPM_0734_2019,
2018. Zech, M., Buggle, B., Leiber-Sauheitl, K., Markovic, S., Glaser, B.,
Hambach, U., Huwe, B., Stevens, T., Sümegi, P., Wiesenberg,
G., and Zöller, L.: Reconstructing Quaternary vegetation history
in the Carpathian Basin, SE Europe, using n-alkane biomarkers
as molecular fossils, Eiszeitalter und Gegenwart, Quat. Sci., 58,
148–155, https://doi.org/10.5167/uzh-76606, 2009. Wetterich, S., Kizyakov, A., Fritz, M., Wolter, J., Mollenhauer, G.,
Meyer, H., Fuchs, M., Aksenov, A., Matthes, H., Schirrmeis-
ter, L., and Opel, T.: The cryostratigraphy of the Yedoma cliff
of Sobo-Sise Island (Lena delta) reveals permafrost dynamics in
the central Laptev Sea coastal region during the last 52 kyr, The
Cryosphere, 14, 4525–4551, https://doi.org/10.5194/tc-14-4525-
2020, 2020a. Zech, M., Andreev, A., Zech, R., Muller, S., Hambach, U., Frechen,
M., and Zech, W.: Quaternary vegetation changes derived from
a loess-like permafrost palaeosol sequence in northeast Siberia
using alkane biomarker and pollen analyses, Boreas, 39, 540–
550, https://doi.org/10.1111/j.1502-3885.2009.00132.x, 2010. Wetterich, S., Meyer, H., Fritz, M., Opel, T., and Schirrmeis-
ter,
L.:
Cryolithology
of
the
Sobo-Sise
Yedoma
cliff
(eastern
Lena
Delta),
PANGAEA
[data
set],
https://doi.org/10.1594/PANGAEA.919470, 2020b. Zhang, X., Bianchi, T., Cui, X., Rosenheim, B., Ping, C.-L.,
Hanna, A., Kanevskiy, M., Schreiner, K., and Allison, M.:
Permafrost Organic Carbon Mobilization From the Watershed
to the Colville River Delta: Evidence From 14C Ramped
Pyrolysis and Lignin Biomarkers: Permafrost OC Transport
in Coville River, Geophys. Res. Lett., 44, 11491–11500,
https://doi.org/10.1002/2017GL075543, 2017. Wetterich, S., Rudaya, N., Nazarova, L., Syrykh, L., Pavlova, M.,
Palagushkina, O., Kizyakov, A., Wolter, J., Kuznetsova, T., Ak-
senov, A., Stoof-Leichsenring, K. C. Haugk et al.: Eroding permafrost cliff carbon characteristics G., Treat, C.,
Fuchs, M., Schädel, C., Ulrich, M., Turetsky, M. R., Keuschnig,
M., Biasi, C., Yang, Y., and Grosse, G.: Permafrost, in: Recar-
bonizing global soils – A technical manual of recommended
management practices, edited by: Food and Agriculture Organi-
zation of the United Nations (FAO), and Intergovernmental Tech-
nical Panel on Soils (ITPS), Food and Agriculture Organization
of the United Nations, Rome, Italy, 127–147, 2021a. Wetterich,
S.,
Kuzmina,
S.,
Andreev,
A. A.,
Kienast,
F.,
Meyer, H., Schirrmeister, L., Kuznetsova, T., and Sierralta,
M.: Palaeoenvironmental dynamics inferred from late Quater-
nary permafrost deposits on Kurungnakh Island, Lena Delta,
Northeast Siberia, Russia, Quat. Sci. Rev., 27, 1523–1540,
https://doi.org/10.1016/j.quascirev.2008.04.007, 2008. Wetterich,
S.,
Schirrmeister,
L.,
Andreev,
A. A.,
Pudenz,
M., Plessen, B., Meyer, H., and Kunitsky, V. V.: Eemian
and
Late
Glacial/Holocene
palaeoenvironmental
records
from
permafrost
sequences
at
the
Dmitry
Laptev
Strait Strauss, J., Laboor, S., Schirrmeister, L., Fedorov, A. N., Fortier,
D., Froese, D., Fuchs, M., Günther, F., Grigoriev, M., Harden, J.,
Hugelius, G., Kanevskiy, M., Kholodov, A., Kunitsky, V., Kraev,
G., Lozhkin, A., Rivkina, E., Shur, Y., Siegert, C., Spektor, V.,
Streletskaya, I., Ulrich, M., Vartanyan, S., Veremeeva, A., Wal- Biogeosciences, 19, 2079–2094, 2022 https://doi.org/10.5194/bg-19-2079-2022 https://doi.org/10.5194/bg-19-2079-2022 C. Haugk et al.: Eroding permafrost cliff carbon characteristics R., Schirrmeister, L., and Fritz,
M.: Paleo-Ecology of the Yedoma Ice Complex on Sobo-Sise Is-
land (EasternLena Delta, Siberian Arctic), Front. Earth Sci., 9,
681511, https://doi.org/10.3389/feart.2021.681511, 2021. Zhang, X., Bianchi, T. S., Hanna, A. J. M., Shields, M. R., Izon, G.,
Hutchings, J. A., Ping, C.-L., Kanevskiy, M., Haghipour, N., and
Eglinton, T. I.: Recent Warming Fuels Increased Organic Carbon
Export From Arctic Permafrost, AGU Adv., 2, e2021AV000396,
https://doi.org/10.1029/2021AV000396, 2021. Wild, B., Andersson, A., Bröder, L., Vonk, J., Hugelius, G., Mc-
Clelland, J. W., Song, W., Raymond, P. A., and Gustafsson, Ö.:
Rivers across the Siberian Arctic unearth the patterns of carbon
release from thawing permafrost, P. Natl. Acad. Sci. USA, 116,
10280–10285, https://doi.org/10.1073/pnas.1811797116, 2019. Biogeosciences, 19, 2079–2094, 2022 https://doi.org/10.5194/bg-19-2079-2022 Biogeosciences, 19, 2079–2094, 2022
|
https://openalex.org/W4235478382
|
https://europepmc.org/articles/pmc6275453?pdf=render
|
English
| null |
The potential for improving cardio-renal outcomes by sodium-glucose co-transporter-2 inhibition in people with chronic kidney disease: a rationale for the EMPA-KIDNEY study
|
Clinical kidney journal
| 2,019
|
cc-by
| 11,981
|
CKJ REVIEW The potential for improving cardio-renal outcomes by
sodium-glucose co-transporter-2 inhibition in people
with chronic kidney disease: a rationale for the
EMPA-KIDNEY study William G. Herrington1,2, David Preiss1,2, Richard Haynes1,2,
Maximilian von Eynatten3, Natalie Staplin1,2,4, Sibylle J. Hauske3,
Jyothis T. George3, Jennifer B. Green5, Martin J. Landray1,2,4,*,
Colin Baigent1,2,* and Christoph Wanner6,* William G. Herrington1,2, David Preiss1,2, Richard Haynes1,2,
Maximilian von Eynatten3, Natalie Staplin1,2,4, Sibylle J. Hauske3,
Jyothis T. George3, Jennifer B. Green5, Martin J. Landray1,2,4,*,
Colin Baigent1,2,* and Christoph Wanner6,* 1Medical Research Council Population Health Research Unit at the University of Oxford, Nuffield Department
of Population Health, University of Oxford, Oxford, UK, 2Clinical Trial Service Unit and Epidemiological
Studies Unit, Nuffield Department of Population Health, University of Oxford, Oxford, UK, 3Boehringer
Ingelheim International, Ingelheim, Germany, 4Li Ka Shing Centre for Health Information and Discovery, Big
Data Institute, University of Oxford, Oxford, UK, 5Duke Clinical Research Institute, Duke University Medical
Center, Durham, NC, USA and 6Wu¨ rzburg University Clinic, Wu¨ rzburg, Germany *These authors contributed equally to this work. Clinical Kidney Journal, 2018, 749–761
doi: 10.1093/ckj/sfy090
Advance Access Publication Date: 25 October 2018
CKJ Review Clinical Kidney Journal, 2018, 749–761 doi: 10.1093/ckj/sfy090
Advance Access Publication Date: 25 October 2018
CKJ Review Received: 22.6.2018. Editorial decision: 24.8.2018
V
C The Author(s) 2018. Published by Oxford University Press on behalf of ERA-EDTA.
This is an Open Access article distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/4.0/),
which permits unrestricted reuse, distribution, and reproduction in any medium, provided the original work is properly cited. V
C The Author(s) 2018. Published by Oxford University Press on behalf of ERA-EDTA.
This is an Open Access article distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/4.0/),
which permits unrestricted reuse, distribution, and reproduction in any medium, provided the original work is properly cited. Correspondence and offprint requests to: William G. Herrington; E-mail: will.herrington@ndph.ox.ac.uk and Christoph Wanne
E-mail: Wanner_C@ukw.de *These authors contributed equally to this work. Received: 22.6.2018. Editorial decision: 24.8.2018 INTRODUCTION In this review we introduce the mechanisms of SGLT-2 inhi-
bition on the kidney and summarize the key clinical evidence
providing a rationale for the testing the effects of SGLT-2 inhibi-
tion on kidney and CV outcomes in people with diabetic kidney
disease. We then introduce the hypothesis that SGLT-2 inhibi-
tion may also have beneficial cardio-renal effects in people with
CKD but without diabetes. Lastly, using randomized data from
the Empagliflozin, Cardiovascular Outcomes, and Mortality in
Type 2 Diabetes (EMPA-REG OUTCOME) trial, we describe what
is currently known about the safety of empagliflozin in T2DM,
both overall and in people with concomitant CKD. In high-income countries, the overall prevalence of chronic kidney
disease (CKD) is 10% [1, 2] and this proportion is expected to rise
as populations age further and diabetes mellitus becomes more
common [3]. Worldwide, diabetic kidney disease accounts for a
large proportion of advanced CKD (i.e. Stages 4 and 5), but the pro-
portion of those without diabetes still ranges from 50 to 70% [4,
5]. CKD can often be a progressive condition, with proteinuria rep-
resenting a significant risk factor for more rapid decline in kidney
function [6]. The avoidance of progressive CKD is important, as
end-stage kidney disease (ESKD) has adverse effects on morbidity
and quality of life, dialysis or transplantation incur substantial so-
cietal costs [7, 8] and low levels of kidney function increase cardio-
vascular (CV) risk and premature mortality [9]. MECHANISMS OF ACTION AND
PHYSIOLOGICAL EFFECTS OF SGLT-2
INHIBITORS ON THE KIDNEY The current standard of care in many forms of CKD is inhibi-
tion of the renin–angiotensin system (RAS) with an angiotensin-
converting enzyme inhibitor (ACEi) or angiotensin II receptor
blocker (ARB). RAS inhibition has been shown to moderately
reduce albuminuria and to slow the rate at which proteinuric kid-
ney diseases progress [10–12]. Compared with RAS inhibition with
a single drug, combination therapy (e.g. ACEi plus ARB) has not
been shown to further delay kidney disease progression but does
increase the risk of serious hyperkalaemia or acute kidney injury
[13]. In people with diabetes, intensification of glycaemic control
has been demonstrated to have moderately beneficial effects on
markers of kidney disease progression compared with ‘standard’
regimens [14–16], and in the long-term this may translate into re-
duced risk of ESKD [17]. Trials of intensification of blood pressure
(BP) lowering also suggest small benefits on CKD progression may
perhaps exist among those with proteinuria [18–21]. Despite these
interventions, however, substantial residual risk of ESKD remains. SGLT-2 is a kidney-specific solute transporter responsible for
the vast majority (80–90%) of kidney tubular glucose reabsorp-
tion under normal physiological conditions [34]. Mutations that
affect
its
function
cause
familial
renal
glycosuria,
which
appears to be a relatively benign condition [35]. In hyperglycae-
mic states, SGLT-2 expression can increase, causing more proxi-
mal tubular glucose and sodium reabsorption [36, 37]. The
reduced sodium delivery to the distal convoluted tubules may
result in a potentially maladaptive hyperfiltration state [30]. SGLT-2 inhibitors cause about half of filtered glucose to be
excreted. This equates to 50–80 g/day under normoglycaemic
and modest hyperglycaemic conditions and perhaps >100 g/day
in people with diabetes and hyperfiltration [38]. Sodium-glucose
co-transporter-1 (SGLT-1) is a distinct co-transporter that is pre-
sent in the intestine, heart, skeletal muscle and kidney. It has a
higher affinity but lower transporting capacity compared with
SGLT-2. Nevertheless, increased SGLT-1 expression or activity
in the distal segment of the proximal tubule in response to
SGLT-2 inhibition may partly explain why a large proportion
(perhaps 50%) of filtered glucose is still reabsorbed in those
treated with an SGLT-2 inhibitor [39]. Lowering low-density lipoprotein cholesterol has been shown
to reduce the incidence of atherosclerotic events in people with
CKD [22], and statin-based regimens are widely recommended for
those at risk [23, 24]. A key feature of CKD, however, is the high
prevalence of non-atherosclerotic heart disease. of CV disease and hyperfiltration (i.e. those with substantially reduced nephron mass and/or albuminuria), irrespective
of whether they have diabetes. of CV disease and hyperfiltration (i.e. those with substantially reduced nephron mass and/or albuminuria), irrespective
of whether they have diabetes. Keywords: cardiovascular, CKD, clinical trial, diabetic kidney disease, SGLT-2 inhibitor ABSTRACT Diabetes is a common cause of chronic kidney disease (CKD), but in aggregate, non-diabetic diseases account for a
higher proportion of cases of CKD than diabetes in many parts of the world. Inhibition of the renin–angiotensin system
reduces the risk of kidney disease progression and treatments that lower blood pressure (BP) or low-density lipoprotein
cholesterol reduce cardiovascular (CV) risk in this population. Nevertheless, despite such interventions, considerable
risks for kidney and CV complications remain. Recently, large placebo-controlled outcome trials have shown that
sodium-glucose co-transporter-2 (SGLT-2) inhibitors reduce the risk of CV disease (including CV death and
hospitalization for heart failure) in people with type 2 diabetes who are at high risk of atherosclerotic disease, and these
effects were largely independent of improvements in hyperglycaemia, BP and body weight. In the kidney, increased
sodium delivery to the macula densa mediated by SGLT-2 inhibition has the potential to reduce intraglomerular
pressure, which may explain why SGLT-2 inhibitors reduce albuminuria and appear to slow kidney function decline in
people with diabetes. Importantly, in the trials completed to date, these benefits appeared to be maintained at lower
levels of kidney function, despite attenuation of glycosuric effects, and did not appear to be dependent on ambient
hyperglycaemia. There is therefore a rationale for studying the cardio-renal effects of SGLT-2 inhibition in people at risk 749 750
|
W.G. Herrington et al. MECHANISMS OF ACTION AND
PHYSIOLOGICAL EFFECTS OF SGLT-2
INHIBITORS ON THE KIDNEY SGLT-2 inhibition in CKD
|
751 Common suggested causes:
- Diabetes
- Obesity
- CKD with reduced nephron mass
- HF
RAS and SGLT-2 inhibition may reduce intraglomerular pressure through their
respective effects on efferent and afferent arteriolar tone. Combining these two
therapeutic approaches may further reduce intraglomerular pressure. RAS inhibition
SGLT-2 inhibition
Combined RAS & SGLT-2 inhibition
Efferent
arteriole
Afferent
arteriole Common suggested causes:
- Diabetes
- Obesity
- CKD with reduced nephron m
- HF
Efferent
arteriole
Afferent
arteriole RAS and SGLT-2 inhibition may reduce intraglomerular pressure through their
respective effects on efferent and afferent arteriolar tone. Combining these two
therapeutic approaches may further reduce intraglomerular pressure. FIGURE 1: Mechanistic concept of the effects of RAS and SGLT-2 inhibition on intraglomerular pressure. FIGURE 1: Mechanistic concept of the effects of RAS and SGLT-2 inhibition on intraglomerular pressure. FIGURE 1: Mechanistic concept of the effects of RAS and SGLT-2 inhibition on intraglomerular pressure. might be to induce an increase in afferent arteriolar tone. Tubuloglomerular feedback and afferent arteriolar tone are atten-
uated in diabetes by up-regulation of tubular SGLTs and other
sodium-exchange transport mechanisms. This leads to increased
tubular sodium reabsorption and consequently to reduced deliv-
ery of sodium to the macula densa, where afferent arteriolar tone
is relaxed in response to decreased adenosine production [41, 43]. SGLT-2 inhibition restores delivery of sodium to the macula
densa, promoting adenosine production and increased afferent
arteriolar tone (Figure 1). Some restoration of blunted tubuloglo-
merular feedback is apparent after a single dose, probably ac-
counting for the acute reduction in GFR and the subsequent
reduction in albuminuria [44] observed in people treated with
SGLT-2 inhibitors, as well as the swift return towards pre-
treatment levels of GFR shortly after drug discontinuation [33,
44]. Importantly, these effects do not appear to be modified by
concomitant use of a RAS blocker [31, 33] or loop diuretic [45]. The
latter observation may be of interest as loop diuretics block Na-K-
Cl co-transporter (NKCC2) channels, which are the main mecha-
nism by which chloride ions enter the macula densa to initiate
tubuloglomerular feedback [46]. It is therefore hypothesized that
additional mechanisms beyond NKCC2 sensing may also be me-
diating the renal haemodynamic effects of SGLT-2 inhibition. undergo
structural
hypertrophy
and
single-nephron
GFR
increases [43, 50]. MECHANISMS OF ACTION AND
PHYSIOLOGICAL EFFECTS OF SGLT-2
INHIBITORS ON THE KIDNEY Intraglomerular hypertension has long been considered to
be a final common pathway for kidney disease progression
shared by many forms of CKD [51]. Support for this hypothesis
includes the observations that for a given level of urinary albu-
min excretion, the risk of ESKD is relatively independent of the
primary cause of kidney disease [52] and RAS blockade appears
effective at slowing both the progression of non-diabetic [10, 53]
and diabetic [11, 12] proteinuric kidney diseases. SGLT-2 inhibi-
tion may restore tubuloglomerular feedback and reduce intra-
glomerular pressure through non-glycaemic mechanisms in
people with diabetes (including those already on RAS blockade)
[31, 33, 44] and has acute effects on GFR in people without dia-
betes [54–56]. It is therefore reasonable to hypothesize that in
people who are at high risk of hyperfiltration (i.e. those with al-
buminuria and/or low GFR), SGLT-2 inhibition may lower intra-
glomerular pressure and be nephroprotective even in the
presence of ambient normoglycaemia. MECHANISMS OF ACTION AND
PHYSIOLOGICAL EFFECTS OF SGLT-2
INHIBITORS ON THE KIDNEY About half of
patients with advanced CKD have abnormal cardiac structure [25,
26], increasing to more than three-quarters by the time dialysis is
initiated [26, 27]. Although ejection fraction is often preserved,
some degree of left ventricular diastolic dysfunction is common
in CKD [25]. Lowering BP may reduce CV disease risk in CKD [28],
but there is a general lack of reliable information about other
treatments that may be effective for the prevention of heart dis-
ease in CKD [29]. Therefore, in addition to testing new interven-
tions that could reduce the risk of CKD progression, there is a
need for more trials of treatments that could further reduce the
types of CV disease commonly experienced by people with CKD. As each reabsorbed molecule of glucose by SGLT-2 is accom-
panied by a sodium ion [30], inhibition of SGLT-2 causes natri-
uresis in addition to the osmotic diuretic effect of glycosuria. This non-glycaemic effect may modify CV risk through reduc-
tions in plasma volume, organ congestion and central and sys-
temic BP [40] and may be particularly beneficial at preventing
heart failure (HF) [31, 32, 41, 42]. Natriuresis and altered tubular handling of filtered sodium
may also have important modulatory effects on glomerular fil-
tration rate (GFR) through reducing intraglomerular pressure
[41, 43]. Hyperfiltration is driven in part by neurohormonal
stimuli that cause either a net reduction in afferent glomerular
arteriolar resistance or a net increase in efferent arteriolar resis-
tance [43]. Angiotensin II is one key mediator of efferent arterio-
lar resistance and ACEi/ARBs effectively reduce intraglomerular
pressure and slow kidney disease progression through a reduc-
tion of efferent vascular resistance [11, 12]. An alternative po-
tential therapeutic strategy to reduce intraglomerular pressure Sodium-glucose co-transporter-2 (SGLT-2) inhibitors were
originally developed to treat hyperglycaemia in people with dia-
betes [30]. Recent large placebo-controlled outcome trials have
shown that empagliflozin and canagliflozin reduce the risk of
CV disease in people with type 2 diabetes mellitus (T2DM) at
high risk of CV disease. Exploratory analyses also suggested
they may reduce kidney disease progression in this population
[31–33]. These CV and kidney effects appeared to be largely in-
dependent of effects on glycaemic control, BP and body weight. RANDOMIZED TRIAL EVIDENCE FOR EFFECTS
OF SGLT-2 INHIBITION ON INTERMEDIATE
FACTORS Attenuation of tubuloglomerular feedback and intraglomer-
ular ‘hypertension’ are not unique to diabetes. Activation of the
sympathetic nervous system has been shown to increase proxi-
mal tubular sodium reabsorption in HF [47] and is recognized in
people with pre-diabetes, pre-hypertension [48] and obesity
[49]. Moreover, among those with reduced nephron mass, which
includes many people with chronically reduced GFR, intrarenal
vasodilatation
may
explain
why
the
remaining
nephrons Effects of SGLT-2 inhibition on glycaemic control and
urinary glucose excretion Effects of SGLT-2 inhibition on body fat and whole-body
fluid homeostasis Short- and medium-term mechanistic trials in T2DM show that
SGLT-2
inhibitors
produce
modest
reductions
in
office-
measured and 24-h ambulatory BP [66, 71], central systolic BP
and central pulse pressure [40]. A systolic BP difference of a sim-
ilar size (4 mmHg) to that measured in these smaller studies
was observed in those allocated empagliflozin versus placebo in
the EMPA-REG OUTCOME trial [32]. Glycosuria induced by SGLT-2 inhibition provides a direct
source of calorie loss, but there are also indirect metabolic
responses induced that are expected to enhance lipolysis. These include a decrease in insulin and an increase in glucagon
secretion, resulting perhaps from both the expected reduction
in plasma glucose and effects on both pancreatic alpha and beta
cells [65]. The metabolic consequence is a shift to more utiliza-
tion of fat for energy production. Modestly increased plasma
glycerol and fatty acid levels (reflecting accelerated lipolysis)
and higher levels of the ketone b-hydroxybutyrate (reflecting
higher liver fat oxidation) are apparent within 2 weeks of start-
ing empagliflozin, and these effects, although perhaps attenu-
ated, are also present in people without diabetes [54]. As is the case for body weight, the effects of SGLT-2 inhibi-
tion on BP are not diminished at lower eGFR levels, at least
within the range of eGFR studied to date. For example, in 2286
people with T2DM randomized to empagliflozin versus placebo,
BP reductions were at least as large even at an eGFR <30 mL/
min/1.73 m2. Similarly, CKD Stages 3A and 3B have not been
shown to modify the BP-lowering effect of dapagliflozin [61, 69]. The persistence of the BP-lowering effects of SGLT-2 inhibition
in CKD is yet to be fully understood, and possibilities include BP
being more responsive to salt mobilization and/or removal in
CKD, increased response to the possible beneficial effects of
SGLT-2 inhibition on arterial stiffness, sympathetic system
overactivity, oxidative stress and endothelial dysfunction [62]
and/or augmentation of the effect of other antihypertensive
medications [62]. Together, these metabolic effects lead to reductions in body
weight [66], including loss of visceral and subcutaneous adipose
tissue [67]. Although there are effects on fluid homeostasis, af-
ter 2 years, weight loss resulting from SGLT-2 inhibition in
T2DM appears attributable in large part to reduced adipose tis-
sue (measured using dual-energy X-ray absorptiometry) [67]. Effects of SGLT-2 inhibition on body fat and whole-body
fluid homeostasis In
the EMPA-REG OUTCOME trial, empagliflozin 10–25 mg/day led
to a sustained difference in weight of about 2 kg (from a mean
of 86 kg) and a 2 cm difference in waist circumference com-
pared with placebo (from a mean of 105 cm) in 7020 people with
T2DM and prior atherosclerotic CV disease [32]. There were al-
most identical-sized reductions observed with canagliflozin
100–300 mg versus placebo among the 10 142 participants in the
Canagliflozin Cardiovascular Assessment Study (CANVAS) and
CANVAS-Renal (CANVAS-R) [31]. However, fat loss is less than
predicted from estimated urinary calorie loss, perhaps because
participants allocated active drug were more likely to increase
calorie intake than those allocated placebo [68]. In people without diabetes, a randomized study of 376 obese
individuals found that 12 weeks of canagliflozin 100–300 mg/
day caused a systolic BP difference of between 1 and 2
mmHg compared with placebo [56]. Few randomized data are
currently available in people with CKD but without diabetes, but
the data from studies in other populations suggest that a mod-
est BP-lowering effect of SGLT-2 inhibition might be expected in
such people. Effects of SGLT-2 inhibition on glycaemic control and
urinary glucose excretion In people with T2DM, trials have established the glucose-
lowering effects of SGLT-2 inhibition, and there appears to be
little difference between empagliflozin, canagliflozin and dapa-
gliflozin
in
this
regard
[57]. In
short-
and
medium-term 752
|
W.G. Herrington et al. mechanistic trials, on average, adding SGLT-2 inhibition to vari-
ous background glucose-lowering therapies reduced glycosy-
lated
haemoglobin
(HbA1c)
by
0.7%
[58],
with
absolute
reductions in HbA1c often larger among those with the highest
baseline HbA1c [59, 60]. mechanistic trials, on average, adding SGLT-2 inhibition to vari-
ous background glucose-lowering therapies reduced glycosy-
lated
haemoglobin
(HbA1c)
by
0.7%
[58],
with
absolute
reductions in HbA1c often larger among those with the highest
baseline HbA1c [59, 60]. increased urinary sodium and electrolyte free-water excretion
[70] (although non-renal effects centrally or in the gut cannot be
excluded). In a mathematical extrapolation of a short-term randomized
trial in healthy individuals, dapagliflozin was shown to cause
proportionally larger reductions in interstitial fluid volume
(480 mL) than blood volume (150 mL), which is in contrast to
the effects of the loop diuretic bumetanide, which caused
smaller reductions in interstitial fluid volume (510 mL) than
blood volume (780 mL) [70]. Relatively larger reductions in in-
terstitial fluid compared with plasma volume, perhaps the re-
sult of combined natriuretic and osmotic diuretic effects of
SGLT-2 inhibition, may have the potential to significantly im-
prove organ congestion in people with fluid overload with less
risk of causing arterial underperfusion or symptomatic dehy-
dration [70], but studies directly measuring changes in intersti-
tial and blood volume in people with CKD and people with HF
are needed to test both the mathematical extrapolation and
these hypotheses. In people with T2DM and CKD, previous pharmacodynamic
studies have consistently shown a linear relationship between
24-h
urinary
glucose
excretion
and
kidney
function
[41]. Consequently, compared with those with preserved GFR, HbA1c
reductions with SGLT-2 inhibition are smaller in people with
lower GFR. In pooled analyses of randomized trials comparing
dapagliflozin or empagliflozin versus placebo among people
with T2DM and estimated GFR (eGFR) levels <60 mL/min/1.73
m2, HbA1c differences were generally attenuated to 0.3 and
0.4%, respectively [61, 62]. In people without diabetes, studies have found that empagli-
flozin 25 mg/day and canagliflozin 300 mg/day each induce 50–
60 g/day of urinary glucose excretion [55, 63], which is approxi-
mately half the corresponding amount induced in people with
diabetes [64]. Effects of SGLT-2 inhibition on albuminuria and short-
term effects on GFR Nevertheless, despite the reduced glycosuric effects of SGLT-
2 inhibition in people with T2DM and CKD, lower eGFR has not
been shown to attenuate these weight-lowering effects, at least
within the range of eGFR studied to date. In placebo-controlled
trials testing dapagliflozin [61, 69] and empagliflozin [62] in
T2DM, differences in weight among those with preserved kid-
ney function were similar to differences observed in those with
CKD Stages 3A, 3B or 4. Part of the preserved effect of SGLT-2 in-
hibition on body weight in CKD may therefore result from In the EMPA-REG OUTCOME trial and CANVAS/CANVAS-R, both
empagliflozin 10–25 mg/day and canagliflozin 100–300 mg/day
were found to reduce albuminuria by between 25 and 50% in
those with T2DM and either micro- or macroalbuminuria, irre-
spective of use of RAS blockade. The albuminuria-lowering
effect was apparent early and maintained throughout the
3–4 years of follow-up [44, 72]. Furthermore, in the EMPA-REG
OUTCOME trial, compared with those allocated to placebo, SGLT-2 inhibition in CKD
|
753 66
68
70
72
74
76
78
Adjusted mean eGFR
(mL min 1.73m2)
0 4
12
28
52
66
80
94
108
122
136
150
164
178
192
Weeks
Empagliflozin
Placebo
Number of patients analysed
Empagliflozin
Placebo
4644
2323
4533
2267
4451
2205
4318
2121
4225
2064
4018
1927
4131
1981
3710
1763
3103
1479
2654
1262
2387
1123
2087
977
1623
731
1037
448
Pre−specified mixed model repeated measures analysis in patients treated with at least one dose of study drug who had a baseline and post−baseline measurement. FIGURE 2: Effect of allocation to empagliflozin versus placebo on Chronic Kidney Disease Epidemiology Collaboration eGFR. Empagliflozin FIGURE 2: Effect of allocation to empagliflozin versus placebo on Chronic Kidney Disease Epidemiology Collaboration eGFR. In obese but otherwise healthy adults with normal kidney
function, canagliflozin has been shown to cause an initial dip in
eGFR, which in this population was 1–2 mL/min/1.73 m2 [56]. Two non-randomized studies in people without diabetes and
with normal kidney function starting empagliflozin have also
noted initial dips in eGFR of 8 to 16 mL/min/1.73 m2, respec-
tively, although the lack of a control arm may mean these re-
flect substantial overestimates of effect [54, 55]. Nonetheless,
this evidence suggests that lowering of intraglomerular pres-
sure can be achieved without the need for ambient hyperglycae-
mia. Effects of SGLT-2 inhibition on albuminuria and short-
term effects on GFR There are no long-term published studies in non-diabetic
populations exploring whether this initial dip in kidney func-
tion with SGLT-2 inhibition is associated with beneficial effects
on subsequent eGFR slopes. albuminuria remained lower among those allocated to empagli-
flozin about 1 month after stopping study treatment, suggesting
that SGLT-2 inhibition prevented diabetes-related structural
changes in the kidney (although this hypothesis has not been
tested by kidney biopsy studies) [44]. The EMPA-REG OUTCOME trial included people with T2DM
with both preserved and modestly reduced kidney function. Commencement of SGLT-2 inhibition led to an initial acute decline
in eGFR of 3mL/min/1.73 m2. The magnitude of this effect was
consistent at daily empagliflozin doses of 10 and 25mg [33]. The
initial dip in eGFR, which is considered an indicator of reduced
intraglomerular pressure, was followed by a marked slower decline
in eGFR compared with those allocated placebo during longer-term
treatment (Figure 2) [33]. These findings have since been replicated
in the CANVAS/CANVAS-R data [73]. Taken together, these data on intermediate clinical parame-
ters show that although the effects on HbA1c are attenuated
among people with T2DM and lower kidney function, important
non-glycaemic effects on fluid balance, body weight, BP, albu-
minuria and markers of intraglomerular pressure are present in
individuals with reduced kidney function, at least down to an
eGFR of 30 mL/min/1.73 m2. Although the effects of SGLT-2 inhi-
bition on such parameters among those without diabetes
remains less well studied to date, there is reason to expect phar-
macological effects that may translate into clinical benefits. Canagliflozin 100–300 mg/day has been compared with the
sulfonylurea glimepiride in 1450 people with T2DM [72]. This
comparison was particularly informative as the reductions in
HbA1c were very similar among those allocated the two differ-
ent classes of glucose-lowering drug. However, after an initial
eGFR dip of 6 mL/min/1.73 m2, the average annual rate of de-
cline in eGFR was 0.5–1.0 mL/min/1.73 m2/year among those al-
located to canagliflozin compared with 3.3 mL/min/1.73 m2/year
among those allocated to glimepiride (in whom no acute change
in eGFR was observed). Such data further support the hypothe-
sis that non-glycaemic effects of SGLT-2 inhibition are central
to any nephroprotective effects. RANDOMIZED TRIAL EVIDENCE FOR EFFECTS
OF SGLT-2 INHIBITION ON CLINICAL EFFICACY
OUTCOMES In a subgroup analysis of the EMPA-REG OUTCOME trial, par-
ticipants with reduced eGFR had a similar-sized initial eGFR dip
followed by a relative slowing in the annual rate of decline in
eGFR [33]. Similarly, data from five medium-term (i.e. 6 month)
placebo-controlled trials of empagliflozin also confirm the ini-
tial dip in eGFR is at least as large in those with an eGFR be-
tween 30 and 60 mL/min/1.73 m2 compared with those with
preserved kidney function. To date, too few people with an
eGFR <30 mL/min/1.73 m2 have been studied to be certain, but
there is some evidence that the initial eGFR dip is also present
at this low level of kidney function [62]. Effects of SGLT-2 inhibition on kidney disease
progression Empagliflozin
(n=4687)
Placebo
(n=2333)
Hazard ratio (95% CI)
p value for
interaction
(A) Traditional kidney disease progression outcome (post−hoc)†
eGFR (mL min 1.73m2)
≥90
7/1043 (0.7%)
15/486 (3.1%)
0.21 (0.09−0.53)
0.18
≥60 to <90
34/2406 (1.4%)
27/1232 (2.2%)
0.61 (0.37−1.02)
≥45 to <60
23/822 (2.8%)
16/416 (3.8%)
0.68 (0.36−1.28)
<45
17/374 (4.5%)
13/189 (6.9%)
0.63 (0.30−1.29)
uACR (mg/g)
<30
22/2766 (0.8%)
26/1376 (1.9%)
0.41 (0.23−0.72)
0.51
≥30 to ≤300
23/1325 (1.7%)
17/671 (2.5%)
0.67 (0.36−1.26)
>300
33/504 (6.5%)
28/260 (10.8%)
0.51 (0.31−0.85)
All participants
81/4645 (1.7%)
71/2323 (3.1%)
0.54 (0.40−0.75) Hazard ratio (95% CI)
p value for
interaction 0.2
0.4
0.6 0.8 1
1.5
Empagliflozin
better
Placebo
better
(B) Incident or worsening nephropathy*
eGFR (mL min 1.73m2)
≥90
90/961 (9%)
71/449 (16%)
0.57 (0.41−0.77)
0.52
≥60 to <90
228/2165 (11%)
156/1105 (14%)
0.69 (0.56−0.84)
≥45 to <60
116/696 (17%)
86/334 (26%)
0.54 (0.41−0.71)
<45
58/269 (22%)
47/145 (32%)
0.62 (0.42−0.91)
uACR (mg/g)
<30
109/2737 (4%)
89/1361 (7%)
0.58 (0.44−0.77)
0.83
≥30 to ≤300
373/1311 (28%)
267/659 (41%)
0.56 (0.48−0.66)
All participants
492/4091 (12%)
360/2033 (18%)
0.61 (0.53−0.70)
eGFR estimated glomerular filtration rate; uACR urinary albumin:creatinine ratio; RRT renal replacement therapy 0.2
0.4
0.6 0.8 1
1.5
Empagliflozin
better
Placebo
better
(B) Incident or worsening nephropathy*
eGFR (mL min 1.73m2)
≥90
90/961 (9%)
71/449 (16%)
0.57 (0.41−0.77)
0.52
≥60 to <90
228/2165 (11%)
156/1105 (14%)
0.69 (0.56−0.84)
≥45 to <60
116/696 (17%)
86/334 (26%)
0.54 (0.41−0.71)
<45
58/269 (22%)
47/145 (32%)
0.62 (0.42−0.91)
uACR (mg/g)
<30
109/2737 (4%)
89/1361 (7%)
0.58 (0.44−0.77)
0.83
≥30 to ≤300
373/1311 (28%)
267/659 (41%)
0.56 (0.48−0.66)
All participants
492/4091 (12%)
360/2033 (18%)
0.61 (0.53−0.70)
eGFR estimated glomer lar filtration rate
ACR
rinar
alb min creatinine ratio RRT renal replacement therap (B) Incident or worsening nephropathy*
2 †Doubling of serum creatinine, initiation of maintenance RRT or death from renal disease. Doubling of serum creatinine, initiation of maintenance RRT or death from renal disease. *New onset of macroalbuminuria, doubling of serum creatinine accompanied by an eGFR ≤45 mL/min/1.73m2,
initiation of maintenance RRT or death from renal disease. A total of 52 participants with missing baseline or post-baseline creatinine data (with no evidence of RRT or renal death)
were excluded from analyses of the post hoc traditional kidney disease progression outcome. Effects of SGLT-2 inhibition on kidney disease
progression In exploratory analyses from two large outcome trials, a hy-
pothesis that SGLT-2 inhibition with empagliflozin or canagli-
flozin has the potential to reduce the risk of kidney disease
progression in people with T2DM was raised. In the EMPA-REG
OUTCOME trial, empagliflozin 10–25 mg/day reduced the inci-
dence of the traditional renal composite outcome of doubling of 754
|
W.G. Herrington et al. 0.2
0.4
0.6 0.8 1
1.5
Empagliflozin
(n=4687)
Placebo
(n=2333)
Hazard ratio (95% CI)
p value for
interaction
Empagliflozin
better
Placebo
better
(A) Traditional kidney disease progression outcome (post−hoc)†
eGFR (mL min 1.73m2)
≥90
7/1043 (0.7%)
15/486 (3.1%)
0.21 (0.09−0.53)
0.18
≥60 to <90
34/2406 (1.4%)
27/1232 (2.2%)
0.61 (0.37−1.02)
≥45 to <60
23/822 (2.8%)
16/416 (3.8%)
0.68 (0.36−1.28)
<45
17/374 (4.5%)
13/189 (6.9%)
0.63 (0.30−1.29)
uACR (mg/g)
<30
22/2766 (0.8%)
26/1376 (1.9%)
0.41 (0.23−0.72)
0.51
≥30 to ≤300
23/1325 (1.7%)
17/671 (2.5%)
0.67 (0.36−1.26)
>300
33/504 (6.5%)
28/260 (10.8%)
0.51 (0.31−0.85)
All participants
81/4645 (1.7%)
71/2323 (3.1%)
0.54 (0.40−0.75)
(B) Incident or worsening nephropathy*
eGFR (mL min 1.73m2)
≥90
90/961 (9%)
71/449 (16%)
0.57 (0.41−0.77)
0.52
≥60 to <90
228/2165 (11%)
156/1105 (14%)
0.69 (0.56−0.84)
≥45 to <60
116/696 (17%)
86/334 (26%)
0.54 (0.41−0.71)
<45
58/269 (22%)
47/145 (32%)
0.62 (0.42−0.91)
uACR (mg/g)
<30
109/2737 (4%)
89/1361 (7%)
0.58 (0.44−0.77)
0.83
≥30 to ≤300
373/1311 (28%)
267/659 (41%)
0.56 (0.48−0.66)
All participants
492/4091 (12%)
360/2033 (18%)
0.61 (0.53−0.70)
eGFR, estimated glomerular filtration rate; uACR, urinary albumin:creatinine ratio; RRT, renal replacement therapy. †Doubling of serum creatinine, initiation of maintenance RRT or death from renal disease. Effects of SGLT-2 inhibition on kidney disease
progression 0.2
0.4
0.6 0.8 1
1.5
Empagliflozin
(n=4687)
Placebo
(n=2333)
Hazard ratio (95% CI)
p value for
interaction
Empagliflozin
better
Placebo
better
(A) Traditional kidney disease progression outcome (post−hoc)†
eGFR (mL min 1.73m2)
≥90
7/1043 (0.7%)
15/486 (3.1%)
0.21 (0.09−0.53)
0.18
≥60 to <90
34/2406 (1.4%)
27/1232 (2.2%)
0.61 (0.37−1.02)
≥45 to <60
23/822 (2.8%)
16/416 (3.8%)
0.68 (0.36−1.28)
<45
17/374 (4.5%)
13/189 (6.9%)
0.63 (0.30−1.29)
uACR (mg/g)
<30
22/2766 (0.8%)
26/1376 (1.9%)
0.41 (0.23−0.72)
0.51
≥30 to ≤300
23/1325 (1.7%)
17/671 (2.5%)
0.67 (0.36−1.26)
>300
33/504 (6.5%)
28/260 (10.8%)
0.51 (0.31−0.85)
All participants
81/4645 (1.7%)
71/2323 (3.1%)
0.54 (0.40−0.75)
(B) Incident or worsening nephropathy*
eGFR (mL min 1.73m2)
≥90
90/961 (9%)
71/449 (16%)
0.57 (0.41−0.77)
0.52
≥60 to <90
228/2165 (11%)
156/1105 (14%)
0.69 (0.56−0.84)
≥45 to <60
116/696 (17%)
86/334 (26%)
0.54 (0.41−0.71)
<45
58/269 (22%)
47/145 (32%)
0.62 (0.42−0.91)
uACR (mg/g)
<30
109/2737 (4%)
89/1361 (7%)
0.58 (0.44−0.77)
0.83
≥30 to ≤300
373/1311 (28%)
267/659 (41%)
0.56 (0.48−0.66)
All participants
492/4091 (12%)
360/2033 (18%)
0.61 (0.53−0.70)
eGFR, estimated glomerular filtration rate; uACR, urinary albumin:creatinine ratio; RRT, renal replacement therapy. *New onset of macroalbuminuria, doubling of serum creatinine accompanied by an eGFR ≤45 mL/min/1.73m2,
i iti ti
f
i t
RRT
d
th f
l di
†Doubling of serum creatinine, initiation of maintenance RRT or death from renal disease. Table 1. Ongoing large SGLT-2 inhibitor clinical trials in CKD and HF populations Other large placebo-controlled SGLT-2 trials in those with T2DM and high CV risk include DECLARE (dapagliflozin 5 or 10 mg) [91], VERTIS CV (ertugliflozin 5 or 15 mg)
[92] and SCORED [93] (sotagliflozin in those with an eGFR of 25–60 mL/min/1.73 m2) which are enrolling 17 276, 8000 and 10 500 people, respectively, and include kid-
ney disease progression endpoints as secondary outcomes. DELIVER (Dapagliflozin 10 mg) in 4700 with preserved LVEF heart failure and SOLOIST-WHF (sotagliflozin)
in 4000 people with heart failure and diabetes are also in development. CREDENCE has been stopped early for efficacy [76] and is likely to report in 2019. Both Dapa-
CKD and EMPA-KIDNEY are event-driven trials and are expected to complete follow-up in around November 2020 and June 2022, respectively. (Table 1), with EMPA-KIDNEY uniquely including people with a
low eGFR (<45 mL/min/1.73 m2) with or without albuminuria. (Table 1), with EMPA-KIDNEY uniquely including people with a
low eGFR (<45 mL/min/1.73 m2) with or without albuminuria. median uACR was 927 mg/g. The treatment comparison is of
canagliflozin 100 mg versus matching placebo [75], and the pri-
mary composite outcome is a doubling of serum creatinine,
ESKD or death from renal or CV causes. Large-scale trials testing
empagliflozin versus placebo (EMPA-KIDNEY) and dapagliflozin
versus placebo (Dapa-CKD) in people with CKD are in progress
[77]. Their designs differ from CREDENCE in that both these stud-
ies are looking at a CKD population with and without diabetes Table 1. Ongoing large SGLT-2 inhibitor clinical trials in CKD and HF populations Ongoing large SGLT-2 inhibitor clinical trials in CKD and HF populations
Key inclusion criteria
Size
Interventions
Primary outcomes
Selected secondary
outcomes
CKD populations
EMPA-KIDNEY: The
Study of Heart and
Kidney Protection
with Empagliflozin [82]
Age 18 years
eGFR 20–45 or eGFR 45–90
mL/min/1.73 m2 with
uACR 200 mg/g
Clinically appropriate
doses of RAS blockade,
unless not tolerated
5000 (1/3 with DM
and 1/3 without DM)
Empagliflozin
10 mg/day versus
placebo
Sustained 40% de-
cline in eGFR,
ESKD or death
from renal or CV
causes
CV death or hospitaliza-
tion for HF
All-cause
hospitalization
All-cause mortality
CREDENCE [74, 75]
Age 30 years
T2DM, HbA1c 6.5–12%
eGFR 30–90 mL/min/1.73
m2
Stable maximally toler-
ated RAS blockade
uACR 300–5000 mg/g
4401
Canagliflozin
100 mg/day versus
placebo
Doubling of creati-
nine, ESKD or
death from renal
or CV causes
CV death or hospitaliza-
tion for HF
Doubling of creatinine,
ESKD or death from a
renal cause
Dapa-CKD [77]
Age 18 years
eGFR 25–75 mL/min/1.73
m2
Stable maximally toler-
ated RAS blockade, if
not contraindicated
uACR 200–5000 mg/g
4000
Dapagliflozin 5 or
10 mg/day versus
placebo
Sustained 50% de-
cline in eGFR,
ESKD or death
from renal or CV
causes
CV death or hospitaliza-
tion for HF
Sustained 50% decline
in eGFR, ESKD or death
from a renal cause
HF populations
EMPEROR-Preserved [88] Age 18 years
Symptomatic chronic HF
with LVEF >40%
NT-proBNP >300 pg/mL
(or >900 if in AF)
Stable dose of oral
diuretic
4100
Empagliflozin
10 mg/day versus
placebo
CV death or hospi-
talization for HF
eGFR slope
Sustained 40% decline
in eGFR or ESKD
EMPEROR-Reduced [89]
Age 18 years
Class II–IV chronic HF
with LVEF 40%
NT-proBNP above a
certain threshold
(stratified by LVEF)
Appropriate doses of
medical therapy and
use of medical devices
2800
Empagliflozin
10 mg/day versus
placebo
CV death or hospi-
talization for HF
eGFR slope
Sustained 40% decline
in eGFR or ESKD
Dapa-HF [90]
Age 18 years
Symptomatic chronic HF
with LVEF 40%
NT-proBNP 600 pg/mL
eGFR 30 mL/min/1.73 m2
Appropriate background
standard of care
4500
Dapagliflozin
10 mg/day versus
placebo
CV death, hospitali-
zation for HF or
urgent HF visit
Sustained 50% decline
in eGFR, ESKD or death
from a renal cause Key inclusion criteria
Size Empagliflozin
10 mg/day versus
placebo
CV death or hospi-
talization for HF
eGFR slope
Sustained 40% decline
in eGFR or ESKD Empagliflozin
10 mg/day versus
placebo
CV death or hospi-
talization for HF
eGFR slope
Sustained 40% decline
in eGFR or ESKD Dapagliflozin
10 mg/day versus
placebo
CV death, hospitali-
zation for HF or
urgent HF visit
Sustained 50% decline
in eGFR, ESKD or death
from a renal cause AF, atrial fibrillation; LVEF, left ventricular ejection fraction; NT-proBNP, N-terminal of the prohormone of brain natriuretic peptide. Table 1. Ongoing large SGLT-2 inhibitor clinical trials in CKD and HF populations Table 1. Ongoing large SGLT-2 inhibitor clinical trials in CKD and HF populations Table 1. Effects of SGLT-2 inhibition on kidney disease
progression Analyses of incident or
worsening nephropathy excluded 769 participants with macroalbuminuria at baseline and a further 127 participants
with missing baseline or follow-up laboratory values (47 creatinine and an additional 80 uACR) plus no evidence of
RRT or renal death (which differs slightly from Wanner et al. [33]). FIGURE 3: Effect of allocation to empagliflozin versus placebo on (A) traditional kidney disease progression outcome (post hoc) and (B) incident or worsening nephropa-
thy, by baseline eGFR and uACR. FIGURE 3: Effect of allocation to empagliflozin versus placebo on (A) traditional kidney disease progression outcome (post hoc) and (B) incident or worsening nephropa-
thy, by baseline eGFR and uACR. creatinine (with eGFR 45 mL/min/1.73 m2), initiation of renal re-
placement therapy (RRT) or renal death by 46% {hazard ratio [HR]
0.54 [95% confidence interval (CI) 0.40–0.75]} [33]. This included a
nominally significant reduction in the necessity to start RRT [HR
0.45 (95% CI 0.21–0.97)]. In the CANVAS/CANVAS-R, compared
with placebo, canagliflozin 100–300 mg/day reduced the risk of
the composite renal outcome of a 40% decline in eGFR, initiation
of RRT or renal death by 40% [HR 0.60 (95% CI 0.47–0.77)] [31]. findings are reinforced by similar observations on a pre-
specified renal composite outcome that further included new-
onset macroalbuminuria (Figure 3B). The first large-scale data on the effects of SGLT-2 inhibition
on diabetic kidney disease progression will be provided by
the Evaluation of the Effects of Canagliflozin on Renal and
Cardiovascular
Outcomes
in
Participants
with
Diabetic
Nephropathy (CREDENCE) [74], which has recently been stopped
early for benefit at the formal interim analysis planned for once
405 participants had experienced a primary outcome [75, 76]. CREDENCE includes 4401 people with T2DM and macroalbumi-
nuria [urinary albumin:creatinine ratio (uACR) 300–5000 mg/g)
and an eGFR of 30–90 mL/min/1.73 m2 on stable RAS blockade. The baseline mean eGFR was 56 mL/min/1.73 m2 and the The EMPA-REG OUTCOME trial enrolled 1800 people with
an eGFR <60 mL/min/1.73 m2, including 560 individuals with
an eGFR <45 mL/min/1.73 m2. The proportional effects of empa-
gliflozin on the traditional renal composite were similar irre-
spective of eGFR at baseline (interaction P ¼ 0.18) or baseline
levels of albuminuria (interaction P ¼ 0.51; Figure 3A) [33]. These SGLT-2 inhibition in CKD
|
755 Effects of SGLT-2 inhibition on CV diseases to reduce this composite by 14% compared with placebo [HR 0.86
(95% CI 0.74–0.99)] [32]. This effect was driven by a highly signifi-
cant 38% [HR 0.62 (95% CI 0.49–0.77)] reduction in CV death. A bene-
ficial effect of SGLT-2 inhibition on CV risk has also been observed
in the CANVAS/CANVAS-R. Canagliflozin 100–300 mg/day reduced
the primary composite outcome of death from CV causes, non-fa-
tal myocardial infarction or non-fatal stroke by 14% [HR 0.86 (95%
CI 0.75–0.97)] [31]. The effect on CV death in the CANVAS/CANVAS-
R [HR 0.87 (95% CI 0.72–1.06)] was directionally consistent, although
more modest, than in the EMPA-REG OUTCOME trial. CANVAS/CANVAS-R also reduced the risk of hospitalization for
HF by about one-third compared with placebo [HR 0.67 (95% CI
0.52–0.87)] [31]. As HF with preserved ejection fraction is common
in people with diabetes (and may even be more common in this
population than HF with reduced ejection fraction [80]), this effect
may have resulted in some reduction in the risk of HF in people
with preserved ejection fraction [25, 81]. This is an important sug-
gestion, as preserved ejection fraction HF, which is common in
CKD, has few proven effective therapies [29]. However, neither of
these trials definitively differentiated the type of HF, so it is not
possible to confirm a reduced risk of preserved ejection fraction
HF. Instead, the Empagliflozin Outcome Trial in Patients with
Chronic
Heart
Failure
with
Preserved
Ejection
Fraction
(EMPEROR-Preserved) and EMPEROR-Reduced trials with empagli-
flozin 10 mg versus placebo are recruiting from these two HF pop-
ulations and including people with and without diabetes. In
addition, the effects of dapagliflozin versus placebo in people
with reduced ejection fraction are being explored in Dapa-HF
(Table 1). The primary endpoints of all three of these dedicated
HF trials include a composite of CV death or hospitalization for
HF [42]. The three large kidney trials of SGLT-2 inhibition have in-
cluded this same composite as a secondary outcome [74, 77, 82, to reduce this composite by 14% compared with placebo [HR 0.86
(95% CI 0.74–0.99)] [32]. This effect was driven by a highly signifi-
cant 38% [HR 0.62 (95% CI 0.49–0.77)] reduction in CV death. A bene-
ficial effect of SGLT-2 inhibition on CV risk has also been observed
in the CANVAS/CANVAS-R. Effects of SGLT-2 inhibition on CV diseases 0.4
0.6
0.8 1
1.5
Empagliflozin
(n=4687)
Placebo
(n=2333)
Hazard ratio (95% CI)
p value for
interaction
Empagliflozin
better
Placebo
better
(A) CV death
eGFR (mL min 1.73m2)
≥90
28/1050 (2.7%)
19/488 (3.9%)
0.70 (0.39−1.25)
0.28
≥60 to <90
69/2423 (2.8%)
70/1238 (5.7%)
0.49 (0.35−0.68)
≥45 to <60
49/831 (5.9%)
30/418 (7.2%)
0.82 (0.52−1.29)
<45
26/381 (6.8%)
18/189 (9.5%)
0.71 (0.39−1.30)
All participants
172/4687 (3.7%)
137/2333 (5.9%)
0.62 (0.49−0.77)
(B) CV death or hospitalization for HF
eGFR (mL min 1.73m2)
≥90
36/1050 (3.4%)
25/488 (5.1%)
0.67 (0.40−1.12)
0.85
≥60 to <90
117/2423 (4.8%)
96/1238 (7.8%)
0.60 (0.46−0.79)
≥45 to <60
71/831 (8.5%)
48/418 (11.5%)
0.74 (0.51−1.07)
<45
41/381 (10.8%)
29/189 (15.3%)
0.67 (0.41−1.07)
All participants
265/4687 (5.7%)
198/2333 (8.5%)
0.66 (0.55−0.79)
(C) All−cause hospitalization
eGFR (mL min 1.73m2)
≥90
343/1050 (33%)
155/488 (32%)
1.03 (0.85−1.24)
0.34
≥60 to <90
842/2423 (35%)
467/1238 (38%)
0.88 (0.79−0.99)
≥45 to <60
357/831 (43%)
203/418 (49%)
0.82 (0.69−0.97)
<45
183/381 (48%)
100/189 (53%)
0.84 (0.66−1.07)
All participants
1725/4687 (37%)
925/2333 (40%)
0.89 (0.82−0.96)
4: Effect of allocation to empagliflozin versus placebo on (A) CV death, (B) CV death or hospitalization for HF and (C) all-cause hospitalization, by baseline eGF Empagliflozin
(n=4687)
Placebo
(n=2333)
Hazard ratio (95% CI)
p value for
interaction
(A) CV death
eGFR (mL min 1.73m2)
≥90
28/1050 (2.7%)
19/488 (3.9%)
0.70 (0.39−1.25)
0.28
≥60 to <90
69/2423 (2.8%)
70/1238 (5.7%)
0.49 (0.35−0.68)
≥45 to <60
49/831 (5.9%)
30/418 (7.2%)
0.82 (0.52−1.29)
<45
26/381 (6.8%)
18/189 (9.5%)
0.71 (0.39−1.30)
All participants
172/4687 (3.7%)
137/2333 (5.9%)
0.62 (0.49−0.77) (B) CV death or hospitalization for HF 0.4
0.6
0.8 1
1.5
Empagliflozin
better
Placebo
better
(C) All−cause hospitalization
eGFR (mL min 1.73m2)
≥90
343/1050 (33%)
155/488 (32%)
1.03 (0.85−1.24)
0.34
≥60 to <90
842/2423 (35%)
467/1238 (38%)
0.88 (0.79−0.99)
≥45 to <60
357/831 (43%)
203/418 (49%)
0.82 (0.69−0.97)
<45
183/381 (48%)
100/189 (53%)
0.84 (0.66−1.07)
All participants
1725/4687 (37%)
925/2333 (40%)
0.89 (0.82−0.96)
RE 4: Effect of allocation to empagliflozin versus placebo on (A) CV death, (B) CV death or hospitalization for HF and (C) all-cause hospitalization, by baseline eGFR. FIGURE 4: Effect of allocation to empagliflozin versus placebo on (A) CV death, (B) CV death or hospitalization for HF and (C) all-cause hospitalization, by baseline eGFR. empagliflozin versus placebo on (A) CV death, (B) CV death or hospitalization for HF and (C) all-cause hospitalization, by baseline eGFR. Effects of SGLT-2 inhibition on CV diseases The primary endpoint in the EMPA-REG OUTCOME trial was a
CV composite of death from CV causes, non-fatal myocardial in-
farction or non-fatal stroke. Empagliflozin 10–25 mg was shown 756
|
W.G. Herrington et al. 0.4
0.6
0.8 1
1.5
Empagliflozin
(n=4687)
Placebo
(n=2333)
Hazard ratio (95% CI)
p value for
interaction
Empagliflozin
better
Placebo
better
(A) CV death
eGFR (mL min 1.73m2)
≥90
28/1050 (2.7%)
19/488 (3.9%)
0.70 (0.39−1.25)
0.28
≥60 to <90
69/2423 (2.8%)
70/1238 (5.7%)
0.49 (0.35−0.68)
≥45 to <60
49/831 (5.9%)
30/418 (7.2%)
0.82 (0.52−1.29)
<45
26/381 (6.8%)
18/189 (9.5%)
0.71 (0.39−1.30)
All participants
172/4687 (3.7%)
137/2333 (5.9%)
0.62 (0.49−0.77)
(B) CV death or hospitalization for HF
eGFR (mL min 1.73m2)
≥90
36/1050 (3.4%)
25/488 (5.1%)
0.67 (0.40−1.12)
0.85
≥60 to <90
117/2423 (4.8%)
96/1238 (7.8%)
0.60 (0.46−0.79)
≥45 to <60
71/831 (8.5%)
48/418 (11.5%)
0.74 (0.51−1.07)
<45
41/381 (10.8%)
29/189 (15.3%)
0.67 (0.41−1.07)
All participants
265/4687 (5.7%)
198/2333 (8.5%)
0.66 (0.55−0.79)
(C) All−cause hospitalization
eGFR (mL min 1.73m2)
≥90
343/1050 (33%)
155/488 (32%)
1.03 (0.85−1.24)
0.34
≥60 to <90
842/2423 (35%)
467/1238 (38%)
0.88 (0.79−0.99)
≥45 to <60
357/831 (43%)
203/418 (49%)
0.82 (0.69−0.97)
<45
183/381 (48%)
100/189 (53%)
0.84 (0.66−1.07)
All participants
1725/4687 (37%)
925/2333 (40%)
0.89 (0.82−0.96)
Effect of allocation to empagliflozin versus placebo on (A) CV death, (B) CV death or hospitalization for HF and (C) all-cause hospitalization, by baseline eGFR. Effects of SGLT-2 inhibition on CV diseases In the EMPA-REG OUTCOME trial, the pre-specified secondary
outcome of hospitalization for HF was reduced by 35% (HR 0.65
(95% CI 0.50–0.85)] [32] while allocation to canagliflozin in the SGLT-2 inhibition in CKD
|
757 0.2 0.4
1
2
4
8
Empagliflozin
(n=4687)
Placebo
(n=2333)
Incidence rate ratio
(95% CI)
p value
trend†
Empagliflozin better
Placebo better
Hypoglycaemia requiring assistance
eGFR ≥60
40/3473 (1.2%)
18/1726 (1.0%)
1.06 (0.61−1.85)
eGFR ≥45 to <60
11/831 (1.3%)
10/418 (2.4%)
0.51 (0.22−1.21)
eGFR <45
12/381 (3.1%)
8/189 (4.2%)
0.78 (0.32−1.92)
0.40
All participants
63/4687 (1.3%)
36/2333 (1.5%)
0.84 (0.56−1.26)
Urinary tract infection*
eGFR ≥60
564/3473 (16.2%) 291/1726 (16.9%)
0.92 (0.80−1.07)
eGFR ≥45 to <60
186/831 (22.4%)
83/418 (19.9%)
1.08 (0.83−1.40)
eGFR <45
92/381 (24.1%)
49/189 (25.9%)
1.03 (0.73−1.46)
0.35
All participants
842/4687 (18.0%) 423/2333 (18.1%)
0.96 (0.85−1.08)
Mycotic genital infection
eGFR ≥60
237/3473 (6.8%)
32/1726 (1.9%)
3.68 (2.54−5.32)
eGFR ≥45 to <60
48/831 (5.8%)
8/418 (1.9%)
2.87 (1.36−6.07)
eGFR <45
16/381 (4.2%)
2/189 (1.1%)
4.26 (0.98−18.55)
0.82
All participants
301/4687 (6.4%)
42/2333 (1.8%)
3.55 (2.57−4.91)
Volume depletion
eGFR ≥60
158/3473 (4.5%)
66/1726 (3.8%)
1.16 (0.87−1.54)
eGFR ≥45 to <60
59/831 (7.1%)
33/418 (7.9%)
0.83 (0.54−1.27)
eGFR <45
22/381 (5.8%)
16/189 (8.5%)
0.70 (0.37−1.33)
0.09
All participants
239/4687 (5.1%)
115/2333 (4.9%)
1.00 (0.80−1.25)
Acute kidney injury
eGFR ≥60
110/3473 (3.2%)
68/1726 (3.9%)
0.77 (0.57−1.04)
eGFR ≥45 to <60
79/831 (9.5%)
53/418 (12.7%)
0.69 (0.49−0.98)
eGFR <45
57/381 (15.0%)
34/189 (18.0%)
0.87 (0.57−1.32)
0.77
All participants
246/4687 (5.2%)
155/2333 (6.6%)
0.76 (0.62−0.93)
Hyperkalemia
eGFR ≥60
46/3473 (1.3%)
36/1726 (2.1%)
0.61 (0.39−0.94)
eGFR ≥45 to <60
33/831 (4.0%)
29/418 (6.9%)
0.53 (0.32−0.86)
eGFR <45
14/381 (3.7%)
13/189 (6.9%)
0.55 (0.26−1.17)
0.72
All participants
93/4687 (2.0%)
78/2333 (3.3%)
0.57 (0.42−0.77)
Bone fracture
eGFR ≥60
122/3473 (3.5%)
59/1726 (3.4%)
0.99 (0.72−1.35)
eGFR ≥45 to <60
39/831 (4.7%)
25/418 (6.0%)
0.73 (0.44−1.20)
eGFR <45
18/381 (4.7%)
7/189 (3.7%)
1.35 (0.57−3.24)
0.97
All participants
179/4687 (3.8%)
91/2333 (3.9%)
0.94 (0.73−1.21)
Lower limb amputation
eGFR ≥60
52/3473 (1.5%)
24/1726 (1.4%)
1.07 (0.66−1.73)
eGFR ≥45 to <60
24/831 (2.9%)
13/418 (3.1%)
0.91 (0.46−1.78)
eGFR <45
12/381 (3.1%)
6/189 (3.2%)
0.97 (0.36−2.58)
0.77
All participants
88/4687 (1.9%)
43/2333 (1.8%)
1.01 (0.70−1.45) 0.2 0.4
1
2
4
8
Empagliflozin
(n=4687)
Placebo
(n=2333)
Incidence rate ratio
(95% CI)
p value
trend†
Empagliflozin better
Placebo better
Hypoglycaemia requiring assistance
eGFR ≥60
40/3473 (1.2%)
18/1726 (1.0%)
1.06 (0.61−1.85)
eGFR ≥45 to <60
11/831 (1.3%)
10/418 (2.4%)
0.51 (0.22−1.21)
eGFR <45
12/381 (3.1%)
8/189 (4.2%)
0.78 (0.32−1.92)
0.40
All participants
63/4687 (1.3%)
36/2333 (1.5%)
0.84 (0.56−1.26)
Urinary tract infection*
eGFR ≥60
564/3473 (16.2%) 291/1726 (16.9%)
0.92 (0.80−1.07)
eGFR ≥45 to <60
186/831 (22.4%)
83/418 (19.9%)
1.08 (0.83−1.40)
eGFR <45
92/381 (24.1%)
49/189 (25.9%)
1.03 (0.73−1.46)
0.35
All participants
842/4687 (18.0%) 423/2333 (18.1%)
0.96 (0.85−1.08)
Mycotic genital infection
eGFR ≥60
237/3473 (6.8%)
32/1726 (1.9%)
3.68 (2.54−5.32)
eGFR ≥45 to <60
48/831 (5.8%)
8/418 (1.9%)
2.87 (1.36−6.07)
eGFR <45
16/381 (4.2%)
2/189 (1.1%)
4.26 (0.98−18.55)
0.82
All participants
301/4687 (6.4%)
42/2333 (1.8%)
3.55 (2.57−4.91)
Volume depletion
eGFR ≥60
158/3473 (4.5%)
66/1726 (3.8%)
1.16 (0.87−1.54)
eGFR ≥45 to <60
59/831 (7.1%)
33/418 (7.9%)
0.83 (0.54−1.27)
eGFR <45
22/381 (5.8%)
16/189 (8.5%)
0.70 (0.37−1.33)
0.09
All participants
239/4687 (5.1%)
115/2333 (4.9%)
1.00 (0.80−1.25)
Acute kidney injury
eGFR ≥60
110/3473 (3.2%)
68/1726 (3.9%)
0.77 (0.57−1.04)
eGFR ≥45 to <60
79/831 (9.5%)
53/418 (12.7%)
0.69 (0.49−0.98)
eGFR <45
57/381 (15.0%)
34/189 (18.0%)
0.87 (0.57−1.32)
0.77
All participants
246/4687 (5.2%)
155/2333 (6.6%)
0.76 (0.62−0.93)
Hyperkalemia
eGFR ≥60
46/3473 (1.3%)
36/1726 (2.1%)
0.61 (0.39−0.94)
eGFR ≥45 to <60
33/831 (4.0%)
29/418 (6.9%)
0.53 (0.32−0.86)
eGFR <45
14/381 (3.7%)
13/189 (6.9%)
0.55 (0.26−1.17)
0.72
All participants
93/4687 (2.0%)
78/2333 (3.3%)
0.57 (0.42−0.77)
Bone fracture
eGFR ≥60
122/3473 (3.5%)
59/1726 (3.4%)
0.99 (0.72−1.35)
eGFR ≥45 to <60
39/831 (4.7%)
25/418 (6.0%)
0.73 (0.44−1.20)
eGFR <45
18/381 (4.7%)
7/189 (3.7%)
1.35 (0.57−3.24)
0.97
All participants
179/4687 (3.8%)
91/2333 (3.9%)
0.94 (0.73−1.21)
Lower limb amputation
eGFR ≥60
52/3473 (1.5%)
24/1726 (1.4%)
1.07 (0.66−1.73)
eGFR ≥45 to <60
24/831 (2.9%)
13/418 (3.1%)
0.91 (0.46−1.78)
eGFR <45
12/381 (3.1%)
6/189 (3.2%)
0.97 (0.36−2.58)
0.77
All participants
88/4687 (1.9%)
43/2333 (1.8%)
1.01 (0.70−1.45) Hypoglycaemia requiring assistance 83]. Effects of SGLT-2 inhibition on CV diseases Canagliflozin 100–300 mg/day reduced
the primary composite outcome of death from CV causes, non-fa-
tal myocardial infarction or non-fatal stroke by 14% [HR 0.86 (95%
CI 0.75–0.97)] [31]. The effect on CV death in the CANVAS/CANVAS-
R [HR 0.87 (95% CI 0.72–1.06)] was directionally consistent, although
more modest, than in the EMPA-REG OUTCOME trial. Given the effects of SGLT-2 inhibition on interstitial/plasma
fluid volume [70] and BP [66, 71], a beneficial effect on cardiac
pre-load and after-load is expected, and a reduction in the inci-
dence of clinical outcomes among those with HF with reduced
or preserved ejection fraction might be anticipated with such
treatments [42]. Hormonal and metabolic effects of SGLT-2 inhi-
bition may also mediate cardiac benefits. Intriguingly, increased
ketone production may have beneficial effects on hypertrophied
or failing hearts [78, 79]. †Test for trend in incidence rate ratio across eGFR categories.
*Pyelonephritis, urosepsis or serious adverse event consistent with urinary tract infection.
Empagliflozin 4/4687 vs placebo 1/2333 participants reported diabetic ketoacidosis.
Empagliflozin 0/4687 vs placebo 0/2333 participants reported Fournier's gangrene (perineal necrotizing fasciitis).
FIGURE 5: Effect of allocation to empagliflozin versus placebo on adverse events, by baseline eGFR. FUNDING W.G.H. is supported by a Medical Research Council and Kidney
Research UK Professor David Kerr Clinician Scientist Award. The EMPA-REG OUTCOME data showed that, as compared
with placebo, empagliflozin increases the frequency of my-
cotic genital infections by 3-fold [HR 3.55 (95% CI 2.57–4.91)]
but did not increase urinary tract infections [HR 0.96 (95% CI
0.85–1.08)]. Unlike dual inhibition of the RAS system, the
combination of RAS blockade and empagliflozin did not cause
serious hyperkalaemia [HR 0.57 (95% CI 0.42–0.77)] or acute
kidney injury [HR 0.76 (95% CI 0.62–0.93)], and all these safety
assessments appeared similar across the range of baseline
eGFRs studied (Figure 5) [13, 33, 86]. Laboratory analyses have
also found that blood concentration of calcium and phos-
phate did not differ in a clinically relevant manner among
those allocated empagliflozin versus placebo [87]. CONFLICT OF INTEREST STATEMENT The Clinical Trial Service Unit and Epidemiological Studies
Unit, which are part of the Nuffield Department of Population
Health, University of Oxford has a staff policy of not accepting
honoraria or consultancy fees except for reimbursement of
expenses to attend scientific meetings. All the authors are
members of the EMPA-KIDNEY Study Steering Committee. WGH reports funding from the UK Medical Research Council
(MRC), Kidney Research UK, British Heart Foundation and
Boehringer Ingelheim. RH, CB and MJL report grants from the
MRC, British Heart Foundation, Boehringer Ingelheim, Merck,
Novartis and Pfizer. MJL also reports grants from Health Data
Research UK and Medicines Company. DP reports grants from
the
National
Institute
for
Health
Research,
Boehringer
Ingelheim and Medicines Company. NS reports a grant from
Boehringer Ingelheim. CW reports honoraria from Boehringer
Ingelheim and Janssen (modest). JBG has received funding
from Merck, AstraZeneca and GlaxoSmithKline, funding and
personal fees from Merck, other support from Boehringer-
Ingelheim and personal fees from Bioscientifica and the
Endocrine
Society. MvE,
SH
and
JG
are
employees
of
Boehringer Ingelheim. The CANVAS/CANVAS-R showed that canagliflozin was also
generally well-tolerated [31]. Like empagliflozin, canagliflozin
caused an excess of genital mycotic infections, but a possible in-
creased risk of lower-limb amputation and bone fracture was
also identified. Neither of these potential hazards were ob-
served in the EMPA-REG OUTCOME trial (Figure 5) or when the
EMPA-REG OUTCOME trial was combined with other placebo-
controlled empagliflozin trials (including >12 000 participants
with T2DM) [87]. Nevertheless, amputations and bone fractures,
in addition to ketoacidosis, are being carefully monitored in the
ongoing SGLT-2 inhibitor trials. SUPPLEMENTARY DATA Supplementary data are available at ndt online. Safety and tolerability of SGLT-2 inhibition In the EMPA-REG OUTCOME trial, empagliflozin was generally
well-tolerated during a median follow-up of just over 3 years. The frequency of adverse events that led to discontinuation of
study treatment and serious adverse events among participants
allocated to empagliflozin was no higher than among those al-
located to placebo [32, 33]. Indeed, there was a significant 11%
reduction in the risk of hospitalization for any cause among
those allocated to empagliflozin compared with placebo [HR
0.89 (95% CI 0.82–0.96); Figure 4C]. Overall in the EMPA-REG OUTCOME trial, there was no sig-
nificant increase in the frequency of hypoglycaemia requiring
assistance among those allocated to empagliflozin as compared
with placebo [HR 0.84 (95% CI 0.56–1.26); Figure 5], but there is a
potential for increased risk of hypoglycaemia with empagliflo-
zin when used in combination with a sulphonylurea or insulin
[87]. Importantly, in studies comprising individuals with normo-
glycaemia, SGLT-2 inhibitors do not alter fasting plasma glucose
levels [55], and so it is not anticipated that SGLT-2 inhibition
will increase hypoglycaemia risk in those without diabetes. ACKNOWLEDGEMENTS We thank Dan Massey for assisting in the analysis of the
EMPA-REG OUTCOME trial data. All currently marketed SGLT-2 inhibitors carry a warning
about diabetic ketoacidosis on their US labels. In the EMPA-REG
OUTCOME trial, ketoacidosis was a rare event (see Figure 5 foot-
note) and so the precise size of the risk of ketoacidosis with
SGLT-2 inhibition in different types of people is currently uncer-
tain. Since the most common cause of ketoacidosis is insuffi-
cient endogenous insulin availability, the risk of ketoacidosis is
expected to be considerably lower in people without diabetes. Effects of SGLT-2 inhibition on CV diseases It should be noted that in post hoc analyses of the EMPA-REG
OUTCOME trial, allocation to empagliflozin reduced the risk of CV
death or hospitalization for HF by 34% [HR 0.66 (95% CI 0.55–0.79)],
a benefit that was similar irrespective of baseline risk of HF [84]. clinical parameter with the largest mediating effect on the re-
duction in CV death [85]. This observation may have particular
relevance in CKD populations, where non-atherosclerotic heart
disease and fluid overload/HF are common [80]. Notably,
in
subgroup
analyses
from
the
EMPA-REG
OUTCOME trial, the proportional effects of empagliflozin on CV
death and on the composite of CV death or hospitalization for Notably,
in
subgroup
analyses
from
the
EMPA-REG
OUTCOME trial, the proportional effects of empagliflozin on CV
death and on the composite of CV death or hospitalization for Exploration of the EMPA-REG OUTCOME data has suggested
that the increase in haematocrit caused by empagliflozin, a sur-
rogate for reductions in plasma volume, was the intermediate 758
|
W.G. Herrington et al. HF were similar irrespective of baseline eGFR (Figure 4A and B)
or the level of albuminuria (Supplementary Figure 1) [33, 86]. HF were similar irrespective of baseline eGFR (Figure 4A and B)
or the level of albuminuria (Supplementary Figure 1) [33, 86]. SGLT-2 inhibition could reduce the risk of kidney disease progres-
sion in CKD. Mechanistically, the effects of SGLT-2 inhibition on
intraglomerular pressure appear to complement the effects of RAS
inhibition (Figure 1) without causing hyperkalaemia or acute kid-
ney injury, so their combination may have the potential to benefit
those at risk of hyperfiltration (i.e., those with reduced eGFR and/
or albuminuria). Moreover, there is good reason to hypothesize
that SGLT-2 inhibition may reduce HF risk, a common condition
in CKD. As SGLT-2 inhibition does not appear to require preserved
kidney function or overt hyperglycaemia to have important phar-
macological effects, another large, prospective, placebo-controlled
trial called EMPA-KIDNEY (The Study of Heart and Kidney
Protection With Empagliflozin) is now planned in order to test de-
finitively whether SGLT-2 inhibition with empagliflozin can re-
duce the risk of cardio-renal outcomes in a broad range of people
with CKD, including individuals with overt albuminuria and those
with low eGFR, irrespective of their level of albuminuria. 1.
Levey AS, Coresh J. Chronic kidney disease. Lancet 2012; 379:
165–80 REFERENCES Clinical
Guidance: Lipid Modification: Cardiovascular Risk Assessment and
the Modification of Blood Lipids for the Primary and Secondary
Prevention of Cardiovascular Disease. http://www.nice.org.uk/
guidance/cg181 (8 April 2018, date last accessed) 8. Kerr M, Bray B, Medcalf J et al. Estimating the financial cost of
chronic kidney disease to the NHS in England. Nephrol Dial
Transplant 2012; 27(Suppl 3): iii73–iii80 9. Matsushita K, van der Velde M, Astor BC, et al. Association of
estimated glomerular filtration rate and albuminuria with
all-cause and cardiovascular mortality in general population
cohorts: a collaborative meta-analysis. Lancet 2010; 375:
2073–2081 25. Park M, Hsu C-Y, Li Y et al. Associations between kidney
function and subclinical cardiac abnormalities in CKD. J Am
Soc Nephrol 2012; 23: 1725–1734 26. Foley RN, Parfrey PS, Kent GM et al. Long-term evolution of
cardiomyopathy in dialysis patients. Kidney Int 1998; 54:
1720–1725 10. Jafar TH, Schmid CH, Landa M, et al. Angiotensin-converting
enzyme inhibitors and progression of nondiabetic renal dis-
ease. A meta-analysis of patient-level data. Ann Intern Med
2001; 135: 73–87 27. Levin A, Thompson CR, Ethier J et al. Left ventricular mass in-
dex increase in early renal disease: impact of decline in he-
moglobin. Am J Kidney Dis 1999; 34: 125–134 11. Lewis EJ, Hunsicker LG, Clarke WR et al. Renoprotective effect
of
the
angiotensin-receptor
antagonist
irbesartan
in
patients with nephropathy due to type 2 diabetes. N Engl J
Med 2001; 345: 851–860 28. Ninomiya T, Perkovic V, Turnbull F et al. Blood pressure low-
ering and major cardiovascular events in people with and
without chronic kidney disease: meta-analysis of rando-
mised controlled trials. BMJ 2013; 347: f5680 12. Brenner BM, Cooper ME, de Zeeuw D et al. Effects of losartan
on renal and cardiovascular outcomes in patients with
type 2 diabetes and nephropathy. N Engl J Med 2001; 345:
861–869 29. Metra
M,
Teerlink
JR. Heart
failure. Lancet
2017;
390:
1981–1995 30. Bakris GL, Fonseca VA, Sharma K et al. Renal sodium-glucose
transport: role in diabetes mellitus and potential clinical
implications. Kidney Int 2009; 75: 1272–1277 13. Fried LF, Emanuele N, Zhang JH et al. Combined angiotensin
inhibition for the treatment of diabetic nephropathy. N Engl J
Med 2013; 369: 1892–1903 31. Neal B, Perkovic V, Mahaffey KW et al. Canagliflozin and car-
diovascular and renal events in type 2 diabetes. N Engl J Med
2017; 377: 644–657 14. Zoungas S, Arima H, Gerstein HC et al. REFERENCES There is a high unmet clinical need to reduce further the risks
of kidney disease progression and CV disease in people with
CKD irrespective of whether they have diabetes. Results from
the
EMPA-REG
OUTCOME
trial
and
the
pooled
CANVAS/
CANVAS-R programme have raised a strong hypothesis that 1. Levey AS, Coresh J. Chronic kidney disease. Lancet 2012; 379:
165–80 2. Stevens PE, O’Donoghue DJ, de Lusignan S et al. Chronic kid-
ney disease management in the United Kingdom: NEOERICA
project results. Kidney Int 2007; 72: 92–99 SGLT-2 inhibition in CKD
|
759 3. Hoerger TJ, Simpson SA, Yarnoff BO et al. The future burden
of CKD in the United States: a simulation model for the CDC
CKD Initiative. Am J Kidney Dis 2015; 65: 403–411 20. Wright JT Jr, Bakris G, Greene T, et al. Effect of blood pressure
lowering and antihypertensive drug class on progression of
hypertensive kidney disease: results from the AASK trial. JAMA 2002; 288: 2421–2431 4. Jha V, Garcia-Garcia G, Iseki K et al. Chronic kidney disease:
global
dimension
and
perspectives. Lancet
2013;
382:
260–272 21. Ettehad D, Emdin CA, Kiran A et al. Blood pressure lowering
for prevention of cardiovascular disease and death: a
systematic review and meta-analysis. Lancet 2016; 387:
957–967 5. US Renal Data System. USRDS. 2014 Annual Data Report: Atlas
of End-Stage Renal Disease in the United States. Bethesda, MD:
National Institutes of Health, National Institute of Diabetes
and Digestive and Kidney Diseases, 2014, 2 22. Herrington W, Emberson J, Mihaylova B, et al. How are the
effects of reducing LDL cholesterol with a statin-based re-
gime influenced by renal function? Meta-analysis of individ-
ual data from 28 randomised trials. Lancet Diabetes Endocrinol
2016; 4: 829–839 6. Fox CS, Matsushita K, Woodward M et al. Associations of kid-
ney disease measures with mortality and end-stage renal
disease in individuals with and without diabetes: a meta-
analysis. Lancet 2012; 380: 1662–1673 23. Kidney
Disease:
Improving
Global
Outcomes
(KDIGO)
Lipid Guideline Development Work Group. KDIGO clinical
practice guideline for lipid management in chronic kid-
ney disease. Kidney Int Suppl 2013; 3: 263–26 7. Kent S, Schlackow I, Lozano-Ku¨ hne J et al. What is the impact
of chronic kidney disease stage and cardiovascular disease
on the annual cost of hospital care in moderate-to-severe
kidney disease? BMC Nephrol 2015; 16: 65 24. National Institute for Health and Clinical Excellence. REFERENCES Cherney DZI, Cooper ME, Tikkanen I et al. Pooled analysis of
phase III trials indicate contrasting influences of renal func-
tion on blood pressure, body weight, and HbA1c reductions
with empagliflozin. Kidney Int 2018; 93: 231–244 46. Lorenz JN, Weihprecht H, Schnermann J et al. Renin release
from isolated juxtaglomerular apparatus depends on mac-
ula densa chloride transport. Am J Physiol 1991; 260(4 Pt 2):
F486–F493 63. Sha S, Devineni D, Ghosh A et al. Canagliflozin, a novel inhib-
itor of sodium glucose co-transporter 2, dose dependently
reduces calculated renal threshold for glucose excretion and
increases urinary glucose excretion in healthy subjects. Diabetes Obes Metab 2011; 13: 669–672 47. Verbrugge FH, Dupont M, Steels P et al. The kidney in conges-
tive heart failure: ‘are natriuresis, sodium, and diuretics re-
ally the good, the bad and the ugly?’. Eur J Heart Fail 2014; 16:
133–142 64. Devineni D, Curtin CR, Polidori D et al. Pharmacokinetics and
pharmacodynamics of canagliflozin, a sodium glucose co-
transporter 2 inhibitor, in subjects with type 2 diabetes mel-
litus. J Clin Pharmacol 2013; 53: 601–610 48. Palatini P. Glomerular hyperfiltration: a marker of early renal
damage in pre-diabetes and pre-hypertension. Nephrol Dial
Transplant 2012; 27: 1708–1714 65. Taylor SI, Blau JE, Rother KI. SGLT2 inhibitors may predis-
pose to ketoacidosis. J Clin Endocrinol Metab 2015; 100:
2849–2852 49. Ogna A, Forni Ogna V, Bochud M et al. Association between
obesity and glomerular hyperfiltration: the confounding ef-
fect of smoking and sodium and protein intakes. Eur J Nutr
2016; 55: 1089–1097 66. Baker WL, Smyth LR, Riche DM et al. Effects of sodium-
glucose co-transporter 2 inhibitors on blood pressure: a sys-
tematic review and meta-analysis. J Am Soc Hypertens 2014;
8: 262–275.e9 50. Helal
I,
Fick-Brosnahan
GM,
Reed-Gitomer
B
et
al. Glomerular hyperfiltration: definitions, mechanisms and
clinical implications. Nat Rev Nephrol 2012; 8: 293–300 67. Ridderstra˚le M, Andersen KR, Zeller C et al. Comparison of
empagliflozin and glimepiride as add-on to metformin in
patients with type 2 diabetes: a 104-week randomised,
active-controlled, double-blind, phase 3 trial. Lancet Diabetes
Endocrinol 2014; 2: 691–700 51. Brenner BM, Lawler EV, Mackenzie HS. The hyperfiltration
theory: a paradigm shift in nephrology. Kidney Int 1996; 49:
1774–1777 52. Haynes R, Staplin N, Emberson J et al. Evaluating the contri-
bution of the cause of kidney disease to prognosis in CKD:
results from the Study of Heart and Renal Protection
(SHARP). REFERENCES Diabetes 2013; 62: 3324–3328 56. Bays HE, Weinstein R, Law G et al. Canagliflozin: effects in
overweight and obese subjects without diabetes mellitus. Obesity (Silver Spring) 2014; 22: 1042–1049 40. Striepe K, Jumar A, Ott C et al. Effects of the selective
sodium-glucose cotransporter 2 inhibitor empagliflozin on
vascular function and central hemodynamics in patients
with type 2 diabetes mellitus. Circulation 2017; 136: 1167–1169 57. Shyangdan DS, Uthman OA, Waugh N. SGLT-2 receptor
inhibitors for treating patients with type 2 diabetes mellitus:
a systematic review and network meta-analysis. BMJ Open
2016; 6: e009417 41. Heerspink HJ, Perkins BA, Fitchett DH et al. Sodium glucose
cotransporter 2 inhibitors in the treatment of diabetes melli-
tus: cardiovascular and kidney effects, potential mecha-
nisms,
and
clinical
applications. Circulation
2016;
134:
752–772 58. Vasilakou D, Karagiannis T, Athanasiadou E et al. Sodium-
glucose cotransporter 2 inhibitors for type 2 diabetes: a sys-
tematic review and meta-analysis. Ann Intern Med 2013; 159:
262–274 42. Butler J, Hamo CE, Filippatos G et al. The potential role and
rationale for treatment of heart failure with sodium-glucose
co-transporter
2
inhibitors. Eur
J
Heart
Fail
2017;
19:
1390–1400 59. Ferrannini E, Ramos SJ, Salsali A et al. Dapagliflozin mono-
therapy in type 2 diabetic patients with inadequate glycemic
control by diet and exercise: a randomized, double-blind,
placebo-controlled, phase 3 trial. Diabetes Care 2010; 33:
2217–2224 43. Tonneijck L, Muskiet MHA, Smits MM et al. Glomerular
hyperfiltration in diabetes: mechanisms, clinical signifi-
cance,
and
treatment. J
Am
Soc
Nephrol
2017;
28:
1023–1039 60. Roden M, Weng J, Eilbracht J et al. Empagliflozin monother-
apy with sitagliptin as an active comparator in patients with
type 2 diabetes: a randomised, double-blind, placebo-con-
trolled, phase 3 trial. Lancet Diabetes Endocrinol 2013; 1:
208–219 44. Cherney DZI, Zinman B, Inzucchi SE et al. Effects of empagli-
flozin on the urinary albumin-to-creatinine ratio in patients
with type 2 diabetes and established cardiovascular disease:
an exploratory analysis from the EMPA-REG OUTCOME
randomised,
placebo-controlled
trial. Lancet
Diabetes
Endocrinol 2017; 5: 610–621 61. Petrykiv S, Sjostrom CD, Greasley PJ et al. Differential effects
of dapagliflozin on cardiovascular risk factors at varying
degrees of renal function. Clin J Am Soc Nephrol 2017; 12:
751–759 45. Perkovic V, de Zeeum D, Mahaffey KW et al. Canagliflozin
and renal outcomes in type 2 diabetes: results from the
CANVAS Program randomised clinical trials. Lancet Diabetes
Endocrinol 2018; 6: 691–704 62. REFERENCES The effects of inten-
sive glucose control on microvascular outcomes in patients
with type 2 diabetes: a meta-analysis of individual partici-
pant data from randomised controlled trials. Lancet Diabetes
Endocrinol 2017; 5: 431–437 32. Zinman B, Wanner C, Lachin JM et al. Empagliflozin, cardio-
vascular outcomes, and mortality in type 2 diabetes. N Engl J
Med 2015; 373: 2117–2128 15. Herrington WG, Preiss D. Tightening our understanding of
intensive glycaemic control. Lancet Diabetes Endocrinol 2017;
5: 405–407 33. Wanner C, Inzucchi SE, Lachin JM et al. Empagliflozin and
progression of kidney disease in type 2 diabetes. N Engl J Med
2016; 375: 323–334 34. Liu JJ, Lee T, DeFronzo RA. Why do SGLT2 inhibitors inhibit
only 30–50% of renal glucose reabsorption in humans? Diabetes 2012; 61: 2199–2204 16. de Boer IH. Kidney disease and related findings in the diabe-
tes control and complications trial/epidemiology of diabetes
interventions and complications study. Diabetes Care 2014;
37: 24–30 35. Santer R, Calado J. Familial renal glucosuria and SGLT2: from
a mendelian trait to a therapeutic target. Clin J Am Soc
Nephrol 2010; 5: 133–141 17. Zoungas S, Chalmers J, Neal B et al. Follow-up of blood-
pressure lowering and glucose control in type 2 diabetes. N Engl J Med 2014; 371: 1392–1406 36. Wang XX, Levi J, Luo Y et al. SGLT2 protein expression is in-
creased in human diabetic nephropathy. J Biol Chem 2017;
292: 5335–5348 18. Klahr S, Levey AS, Beck GJ et al. The effects of dietary protein
restriction and blood-pressure control on the progression of
chronic renal disease. Modification of Diet in Renal Disease
Study Group. N Engl J Med 1994; 330: 877–884 37. Alicic RZ, Johnson EJ, Tuttle KR. SGLT2 inhibition for the pre-
vention and treatment of diabetic kidney disease: a review. Am J Kidney Dis 2018; 72: 267–277 19. Lv J, Ehteshami P, Sarnak MJ et al. Effects of intensive blood
pressure lowering on the progression of chronic kidney dis-
ease: a systematic review and meta-analysis. CMAJ 2013;
185: 949–957 38. Vallon V, Thomson SC. Targeting renal glucose reabsorption
to treat hyperglycaemia: the pleiotropic effects of SGLT2 in-
hibition. Diabetologia 2017; 60: 215–225 760
|
W.G. Herrington et al. and individuals with type 2 diabetes. Diabetes 2017; 66:
1999–2006 39. Abdul-Ghani MA, DeFronzo RA, Norton L. Novel hypothesis
to explain why SGLT2 inhibitors inhibit only 30–50% of fil-
tered glucose load in humans. REFERENCES Am J Kidney Dis 2014; 64: 40–48 68. Ferrannini G, Hach T, Crowe S et al. Energy balance after
sodium-glucose cotransporter 2 inhibition. Diabetes Care
2015; 38: 1730–1735 53. Perico N, Ruggenenti P, Remuzzi G. ACE and SGLT2 inhibi-
tors: the future for non-diabetic and diabetic proteinuric re-
nal disease. Curr Opin Pharmacol 2017; 33: 34–40 69. Kohan DE, Fioretto P, Tang W et al. Long-term study of
patients with type 2 diabetes and moderate renal impair-
ment shows that dapagliflozin reduces weight and blood
pressure but does not improve glycemic control. Kidney Int
2014; 85: 962–971 54. Ferrannini E, Baldi S, Frascerra S et al. Shift to fatty substrate
utilization in response to sodium-glucose cotransporter 2 in-
hibition in subjects without diabetes and patients with type
2 diabetes. Diabetes 2016; 65: 1190–1195 70. Hallow KM, Helmlinger G, Greasley PJ et al. Why do SGLT2
inhibitors reduce heart failure hospitalization? A differential
volume regulation hypothesis. Diabetes Obes Metab 2018; 20:
479–487 55. Al-Jobori H, Daniele G, Cersosimo E et al. Empagliflozin and
kinetics of renal glucose transport in healthy individuals SGLT-2 inhibition in CKD
|
761 71. Mancia G, Cannon CP, Tikkanen I et al. Impact of empagliflo-
zin on blood pressure in patients with type 2 diabetes melli-
tus
and
hypertension
by
background
antihypertensive
medication. Hypertension 2016; 68: 1355–1364 agliflozin-be-studied-chronic-kidney-disease
(29
August
2017, date last accessed) 84. Fitchett D, Butler J, van de Borne P et al. Effects of empa-
gliflozin on risk for cardiovascular death and heart failure
hospitalization across the spectrum of heart failure risk
in the EMPA-REG OUTCOME
V
R trial. Eur Heart J 2018; 39:
363–370 72. Heerspink HJ, Desai M, Jardine M et al. Canagliflozin slows
progression of renal function decline independently of gly-
cemic effects. J Am Soc Nephrol 2017; 28: 368–375 73. Neuen BL, Ohkuma T, Neal B, et al. Canagliflozin and Renal
Outcomes in Patients with Chronic Kidney Disease. Poster at the
National Kidney Federation Spring Clinical Meetings - April 10-14
2018,
Austin,
Texas. http://jgl_nkf2018.scientificpresenta
tions.org/Neuen_JGL63181.pdf
(13
April
2018,
date
last
accessed) 85. Inzucchi SE, Zinman B, Fitchett D et al. How does empagliflo-
zin reduce cardiovascular mortality? insights from a media-
tion analysis of the EMPA-REG OUTCOME trial. Diabetes Care
2018; 41: 356–363 86. Wanner C, Lachin JM, Inzucchi SE et al. Empagliflozin and
clinical outcomes in patients with type 2 diabetes, estab-
lished cardiovascular disease and chronic kidney disease. REFERENCES Circulation 2017; 132: 119–129 74. Evaluation
of
the
Effects
of
Canagliflozin
on
Renal
and
Cardiovascular
Outcomes
in
Participants
with
Diabetic
Nephropathy (CREDENCE). https://clinicaltrials.gov/ct2/show/
NCT02065791 (29 August 2017, date last accessed) 87. Kohler S, Zeller C, Iliev H et al. Safety and tolerability of
empagliflozin in patients with type 2 diabetes: pooled analy-
sis of Phase I–III clinical trials. Adv Ther 2017; 34: 1707–1726 75. Jardine MJ, Mahaffey KW, Neal B et al. The Canagliflozin and
Renal Endpoints in Diabetes with Established Nephropathy
Clinical Evaluation (CREDENCE) study rationale, design, and
baseline characteristics. Am J Nephrol 2017; 46: 462–472 88. Empagliflozin Outcome Trial in Patients with Chronic Heart Failure
with Preserved Ejection Fraction (EMPEROR-Preserved). https://
clinicaltrials.gov/ct2/show/NCT03057951 (29 August 2017,
date last accessed) 76. Phase
3
CREDENCE
Renal
Outcomes
Trial
of
INVOKANA
(Canagliflozin) is Being Stopped Early for Postive Efficacy Findings. https://www.janssen.com/phase-3-credence-renal-out
comes-trial-invokanar-canagliflozin-being-stopped-early-
positive-efficacy (6 August 2018, date last accessed) 89. Empagliflozin Outcome Trial in Patients with Chronic Heart Failure
with Reduced Ejection Fraction (EMPEROR-Reduced). https://clini
caltrials.gov/ct2/show/NCT03057977 (29 August 2017, date
last accessed) 90. Study to Evaluate the Effect of Dapagliflozin on the Incidence of
Worsening Heart Failure or Cardiovascular Death in Patients With
Chronic Heart Failure (Dapa-HF). https://clinicaltrials.gov/ct2/
show/NCT03036124 (29 August 2017, date last accessed) 77. A Study to Evaluate the Effect of Dapagliflozin on Renal Outcomes
and Cardiovascular Mortality in Partients with Chronic Kidney
Disease
(Dapa-CKD). https://clinicaltrials.gov/ct2/show/
NCT03036150 (29 August 2017, date last accessed) 78. Staels B. Cardiovascular protection by sodium glucose
cotransporter 2 inhibitors: potential mechanisms. Am J Med
2017; 130: S30–S39 91. Multicenter Trial to Evaluate the Effect of Dapagliflozin on the
Incidence of Cardiovascular Events (DECLARE-TIMI58). https://
clinicaltrials.gov/ct2/show/NCT01730534 (29 August 2017,
date last accessed) 79. Aubert G, Martin OJ, Horton JL et al. The failing heart relies
on ketone bodies as a fuel. Circulation 2016; 133: 698–705 92. Cardiovascular Outcomes Following Ertugliflozin Treatment in
Type 2 Diabetes Mellitus Participants With Vascular Disease, The
VERTIS CV Study (MK-8835-004). https://clinicaltrials.gov/ct2/
show/NCT01986881 (12 September 2017, date last accessed) 80. Bui AL, Horwich TB, Fonarow GC. Epidemiology and risk pro-
file of heart failure. Nat Rev Cardiol 2011; 8: 30–41 81. Fang JC. Heart failure with preserved ejection fraction: a kid-
ney disorder? Circulation 2016; 134: 435–437 93. Effect of Sotagliflozin on Cardiovascular and Renal Events in Patients
With Type 2 Diabetes and Moderate Renal Impairment Who Are
at Cardiovascular Risk (SCORED). https://clinicaltrials.gov/ct2/
show/NCT03315143 (19 June 2018, date last accessed) 82. REFERENCES www.empakidney.org (10 September 2018, date last accessed) 83. Empagliflozin (JardianceV
R ) to be studied in chronic kidney disease. https://www.boehringer-ingelheim.com/press-release/emp
|
https://openalex.org/W4248703955
|
https://acp.copernicus.org/articles/20/9115/2020/acp-20-9115-2020.pdf
|
English
| null |
Representation of the Equatorial Stratopause Semiannual Oscillation in Global Atmospheric Reanalyses
| null | 2,020
|
cc-by
| 13,068
|
Representation of the equatorial stratopause semiannual oscillation
in global atmospheric reanalyses Yoshio Kawatani1, Toshihiko Hirooka2, Kevin Hamilton3,4, Anne K. Smith5, and Masatomo Fujiwara6
1Japan Agency for Marine-Earth Science and Technology, Yokohama, 236-0001, Japan
2Faculty of Science, Kyushu University, Fukuoka, 819-0395, Japan
3International Pacific Research Center, University of Hawaii, Honolulu, HI 96822, USA
4Department of Atmospheric Sciences, University of Hawaii, Honolulu, HI 96822, USA
5National Center for Atmospheric Research, Boulder, CO 80307, USA
6Faculty of Environmental Earth Science, Hokkaido University, Sapporo, 060-0810, Japan Yoshio Kawatani1, Toshihiko Hirooka2, Kevin Hamilton3,4, Anne K. Smith5, and Masatomo Fujiwara6
1Japan Agency for Marine-Earth Science and Technology, Yokohama, 236-0001, Japan
2Faculty of Science, Kyushu University, Fukuoka, 819-0395, Japan
3International Pacific Research Center, University of Hawaii, Honolulu, HI 96822, USA
4Department of Atmospheric Sciences, University of Hawaii, Honolulu, HI 96822, USA
5National Center for Atmospheric Research, Boulder, CO 80307, USA
6Faculty of Environmental Earth Science, Hokkaido University, Sapporo, 060-0810, Japan Correspondence: Yoshio Kawatani (yoskawatani@jamstec.go.jp) Received: 23 January 2020 – Discussion started: 9 March 2020
Revised: 21 June 2020 – Accepted: 29 June 2020 – Published: 31 July 2020 Received: 23 January 2020 – Discussion started: 9 March 2020
Revised: 21 June 2020 – Accepted: 29 June 2020 – Published: 31 July 2020 Abstract. This paper reports on a project to compare the rep-
resentation of the semiannual oscillation (SAO) in the equa-
torial stratosphere and lower mesosphere within six major
global atmospheric reanalysis datasets and with recent satel-
lite Sounding of the Atmosphere Using Broadband Emis-
sion Radiometry (SABER) and Microwave Limb Sounder
(MLS) observations. All reanalyses have a good represen-
tation of the quasi-biennial oscillation (QBO) in the equato-
rial lower and middle stratosphere and each displays a clear
SAO centered near the stratopause. However, the differences
among reanalyses are much more substantial in the SAO re-
gion than in the QBO-dominated region. The degree of dis-
agreement among the reanalyses is characterized by the stan-
dard deviation (SD) of the monthly mean zonal wind and
temperature; this depends on latitude, longitude, height, and
time. The zonal wind SD displays a prominent equatorial
maximum that increases with height, while the temperature
SD reaches a minimum near the Equator and is largest in
the polar regions. Along the Equator, the zonal wind SD is
smallest around the longitude of Singapore, where consis-
tently high-quality near-equatorial radiosonde observations
are available. Representation of the equatorial stratopause semiannual oscillation
in global atmospheric reanalyses Interestingly, the near-Singapore minimum in
SD is evident to at least ∼3 hPa, i.e., considerably higher
than the usual ∼10 hPa ceiling for in situ radiosonde obser-
vations. Our measurement of the agreement among the re-
analyses shows systematic improvement over the period con-
sidered (1980–2016), up to near the stratopause. Character-
istics of the SAO at 1 hPa, such as its detailed time varia- tion and the displacement off the Equator of the zonal wind
SAO amplitude maximum, differ significantly among the re-
analyses. Disagreement among the reanalyses becomes still
greater above 1 hPa. One of the reanalyses in our study also
has a version produced without assimilating satellite obser-
vations, and a comparison of the SAO in these two versions
demonstrates the very great importance of satellite-derived
temperatures in the realistic analysis of the tropical upper
stratospheric circulation. Atmos. Chem. Phys., 20, 9115–9133, 2020
https://doi.org/10.5194/acp-20-9115-2020
© Author(s) 2020. This work is distributed under
the Creative Commons Attribution 4.0 License. 1
Introduction The semiannual oscillation (SAO) is an alternation of the
equatorial zonal wind between easterlies and westerlies with
a period of 6 months and is observed from the upper strato-
sphere to above the mesosphere. The SAO was first de-
tected in rocketsonde observations of the zonal wind near the
Equator (Reed, 1965, 1966; Hopkins, 1975) and these ob-
servations indicate that the SAO amplitude has two peaks:
one near the stratopause (∼1 hPa) and the other near the
mesopause (∼0.01 hPa; Hirota, 1978, 1980). The present pa-
per focuses on the stratopause SAO. Below the SAO region,
the mean wind variations are dominated by the stratospheric
quasi-biennial oscillation (QBO). The QBO and SAO zonal
wind variations have some similarities, notably a consistent
downward progression of the wind reversals and the forma-
tion of strong vertical shear zones. The peak SAO harmonic Y. Kawatani et al.: Representation of the equatorial stratopause semiannual oscillation Even in
assimilating the same observations, differences in the fore-
cast model and data assimilation technique of various obser-
vational datasets will lead to differences in their representa-
tions of atmospheric fields, including those of the mean state,
variability, and long-term trend. Kawatani et al. (2016) com-
pared the representation of the monthly mean zonal wind in
the equatorial stratosphere up to 10 hPa among major global
atmospheric reanalysis datasets. It was found that differ-
ences among reanalyses in the zonal wind depend signifi-
cantly on the number of in situ radiosonde observations, the
QBO phase, and the representation of extratropical quasi-
stationary planetary waves propagating toward the Equator. Observations of the winds and temperatures in the up-
per stratosphere and lower mesosphere are limited compared
with the lower and middle stratosphere. There are ∼220
radiosonde stations within 10◦S–10◦N included in the In-
tegrated Global Radiosonde Archive (IGRA; Durre et al.,
2006; NCDC, 2016), but only a fraction of these stations re-
ported a large number observations in the stratosphere. The
number of radiosonde observations decreases significantly
with height (Figs. 5 and 15 of Kawatani et al., 2016) and in
the IGRA database there are no observations above the usual
10 hPa ceiling for weather balloons. Rocketsondes can pro-
vide in situ measurements of the wind and temperature in the
upper stratosphere and lower mesosphere, but these obser-
vations are available only for a few locations and for limited
periods (Hirota, 1980; Garcia et al., 1997; Baldwin and Gray,
2005). Kawatani et al. (2016) noted that global meteorological
analysis processes are particularly challenging in the tropical
middle atmosphere, even in the lower stratosphere. This can
be attributed to the following reasons: the relative paucity
of in situ data (especially in the eastern and central Pacific
area with few stations; see their Fig. 5), the weaker constraint
connecting the winds and temperatures because of the small
Coriolis parameter near the Equator, and the relatively coarse
vertical resolution of satellite remote-sensing temperature re-
trievals compared to the thin regions of large vertical shear
of the zonal wind characteristic of both the QBO and SAO. As noted earlier, the lack of observations is even more pro-
nounced in the stratosphere above 10 hPa, and thus in the
SAO region we can perhaps anticipate more dependence on
the dynamical models used in the assimilation procedure. Y. Kawatani et al.: Representation of the equatorial stratopause semiannual oscillation Manney et al. (1996) indicated that the balance wind calcu-
lated from reanalysis geopotential height tends to be slower
than the reanalysis zonal wind. As the balanced gradient
wind equation is not valid near the Equator, Smith et al. (2017) estimated equatorial wind by the cubic spline interpo-
lation of the balance winds at and poleward of latitude ±8◦
for SABER and ±6◦for MLS. They assessed the reliabil-
ity of their estimations in the lower stratosphere and upper
mesosphere zonal wind by comparing with in situ radiosonde
observations in the lower stratosphere and radar-measured
meteor winds in the upper mesosphere at Ascension Island
(8◦S). amplitude determined by Hirota (1978) from rocket obser-
vations is over 30 ms−1, which is somewhat larger than the
corresponding peak QBO amplitude. Unlike the QBO, which
has exhibited periods among individual cycles ranging from
∼18 to 34 months, the SAO is clearly locked to the seasonal
cycle. Results from simple models (Dunkerton, 1979; Holton
and Wehrbein, 1980), diagnostic studies (Hamilton, 1986;
Hitchman and Leovy, 1986; Ray et al., 1998), and compre-
hensive general circulation models (Hamilton and Mahlman,
1988; Jackson and Gray, 1994) have shown that the SAO is
driven by a combination of cross-equatorial meridional ad-
vection of mean momentum, the transports of zonal momen-
tum by vertically propagating equatorial and gravity waves,
and the wave forcing from extratropical quasistationary plan-
etary waves. amplitude determined by Hirota (1978) from rocket obser-
vations is over 30 ms−1, which is somewhat larger than the
corresponding peak QBO amplitude. Unlike the QBO, which
has exhibited periods among individual cycles ranging from
∼18 to 34 months, the SAO is clearly locked to the seasonal
cycle. Results from simple models (Dunkerton, 1979; Holton
and Wehrbein, 1980), diagnostic studies (Hamilton, 1986;
Hitchman and Leovy, 1986; Ray et al., 1998), and compre-
hensive general circulation models (Hamilton and Mahlman,
1988; Jackson and Gray, 1994) have shown that the SAO is
driven by a combination of cross-equatorial meridional ad-
vection of mean momentum, the transports of zonal momen-
tum by vertically propagating equatorial and gravity waves,
and the wave forcing from extratropical quasistationary plan-
etary waves. Global atmospheric analyses that assimilate all available
satellite remote sensing and in situ observations are another
potential source of information regarding the SAO. Published by Copernicus Publications on behalf of the European Geosciences Union. Published by Copernicus Publications on behalf of the European Geosciences Union. 9116 Y. Kawatani et al.: Representation of the equatorial stratopause semiannual oscillation Y. Kawatani et al.: Representation of the equatorial stratopause semiannual oscillation 9117 contribution of satellite observations, we also analyze JRA-
55C, which assimilated conventional data only (Kobayashi
et al., 2014). Data from January 1979 to December 2016 are
used for ERA-I, ERA5, and JRA-55, and data from January
1980 to December 2016 are used for MERRA-2 (MERRA-
2 data are available from January 1980). Data are available
from January 1979 to February 2016 for MERRA, JRA-55C
is available until December 2012, and NCEP-CFSR is avail-
able from January 1979 to December 2010. As MERRA-2
data are available after January 1980 and NCEP-CFSR ends
in December 2010, monthly mean data from January 1980 to
December 2010 are used to compare the reanalyses’ clima-
tology. (in preparation) are now investigating the momentum bud-
gets and compare the mechanism among reanalyses, which
is also one of the S-RIP papers (personal communication). Detailed information, such as assimilated satellite datasets
used in each reanalysis, was provided by the S-RIP
project, notably summarized in Fujiwara et al. (2017) and
Wright et al. (2020; https://jonathonwright.github.io/pdf/
S-RIPChapter2E.pdf). However, as discussed in Kawatani
et al. (2016), it is not feasible to determine exactly what ob-
servational data were actually assimilated at each data assim-
ilation analysis step; these complications make it difficult to
conclusively attribute all the differences seen among the re-
analyses products. At the SAO altitudes, observational data
available to be assimilated are particularly limited. In addi-
tion, as described in Sect. 2, the treatment of sponge lay-
ers near the model tops is also different among reanalyses;
these factors may be expected to result in significantly larger
differences among reanalyses compared to those from QBO
altitudes described by Kawatani et al. (2016). This paper
will discuss the results of our detailed intercomparison and
will help identify the uncertainties in the reanalyses and how
uncertainties change with time as the satellite data sources
evolve. Monthly mean zonal wind and temperature data analyzed
in this study were computed from daily means. Daily mean
data are the average of the instantaneous 00:00, 06:00, 12:00,
and 18:00 UTC analyses in ERA-I, JRA-55, JRA-55C, and
NCEP-CFSR. In ERA5, the daily means are calculated
from instantaneous hourly data, from 00:00 to 23:00 UTC. For MERRA and MERRA-2, daily means consisting of
instantaneous 3-hourly “asm” output are used (for more
details on “asm”, see https://gmao.gsfc.nasa.gov/reanalysis/
MERRA-2/docs/ANAvsASM.pdf). Y. Kawatani et al.: Representation of the equatorial stratopause semiannual oscillation Consequently, the solar
diurnal (24 h) and semidiurnal (12 h) tides should be elimi-
nated in monthly mean data in each case, but the effect of
higher-order tides (e.g., 8 h, 6 h) could still be present in the
monthly means for those reanalyses with 6-hourly instanta-
neous data. However, the effects of these tides on the zonal
mean should be extremely small, at least at altitudes analyzed
in this study (up to 1 hPa for all reanalysis comparisons and
0.1 hPa for the MERRA vs. MERRA-2 comparison). One of the reanalysis datasets considered in our project,
the JRA-55 reanalyses produced by the Japan Meteorolog-
ical Agency, also has a version produced without assimilat-
ing satellite observations (JRA-55C). Kobayashi et al. (2014)
compared the JRA-55 and JRA-55C reanalyses and found
that the tropical zonal wind difference between the JRA-55
and JRA-55C is larger in the upper stratosphere than in the
lower stratosphere. The present paper will include a more
detailed comparison between JRA-55 and JRA-55C, allow-
ing a direct assessment of the importance of satellite-derived
temperatures in the realistic analysis of the tropical upper
stratospheric circulation. p
In global models, sponge layers are commonly set near
the upper boundary in order to reduce unrealistic reflec-
tion of vertically propagating waves from the model top. The formulation of sponge layers differs among the reanaly-
sis operational models. Wright et al. (https://jonathonwright. github.io/pdf/S-RIPChapter2E.pdf) summarized the details
of sponge layers and the placement of the model top level. The model tops are ∼0.266 hPa in NCEP-CFSR; 0.1 hPa in
ERA-I and JRA-55; and 0.01 hPa in ERA5 MERRA, and
MERRA-2. Sponge layers in ERA-I and ERA5 are applied
above 10 hPa by adding an additional function to the hori-
zontal diffusion terms, whose strength varies with wavenum-
ber and model level. In addition, ERA-I includes Rayleigh
friction but ERA5 does not. JRA-55 includes a sponge layer
that gradually enhances the horizontal diffusion coefficient
with height above 100 hPa and Rayleigh friction is also used
above 50 hPa. MERRA and MERRA-2 implement sponge
layers above ∼0.24 hPa by increasing the horizontal diver-
gence damping coefficient. NCEP-CSFR applies Rayleigh
damping above ∼2 hPa in addition to employing a height-
dependent horizontal diffusion coefficient. Indeed, the treat-
ment of sponge layers differs quite considerably among the
reanalysis models. Y. Kawatani et al.: Representation of the equatorial stratopause semiannual oscillation It is difficult to quantitatively estimate
how different sponge layers affect representations of the cir- This paper is outlined as follows: Sect. 2 briefly describes
the reanalysis products evaluated and the satellite observa-
tion data employed, Sect. 3 investigates differences in the
overall patterns of tropical zonal wind and temperatures
among reanalyses, Sect. 4 discusses the similarities and dif-
ferences of the SAO among the reanalyses, and conclusions
are summarized in Sect. 5. Y. Kawatani et al.: Representation of the equatorial stratopause semiannual oscillation A unique opportunity to observe global winds in the
middle atmosphere was provided by the High-Resolution
Doppler Imager (HRDI) on the Upper Atmosphere Research
Satellite (UARS) during 1992–1996, but the HRDI data are
not accurate in the 40–60 km range (Garcia et al., 1997; Ray
et al., 1998). The stratopause region is also hard to observe
with ground-based radars as it is above the usual ceiling for
atmospheric profilers and below the region that can be ob-
served with meteor winds and medium-frequency radar tech-
niques. The temperature in the stratopause region has been ob-
served by satellite-based radiometers since 1979, and for
about the last 2 decades there have been specialized limb-
sounding instruments deployed that provide higher-quality
temperature retrievals in the middle atmosphere. Recently,
Smith et al. (2017) derived the zonal mean zonal winds via
the balance wind relationship using the geopotentials de-
rived from the temperature retrievals from the Thermosphere,
Ionosphere, Mesosphere Energetic and Dynamics (TIMED)
Sounding of the Atmosphere Using Broadband Emission Ra-
diometry (SABER) instrument for 15 years and the Aura Mi-
crowave Limb Sounder (MLS) for 12.5 years. These datasets
have the advantages of long data records with no gaps and
continuous vertical coverage through the middle atmosphere. The present paper reports on one of the studies contribut-
ing to the SPARC Reanalysis Intercomparison Project (S-
RIP; Fujiwara et al., 2017), a project that focuses on evaluat-
ing reanalysis output for the middle atmosphere. We compare
the representation of the near-stratopause SAO in several
contemporary reanalysis products and further evaluate the
reanalyses by comparing them with the Smith et al. (2017)
winds derived more directly from satellite temperature obser-
vations. We have restricted our attention to the period starting
in 1979, when NOAA operational satellite radiance observa-
tions became available and were incorporated as an impor-
tant data source in all reanalyses. We do not focus on the driv-
ing mechanism of the SAO in reanalyses. Tomikawa et al. Atmos. Chem. Phys., 20, 9115–9133, 2020 https://doi.org/10.5194/acp-20-9115-2020 Y. Kawatani et al.: Representation of the equatorial stratopause semiannual oscillation 3
Differences of the overall patterns of tropical zonal
wind and temperature among reanalyses Figure 1 shows the time–height variation of monthly mean
zonal mean zonal wind over the Equator derived from
SABER and MLS observations and in each reanalysis from
2002 to 2016. All reanalyses clearly capture the basic fea-
tures of the QBO, including the cycle-to-cycle variation in
period and amplitude. The SAO is also represented in all re-
analyses, and in each case the SAO zonal wind is qualita-
tively similar to that derived from the satellite observations. It is evident, however, that the differences among reanalyses
are more pronounced at the SAO region than in the lower
and middle stratosphere. The substantially weaker amplitude
of the SAO in JRA-55C compared to that of JRA-55 indi-
cates the importance of satellite data for the representation
of the SAO. In the rest of this section, results of JRA-55C
are omitted in order to compare reanalyses under the same
conditions (i.e., reanalyses assimilating satellite data). The representation of the QBO and SAO in ERA5
has also been discussed recently in conferences and in-
formal reports (https://confluence.ecmwf.int/display/CKB/
ERA5{%}3A+The+QBO+and+SAO; Shepherd et al. 2018). ERA5 data above 1 hPa are not available at the time of this
writing, but these recent investigators did have access to
ERA5 at higher levels for the limited period 2008–2017. They concluded that the QBO at altitudes from about 50
to 5 hPa and the SAO from 5 to 0.5 hPa in ERA5 are
close to those in ERA-I. However, the representation of the
SAO above 0.5 hPa is very different between ERA5 and
ERA-I. Shepherd et al. (2018) reported that the operational
global model used to generate ERA5 simulates unrealisti-
cally strong westerlies (spurious equatorial mesospheric jet)
around 0.1 hPa during October and May. y
g
Figure 2 displays vertical profiles of the climatological an-
nual and zonal mean zonal wind and temperature over the
Equator. Climatology is calculated from January 1980 to De-
cember 2010 for the reanalyses, whereas it is calculated for
more recent years for the two satellite datasets after SABER
and MLS data were available (from January 2002 and from
January 2005, for SABER and MLS, respectively, until De-
cember 2016). For temperature, reanalyses agree well with
the satellite climatology and differences among reanalyses
are relatively small (the exception is JRA-55 above ∼0.5 hPa
where it is an outlier, possibly due to the effects of the artifi-
cial sponge layer above 1 hPa). Y. Kawatani et al.: Representation of the equatorial stratopause semiannual oscillation Y. Kawatani et al.: Representation of the equatorial stratopause semiannual oscillatio 9118 from all the datasets onto a common 1.5◦longitude–latitude
grid on 41 standard pressure levels from 1000 to 0.1 hPa. Note again that MERRA and MERRA-2 provide data on
pressure levels up to 0.1 hPa, while the others provide pres-
sure level data only up to 1 hPa. In addition, however, model
level data above 1 hPa are available for the ERA-I, JRA-
55, and JRA-55C reanalyses, and we interpolated these data
to pressure levels allowing extension of the data to 0.1 hPa. Only data up to 1 hPa are analyzed here for NCEP-CFSR and
ERA5. culation among the different reanalyses. Six reanalyses pro-
vide pressure level data up to at least 1 hPa, and thus we as-
sume that their sponge layers were designed not to strongly
affect the representation of the large-scale features of the cir-
culation up to 1 hPa. p
There are some particular concerns noted earlier about the
tropical middle atmosphere representation in the MERRA-
2 reanalyses. The dynamical model used in producing the
MERRA-2 reanalyses is able to simulate a spontaneous QBO
in the tropical stratosphere because it includes strong pa-
rameterized momentum fluxes from non-orographic gravity
waves (Fig. 3 of Molod et al., 2015). Kawatani et al. (2016)
showed that MERRA-2 exhibits spurious semi-annual varia-
tions of the 10 hPa zonal wind in the 1980s and in late 1993. The downward propagation of the westerly SAO phase is un-
realistically enhanced during these periods, possibly because
of overly strong gravity wave forcing. Coy et al. (2016) also
noted that MERRA-2 appears to overemphasize the annual
cycle before 1995. 2
Reanalysis and satellite observation data We analyzed the monthly mean zonal wind and tempera-
ture in six sets of global reanalysis data, which are available
through at least 2010 and extend at least up to 1 hPa altitude. Relatively old reanalyses used in Kawatani et al. (2016), i.e.,
NCEP1, NCEP2, ERA-40, and JRA-25, were not analyzed
here; instead, we analyze ERA-I (Dee et al., 2011), ERA5
(Hersbach et al., 2019), JRA-55 (Kobayashi et al., 2015),
MERRA (Rienecker et al., 2011), MERRA-2 (Gelaro et al.,
2017), and NCEP-CFSR (Saha et al., 2010). To assess the Atmos. Chem. Phys., 20, 9115–9133, 2020 https://doi.org/10.5194/acp-20-9115-2020 Y. Kawatani et al.: Representation of the equatorial stratopause semiannual oscillation 3
Differences of the overall patterns of tropical zonal
wind and temperature among reanalyses In contrast, the spread among
the reanalyses is quite large for the equatorial zonal wind. MERRA-2 shows a westerly bias compared to other reanal-
yses and observation above 20 hPa. Above 10 hPa, differ-
ences among both reanalyses and anomalies from SABER
and MLS become significantly larger. Above 1 hPa, zonal
winds in ERA-I and JRA-55 are fairly weak and those in
MERRA and MERRA-2 trend toward zero above ∼0.5 hPa,
presumably due to effects of sponge layers, while satellite
observations represent stronger climatological westerlies at
0.1–1 hPa. The long-term mean of the satellite zonal winds
showing mean easterlies in the middle stratosphere and west-
erlies in the lower mesosphere is in good overall agreement For comparison with the reanalyses we will use the two
zonal mean temperature and wind datasets derived by Smith
et al. (2017), one for January 2002–December 2016 based
on SABER measurements and another for August 2004–
December 2016 based on MLS measurements. Note here that
MERRA-2 is the only reanalysis that assimilates temperature
data from Aura MLS but only at pressures less than 5 hPa,
and that none of the reanalyses assimilate SABER data (Fu-
jiwara et al. 2017). Climatological fields are calculated using data from 2002
to 2016 for SABER and from 2005 to 2016 for MLS (when
the MLS data are included in the assimilation over each en-
tire year). Note that different time periods are used among
reanalyses (1980–2010), SABER (2002–2016), and MLS
(2005–2016) for the comparison of climatology. We have
confirmed that the overall characteristics of climatology dis-
cussed in this paper are not significantly affected by the
somewhat different periods analyzed. The various reanalysis and satellite datasets are provided
on a variety of horizontal grids and vertical pressure level
structures. We have interpolated the winds and temperatures Atmos. Chem. Phys., 20, 9115–9133, 2020 https://doi.org/10.5194/acp-20-9115-2020 Y. Kawatani et al.: Representation of the equatorial stratopause semiannual oscillation Y. Kawatani et al.: Representation of the equatorial stratopause semiannual oscillation 9120 Figure 2. Vertical profiles of the climatological annual and zonal mean (a) zonal wind, and (b) temperature over the Equator for SABER
(solid black), MLS (dashed black), ERA-I (blue), ERA5 (light blue), JRA-55 (purple), MERRA (pink), MERRA-2 (red), and NCEP-CFSR
(green). Climatology is calculated from 1980 to 2010 in the reanalyses, from 2002 to 2016 in SABER, and from 2005 to 2016 in MLS. Figure 2. Vertical profiles of the climatological annual and zonal mean (a) zonal wind, and (b) temperature over the Equator for SABER
(solid black), MLS (dashed black), ERA-I (blue), ERA5 (light blue), JRA-55 (purple), MERRA (pink), MERRA-2 (red), and NCEP-CFSR
(green). Climatology is calculated from 1980 to 2010 in the reanalyses, from 2002 to 2016 in SABER, and from 2005 to 2016 in MLS. Figure 3 shows the time series of monthly mean zonal
mean equatorial zonal wind in the satellite-derived datasets
(black lines) and in each reanalysis. At 1 hPa, all reanalyses
(ERA-I, ERA5, JRA-55, MERRA, MERRA2, and NCEP-
CFSR) represent the SAO with qualitative agreement with
the satellite-derived winds. However, there are substantial
differences among the reanalyses in the individual months. MERRA-2 sometimes represents significantly stronger west-
erly extremes than other reanalyses (e.g., during 1980 and
1989). ERA-I and ERA5 also sometimes have stronger west-
erly extremes (e.g., in 1991 and 1996). At 0.1 hPa the dis-
agreement among reanalyses (ERA-I, JRA-55, MERRA and
MERRA-2) is larger. MERRA winds show much stronger
easterlies than MERRA-2 until ∼1998, and these easterly
extremes become smaller later on, as in MERRA-2, al-
though the SAO phases between MERRAs are quite different
(see also Fig. 9 shown later). At 0.1 hPa the winds derived
from both the SABER and MLS satellite observations have
a strong disagreement with all the reanalyses, which dis-
play much weaker westerlies in the annual, long-term mean
(Fig. 2). mean over all N reanalyses. The SD is calculated for each
month using the monthly mean zonal wind. We calculate
the SD among all six reanalyses from 1980 to 2010, the SD
among the five reanalyses (excluding NCEP-CFSR and ex-
tended until 2015), and the SD between the two MERRA
versions (MERRA and MERRA-2) from 1980 to 2015. Atmos. Chem. Phys., 20, 9115–9133, 2020 Y. Kawatani et al.: Representation of the equatorial stratopause semiannual oscillation Y. Kawatani et al.: Representation of the equatorial stratopause semiannual oscillation 9119 Y. Kawatani et al.: Representation of the equatorial stratopause semiannual oscillation me–height section of the equatorial zonal mean zonal wind in the (a) SABER and (b) MLS satellite-derived data
al. (2017) and (c–i) each reanalysis from January 2002 to December 2016. The color intervals are 10 ms−1. computed from rocketsonde observations at low
Hitchman and Leovy, 1986). We also have drawn
December 2010 (i.e., years that are available i
and two satellites) and confirmed the results Figure 1. Time–height section of the equatorial zonal mean zonal wind in the (a) SABER and (b) MLS satellite-derived datasets as processed
by Smith et al. (2017) and (c–i) each reanalysis from January 2002 to December 2016. The color intervals are 10 ms−1. Figure 1. Time–height section of the equatorial zonal mean zonal wind in the (a) SABER and (b) MLS satellite-derived datasets as processed
by Smith et al. (2017) and (c–i) each reanalysis from January 2002 to December 2016. The color intervals are 10 ms−1. December 2010 (i.e., years that are available in all reanalyses
and two satellites) and confirmed the results are not signifi-
cantly affected by the choice of periods. December 2010 (i.e., years that are available in all reanalyses
and two satellites) and confirmed the results are not signifi-
cantly affected by the choice of periods. with that computed from rocketsonde observations at low
latitudes (Hitchman and Leovy, 1986). We also have drawn
these vertical profiles using averages from January 2005 to Atmos. Chem. Phys., 20, 9115–9133, 2020 https://doi.org/10.5194/acp-20-9115-2020 Atmos. Chem. Phys., 20, 9115–9133, 2020 Y. Kawatani et al.: Representation of the equatorial stratopause semiannual oscillation The
SD among both six and five reanalyses is discussed up to
1 hPa, whereas the SD between MERRA and MERRA-2 is
discussed up to 0.1 hPa (i.e., the maximum altitude of pres-
sure data provided). The time–height section of the zonal mean SD of the equa-
torial zonal wind and temperature among the five reanaly-
ses and only between MERRA and MERRA-2 is shown in
Fig. 4. Three-year running mean of annual mean SD is per-
formed on the SD. Figure 4c and f show the linear regression
trend in the SD at each level over the whole record. Trends
are expressed in terms of percentage per year at each level
(note that linear trends are not always appropriate in this case,
as the SD sometimes drops with the introduction of new ob-
servational data in the assimilation, as shown in Kawatani
et al., 2016). Both zonal wind and temperature SDs among five reanal-
yses decrease with time, a trend that is particularly clear at
levels from 70 to ∼2 hPa. As discussed in Kawatani et al. (2016), one possible reason for the reduction of the SD over
time is the upgrading of satellite radiance observations. From
1979 to 2006, the Television InfraRed Operational Satellite
(TIROS) Operational Vertical Sounder (TOV), Stratospheric
Sounding Unit (SSU), and Microwave Sounding Unit (MSU) To quantify the spread among reanalyses, the standard de-
viation (SD) among the reanalyses is calculated as follows: SD =
v
u
u
t
N
X
i
(ui −[u])2/N,
(1) (1) where i represents the individual datasets from among the
N datasets included. The square brackets [...] denote the Atmos. Chem. Phys., 20, 9115–9133, 2020 https://doi.org/10.5194/acp-20-9115-2020 Y. Kawatani et al.: Representation of the equatorial stratopause semiannual oscillation epresentation of the equatorial stratopause semiannual oscillation 9121 atani et al.: Representation of the equatorial stratopause semiannual oscillation
91
. (a, b) Time variations of the equatorial zonal wind at (a) 1 hPa and (b) 0.1 hPa for SABER, MLS, and each reanalysis. ailable After May 1998 data from the Advanced Mi
creased Kawatani et al (2016) also showed the evoluti Figure 3. (a, b) Time variations of the equatorial zonal wind at (a) 1 hPa and (b) 0.1 hPa for SABER, MLS, and each reanalysis. Figure 3. (a, b) Time variations of the equatorial zonal wind at (a) 1 hPa and (b) 0.1 hPa for SABER, MLS, and each reanalysis. Figure 3. Y. Kawatani et al.: Representation of the equatorial stratopause semiannual oscillation Time–height cross section of zonal mean annual mean standard deviation of (a, d) zonal wind and (b, e) temperature among (a, b)
five reanalyses (ERA-I ERA5 JRA-55 MERRA MERRA-2) and (d e) between MERRA and MERRA-2 over the Equator from 1980 to Figure 4. Time–height cross section of zonal mean annual mean standard deviation of (a, d) zonal wind and (b, e) temperature among (a, b)
five reanalyses (ERA-I, ERA5, JRA-55, MERRA, MERRA-2) and (d, e) between MERRA and MERRA-2 over the Equator from 1980 to
2015. The 3-year running mean values are plotted in each case. The color intervals are 1, 1.5, 2, 3, 5, 7, 9, and 11 ms−1 for zonal wind and
0.2, 0.4 0.6, 0.8, 1, 1.5, 2, and 2.5 K for temperature. (c, f) The linear regression trends in the standard deviation (in percent per year) over
time with height in (blue) zonal wind and (red) temperature (c) among five reanalyses and (f) between MERRAs. Filled circles indicate that
the trend at that level is different from zero with a statistical significance P < 0.05. Error bars denote P < 0.05 confidence intervals in the
trend estimates. Next, we explore spatial variation of SD among reanaly-
ses. Figure 5 displays the latitude–height distributions of the
zonal means of the zonal wind and temperature SD among
six reanalyses averaged from 1980 to 2010, as well as the
SD between MERRA and MERRA-2. In the SD among the
six reanalyses (Fig. 5a and b), the largest values are on the
Equator and increase from the upper troposphere to the up-
per stratosphere so that the largest values occupy a wedge-
shaped region shown clearly in the Fig. 5. The temperature
SD shows an opposite structure, having a minimum SD in the
lower latitudes and becoming larger at higher latitudes. Near
the Equator small differences in temperature (say between
two reanalysis datasets) can be expected to result in large
differences in the thermal wind shear. The zonal wind and
temperature SDs between MERRA and MERRA-2 (Fig. 5c
and d) show qualitatively similar structures. The local max-
ima of zonal wind SD around 50◦S and 50◦N at 0.1–0.3 hPa
arise from the different shape of the mesospheric jets be-
tween MERRA and MERRA-2 (not shown). This may result
from the inclusion of parameterized non-orographic gravity wave drag in MERRA-2. Y. Kawatani et al.: Representation of the equatorial stratopause semiannual oscillation (a, b) Time variations of the equatorial zonal wind at (a) 1 hPa and (b) 0.1 hPa for SABER, MLS, and each reanalysis. creased. Kawatani et al. (2016) also showed the evolution
of the number of available monthly mean near-equatorial ra-
diosonde observations at 10–70 hPa. The number of available
radiosonde observations generally increased with time at all
levels, which also likely contributed to the decreasing trend
of SD among reanalyses from 70 to 10 hPa. The SDs between
MERRAs show a similar decrease with time, although time
variations of both zonal wind and temperature SDs above
1 hPa are more complicated. were available. After May 1998, data from the Advanced Mi-
crowave Sounding Unit (AMSU) became available. Satellite
radiance data will presumably affect the assimilated temper-
atures in the stratosphere but will also have considerable in-
fluence on the wind (see Iida et al., 2014). Near the Equator
thermal wind shears are particularly sensitive to small errors
in observed temperatures. After 2000, the amount of avail-
able satellite data increased greatly (e.g., Kobayashi et al.,
2015), and the contributions of satellite radiance data to bet-
ter representation of the tropical winds also presumably in- Atmos. Chem. Phys., 20, 9115–9133, 2020 https://doi.org/10.5194/acp-20-9115-2020 Y. Kawatani et al.: Representation of the equatorial stratopause semiannual oscillation 9122 p
q
p
Figure 4. Time–height cross section of zonal mean annual mean standard deviation of (a, d) zonal wind and (b, e) temperature among (a, b)
five reanalyses (ERA-I, ERA5, JRA-55, MERRA, MERRA-2) and (d, e) between MERRA and MERRA-2 over the Equator from 1980 to
2015. The 3-year running mean values are plotted in each case. The color intervals are 1, 1.5, 2, 3, 5, 7, 9, and 11 ms−1 for zonal wind and
0.2, 0.4 0.6, 0.8, 1, 1.5, 2, and 2.5 K for temperature. (c, f) The linear regression trends in the standard deviation (in percent per year) over
time with height in (blue) zonal wind and (red) temperature (c) among five reanalyses and (f) between MERRAs. Filled circles indicate that
the trend at that level is different from zero with a statistical significance P < 0.05. Error bars denote P < 0.05 confidence intervals in the
trend estimates. Figure 4. 4
Similarities and differences of the semiannual
oscillation among the reanalyses duced SD (negative SD anomaly in Fig. 5c) extends up to
∼3 hPa. This suggests that the influence of in situ obser-
vations near the Equator on the reanalyses extends to alti-
tudes considerably above the actual observation heights. At
1 hPa, the SD in the Eastern Hemisphere is slightly smaller
than that in the Western Hemisphere. Another region of small
SD over South America around 50◦W does not extend as
high as it does over the Singapore, because the observational
density at 10 hPa is significantly lower over South Amer-
ica, and the higher density is limited at 50–70 hPa (Fig. 5 in
Kawatani et al., 2016). The zonal wind SD between MERRA
and MERRA-2 (Fig. 6b and d) also show a similar struc-
ture; however, the reason for the sudden increase of the zonal
asymmetry of the SD at ∼0.5 hPa is unclear. duced SD (negative SD anomaly in Fig. 5c) extends up to
∼3 hPa. This suggests that the influence of in situ obser-
vations near the Equator on the reanalyses extends to alti-
tudes considerably above the actual observation heights. At
1 hPa, the SD in the Eastern Hemisphere is slightly smaller
than that in the Western Hemisphere. Another region of small
SD over South America around 50◦W does not extend as
high as it does over the Singapore, because the observational
density at 10 hPa is significantly lower over South Amer-
ica, and the higher density is limited at 50–70 hPa (Fig. 5 in
Kawatani et al., 2016). The zonal wind SD between MERRA
and MERRA-2 (Fig. 6b and d) also show a similar struc-
ture; however, the reason for the sudden increase of the zonal
asymmetry of the SD at ∼0.5 hPa is unclear. In this section, we focus on the similarities and differences
of the seasonal cycle (dominated of course at low latitudes
by the semiannual component) among reanalyses. Figure 7a
shows the long-term mean annual cycle of zonal mean equa-
torial zonal wind at 1 hPa, as derived from SABER and MLS
observations and in each reanalysis. Some differences are
seen between the winds based on SABER and MLS, e.g.,
stronger westerly maxima (around April and October) and
weaker easterly maxima (around January and July) in the
MLS-derived winds compared to the SABER-derived winds
(also consistent with the long-term annual mean equatorial
zonal wind shown in Fig. 2a). Y. Kawatani et al.: Representation of the equatorial stratopause semiannual oscillation 9123 Figure 5. Latitude–height cross sections of climatological zonal mean standard deviation of (a, c) zonal wind and (b, d) temperature (a, b)
among six reanalyses and (c, d) between MERRA and MERRA-2. The color intervals are 0.5, 1, 2, 3, 4, 5, 6, 7, 8 ms−1 and K for zonal
wind and temperature, respectively. Shading indicates values larger than 0.5 ms−1 and 0.5 K. Figure 5. Latitude–height cross sections of climatological zonal mean standard deviation of (a, c) zonal wind and (b, d) temperature (a, b)
among six reanalyses and (c, d) between MERRA and MERRA-2. The color intervals are 0.5, 1, 2, 3, 4, 5, 6, 7, 8 ms−1 and K for zonal
wind and temperature, respectively. Shading indicates values larger than 0.5 ms−1 and 0.5 K. Y. Kawatani et al.: Representation of the equatorial stratopause semiannual oscillation The observational constraints in the
lower mesosphere are much weaker and the representation of
the zonal winds may depend strongly on the model configu-
ration used in each reanalysis. Figure 6a shows the horizontal distributions of the zonal
wind SD among six reanalyses at 1 hPa. The SD shows
a fairly zonally uniform structure. Kawatani et al. (2016)
showed the zonal wind SD from 70 to 10 hPa, and noted that
the SD becomes more zonally uniform as height increases. The 0.1 hPa SD between the MERRAs also shows a fairly
zonally uniform structure (Fig. 6b). Figure 6c and d display
the longitude–height cross section of zonal wind SD after
the zonal mean of the SD is subtracted. This shows that there
actually are some systematic zonal variations in the SD. No-
tably positive SD anomalies are found in the central Pacific,in
the lower to middle stratosphere, where in situ observations
are few (Fig. 5 of Kawatani et al., 2016). The smallest SD
is seen near Singapore (1.4◦N, 104◦E), where high-quality
radiosonde observations up to 10 hPa are consistently avail-
able (Naujokat, 1986). It is interesting that the region of re- Atmos. Chem. Phys., 20, 9115–9133, 2020 https://doi.org/10.5194/acp-20-9115-2020 Y. Kawatani et al.: Representation of the equatorial stratopause semiannual oscillation Atmos. Chem. Phys., 20, 9115–9133, 2020 4
Similarities and differences of the semiannual
oscillation among the reanalyses The reanalysis zonal winds
shown in Fig. 7a in most months are within ∼10 ms−1 of
one of the satellite-derived values, with the exception of
MERRA-2, which has clear westerly biases in all months
compared to satellite-derived winds and actually displays
westerlies for the climatological values in July (i.e., during Atmos. Chem. Phys., 20, 9115–9133, 2020 https://doi.org/10.5194/acp-20-9115-2020 Y. Kawatani et al.: Representation of the equatorial stratopause semiannual oscillation 9124 Figure 6. (a, b) Horizontal maps of climatological zonal wind standard deviation (a) among the six reanalyses at 1 hPa and (b) between
MERRA and MERRA-2 at 0.1 hPa. (c, d) Longitude–height cross section of the anomaly of the zonal wind standard deviation from its zonal
mean over the Equator (c) among the six reanalyses and (d) between MERRA and MERRA-2. The color intervals are 1 ms−1 for (a, b) and
0.2 ms−1 for (c, d). Figure 6. (a, b) Horizontal maps of climatological zonal wind standard deviation (a) among the six reanalyses at 1 hPa and (b) between
MERRA and MERRA-2 at 0.1 hPa. (c, d) Longitude–height cross section of the anomaly of the zonal wind standard deviation from its zonal
mean over the Equator (c) among the six reanalyses and (d) between MERRA and MERRA-2. The color intervals are 1 ms−1 for (a, b) and
0.2 ms−1 for (c, d). the second easterly phase of the SAO, as shown in other
datasets). of JRA-55C are included again to compare it to JRA-55). The
patterns are similar between satellite-derived winds based on
MLS and SABER, as already shown in Smith et al. (2017,
their Fig. 5). The easterlies extend into the tropics from the
summer hemisphere during the solstices; the westerlies exist
at all latitudes during the equinox season. This characteristic
is qualitatively represented by all reanalyses. MERRA-2 fea-
tures significantly stronger equatorial westerlies around both
equinoxes. This can be explained by the fact that the westerly
phase of the SAO is believed to be driven mainly by the atmo-
spheric waves (Dunkerton, 1982), and the dynamical model
used in MERRA-2 includes quite strong parameterized mo-
mentum fluxes from non-orographic gravity waves (Fig. 3 of
Molod et al., 2015). In contrast, JRA-55 and NCEP-CFSR
represent weaker westerlies during equinoxes. Y. Kawatani et al.: Representation of the equatorial stratopause semiannual oscillation Figure 7. (a) Climatological mean annual cycle of zonal mean zonal
wind over the Equator at 1 hPa for SABER, MLS, and each reanal-
ysis; (b) cycle-to-cycle variability of equatorial zonal mean zonal
wind calculated as the interannual standard deviation in each calen-
dar month. Climatology and standard deviation are calculated from
1980 to 2010 in the reanalyses, from 2002 to 2016 in SABER, and
from 2005 to 2016 in MLS. At 0.1 hPa, both the SABER- and MLS-derived equatorial
zonal winds (Fig. 9a and b) are westerly throughout the year,
while both MERRA and MERRA-2 show easterly phases,
although the timing of easterly appearance is different (De-
cember and June for MERRA and April and October for
MERRA-2). ERA5 data above 1 hPa are not currently avail-
able publicly, but Shepherd et al. (2018) did have access to
ERA5 at higher levels for the period 2008–2017, and they
note that ERA5 has westerlies throughout the year in the
long-term mean wind field at 0.1 hPa averaged over 5◦S–
5◦N, which is more similar to the SABER- and MLS-derived
winds and very different from ERA-I (Fig. 9c). Figure 10 is the same as Fig. 9, except that long-term an-
nual mean climatology (i.e., the mean zonal wind in Fig. 2a)
has been removed at each level. In this figure, the SAO east-
erly and westerly phases are more clearly visible. A stronger
SAO easterly in the first cycle compared to the second cycle
is seen in both observations and all reanalyses. Differences
among reanalyses are still obvious in this figure. To investigate the SAO in more detail, SAO components
were extracted from each of the datasets considered here and
then the SAO amplitude was calculated as follows: Figure 7. (a) Climatological mean annual cycle of zonal mean zonal
wind over the Equator at 1 hPa for SABER, MLS, and each reanal-
ysis; (b) cycle-to-cycle variability of equatorial zonal mean zonal
wind calculated as the interannual standard deviation in each calen-
dar month. Climatology and standard deviation are calculated from
1980 to 2010 in the reanalyses, from 2002 to 2016 in SABER, and
from 2005 to 2016 in MLS. Y. Kawatani et al.: Representation of the equatorial stratopause semiannual oscillation SAO amplitude =
√
2 ×
v
u
u
t
N
X
i
u′
i
2/N,
(2) SAO amplitude =
√
2 ×
v
u
u
t
N
X
i
u′
i
2/N,
(2) (2) where u′ is the monthly mean zonal wind of the filtered SAO
component, and the sum is over all the months in the time
series for each dataset. Figure 11 shows the latitude–height cross section of SAO
amplitude in zonal wind. The observed SAO amplitude has
its maximum off the Equator in the Southern Hemisphere,
around 7.5◦S in the SABER-derived dataset and 12◦S in the
MLS-derived dataset. These results are in basic agreement
with the earlier study of Hopkins (1975), who analyzed the
SAO amplitude determined from roughly 10 years of rock-
etsonde data at several stations and concluded that the SAO
amplitude near the stratopause peaked around 10–15◦S. The
SAO amplitude from the MLS-derived winds has a more pro-
nounced asymmetry between the Northern Hemisphere and
Southern Hemisphere, notably with the region of large am-
plitude near the stratopause appearing thinner on the northern
side of the Equator, again reminiscent of the pattern in the
Hopkins (1975) result. Note again that the near-equatorial
winds in our satellite-derived datasets were estimated by cu-
bic spline interpolation of the balance winds at and poleward
of latitude ±8◦for SABER and ±6◦for MLS (Smith et al.,
2017) and that this must introduce uncertainty in the wind
estimates near the Equator in each case. over the Equator in each of the datasets. The satellite-derived
winds from SABER and MLS (Fig. 9a and b) agree fairly
well, particularly below about 0.3 hPa and these results agree
reasonably well with the time–height section of the SAO de-
termined from rocketsonde observations (Garcia et al., 1997;
Smith et al., 2017). The equinoctial maxima of the westerly
peaks are above 0.1 hPa, while the easterly wind peaks are
at ∼1 hPa. The satellite-derived winds clearly show that the
SAO cycle in the first half of the year (boreal winter and
spring) is stronger than that in the second half (austral win-
ter and spring), a feature that has been apparent since the
earliest rocketsonde studies of the SAO (Reed, 1966). All of
the reanalysis zonal wind datasets differ considerably from
the satellite-derived winds, and the reanalyses differ signifi-
cantly among themselves as well. 4
Similarities and differences of the semiannual
oscillation among the reanalyses Comparing
JRA-55 to JRA-55C, both easterlies and westerlies are sig-
nificantly weaker in JRA-55C, once again demonstrating the
significant contribution of satellite temperature observations
to the representation of a realistically strong SAO. Figure 7b shows a measure of the interannual variability
of equatorial zonal mean zonal wind in observations and re-
analysis. This variability is characterized by the SD for each
calendar month. For example, January SD in each reanalysis
is calculated from the monthly mean zonal wind data from
1980 to 2010 (31 data points). Note that the analyzed peri-
ods for the satellite-derived winds are shorter than those of
the reanalyses (after January 2002 and 2005 for MLS and
SABER, respectively, until December 2016); reanalyses gen-
erally have a larger SD compared to satellite-derived winds. The SD in MERRA-2 varies significantly among months, be-
ing larger during the easterly phase of the SAO than during
the westerly phase. The interannual SD of ERA-I, ERA5,
JRA-55, and NCEP-CFSR show little variability through the
year, as do the satellite-derived winds from both SABER and
MLS. Figure 8 shows the time–latitude cross sections of the cli-
matological annual cycle of the zonal mean zonal wind at
1 hPa for individual datasets (in the rest of this section, results Figure 9 shows time–height cross sections of the clima-
tological mean annual cycle of the zonal mean zonal wind Atmos. Chem. Phys., 20, 9115–9133, 2020 https://doi.org/10.5194/acp-20-9115-2020 9125 : Representation of the equatorial stratopause semiannual oscillation Y. Kawatani et al.: Representation of the equatorial stratopause semiannual oscillation Atmos. Chem. Phys., 20, 9115–9133, 2020 Y. Kawatani et al.: Representation of the equatorial stratopause semiannual oscillation The SAO descending west-
erlies in the SABER- and MLS-derived data penetrate down
to ∼5–7 hPa in the first cycle of the year and ∼3 hPa in
the second cycle. The westerlies do not penetrate quite as
far down in ERA-I, ERA5, JRA-55, JRA-55C, and NCEP-
CFSR. This westerly penetration seen in the satellite data is
reasonably well represented in MERRA, while MERRA-2
displays an even further downward penetration of the wester-
lies. MERRA-2 is also the only dataset considered that does
not display easterly mean winds at 1 hPa in July and August. The results for the SAO zonal wind amplitude for each
of the reanalyses (Fig. 11c–i) show the overall pattern of
a peak near the low-latitude stratopause, but the details dif-
fer quite substantially among the different datasets. Notably,
the degree of interhemispheric asymmetry differs: the max-
imum latitude of the SAO amplitude at 1 hPa is at 3◦S https://doi.org/10.5194/acp-20-9115-2020 Atmos. Chem. Phys., 20, 9115–9133, 2020 Y. Kawatani et al.: Representation of the equatorial stratopause semiannual oscillation Y. Kawatani et al.: Representation of the equatorial stratopause semiannual oscillation 9126 Figure 8. Time–latitude sections of climatological mean annual cycle of the zonal mean zonal wind for (a) SABER, (b) MLS, and (c–i) each
reanalysis at 1 hPa. The contour intervals are 10 ms−1. Climatology is calculated from 1980 to 2010 in the reanalyses, from 2002 to 2016 in
SABER, and from 2005 to 2016 in MLS. Figure 8. Time–latitude sections of climatological mean annual cycle of the zonal mean zonal wind for (a) SABER, (b) MLS, and (c–i) each
reanalysis at 1 hPa. The contour intervals are 10 ms−1. Climatology is calculated from 1980 to 2010 in the reanalyses, from 2002 to 2016 in
SABER, and from 2005 to 2016 in MLS. in ERA-I, 4.5◦S in ERA5, 12◦S in JRA-55, 3◦S in JRA-
55C, 4.5◦S in MERRA, 9.5◦S in MERRA-2, and 13.5◦S in
NCEP-CFSR. The peak SAO wind amplitude is larger than
in the satellite-derived datasets (Fig. 11a and b) in ERA-I,
ERA5, and MERRA but smaller in NCEP-CFSR. The SAO
peak amplitude is much weaker and occurs at a lower alti-
tude in the JRA-55C vs. the JRA-55 reanalyses. Poleward of
20◦in both hemispheres, the SAO amplitude becomes larger
with latitude (e.g., from 20◦S to 30◦S and/or 20◦to 30◦N,
from 0.3 to 1 hPa) in both MERRA and MERRA-2 (Fig. Y. Kawatani et al.: Representation of the equatorial stratopause semiannual oscillation Time–height sections of climatological mean annual cycle of the zonal mean zonal wind over the Equator for (a) SABER, (b
MLS, and (c–i) each reanalysis. Climatology is calculated from 1980 to 2010 in the reanalyses, from 2002 to 2016 in SABER, and from
2005 to 2016 in MLS. The contour intervals are 5 ms−1. Figure 9. Time–height sections of climatological mean annual cycle of the zonal mean zonal wind over the Equator for (a) SABER, (b)
MLS, and (c–i) each reanalysis. Climatology is calculated from 1980 to 2010 in the reanalyses, from 2002 to 2016 in SABER, and from
2005 to 2016 in MLS. The contour intervals are 5 ms−1. 2 in 1980–2003 and in 2005–2016. It is evident that over
∼2–0.5 hPa the MERRA-2 SAO amplitude over the Equator
became closer to the SABER and MLS observations after the
MLS temperature was assimilated, although the significantly
smaller amplitude at 0.1 hPa is not improved. Notably, the
MERRA-2 representation of the SAO amplitude near 0.1 hPa
and 40◦S and 40◦N improves in the period with MLS data
included in the assimilation. reanalyses and is quite weak in the ERA-I reanalyses. These
discrepancies may be attributable to the sponge layers with
strong numerical dumping of wave components above 1 hPa
for ERA-I and JRA-55. The large SAO amplitudes near 0.1
and 40◦S and 40◦N have no counterpart in any of the reanal-
ysis datasets except possibly MERRA and MERRA-2 near
40◦S. The amplitudes shown for the reanalysis datasets in Fig. 12
are calculated using all the data from January 1980 to De-
cember 2010. Here, it should be noted that MERRA-2 is the
only reanalysis that assimilates the MLS temperature above
5 hPa. Thus, some improvement is expected for MERRA-2
in the SAO temperature field after the MLS data were assim-
ilated starting in August 2004. Figure 13a and b shows the
same quantity as Fig. 12 but for MERRA-2 in 1980–2003
(i.e., pre-MLS years) and in 2005–2016 (i.e., when the MLS
data are included in the assimilation over each entire year). Figure 13c shows vertical profiles of the temperature SAO
amplitude over the Equator for SABER, MLS, and MERRA- Y. Kawatani et al.: Representation of the equatorial stratopause semiannual oscillation 11g
and h), features that find no support in the corresponding re-
sult from the satellite-derived datasets (Fig. 11a and b). ture over the Equator is found at 2–3 hPa, with another peak
at 0.1 hPa in both SABER and MLS. Large SAO compo-
nents are also found at 40◦S and 40◦N at 0.1 hPa in both
the SABER and MLS observations. The temperature SAO amplitude at 2 hPa over the Equator
is approximately 4.3 K (SABER), 4.2 K (MLS), 5.6 K (ERA-
I), 4.7 K (ERA5), 4.0 K (JRA-55), 2.2 K (JRA-55C), 4.2 K
(MERRA), 3.9 K (MERRA-2), and 2.9 K (NCEP-CFSR). ERA-I and ERA5 overestimated the temperature SAO am-
plitude compared to the satellite observations, while the
amplitude is significantly underestimated in NCEP-CFSR. The much weaker SAO apparent in JRA-55C than JRA-55
(Fig. 11f vs. Fig. 11e) again demonstrates the key role as-
similated satellite data play in defining the SAO in reanaly-
ses. For the second amplitude maximum over the Equator at
0.1 hPa, MERRA has a larger peak value and meridionally
wider structure compared with the amplitude computed from
the satellite datasets, while MERRA-2 significantly under-
estimates this maximum. The 0.1 hPa equatorial maximum
seen in the satellite datasets is not apparent in the JRA-55 Figure 12 shows the same amplitude estimate as in Fig. 11,
but for the SAO in temperature. In contrast to the zonal wind,
the temperature SAO amplitude is equatorially centered and
appears to be distributed more symmetrically about the Equa-
tor. Different latitudinal gradients of the temperature, which
are larger in the Southern Hemisphere, result in a southward
shift of the maximum zonal wind amplitude in the satellite-
derived datasets. The maximum SAO amplitude of tempera- Atmos. Chem. Phys., 20, 9115–9133, 2020 https://doi.org/10.5194/acp-20-9115-2020 Y. Kawatani et al.: Representation of the equatorial stratopause semiannual oscillation Y. Kawatani et al.: Representation of the equatorial stratopause semiannual oscillation 9127 p
q
p Figure 9. Time–height sections of climatological mean annual cycle of the zonal mean zonal wind over the Equator for (a) SABER, (b)
MLS, and (c–i) each reanalysis. Climatology is calculated from 1980 to 2010 in the reanalyses, from 2002 to 2016 in SABER, and from
2005 to 2016 in MLS. The contour intervals are 5 ms−1. Figure 9. 5
Summary and concluding remarks The systematic observation of the equatorial middle atmo-
sphere began with balloon-borne radiosondes in the 1950s,
rocketsondes and radars in the 1960s, and satellite radiance
measurements in the late 1970s. These observations turned
up interesting and unexpected phenomena, notably the oc-
currence of well-defined, very low-frequency oscillations of
the zonal mean circulation at various altitudes: the QBO in
the lower and middle stratosphere and the SAO with large https://doi.org/10.5194/acp-20-9115-2020 Atmos. Chem. Phys., 20, 9115–9133, 2020 Y. Kawatani et al.: Representation of the equatorial stratopause semiannual oscillation Y. Kawatani et al.: Representation of the equatorial stratopause semiannual oscillation 9128 Figure 10. The same as Fig. 9 but with annual mean climatological zonal winds (shown in Fig. 2a) removed at each level. Figure 10. The same as Fig. 9 but with annual mean climatological zonal winds (shown in Fig. 2a) removed at each level. among the reanalyses was largest on the Equator and grew
with height (at least up to 10 hPa). amplitude in the stratopause region and then again in the
mesopause region. The tropical middle atmosphere has its
own unique dynamics and is known to be particularly chal-
lenging for realistic simulation by comprehensive general
circulation models (e.g., Butchart et al., 2018). Global re-
analyses have proven to be very valuable in efforts to un-
derstand the dynamics of the atmosphere and the climate
system, but the accurate representation of the atmospheric
flow in gridded analyses is particularly challenging for the
tropical middle atmosphere. As discussed in the introduction
of Kawatani et al. (2016), the initial attempts in the 1980s
and 1990s at global analyses based on variational assimila-
tion into dynamical model integrations produced results with
very poor representation of the tropical stratosphere. More
modern reanalysis datasets have produced much more satis-
factory results, at least in the lower and middle stratosphere
where a quite realistic QBO is apparent in the analyzed zonal
wind and temperature fields. Kawatani et al. (2016) com-
pared the monthly mean zonal wind field up to 10 hPa in
several reanalysis datasets and found that the disagreement In the present paper we assessed the representation of the
zonal wind and temperature fields near the Equator in the
region 10–0.1 hPa, where the dominant feature is the near-
stratopause SAO. One complication of examining this re-
gion is the nearly complete lack of relevant in situ mea-
surements to compare with. 5
Summary and concluding remarks While the SAO was first discov-
ered in rocketsonde wind measurements, Baldwin and Gray
(2005) found that rocketsonde data at stations within 10◦of
the Equator that were frequent enough to produce reasonable
monthly mean zonal winds were only available from 1965 to
1983. There is very little overlap with the reanalysis datasets
we evaluated, as our interest is confined to reanalyses after
1979 when satellite radiometer data are included in the as-
similations. In our project we compared, where possible, the
reanalyses to observed limb-sounding radiometer data from
SABER and MLS, using both the temperature retrievals and
the analysis of Smith et al. (2017), who applied a dynamical
balance relation to compute zonal winds from the geopoten-
tials derived from the SABER and MLS temperatures. Atmos. Chem. Phys., 20, 9115–9133, 2020 https://doi.org/10.5194/acp-20-9115-2020 Y. Kawatani et al.: Representation of the equatorial stratopause semiannual oscillation Y. Kawatani et al.: Representation of the equatorial stratopause semiannual oscillation
9129 .: Representation of the equatorial stratopause semiannual oscillation 9129 p
q
p
Figure 11. Latitude–height cross section of amplitude of the SAO in zonal wind. Color intervals are 2 ms−1. The amplitude is calculated
from 1980 to 2010 in the reanalyses, from 2002 to 2016 in SABER, and from 2005 to 2016 in MLS. Figure 11. Latitude–height cross section of amplitude of the SAO in zonal wind. Color intervals are 2 ms−1. The amplitude is calculated
from 1980 to 2010 in the reanalyses, from 2002 to 2016 in SABER, and from 2005 to 2016 in MLS. Figure 11. Latitude–height cross section of amplitude of the SAO in zonal wind. Color intervals are 2 ms−1. The amplitude is calculated
from 1980 to 2010 in the reanalyses, from 2002 to 2016 in SABER, and from 2005 to 2016 in MLS. Our study evaluated six major global atmospheric reanal-
ysis datasets that have been widely applied in studies of
the middle atmosphere (ERA-I, ERA5, JRA-55, MERRA,
MERRA-2, and NCEP-CFSR). All six reanalyses have
a good representation of the QBO in the equatorial lower
and middle stratosphere, and each displays a clear SAO cen-
tered near the stratopause. However, the differences among
reanalyses are much more substantial in the SAO region than
in the QBO-dominated lower and middle stratosphere. We
followed Kawatani et al. (2016) in characterizing the de-
gree of disagreement among the reanalyses using the SD of
the monthly mean zonal wind and temperature. Atmos. Chem. Phys., 20, 9115–9133, 2020 Y. Kawatani et al.: Representation of the equatorial stratopause semiannual oscillation Y. Kawatani et al.: Representation of the equatorial stratopause semiannual oscillation Y. Kawatani et al.: Representation of the equatorial stratopause semiannual oscillation 9130 Figure 12. The same as Fig. 11 but for the SAO in temperature. Color intervals are 0.4 K. Figure 12. The same as Fig. 11 but for the SAO in temperature. Color intervals are 0.4 K. Figure 12. The same as Fig. 11 but for the SAO in temperature. Color intervals are 0.4 K. to the observational density (i.e., significantly fewer observa-
tions in the polar regions). provement in the actual observations available for assimila-
tion. The temperature sounders on NOAA operational satel-
lites were upgraded significantly in May 1998, from the
SSU to the Advanced Microwave Sounding Unit (AMSU). Kawatani et al. (2016) showed an improvement around 1998
of the QBO representation up to 10 hPa and a reduction in
the wind and temperature SD. The present paper shows that
the long-term decreasing trend of the SD is seen up to at least
∼2 hPa. We also examined the representation of the upper strato-
spheric winds and temperatures off the Equator. All of the
reanalyses represent features of the time–latitude variation
of the zonal mean zonal wind at 1 hPa that agree broadly
with the satellite-derived datasets (Fig. 8). Notably, wester-
lies are present at all latitudes during the equinox seasons,
while easterlies are clear in the summer hemisphere and ex-
tend to the Equator around the solstices. In addition, the am-
plitude of the SAO in wind and temperature as a function of
height and latitude was computed from each dataset (Figs. 11
and 12). The zonal wind SAO amplitude in each case shows
a maximum near 1 hPa and at low latitudes, but the details
vary somewhat from dataset to dataset. All the results show
that the peak SAO is displaced off the Equator somewhat into
the Southern Hemisphere (in agreement with earlier studies
using rocketsonde measurements), but the extent of the dis-
placement and the values of the peak SAO amplitude vary
considerably among the reanalyses. The latitude–height cross sections of the SD display oppo-
site structures between zonal wind and temperature (Fig. 5). The zonal wind SD has a prominent equatorial maximum, in-
dicating the particularly challenging nature of the reanalysis
problem in the low-latitude stratosphere, where the Coriolis
parameter is small and in situ observations are sparse. 5
Summary and concluding remarks The zonal
wind SD has an equatorial maximum that increases with
height. Above 10 hPa even the long-term annual mean equa-
torial wind differs substantially among the reanalyses. Above
10 hPa there are also substantial differences in the equatorial
zonal wind fields between each of the reanalyses and the two
satellite-derived wind datasets. the SD characterizing the disagreement among reanalyses. Notably, this was tied to the availability of in situ balloon
measurements: the smallest SD was around the longitude of
Singapore, where consistently high-quality near-equatorial
radiosonde observations are available, and the largest val-
ues were around the eastern and central Pacific, where there
are virtually no stratospheric in situ wind observations close
to the Equator. Kawatani et al. (2016) found that the zonal
contrast is reduced as height increases, but that a significant
variation is still found at 10 hPa. In the present study, we
extended this analysis to the 0.1 hPa level. Interestingly the
near-Singapore minimum in SD is evident to at least ∼3 hPa
(Fig. 6), i.e., considerably higher than the usual ∼10 hPa
ceiling for weather balloons. Over the Equator we showed that there are overall long-
term trends for both the zonal wind and temperature SD to
become smaller and this is seen clearly from 70 to ∼2 hPa
(Fig. 4). Since within each reanalysis the dynamical atmo-
spheric model and the analysis procedures are fixed, im-
provement with time must be related to an increase or im- In their study of low-latitude winds below 10 hPa,
Kawatani et al. (2016) found substantial zonal variation in Atmos. Chem. Phys., 20, 9115–9133, 2020 https://doi.org/10.5194/acp-20-9115-2020 Y. Kawatani et al.: Representation of the equatorial stratopause semiannual oscillation The Japan Meteorological Agency created a complemen-
tary dataset, JRA-55C, by repeating the assimilation proce-
dure for JRA-55 without including any satellite observations. We compared several aspects of the SAO in the upper strato-
sphere between JRA-55 and JRA-55C. Very large differences
were found, and the stratopause SAO in JRA-55C is indeed
very weak (Figs. 11f and 12f). We conclude that the JRA-
55C reanalyses are not useful representations of the SAO in
the tropical upper stratosphere but that the comparison with
JRA-55 rather directly shows the great importance of satel-
lite radiance observations to defining the SAO in this region. Figure 13. (a, b) Latitude–height cross section of SAO temperature
amplitude for MERRA from (a) 1980 to 2003 and (b) 2005 to 2016. (c) Vertical profiles of SAO temperature amplitude over the Equa-
tor for SABER (solid black), MLS (dashed black), MERRA-2 from
1980–2003 (dashed red), and MERRA-2 from 2005 to 2016 (solid
red). To sum up, we have examined of the SAO in the tropi-
cal upper stratosphere as represented in sophisticated global
reanalysis datasets. The reanalyses are able to represent the
basic features of the stratopause SAO, but there are large un-
certainties even for the very basic fields (temperatures and
zonal winds) that we examined. The reanalyses in the tropi-
cal upper stratosphere must be regarded as much less reliable
than those for the region below 10 hPa where there are at least
some high-quality in situ observations of wind and tempera-
ture near the Equator each day. On the positive side, we found
that up to at least ∼2 hPa the reanalyses agreed better among
themselves as the satellite radiance data used in the assimila-
tions improved. Next generation reanalyses are now appear-
ing, such as ERA5 and JRA-3Q (Shinya Kobayashi, personal
communication, 2020), which extend up to the mesopause. We expect that a comparison of multiple reanalyses could be
extended up to at least 0.01 hPa in the near future. Data availability. References for each reanalysis dataset are: ERA-
I (Dee et al., 2011), ERA5 (Hersbach et al., 2019), JRA-55
(Kobayashi et al., 2015), MERRA (Rienecker et al., 2011),
MERRA-2 (Gelaro et al., 2017), and NCEP-CFSR (Saha et al.,
2010). Summary descriptions of the reanalysis datasets can also
be found via the S-RIP website (https://s-rip.ees.hokudai.ac.jp/
resources/links.html). Reanalysis and satellite data (Smith et al. Y. Kawatani et al.: Representation of the equatorial stratopause semiannual oscillation The
zonal wind SD in low latitudes becomes larger with height,
showing a wedge-shaped structure. In the mid-to-high lati-
tudes, the zonal wind SD becomes smaller. In contrast, the
temperature SD has minimum values over the Equator and
maximum values in the polar region, which might be related Atmos. Chem. Phys., 20, 9115–9133, 2020 https://doi.org/10.5194/acp-20-9115-2020 9131 Y. Kawatani et al.: Representation of the equatorial stratopause semiannual oscillation 9132 assimilation system, Q. J. Roy. Meteorol. Soc., 137, 553–597,
https://doi.org/10.1002/qj.828, 2011. assimilation system, Q. J. Roy. Meteorol. Soc., 137, 553–597,
https://doi.org/10.1002/qj.828, 2011. Acknowledgements. We express our gratitude to the scientific guid-
ance and sponsorship of the World Climate Research Programme
(WCRP) coordinated in the framework of Stratosphere-troposphere
Processes And their Role in Climate (SPARC) and the SPARC Re-
analysis Intercomparison Project (S-RIP). We acknowledge the re-
analysis centers for providing their support and data products. The
GFD-DENNOU Library and GrADS were used to draw the figures. Dunkerton, T. J.: On the role of the Kelvin waves in the westerly
phase of the semiannual zonal wind oscillation, J. Atmos. Sci.,
36, 32–41, 1979. Dunkerton, T. J.: Theory of the mesopause semiannual oscillation,
J. Atmos. Sci., 39, 2681–2690, https://doi.org/10.1175/1520-
0469(1982)039<2681:TOTMSO>2.0.CO;2, 1982. Durre, I., Russell, S. V., and David, B. W.: Overview of the inte-
grated global radiosonde archive, J. Climate, 19, 53–68, 2006. Financial support. Yoshio Kawatani was supported by Japan So-
ciety for Promotion of Science (JSPS) KAKENHI (grant nos. JP15KK0178, JP17K18816, and JP18H01286) and by the En-
vironment Research and Technology Development Fund (JP-
MEERF20192004) of the Environmental Restoration and Conser-
vation Agency of Japan. Masatomo Fujiwara was also supported
by JSPS KAKENHI JP18H01286. Toshihiko Hirooka and Yoshio
Kawatani were supported by JSPS KAKENHI JP16H04052 and
17H01159. Yoshio Kawatani and Kevin Hamilton were supported
by the Japan Agency for Marine-Earth Science and Technology
(JAMSTEC) through its sponsorship of research at the International
Pacific Research Center. The National Center for Atmospheric Re-
search is a major facility sponsored by the National Science Foun-
dation under cooperative agreement no. 1852977. Fujiwara, M., Wright, J. S., Manney, G. L., Gray, L. J., Anstey,
J., Birner, T., Davis, S., Gerber, E. P., Harvey, V. L., Hegglin,
M. I., Homeyer, C. R., Knox, J. A., Krüger, K., Lambert, A.,
Long, C. S., Martineau, P., Molod, A., Monge-Sanz, B. M., San-
tee, M. L., Tegtmeier, S., Chabrillat, S., Tan, D. G. H., Jack-
son, D. R., Polavarapu, S., Compo, G. P., Dragani, R., Ebisuzaki,
W., Harada, Y., Kobayashi, C., McCarty, W., Onogi, K., Paw-
son, S., Simmons, A., Wargan, K., Whitaker, J. S., and Zou,
C.-Z.: Introduction to the SPARC Reanalysis Intercomparison
Project (S-RIP) and overview of the reanalysis systems, At-
mos. Chem. Phys., 17, 1417–1452, https://doi.org/10.5194/acp-
17-1417-2017, 2017. Garcia, R. R., Dunkerton, T. J., Lieberman, R. S., and Vincent,
R. References Baldwin, M. P. and Gray, L. J.: Tropical stratospheric zonal
winds in ECMWF ERA-40 reanalysis, rocketsonde data,
and rawinsonde data, Geophys. Res. Lett., 32, L09806,
https://doi.org/10.1029/2004GL022328, 2005. Butchart, N., Anstey, J. A., Hamilton, K., Osprey, S., McLandress,
C., Bushell, A. C., Kawatani, Y., Kim, Y.-H., Lott, F., Scinocca,
J., Stockdale, T. N., Andrews, M., Bellprat, O., Braesicke, P.,
Cagnazzo, C., Chen, C.-C., Chun, H.-Y., Dobrynin, M., Garcia,
R. R., Garcia-Serrano, J., Gray, L. J., Holt, L., Kerzenmacher,
T., Naoe, H., Pohlmann, H., Richter, J. H., Scaife, A. A., Schen-
zinger, V., Serva, F., Versick, S., Watanabe, S., Yoshida, K., and
Yukimoto, S.: Overview of experiment design and comparison
of models participating in phase 1 of the SPARC Quasi-Biennial
Oscillation initiative (QBOi), Geosci. Model Dev., 11, 1009–
1032, https://doi.org/10.5194/gmd-11-1009-2018, 2018. Hamilton, K.: Dynamics of the stratospheric semiannual oscillation,
J. Meteorol. Soc. Jpn., 64, 227–244, 1986. Hamilton, K. and Mahlman, J. D.: General circulation model simu-
lation of the semiannual oscillation in the tropical middle atmo-
sphere, J. Atmos. Sci., 45, 3212–3235, 1988. Hersbach, H., Bell, B., Paul, B., András, H., Sabater, J. M., Nicolas, J., Radu, R., Schepers, D., Simmons, A.,
Soci, C., and Dee, D.: Global reanalysis: goodbye ERA-
Interim,
hello
ERA-5,
ECMWF
Newslett.,
159,
17–24,
https://doi.org/10.21957/vf291hehd7, 2019. Hirota, I.: Equatorial waves in the upper stratosphere and meso-
sphere in relation to the semiannual oscillation of the zonal wind,
J. Atmos. Sci., 35, 714–722, 1978. Coy, L., Wargan, K., Molod, A., McCarty, W., and Pawson, S: Struc-
ture and dynamics of the quasi-biennial oscillation in MERRA-2,
J. Climate, 29, 5339–5354, https://doi.org/10.1175/JCLI-D-15-
0809.1, 2016. Hirota, I.: Observational evidence of the semiannual oscillation in
the tropical middle atmosphere – A review, Pure Appl. Geophys.,
118, 217–238, https://doi.org/10.1007/BF01586452, 1980. Dee, D. P., Uppala, S. M., Simmons, A. J., Berrisford, P., Poli,
P., Kobayashi, S., Andrae, U., Balmaseda, M. A., Balsamo, G.,
Bauer, P., Bechtold, P., Beljaars, A. C. M., van de Berg, I., Biblot,
J., Bormann, N., Delsol, C., Dragani, R., Fuentes, M., Greer, A. J., Haimberger, L., Healy, S. B., Hersbach, H., Holm, E. V., Isak-
sen, L., Kallberg, P., Kohler, M., Matricardi, M., McNally, A. P.,
Mong-Sanz, B. M., Morcette, J.-J., Park, B.-K., Peubey, C., de
Rosnay, P., Tavolato, C., Thepaut, J. N., and Vitart, F.: The ERA-
Interim reanalysis: Configuration and performance of the data Hitchman, M. H. and Leovy, C. Y. Kawatani et al.: Representation of the equatorial stratopause semiannual oscillation A.: Climatology of the semiannual oscillation of the trop-
ical middle atmosphere, J. Geophys. Res., 102, 26019–26032,
https://doi.org/10.1029/97JD00207, 1997. Review statement. This paper was edited by Martin Dameris and
reviewed by Gloria Manney and one anonymous referee. Review statement. This paper was edited by Martin Dameris and
reviewed by Gloria Manney and one anonymous referee. Gelaro, R., McCarty, W., Suárez, M. J., Todling, R., Molod, A.,
Takacs, L., Randles, C. A., Darmenov, A., Bosilovich, M. G., Re-
ichle, R., Wargan, K., Coy, L., Cullather, R., Draper, C., Akella,
S., Buchard, V., Conaty, A., da Silva, A. M., Gu, W., Kim, G.-
K., Koster, R., Lucchesi, R., Merkova, D., Nielsen, J. E., Par-
tyka, G., Pawson, S., Putman, W., Rienecker, M., Schubert, S. D.,
Sienkiewicz, M., and Zhao, B.: The Modern-Era Retrospective
Analysis for Research and Applications, Version 2 (MERRA-2),
J. Climate, 30, 5419–5454, https://doi.org/10.1175/JCLI-D-16-
0758.1, 2017. Y. Kawatani et al.: Representation of the equatorial stratopause semiannual oscillation 2017) used in this study can be also inquired about by contacting
the authors. Data availability. References for each reanalysis dataset are: ERA-
I (Dee et al., 2011), ERA5 (Hersbach et al., 2019), JRA-55
(Kobayashi et al., 2015), MERRA (Rienecker et al., 2011),
MERRA-2 (Gelaro et al., 2017), and NCEP-CFSR (Saha et al.,
2010). Summary descriptions of the reanalysis datasets can also
be found via the S-RIP website (https://s-rip.ees.hokudai.ac.jp/
resources/links.html). Reanalysis and satellite data (Smith et al. 2017) used in this study can be also inquired about by contacting
the authors. Figure 13. (a, b) Latitude–height cross section of SAO temperature
amplitude for MERRA from (a) 1980 to 2003 and (b) 2005 to 2016. (c) Vertical profiles of SAO temperature amplitude over the Equa-
tor for SABER (solid black), MLS (dashed black), MERRA-2 from
1980–2003 (dashed red), and MERRA-2 from 2005 to 2016 (solid
red). Figure 13. (a, b) Latitude–height cross section of SAO temperature Author contributions. YK and TH designed the study. YK analyzed
the data. YK and KH wrote the manuscript. AS calculated the zonal
wind in SABER and MLS. MF investigated details of each reanaly-
sis. All authors reviewed and edited the paper. Four of the reanalyses considered here provide data val-
ues to the 0.1 hPa level, while two others (CFSR and ERA5)
have provided data only to the 1 hPa level. The reanalysis
fields above 1 hPa in the JRA-55 and ERA-I dataset have to
be regarded with some caution, however, as it is known that
the dynamical models used in these assimilations have strong
sponge layers with artificially strong damping above 1 hPa. In general, we find significant uncertainties in the reanalyses
in the 1–0.1 hPa layer. One example is apparent in Fig. 11,
where the amplitude of the zonal wind SAO at the Equator
and 0.1 hPa is about 2.5 times larger in MERRA than in the
ERA-I reanalyses. Competing interests. The authors declare that they have no conflict
of interest. Special issue statement. This article is part of the special is-
sue “The SPARC Reanalysis Intercomparison Project (S-RIP)
(ACP/ESSD inter-journal SI)”. It does not belong to a conference. https://doi.org/10.5194/acp-20-9115-2020 Atmos. Chem. Phys., 20, 9115–9133, 2020 Y. Kawatani et al.: Representation of the equatorial stratopause semiannual oscillation 9133 Naujokat, B.: An update of the observed quasi-biennial oscillation
of the stratospheric zonal wind over the tropics, J. Atmos. Sci.,
43, 1873–1877, 1986. Hopkins,
R. H.:
Evidence
of
polar–tropical
coupling
in
upper
stratospheric
zonal
wind
anomalies,
J. At-
mos. Sci.,
32,
712–719,
https://doi.org/10.1175/1520-
0469(1975)032<0712:EOPTCI>2.0.CO;2, 1975. Ray, E. A., Alexander, M. J., and Holton, J. R.: An analysis
of the structure and forcing of the equatorial semiannual os-
cillation in zonal wind, J. Geophys. Res., 103, 1759–1774,
https://doi.org/10.1029/97JD02679, 1998. Iida, C., Hirooka, T., and Eguchi, N.: Circulation changes in the
stratosphere and mesosphere during the stratospheric sudden
warming event in January 2009, J. Geophys. Res. Atmos., 119,
7104–7115, 2014. Reed, R. J.: The quasi-biennial oscillation of the atmosphere be-
tween 30 and 50 km over Ascension Island, J. Atmos. Sci., 22,
331–333, 1965. Jackson, D. R. and Gray, L.J.: Simulation of the semi-annual oscil-
lation of the equatorial middle atmosphere using the Extended
UGAMP General Circulation Model, Q. J. Roy. Meteorol. Soc.,
120, 1559–1588, 1994. Reed, R. J.: Zonal wind behavior in the equatorial stratosphere and
lower mesosphere, J. Geophys. Res., 71, 4223–4233, 1966. Kawatani, Y., Hamilton, K., Miyazaki, K., Fujiwara, M., and
Anstey, J. A.: Representation of the tropical stratospheric zonal
wind in global atmospheric reanalyses, Atmos. Chem. Phys., 16,
6681–6699, https://doi.org/10.5194/acp-16-6681-2016, 2016. Rienecker, M. M., Suarez, M. J., Gelaro, R., Todling, R., Bacmeis-
ter, J., Liu, E., Bosilovich, M. G., Schubert, S. D., Takacs, L.,
Kim, G-K., Bloom, S., Chen, J., Collins, D., Conaty, A., da Silva,
A., Gu, W., Joiner, J., Koster, R. D., Lucchesi, R., Molod, A.,
Owens, T., Pawson, S., Pegion, P., Redder, C. R., Reichle, R.,
Robertson, F. R., Ruddick, A. G., Sienkiewicz, M., and Woollen,
J.: MERRA: NASA’s Modern-Era Retrospective Analysis for
Research and Applications, J. Climate, 24, 3624–3648, 2011. Kobayashi, C., Endo, H., Ota, Y., Kobayashi, S., Onoda, H.,
Harada, Y., Onogi, K., and Kamahori, H.: Preliminary results
of the JRA-55C, an atmospheric reanalysis assimilating conven-
tional observations only, Sci. Online Lett. Atmos., 10, 78–82,
https://doi.org/10.2151/sola.2014-016, 2014. Robertson, F. R., Ruddick, A. G., Sienkiewicz, M., and Woollen, Saha, S., Nadiga, S., Thiaw, C.,Wang, J.,Wang,W., Zhang, Q., van
den Dool, H. M., Pan, H.-L., Moorthi, S., Behringer, D., Stokes,
D., Peña, M., Lord, S., White, G., Ebisuzaki, W., Peng, P., and
Xie, P: The NCEP Climate Forecast System Reanalysis, B. Am. Meteorol. Soc., 91, 1015–1057, 2010. References B.: Evolution of the zonal mean
state in the equatorial middle atmosphere during October 1978–
May 1979, J. Atmos. Sci., 43, 3159–3176, 1986. Holton, J. R. and Wehrbein, W. M.: A numerical model of the zonal
mean circulation of the middle atmosphere, Pure Appl. Geo-
phys., 118, 284–306, 1980. https://doi.org/10.5194/acp-20-9115-2020 Atmos. Chem. Phys., 20, 9115–9133, 2020 Y. Kawatani et al.: Representation of the equatorial stratopause semiannual oscillation Kobayashi, C., Ota, Y., Harada, Y., Ebita, A., Moriya, M., Hi-
rokatsu, O., Onogi, K., Kamahori, H., Kobayashi, C., Endo, H.,
Miyaoka, K. and Takahashi, K.: The JRA-55 Reanalysis: general
specifications and basic characteristics, J. Meteorol. Soc. Jpn.,
93, 5–48, 2015. Shepherd, T. G., Polichtchouk, I., Hogan, R. J., and Simmons, A. J.:
Report on Stratosphere Task Force, in ECMWF Technical Mem-
orandum, 824, June 2018, available at: https://www.ecmwf.int/
en/elibrary/18259-report-stratosphere-task-force
(last
access:
15 December 2019), 2018. Manney, G. L., Swinbank, R., Massie, S. T., Gelman, M. E., Miller,
A. J., Nagatani, R., O’Neill, A., Zurek. R. W.: Comparison of
U.K. Meteorological Office and U.S. National Meteorological
Center stratospheric analyses during northern and southern win-
ter, J. Geophys. Res., 101, 10311–10344, 1996. Smith, A. K., Garcia, R. R., Moss, A. C., and Mitchell, N. J.: The
semiannual oscillation of the tropical zonal wind in the middle at-
mosphere derived from satellite geopotential height retrievals, J. Atmos. Sci., 74, 2413–2425, https://doi.org/10.1175/JAS-D-17-
0067.1, 2017. Molod, A., Takacs, L., Suarez, M., and Bacmeister, J.: Development
of the GEOS-5 atmospheric general circulation model: evolution
from MERRA to MERRA2, Geosci. Model Dev., 8, 1339–1356,
https://doi.org/10.5194/gmd-8-1339-2015, 2015. NCDC:
https://www.ncdc.noaa.gov/data-access/weather-balloon/
integrated-global-radiosonde-archive, last access: 10 May 2016. Atmos. Chem. Phys., 20, 9115–9133, 2020 Atmos. Chem. Phys., 20, 9115–9133, 2020 https://doi.org/10.5194/acp-20-9115-2020
|
https://openalex.org/W3111323460
|
https://orbi.uliege.be/bitstream/2268/254017/1/energies-13-06628.pdf
|
English
| null |
Modelling the Integration of Residential Heat Demand and Demand Response in Power Systems with High Shares of Renewables
|
Energies
| 2,020
|
cc-by
| 10,280
|
In order to meet
this target, the integration of high shares of Renewable Energy Sources (RESs) in the energy system is
of primary importance. Nevertheless, the large deployment of variable renewable sources such as
wind and photovoltaic power will pose important challenges in terms of power management. For this
reason, increasing the system flexibility will be crucial to ensure the security of supply in future power
systems. This work investigates the flexibility potential obtainable from the diffusion of Demand
Response (DR) programmes applied to residential heating for different renewables penetration and
power system configuration scenarios. To that end, a bottom-up model for residential heat demand
and flexible electric heating systems (heat pumps and electric water heaters) is developed and directly
integrated into Dispa-SET, an existing unit commitment optimal dispatch model of the power system. The integrated model is calibrated for the case of Belgium and different simulations are performed
varying the penetration and type of residential heating technologies installed renewables capacity Keywords: demand response; buildings; flexibility; renewables; heating systems; heat pumps;
electric water heaters; energy modelling
Received: 18 November 2020; Accepted: 10 December 2020; Published: 15 December 2020 Abstract: The EU aims to become the world’s first climate-neutral continent by 2050. In order to meet
this target, the integration of high shares of Renewable Energy Sources (RESs) in the energy system is
of primary importance. Nevertheless, the large deployment of variable renewable sources such as
wind and photovoltaic power will pose important challenges in terms of power management. For this
reason, increasing the system flexibility will be crucial to ensure the security of supply in future power
systems. This work investigates the flexibility potential obtainable from the diffusion of Demand
Response (DR) programmes applied to residential heating for different renewables penetration and
power system configuration scenarios. To that end, a bottom-up model for residential heat demand
and flexible electric heating systems (heat pumps and electric water heaters) is developed and directly
integrated into Dispa-SET, an existing unit commitment optimal dispatch model of the power system. The integrated model is calibrated for the case of Belgium and different simulations are performed
varying the penetration and type of residential heating technologies, installed renewables capacity
and capacity mix. Results show that, at country level, operational cost could be reduced up to e35
million and curtailment up to 1 TWh per year with 1 million flexible electric heating systems installed. These benefits are significantly reduced when nuclear power plants (non-flexible) are replaced by
gas-fired units (flexible) and grow when more renewable capacity is added. Moreover, when the
number of flexible heating systems increases, a saturation effect of the flexibility is observed. Abstract: The EU aims to become the world’s first climate-neutral continent by 2050. In order to meet
this target, the integration of high shares of Renewable Energy Sources (RESs) in the energy system is
of primary importance. Nevertheless, the large deployment of variable renewable sources such as
wind and photovoltaic power will pose important challenges in terms of power management. For this
reason, increasing the system flexibility will be crucial to ensure the security of supply in future power
systems. This work investigates the flexibility potential obtainable from the diffusion of Demand
Response (DR) programmes applied to residential heating for different renewables penetration and
power system configuration scenarios. To that end, a bottom-up model for residential heat demand Abstract: The EU aims to become the world’s first climate-neutral continent by 2050. energies energies energies
Article
Modelling the Integration of Residential Heat
Demand and Demand Response in Power Systems
with High Shares of Renewables
Chiara Magni 1,2,*, Alessia Arteconi 1,2,3, Konstantinos Kavvadias 4 and Sylvain Quoilin 1,2,5
1
KU Leuven, Smart Energy Systems Research Unit Campus Geel, 2440 Geel, Belgium;
alessia.arteconi@kuleuven.be (A.A.); sylvain.quoilin@kuleuven.be (S.Q.)
2
EnergyVille, Thor Park, 3600 Genk, Belgium
3
Dipartimento di Ingegneria Industriale e Scienze Matematiche, Università Politecnica delle Marche,
60121 Ancona, Italy
4
Laboratory of Process Analysis and Design, NTUA National Technical University of Athens,
15780 Athens, Greece; kavadias@mail.ntua.gr
5
Faculty of Applied Sciences, University of Liège, 4000 Liège, Belgium
*
Correspondence: chiara.magni@kuleuven.be
Received: 18 November 2020; Accepted: 10 December 2020; Published: 15 December 2020
energies
Article
Modelling the Integration of Residential Heat
Demand and Demand Response in Power Systems
with High Shares of Renewables
Chiara Magni 1,2,*, Alessia Arteconi 1,2,3, Konstantinos Kavvadias 4 and Sylvain Quoilin 1,2,5
1
KU Leuven, Smart Energy Systems Research Unit Campus Geel, 2440 Geel, Belgium;
alessia.arteconi@kuleuven.be (A.A.); sylvain.quoilin@kuleuven.be (S.Q.)
2
EnergyVille, Thor Park, 3600 Genk, Belgium
3
Dipartimento di Ingegneria Industriale e Scienze Matematiche, Università Politecnica delle Marche,
60121 Ancona, Italy
4
Laboratory of Process Analysis and Design, NTUA National Technical University of Athens,
15780 Athens, Greece; kavadias@mail.ntua.gr
5
Faculty of Applied Sciences, University of Liège, 4000 Liège, Belgium
*
Correspondence: chiara.magni@kuleuven.be
Received: 18 November 2020; Accepted: 10 December 2020; Published: 15 December 2020
www.mdpi.com/journal/energies Modelling the Integration of Residential Heat
Demand and Demand Response in Power Systems
with High Shares of Renewables 3
Dipartimento di Ingegneria Industriale e Scienze Matematiche, Università Politecnica delle Marche
60121 Ancona, Italy 5
Faculty of Applied Sciences, University of Liège, 4000 Liège, Belgium 5
Faculty of Applied Sciences, University of Liège, 4 *
Correspondence: chiara.magni@kuleuven.be *
Correspondence: chiara.magni@kuleuven.be 1. Introduction As climate and environmental concern grows, many authorities have set ambitious targets for
emission reduction and renewable technologies integration in the power sector. In this context, with the
presentation of the European Green Deal [1] at the end of 2019, the European Commission announced
its ambition to be the first continent reaching carbon neutrality by 2050. The intermediate key-targets
for 2030 include 55% cuts in greenhouse gas emissions (from 1990 levels) and at least a 32% share
of renewables in the energy system [2], of which more than 50% is provided by non-dispatchable
intermittent sources such as solar and wind [3]. www.mdpi.com/journal/energies www.mdpi.com/journal/energies Energies 2020, 13, 6628; doi:10.3390/en13246628 Energies 2020, 13, 6628 2 of 19 While increasing the share of renewable sources in the power production mix has evident
advantages, their unpredictability and variability poses significant challenges in terms of supply
security, requiring additional flexibility for grid balancing. Traditionally, power systems’
flexibility requirements are fulfilled through flexible generators and large-scale storage facilities
(e.g., pumped hydro), but recently, thanks to digitalization and automation technologies, demand
response (DR) strategies are getting more and more attention. DR strategies have been studied extensively in the recent years. Different demand response
schemes have been proposed (direct load control, curtailable load, demand-side bidding, time-of-use
tariffs, peak pricing, real-time pricing) and are extensively described in [4]. Lynch et al. [5] prove that their diffusion could help mitigating the future challenges for grid
operators in terms of flexibility and renewables integration by reducing the total amount of generation
capacity investment required to ensure electricity system security. Dupont et al. [6] demonstrate DR benefits for the power system in terms of cost reduction,
higher reliability, and reduced carbon emission. Pina et al. [7], Moura et al. [8] and McPerson et al. [9]
assess the impact of DR on enabling the integration of high shares of intermittent renewable generation
in the power system. Eventually, Gils [10] presents an assessment of the theoretical DR potential
at European level considering the contribution of all consumers sectors (industrial, commercial and
residential). The results obtained show that the aggregated demand side flexibility could ensure a
minimum load reduction of 61 GW and a minimum load increase of 68 GW in every hour of the year. Among the variety of demand side sources available, a promising option is related to residential
heating technologies. 1. Introduction The household sector accounts for around 26% of the final energy demand
in EU28 and more than 70% of this share corresponds to the sole heat demand composed by space
heating (SH) and domestic hot water (DHW) [3]. Thanks to its high energy intensity and to the
expected substantial electrification, this sector has an important potential in terms of energy savings
and load shifting [11], thus enabling the integration of more intermittent renewables in the power grid. Nevertheless, the study of DR in the case of domestic heating demand cannot fail to consider the users’
thermal comfort which has to be ensured. This can be achieved through thermal storage technologies,
which allow the decoupling of heat demand from the electric demand. Such a disconnection between
electricity and heat load profile can be ensured in the case of the residential sector without the need for
investments in separate storage units thanks to the inherent thermal storage of the system (both in the
building envelope [12,13] and in the hot water tank [14]). Since the electrification of heating systems is expected to play a major role for the energy
transition [15], it is fundamental to investigate how DR large scale deployment will affect the
energy system. In this regard, some works in literature attempt to estimate the DR potential of residential heating
through a detailed representation of the demand side through state space equations or building
simulation tools. For example, Vivian et. al [16] aim to determine the peak shaving obtainable from a
pool of buildings provided with heat pumps, and Sperber et. al [17] estimate the shift load potential
of space heating for a future German building stock. In these two cases, the demand response is
only guided by an external input price signal, without a real representation of the characteristics and
dynamics of the energy system. However, as emphasized by Bruninx et al. [18] and Patteeuw et al. [19], the use of an integrated
model including both supply and demand side is paramount to assess the interaction between flexible
electric heating systems and the power system. One example is provided by Papaefthymiou et al. [20]:
detailed thermodynamic simulations are performed through the dynamic thermal simulation software
TRNSYS for a set of building archetypes representative of the German building stock. The model for
building thermal behaviour is then coupled to a high-level electricity market model. 1. Introduction Nevertheless, as underlined by Sperber et al. [17], the application of a detailed thermal simulation
model is too computationally expensive to efficiently extend the work to a higher number of cases and
simulations. As stated in the same work [17], a good compromise between accuracy and computational Energies 2020, 13, 6628 3 of 19 efficiency consists in the use of linear state space model for building thermodynamics. Linear state
space models are applied in literature by different authors as described hereafter. Hedegaard et al. [21] assess the effect of heat pumps deployment for the integration of high shares
of wind power in Denmark by applying an energy system model which optimizes both investments
and operational costs, though simplifying the model assuming a constant coefficient of performance
(COP) for heat pumps and neglecting solar transmission. Patteeuw et al. [19] present an integrated model with high detail at system level for short term
demand response of flexible electric heating systems. In this work, the optimization function minimizes
the overall operational cost of the system subjected to both electricity supply and heating systems
constraints. The integrated model is later applied by the same author in [22] to assess the potential
benefits offered by demand response with heat pumps in terms of reducing costs, peak shaving and
CO2 emissions reduction for the overall system. The focus is on the demand side, where it is shown that
CO2 abatement cost is strongly influenced by different factors at the building level. The supply-side
characteristics are kept unvaried. Arteconi et al. [23] perform a Belgian case study assessing the effect of different penetration of
demand response with electric heating systems coupled to thermal energy storage while varying the
renewable penetration on the generation side. Heat pumps and electric resistance heaters are coupled
with building envelope and hot water tank storage. The model minimizes the operational cost of the
system by combining a merit order optimization model of the electricity generation side with a detailed
representation of the demand side. On the supply side, the minimum and maximum capacities of
the generation units are taken into account, but ramping constraints, start-up costs and minimum on
and off- times are neglected. These assumptions might lead to unrealistic power plant operation and
thereby significantly affect the results, as stated in [19]. 1. Introduction Energies 2020, 13, 6628 4 of 19 The remainder of the manuscript is structured as follows: Section 2 describes the methodology
applied for the integrated model; in Section 3, simulations and results are presented and discussed. Finally, Section 4 concludes illustrating some final considerations. 2.1. Supply Side The supply side model is based on the unit commitment optimal dispatch model Dispa-SET [25]
that represents the short term operation of large-scale power systems with high level of detail. The model is expressed as a mixed-integer linear programming (MILP) optimization problem with
binary variables representing the commitment status of each unit δ. The MILP objective function
minimizes the total operational cost over the optimization period under the assumption of demand
inelastic to price signal. To that aim, the system is considered to be managed by a central operator with
full information on generation units characteristics, the transmission network and the demand at each
hour for each node. The model features are: minimum and maximum power for each unit, power plant ramping
limits, reserves up and down, minimum up/down times, load shedding, curtailment, power-to-heat,
pumped-hydro storage, non-dispatchable units (e.g. wind turbines, run-of-river, etc.), start-up costs
and ramping costs. The main constraints of the system are represented by the demand-supply balance and by limits
related to the operations of the power plants. In the following paragraphs the main variables and equations of the model are illustrated and
briefly explained. 1. System Cost 2. Methodology In this work, an integrated model for heat demand and demand response integration of residential
thermal units in the energy system is developed and applied to the case study of the Belgian power
system. To do that, the existing unit-commitment and optimal dispatch model Dispa-SET [25] is
coupled with a new demand side model designed for the aim of this work. In this chapter the existing
supply side model is reported (Section 2.1), followed by an accurate description of the demand side
model (Section 2.2); finally, the integration between the two is illustrated (Section 2.3). 1. Introduction In general, while different effective models have been developed to represent the demand side,
the majority of the studies fail to integrate these latter in realistic representation of large scale power
systems. Due to this, the interaction between distributed flexibility sources and energy systems at
regional or national level still remains unclear; in particular, without a detailed representation of the
supply side and its constraints it is not possible neither to evaluate how the energy mix and overall
system characteristics of a specific region affect the flexibility potential, neither to compare DR with
other flexibility sources. For these reasons, the objective of this work is twofold: first, to develop an integrated model able
to realistically represent both supply and demand constraints while maintaining enough computational
efficiency to perform a variety of simulations; second, to assess the impact on the energy system of the
spread of DR applied to heating technologies and comprehensively investigate how the results are
affected by the characteristics of the generation mix. The latter is of utmost importance for the case
study under investigation in this work, since Belgium announced its nuclear phase-out to take place
by the end of 2025. With a domestic electricity supply relying for more than 40% on nuclear energy in
the past few years [24], it is fundamental to understand how to ensure security of supply in the short
and long term with the national system undergoing such substantial change. Finally, DR is compared
in terms of performances to hydro pumped storage as a source of flexibility. To that end, a heat demand model is developed and directly integrated into an open-source unit
commitment and dispatch model described in [25]. The considered residential heat demand consists in
space heating demand and domestic hot water demand and is coupled to the power system through
flexible electric heating devices (heat pumps and domestic hot water heaters). The integrated model is parametrized for the Belgian current building stock and power system. g
p
g
g
p
y
In order to assess the potential benefits of the flexible heat demand, several simulations are
performed considering different flexible electric heating system penetrations and typologies. Moreover,
the installed renewable capacity and capacity mix is modified to investigate how the generation side
composition affects the previous results. Finally, flexible heating systems potential benefits for the grid
are compared with the ones offered by the existing pumped hydro storage units. 2.1.2. Demand-Related Constraints The main constraint for the system is the supply-demand balance, which has to be met for each
period i and zone n in the day-ahead market representation: ∑
u
(Pu,i · Locationu,n) + ∑
l
(Flowl,i · LineNodel,n)
= Demandn,i + ∑
r
StorageInputs,i · Locations,n
−ShedLoadn,i
(2) (2) −ShedLoadn,i
(2) where Pu,i is the power output of the unit u at period i, Flowl,i is the cross border flow related to
line l at the period i, Demandn,i is the total load for zone n at time step i and is given as an input to
the model. StorageInputs,i is the charging input for the storage unit s and ShedLoadn,i represents the
amount of load shedded to hour i (this variable is associated to a cost and is limited by an upper
constraint). Locationu,n is a binary variable which is equal to 1 only when the unit u is located in the
zone n, while LineNodel,n assume the value of +1 or -1 depending on the direction of the flow on line l
with respect to node n. Besides the supply-demand balance also the secondary and tertiary reserve requirements
(upwards and downwards), which are calculated in Dispa-SET through empirical formulations,
must be met for each node and time-step as well. This is ensured through the reservation of a certain
amount of the power plants available capacity. 2.1.1. System Cost The total system cost is the main optimization variable of the system and is defined as follows: SystemCosti = ∑
u,n u,n
h
CostStartUpu,i + CostShutDownu,i
+ CostFixedu · δu,i + CostVariableu,i · Pu,i
+ CostRampUpu,i + CostRampDownu,i
+ PriceTransimissioni,l · Flowi,l
+ CostLoadSheddingi,n · ShedLoadi,n
i
(1) h
CostStartUpu,i + CostShutDownu,i (1) with δu,i is the commitment status of the unit (the variable is equal to 1 if unit u is online at time step i,
0 if not) and ShedLoadn,i represents the amount of load shedded to hour i resulting from contractual
arrangements between generators and industrial sector consumers. with δu,i is the commitment status of the unit (the variable is equal to 1 if unit u is online at time step i,
0 if not) and ShedLoadn,i represents the amount of load shedded to hour i resulting from contractual
arrangements between generators and industrial sector consumers. The variable production costs are expressed in EUR/MWh and are determined by fuel and
emission prices corrected by the efficiency and the emission rate of the unit. The start-up and shut-down
costs are positive variables related to the number of startups/shutdowns for each power plant at every
time step. Ramping costs are related to the increase/decrease of power production between subsequent Energies 2020, 13, 6628 5 of 19 time steps for each power plant and are defined as positive variables for the system. Renewable units
are enforced committed when available. time steps for each power plant and are defined as positive variables for the system. Renewable units
are enforced committed when available. 2.2. Demand Side The demand side model aims to represent the residential electric demand for space heating (SH)
and domestic hot water (DHW) with a bottom up approach. The demand side model aims to represent the residential electric demand for space heating (SH)
and domestic hot water (DHW) with a bottom up approach. To that aim, the Belgian building stock is first analyzed and its thermal behaviour is modelled
through state space models for a certain number of building archetypes. Second, the demand for space
heating and domestic hot water is described by a set of comfort constraints (time series of set point
temperatures and DHW demand) representing different users types. Then, the electric demand needed
to satisfy the users’ needs is evaluated combining comfort constraints and thermophysical properties
of the buildings with models representing electric heating technologies. To this regard, two different
technologies are modelled: heat pumps, which are assumed to provide both SH and DHW, and electric
water heaters, which provide only DHW. Finally, depending on the number of buildings provided
with heat pumps and electric water heaters, the electric demands are aggregated. The demand side and supply side models are coupled through the electric demand of the heating
devices, which is computed considering both flexible and non flexible heating devices. In the base case, the power consumption of the heating devices is computed for each time
step according to an ON-OFF control strategy: the heating system goes ON at full load when the
temperature falls below its lower limit and stops when the temperature is in the defined acceptable
range again. When the the space heating and the hot water temperature are lower than their minimum
limit, priority is given to domestic hot water. Once the hourly load is determined, this latter is summed
to the hourly electricity demand as an input for the supply side model. In the flexible devices scenarios instead the power consumption is calculated by considering the
full flexibility potential of the the heating systems: the heat pump is allowed to work at partial load as
described in Section 2.2.5 and the functioning of the electric water heater storage tank is optimized. Moreover, a relaxation variable is introduced to add a certain tolerance (around 1 ◦C) to the ambient
temperature set point; in this way, the thermal inertia of the building and heating systems can be fully
exploited. 2.2. Demand Side In this case, the power consumption of the heating systems is an optimization variable for
the integrated demand and supply side model with the aim to minimize the total systems costs while
respecting the comfort constraints for the dwellers. 2.1.3. Power Plants and Storage-Related Constraints The power output of the units is limited by their must-run or stable generation level, by the
available capacity and by the ramping capabilities of the units as described by Equation (3). PowerMustRunu,i · δu,i
≤Pu,i ≤
α · PowerCapacityu · δu,i
(3) ≤Pu,i ≤ (3) α · PowerCapacityu · δu,i
(3) where α is the share of available capacity taking outages and time-dependent (renewable) generation
into account. The power output in a given period also depends on the ramping capabilities of the units. If the unit is shut down, the ramping capability is given by the maximum start up ramp, while if the
unit is online the limit is defined by the maximum ramp up rate. Pu,i ≤ Pu,i ≤ u,i ≤
Pu,i−1 + δu,i−1 · RampUpMaximumu
+ (1 −δu,i−1) · RampStartUpMaximumu
(4) Pu,i−1 + δu,i−1 · RampUpMaximumu
+ (1 −δu,i−1) · RampStartUpMaximumu
(4) Pu,i−1 + δu,i−1 · RampUpMaximumu + (1 −δu,i−1) · RampStartUpMaximumu
(4) (4) Similarly, the ramp down capability is limited by the maximum ramp down or the maximum shut
down ramp rate. Another factor which limits the operation of the generation units is the amount of time the unit
has been running or stopped. In order to avoid excessive ageing of the generators, or because of
their physical characteristics, once a unit is started up (shut down), it cannot be shut down (started)
immediately. To ensure this, a minimum online time and shut-down time are defined for each unit. In addition to the above mentioned constraints, generation units with energy storage capabilities
(mostly large hydro reservoirs and pumped hydro storage units) must meet the restrictions related 6 of 19 Energies 2020, 13, 6628 to the amount of energy stored. These restrictions include the storage capacity, inflow, outflow,
charging, charging capacity, charge/discharge efficiencies, etc. Discharging is considered as the
standard operation mode and is therefore linked to the P variable, common to all units. 2.2.2. Users Demand Profiles In addition to the thermophysical characteristics of the dwellings, the user consumption profiles
are needed to characterize the final demand profiles for SH and DHW. The space heating demand is defined through temperature set point profiles which are generated
randomly according to three conventional profiles. For DHW, water consumptions profiles are
generated from a database developed by Georges et al. [27] with the additional constraint to always
maintain the domestic hot water temperature between a minimum of 50 ◦C and a maximum of 65 ◦C in
the hot water tank. The demand profiles are then clustered for each insulation level using the validated
methodology proposed in Georges et al. [28]. The space heating demand profiles are aggregated in four clusters and the domestic hot water
profiles are aggregated in five clusters, resulting in a number of 20 aggregated demands per house
typology (40 in total). The clustering methodology for each demand type is based on the “averaging
method” presented in [27]. The number of aggregated demands per house typology was selected as a
good trade-off between complexity and model tractability. 2.2.1. Belgian Building Stock In order to obtain a bottom-up representation of the residential heat demand, the current
Belgian building stock is investigated and represented through state-space models able to describe its
thermal behaviour. The building archetypes and relative data regarding their thermophysical characteristics refer to
Gendebien et al. [26], who propose a tree-structure for the characterization of the residential building
stock of Belgium clustering the dwellings according to their geometry, year of construction, insulation
level, etc. Out of the archetypes proposed in [26], two house typologies are considered. They are
characterized by the same geometry, representative of a typical two-story free-standing house with a
heavy concrete structure and built after 1991. Two different and relatively high insulation levels and
air tightness are considered. For the first house typology, which is considered to be representative
for the 75% of the buildings included in the simulations, the thermal transmittance (U-value) is
U = 0.458 W/m2K and the value of air changes per hour (ACH) is n50 = 6 ACH at 50 Pa. For the
second house type (representing 25% of the buildings considered) the parameters adopted are
U = 0.305 W/m2K, n50 = 3 ACH. The chosen values represent the characteristics of two building 7 of 19 Energies 2020, 13, 6628 archetypes which well represent the majority of the Belgian building stock. This choice is based
on the assumption that electric heating systems are typically installed in recent and well-insulated
free-standing houses. g
The thermophyisical characteristics (thermal resistance, infiltration rate, etc.) of the dwellings
serve as parameters for the SH state space models. 2.2.3. State Space Models To represent the heat transfers occurring in the dwelling and in the hot water tank accurately,
state space models are used. The general form of state space models is Tt = ATt−1 + BUt
(5) (5) where Tt is the vector of temperatures at time step t, Ut is the vector of influence variables at time step
t and A and B are matrices containing the thermophysical parameters. where Tt is the vector of temperatures at time step t, Ut is the vector of influence variables at time step
t and A and B are matrices containing the thermophysical parameters. Space Heating In the case of space heating, the inside air temperature Ta, the wall temperature Twl and the
floor temperature T f are the state variables while the outside temperature To, the solar irradiation I,
the internal gains ˙Qi and the heating ˙Qh are considered as exogenous parameters. The space heating
state space model is represented in Figure 1. The system of differential equations is defined based on
an equilibrium equation for each temperature node that can be written as: Figure 1. Space heating state space model representation. Figure 1. Space heating state space model representation. Energies 2020, 13, 6628 8 of 19 Ca dTa
dt
=
Ra,wl(Twl
t
−Ta
t ) + Ra,f (T f
t −Ta
t )
+ ˙min f cp,a(To
t −Ta
t ) + τAIt + ˙Qi
t + ˙Qh
t
Cwl dTwl
dt
=
Ra,wl(Ta
t −Twl
t ) + Rwl,o(To
t −Twl
t )
Cf dT f
dt
=
Ra,f (Ta
t −T f
t ) + R f,o(To
t −T f
t )
(6) Ca dTa
dt
=
Ra,wl(Twl
t
−Ta
t ) + Ra,f (T f
t −Ta
t )
+ ˙min f cp,a(To
t −Ta
t ) + τAIt + ˙Qi
t + ˙Qh
t (6) Cwl dTwl
dt
=
Ra,wl(Ta
t −Twl
t ) + Rwl,o(To
t −Twl
t )
(6) Cf dT f
dt
=
Ra,f (Ta
t −T f
t ) + R f,o(To
t −T f
t ) where Rx,y is the thermal resistance between x and y, ˙min f is the mass infiltration rate of outside air, cp,a
is the specific thermal capacity of air, τ is the transmittance of the windows and A the windows area. where Rx,y is the thermal resistance between x and y, ˙min f is the mass infiltration rate of outside air, cp,a
is the specific thermal capacity of air, τ is the transmittance of the windows and A the windows area. In order to get a linear state space model, a linearization of the time derivatives over the desired
period is done through the central finite difference method. Linearizing and rearranging the terms,
Equation (6) can be written in the same form of Equation (5) thus obtaining explicitly the state space
parameters matrices A and B. 2.2.4. Hot Water Tank Regarding DHW, the hot water tank is modelled through a state-space model to describe Tw
under the hypothesis of isothermal hot water tank. The influence of the outside temperature To, the city water temperature Tcw and the heating
˙Qh is taken into account. The domestic hot water state space model is represented in Figure 2. The equilibrium equation for the hot water tank temperature can be written as: Cw
dTw
dt
= AUloss(To
t −Tw
t ) + cp,w ˙mw(Tcw
t
−Tw
t ) + ˙Qh
t
(7) (7) where Cw is the thermal capacity of the water tank, AUloss the overall heat transfer coefficient between
the inside and outside of the tank, cp,w the specific thermal capacity of water and ˙mw the specific hot
water consumption determined by the domestic hot water demand. Figure 2. Domestic hot water state space model representation. Figure 2. Domestic hot water state space model representation. Rearranging the terms, the state space parameters matrices A and B can be determined. 2.2.5. Heating Systems Heat Pumps In this work, variable-speed heat pumps are considered. The model describing heat pumps
is a linear empirical model based on the ConsomClim method [29] which was obtained by fitting
monitoring data relative to a high number of heat pumps types and operating conditions. The same
model is used for all the heat pumps, with the only difference of nominal capacity which is determined
depending on the insulation level and the supply temperature which is different for space heating and
for domestic hot water heating. The nominal characteristics of the heat pumps are the following: 9 of 19 Energies 2020, 13, 6628 •
Outside temperature: 7 ◦C •
Exhaust temperature: 35 ◦C •
Capacity: 11.2 kW/14 kW
•
COP: 3.95 •
Capacity: 11.2 kW/14 kW
•
COP: 3.95 The heat pumps can operate both in space heating mode or in domestic hot water heating mode
but not simultaneously. The heat pumps’ full load capacity and COP depend on the outside temperature and on the
heating system supply temperatures (45 ◦C for SH and 60 ◦C for DHW) according to: DT
=
Tout
Tsu
−Tout,n
Tsu,n
COPf l
=
COPn
C0 + C1DT + C2DT2
˙Q f l
=
[D0 + D1(Tout −Tout,n)
+
D2(Tsu −Tsu,n)] ˙Qn
˙Wf l
=
˙Q f l
COPf l
(8) DT
=
Tout
Tsu
−Tout,n
Tsu,n
COPf l
=
COPn
C0 + C1DT + C2DT2
˙Q f l
=
[D0 + D1(Tout −Tout,n)
+
D2(Tsu −Tsu,n)] ˙Qn
˙Wf l
=
˙Q f l
COPf l
(8) DT
=
Tout
Tsu
−Tout,n
Tsu,n
COPf l
=
COPn
C0 + C1DT + C2DT2
˙Q f l
=
[D0 + D1(Tout −Tout,n)
+
D2(Tsu −Tsu,n)] ˙Qn
˙Wf l
=
˙Q f l
COPf l
(8) (8) where Cx and Dx are parameters specific to the heat pump design, Tsu and Tsu,n are the heating system
supply temperatures (effective and nominal), COPn the nominal performance and ˙Qn the nominal
heat capacity. The part-load electrical consumption model of the heat pumps differs in space heating and in
water heating. Operating at part load affects the performance (COP) in SH mode but not in DHW mode. In space heating mode, the heat pump consumption at part load is modelled using a piecewise
linear approximation and limited by the consumption at full load performance as described by
Equation (9). Heat Pumps Wr ≥0, 77Qr
Wr ≥0, 69 + ((K2 −K1) + 1, 5(1 −K2))(Qr −0, 75)
Wr ≤Qr
(9) (9) where Qr = ˙Q/ ˙Q f l and Wr = ˙W/ ˙Wf l are respectively the heat and electric part load ratios, while the
coefficients K1 and K2 are empirically determined [29]. In addition, the heat pump comprises an
additional electric heater for space heating. The capacity of this additional heater is 3 kW. This
represent a novel modelling approach which allows to describe the behaviour heat pumps working at
partial load in MILP problems in the case of aggregated demand. In DHW mode, the performance is supposed constant at part load and is equal to the full
load performance: ˙W =
˙Q
COPf l
(10) (10) The total consumption of the heat pumps is the addition of the space heating consumption and
the domestic hot water consumption. An additional binary variable y is used to account for the
non-simultaneous working of the two modes by limiting their capacities by a fraction y for space
heating and (1 −y) for domestic hot water. Water Heaters The water heaters are electric resistance heaters and provide only domestic hot water heating. Their capacity and performance do not depend on external parameters and they can operate at part Energies 2020 13 6628
10 of 19 10 of 19 Energies 2020, 13, 6628 load with a constant performance. The coefficient of performance of the water heaters is thus at all
time equal to unity. load with a constant performance. The coefficient of performance of the water heaters is thus at all
time equal to unity. The water heaters capacity is calculated so that they are able to maintain the tank temperature its
minimum allowed value ˙Qd = maxh
n
˙Q : ∆Th∥Tw,min = 0
o
(11)
˙Qmin, ˙Qmax
= [1 kW, 4 kW]. ˙Qd = maxh
n
˙Q : ∆Th∥Tw,min = 0
o
(11) (11) and within the limited range
˙Qmin, ˙Qmax
= [1 kW, 4 kW]. and within the limited range
˙Qmin, ˙Qmax
= [1 kW, 4 kW]. and within the limited range
˙Qmin, ˙Qmax
= [1 kW, 4 kW]. 2.3. Integrated Model The integrated model presented in this work is obtained from the integration of the demand
side model (space and water heating) introduced in Section 2.2 into the existing supply side model
Dispa-SET for the power system that was presented in Section 2.1. The mathematical program is
formulated as follows: min
System cost
s.t. supply-demand balance
supply side constraints
demand side constraints
(12) min (12) A simple scheme representing the main modules of the integrated model is available in Figure 3. A simple scheme representing the main modules of the integrated model is available in Figure 3. The conjunction between the two models is realized by introducing power-to-heat options to the
system (heat pumps and electric water heaters) and by the add of the thermal demand to Equation (2). Equation (13) represents the demand-side balance constraint for the integrated model. ∑
u
(Pu,i · Locationu,n) + ∑
l
(Flowl,i · LineNodel,n) ∑
u
(Pu,i · Locationu,n) + ∑
l
(Flowl,i · LineNodel,n)
= Demandn,i + DemandHeatDevices,n,i + ∑
r
StorageInputs,i · Locations,n
−ShedLoadn,i
(13) = Demandn,i + DemandHeatDevices,n,i + ∑
r
StorageInputs,i · Locations,n
Sh dL
d
(13) = Demandn,i + DemandHeatDevices,n,i + ∑
r
StorageInputs,i · Locations,n
−ShedLoadn,i
(13) (13) where DemandHeatDevices,n,i is the electric load associated to the users thermal demand. This latter is
alternatively defined as an exogenous variable (when the thermal units are considered non-flexible) or
as an optimization variable (in the case of flexible heat pumps or electric water heaters). Figure 3. Integrated model flow chart. Figure 3. Integrated model flow chart. 11 of 19 Energies 2020, 13, 6628 All simulations are performed for a whole year with a time step of 1 hour. To reduce the
computational burden, the problem was split into smaller optimization problems that are run
recursively throughout the year. In order to avoid end-of-horizon artefacts (such as emptying of
the hydro reservoirs) with this rolling horizon approach, a look-ahead period is used but discarded for
the following optimization horizon [30]. In addition, the optimization is partially relaxed: the solution
has to be close enough to the optimum relaxed (linear) solution obtained. An optimality criterion is
fixed as the distance between the best relaxed solution and the MILP solution. Here, the optimality
gap criterion is set at 1%. 3.1. Scenarios Starting from the base case scenario, a certain number of simulations was performed in order to
assess the potential of heat pumps and resistance water heaters in terms of flexibility, RES integration
and total system costs at country level. The dependency of these results on the power system
configuration was also analyzed. The base case scenario, including the characteristics of the capacity
mix, of the system costs and of the electric demand, was defined as the one provided by Dispa-SET for
Belgium in 2015 with the total installed capacity composed for around one third by nuclear power,
one third by gas-fired units and one quarter from variable renewable sources (16% sun and 10% wind). The other scenarios were obtained through variations both on the demand (diffusion of electric heating
technologies) and supply side (variations in the capacity mix) of the system and can be summed
up as described in Table 1. Simulations were performed for six different supply-side configurations:
the renewable capacity was varied from one to three times the current renewable capacity (R1, R2,
R3). Moreover, for the power system configuration, two possibilities were evaluated considering
the current scenario, which included a relevant share of nuclear power plants, and an alternative
configuration with nuclear capacity substituted by gas-fired units, which are typically more flexible in
their operations. Table 1. Parametric analysis scenarios. Heating technology
HP—WH
Number of flexible units
0—0.2—0.4—0.6—0.8—1 M
Renewable capacity
R1—R2—R3
Non-renewable capacity
Nuclear power (non-flexible)—gas-fired units (flexible) Table 1. Parametric analysis scenarios. Heating technology
HP—WH
Number of flexible units
0—0.2—0.4—0.6—0.8—1 M
Renewable capacity
R1—R2—R3
Non-renewable capacity
Nuclear power (non-flexible)—gas-fired units (flexible) Table 1. Parametric analysis scenarios. On the demand-side, simulations included a sensitivity analysis on the number of installed
electric heating units (heat pumps and water heaters alternatively) which was varied between zero
(non-flexible demand) to one million units, with the national building stock accounting for a total of
5.6 million dwellings based on the latest data available [31] . On the demand-side, simulations included a sensitivity analysis on the number of installed
electric heating units (heat pumps and water heaters alternatively) which was varied between zero
(non-flexible demand) to one million units, with the national building stock accounting for a total of
5.6 million dwellings based on the latest data available [31] . 3. Simulations and Results In this section, the analyzed scenarios are described. Then, performance indicators are defined
and the results are finally presented and discussed. It is important to note that all simulations were
performed under a “perfect forecast” assumption: all the parameters of the model were considered to
be predicted with no error for the analyzed period of time. For this reason, the benefits of the flexible
heating devices assessed in this work have to be considered as an upper bound of the benefits of
residential DR, limited in practice by forecast errors and sub-optimal control. 2.3. Integrated Model The constraints of the system are represented by the sum of all the demand side and supply side
model constraints illustrated in the previous sections, while the objective function is the total system
cost (Equation (1)). 3.2. Performance Indicators The main indicators employed for the comparison of different scenarios were the followin The main indicators employed for the comparison of different scenarios were the following:
•
Energy system operational costs
•
Total load variation
•
Curtailment
•
Roundtrip storage efficiency (ratio of the useful energy to the total energy stored) •
Energy system operational costs
•
Total load variation
•
Curtailment
•
Roundtrip storage efficiency (ratio o •
Energy system operational costs •
Roundtrip storage efficiency (ratio of the useful energy to the total energy stored) All the indicators were computed on the basis of 1 year simulations and were descriptive of the
Belgian energy system. 3.3. Results 3.1. Scenarios Moreover, additional scenarios were analyzed to compare the performances of the thermal
storage capacity achievable with heating-electricity sector coupling solutions with the one offered 12 of 19 Energies 2020, 13, 6628 by the existing hydro pumped storage units. To this aim, a new base case scenario was formulated
including no storage units. This latter was compared with simulations including alternatively hydro
pumped storage (with an installed capacity equal to the one already installed), flexible heat pumps
(1 M) and flexible water heaters (1 M). The main features of the simulations are described in Table 2. Table 2. Storage units comparison scenarios Table 2. Storage units comparison scenarios
Storage technology
No storage units—HPHS (1308MW—HP (1 M)—WH (1 M)
Renewable capacity
R3
Non-renewavble capacity
Nuclear power (non-flexible) Operational Cost The total operational costs of the system with non-flexible heating devices is shown in Table 3. It can be seen that an increase in the renewable capacity made the operational costs decrease
significantly. On the other hand, the shift from nuclear power to gas-fired units led to more flexibility
but also to higher operational costs. Figure 5 shows the operational costs reduction which could be obtained through the progressive
diffusion of flexible water heaters. In the system configuration including nuclear power, Figure 5a shows that the more water heaters
were made flexible, the higher the benefit in cost reduction. Moreover, increasing the renewable
capacity also increased this cost reduction. A higher renewables penetration increased the system’s need for flexibility as well as the savings
obtainable by flexible heaters diffusion. As a general trend, the increase in cost benefit seemed to
follow a sub-linear trend with respect to the diffusion of flexible units; in fact, when water heaters’
flexibility was unlocked, the system flexibility increased and its flexibility need decreased, reducing
the marginal benefit linked to an additional flexible unit. Saturation was however not reached in these
simulations (i.e., additional flexible water heaters always benefited the operational cost). Table 3. Total operational cost for the different base-case scenarios [Me]. Nuclear Power (Non-Flexible)
Gas-Fired Units (Flexible)
R1
711
2362
R2
544
2026
R3
457
1705
(a) Nuclear power (non-flexible)
(b) Gas-fired units (flexible)
Figure 5. Total cost reduction (water heaters). Table 3. Total operational cost for the different base-case scenarios [Me]. Nuclear Power (Non-Flexible)
Gas-Fired Units (Flexible) Figure 5. Total cost reduction (water heaters). In the case of shift from nuclear to gas-fired units, Figure 5b shows results comprised between e2
million of cost increase (R1) and e2.5 million of potential savings (R3). However, the obtained results
fell below the solver precision and could therefore not be considered as significant. In those scenarios,
because of the flexible resources already present in the system, the spread of flexible electric heaters
did not have a significant influence on the system costs. 3.3.2. Water Heater Scenarios In the following scenarios, the impact on the energy system deriving from the diffusion of flexible
water heaters was analyzed. Contrary to heat pumps, these units only provided DHW (cfr Section 2). O
ti
l C
t In the following scenarios, the impact on the energy system deriving from the diffusion of flexible
water heaters was analyzed. Contrary to heat pumps, these units only provided DHW (cfr Section 2). 3.3.1. Base Case Scenario In the base case scenario the share of renewables was equal to 11.9% with 1.2% curtailment. Figure 4a shows the optimal dispatch in some summer days which were most critical for
curtailment. It has to be noted that in the base case scenario, curtailment did not arise from an
excess in renewable availability compared to the demand but was due to the non-flexibility of the
nuclear power plants. Base load units were unable to follow the residual load variations induced by
renewable generation. This explains the fact that some curtailment took place in night time due to the
reduced load in these hours. In the base case scenario, pumped hydro storage and other storage units
provided most of the flexibility to the system. These units appeared to absorb a significant part of the
excess energy and thereby reduce curtailment. Figure 4b shows a typical winter day. Significant differences between summer and winter were
observed. First, no curtailment occurred. Then, the generation of nuclear and coal units was more
stable. Finally there was a significant increase in non-nuclear conventional generation in particular
gas fired power plants due to the higher demand (between 1 and 2 GW more then in summer). These
effects arose without significant reduction in variable renewable generation. This load increase allowed
low-cost non-flexible units like nuclear power plants and coal units to work at full load. Flexibility
was provided mainly by gas fired combined cycle power plants. (a) Summer case
(b) Winter case
Figure 4. Optimal dispatch. (b) Winter case (a) Summer case Figure 4. Optimal dispatch. 13 of 19 Energies 2020, 13, 6628 Total Load Variation The load variation caused by the flexibility activation of the water heaters is shown in Figure 6. The penetration of flexible devices seemed to lead to an increase in the total load (mainly caused by 14 of 19 Energies 2020, 13, 6628 the thermal losses of the thermal storage); however, the observed variations could not be considered
relevant being lower then the solver precision. Therefore, it can be concluded that the activation of
flexible water heaters did not significantly increase the total electricity load. (a) Nuclear power (non-flexible)
(b) Gas-fired units (flexible)
Figure 6. Total load variation (water heaters). (a) Nuclear power (non-flexible) (b) Gas-fired units (flexible)
i
i
(
h
) Figure 6. Total load variation (water heaters). Curtailment Figure 7 shows the curtailment variation when flexible water heaters were added, demonstrating,
in the low-flexibility case, that it decreased significantly with the introduction of more flexible
devices. The variation increased with the increasing penetration of renewables in the system. In the
high flexibility case (shift from nuclear to gas) the system had enough flexibility to cope with the
intermittency of renewables and the introduction of flexible electric heaters had a negligible effect on
the total curtailment. (a) Nuclear power (non-flexible)
(b) Gas-fired units (flexible)
Figure 7. Curtailment (water heaters). (b) Gas-fired units (flexible)
nt (water heaters) Figure 7. Curtailment (water heaters). Figure 7. Curtailment (water heaters). Figure 8 shows a typical day where curtailment occurred (R2—nuclear power configuration) for
three penetration levels of flexible water heaters. When 0.4 M flexible heaters were introduced, thermal storage was able to offset all the curtailed
power of the first period and a part of the second period. Increasing the number of flexible heaters to
0.8 million further reduced curtailment, but without completely offsetting it. 15 of 19 Energies 2020, 13, 6628 (a) No flexible WH
(b) 0.4 M flexible WH
(c) 0.8 flexible WH
Figure 8. Curtailment reduction for different numbers of flexible water heaters (R2, nuclear power
(non-flexible), September 16) (MW). (c) 0.8 flexible WH (a) No flexible WH (b) 0.4 M flexible WH (a) No flexible WH (b) 0.4 M flexible WH Figure 8. Curtailment reduction for different numbers of flexible water heaters (R2, nuclear power
(non-flexible), September 16) (MW). Operational Cost The total operational cost reduction is shown in Figure 9. Compared with the water heaters cases,
the cost reduction was larger and became more significant with the shift from nuclear to gas-fired units. When heat pumps flexibility was “activated”, cost reduction occurred for two reasons: the increase
in flexibility (like for water heaters) and the heat pump consumption reduction thanks to an improved
control strategy which optimized its part-load performance over time. In the gas-fired units configuration, the flexibility provided by the heating systems had no
significant effect on the total cost. In these cases, the cost reduction observed was entirely due to
the consumption reduction. This consumption reduction remained constant for different renewable
capacities installed. For the nuclear power case, cost reduction became significant and the total cost reduction was
more important. (a) Nuclear power (non-flexible)
(b) Gas-fired units (flexible)
Figure 9. Total cost reduction (heat pumps). Figure 9. Total cost reduction (heat pumps). 3.3.3. Heat Pumps Scenarios In this section, the results for flexible heat pumps scenarios are discussed and compared to the
results of the electric heater case. Total Load Variation Figure 10a represents the obtained results for the total load variation in the case of flexible heat
pumps. As for the water heaters simulations, the total power consumption was not significantly
affected by the introduction of flexible heat pumps: in both nuclear and gas-fires units case,
the variation was lower than the simulation accuracy (equal to 1%). Energies 2020, 13, 6628 16 of 19 (a) Nuclear power (non-flexible)
(b) Gas-fired units (flexible)
Figure 10. Total load variation (heat pumps). (a) Nuclear power (non-flexible) Figure 10. Total load variation (heat pumps). 3.4. Comparison between Storage Technologies In this paragraph, the storage and overall system performances of flexible water heaters and
heat pumps are compared with the ones offered by existing hydro pumped storage units. To this aim,
four simulations were performed, as described in Table 2. The main results are resumed in Table 4. Table 4. Storage technologies performances. HPHS
HP
WH
Installed Storage Capacity [MW]
5886
24,848
3489
Avoided curtailment [MWh]
1,684,580
1,462,927
1,046,394
Specific avoided curtailment [MWh]
286.2
58.9
299.9
Total stored energy [MWh/MW]
1,863,084
6,422,305
1,319,002
Specific stored energy [MWh/MW]
316.5
258.5
378.0
Storage peak [%]
100
100
100
Roundrip efficiency [%]
0.74
0.90
0.68 Table 4. Storage technologies performances. Avoided curtailment is expressed with respect to the base case scenario, which includes an amount
of curtailed power equal to 2,995,154 MWh. Results showed that pumped hydro storage (HPHS) and flexible water heaters were the most
performing technologies in terms of specific avoided curtailment (expressed as the ratio between
avoided curtailment and virtual/installed energy storage capacity). Roundtrip efficiency was higher
in the case of heat pumps, while the flexible water heaters’ storage roundtrip efficiency was similar to
the one of hydro pumped storage units. In addition to the scenarios described, more simulations were performed in order to find the
equivalent installed capacity of HPHS able to provide the same results of flexible thermal units in terms
of avoided curtailment. Results showed that 3105 MWh of HPHS storage capacity would provide the
same curtailment reduction as 1 M flexible WH scenarios, while for the case of flexible HP, 4761 MWh
would be necessary. In general, results proved the effectiveness of heating-electricity sector coupling technologies
in capturing the inherent thermal storage potential of the thermal device itself and of the building
envelope, ensuring in this way a valid alternative to traditional energy storage units as pumped
hydro storage. Nevertheless, even if the flexibility potential of electric heating technologies came with no
additional investment costs, its practical exploitation required great effort in terms of data forecasting
and management of the service compared to more traditional flexibility sources like HPHS. Energies 2020, 13, 6628 17 of 19 17 of 19 4. Conclusions Funding: This research received no external funding. Funding: This research received no external funding. Funding: This research received no external funding. Acknowledgments: The authors would like to acknowledge Elodie Ciciriello’s contibution in the design of an
early version of presented modelling framework. Acknowledgments: The authors would like to acknowledge Elodie Ciciriello’s contibution in the design of an
early version of presented modelling framework. Conflicts of Interest: The authors declare no conflict of interest. Conflicts of Interest: The authors declare no conflict of interest. 4. Conclusions This paper presents an innovative modelling framework in which building and DHW models are
directly integrated within a well-established unit commitment and optimal dispatch model. The demand side heat model is calibrated for the case of the current Belgian building stock. The demand for heating is linked to the electricity demand through accurate state space models and
electric heating systems models for flexible resistance water heaters and flexible air-to-water heat
pumps. The combined model allows optimizing both the operation of the power system and the
smart management of the flexible units. The optimization if performed under a perfect foresight
hypothesis and should therefore be considered as representative of an upper limit of the available
flexible resources through demand response. A parametric analysis was performed assessing the influence of the flexible devices penetration,
the renewable capacity and the composition of the capacity mix. Results show that significant benefits
can be achieved with demand response management. Operational cost benefits up to e16 per water
heater and e35 per heat pump are assessed. Thermal storage has the potential to offset up to 1 TWh
of curtailment with 1 million electric heating systems (1 MWh of curtailment reduction per electric
device) over the optimization period when the flexibility needs are the most important (high renewable
capacity and nuclear power configuration). It is important to note that these benefits are reduced
significantly when the shift from nuclear to gas-fired units (more flexible) occurs thanks to a higher
ability of the system to follow renewables generation. It can therefore be concluded that demand response applied to flexible electric heating devices
offers significant flexibility resource to the system. This is especially true if more renewables are added
to the system and if the power plant fleet is relatively non-flexible. The results highlight the importance
of considering not only the demand-side and the renewables penetration in the system, but also the
existing capacity mix and its characteristics when dealing with the flexibility potential offered by
demand response. This can only be achieved through integrated models, and thereby justify ex post
the relevance of the proposed modelling framework. Author Contributions: conceptualization, C.M., S.Q. and K.K.; methodology, C.M. and S.Q.; data curation, C.M. and S.Q.; writing—original draft preparation, C.M.; writing—review and editing, S.Q., A.A. and K.K.; supervision,
S.Q., A.A. and K.K. All authors have read and agreed to the published version of the manuscript. References Reduced-order models for assessing demand response with heat
pumps—Insights from the German energy system. Energy Build. 2020, 223, 110144. 18. Bruninx, K.; Patteeuw, D.; Delarue, E.; Helsen, L.; D’haeseleer, W. Short-term demand response of flexible
electric heating systems: The need for integrated simulations. In Proceedings of the 2013 10th International
Conference on the European Energy Market (EEM), Stockholm, Sweden, 27–31 May 2013; pp. 1–10. 19. Patteeuw, D.; Bruninx, K.; Arteconi, A.; Delarue, E.; D’haeseleer, W.; Helsen, L. Integrated modeling of active
demand response with electric heating systems coupled to thermal energy storage systems. Appl. Energy
2015, 151, 306–319. 20. Papaefthymiou, G.; Hasche, B.; Nabe, C. Potential of heat pumps for demand side management and wind
power integration in the German electricity market. IEEE Trans. Sustain. Energy 2012, 4, 636–642. 21. Hedegaard, K.; Münster, M. Influence of individual heat pumps on wind power integration–Energy system
investments and operation. Energy Convers. Manag. 2013, 75, 673–684. 22. Patteeuw, D.; Reynders, G.; Bruninx, K.; Protopapadaki, C.; Delarue, E.; D’haeseleer, W.; Saelens, D.;
Helsen, L. CO2-abatement cost of residential heat pumps with active demand response: Demand-and
supply-side effects. Appl. Energy 2015, 156, 490–501. pp y
pp
gy
23. Arteconi, A.; Patteeuw, D.; Bruninx, K.; Delarue, E.; D’haeseleer, W.; Helsen, L. Active demand response
with electric heating systems: Impact of market penetration. Appl. Energy 2016, 177, 636–648. 4. IEA. Data and Statistics, Belgium; IEA: Paris, France, 2019. 24. IEA. Data and Statistics, Belgium; IEA: Paris, France, 2019. 25. Quoilin, S.; Hidalgo Gonzalez, I.; Zucker, A. Modelling Future EU Power Systems Under High Shares of
Renewables: The Dispa-SET 2.1 Open-Source Model; Technical Report; Joint Research Centre (JRC): Ispra, Italy;
EU: Brussels, Belgium, 2017; doi:10.2760914770. 26. Gendebien, S.; Georges, E.; Bertagnolio, S.; Lemort, V. Methodology to characterize a residential building
stock using a bottom-up approach: A case study applied to Belgium. Int. J. Sustain. Energy Plan. Manag. 2015, 4, 71–88. 27. Georges, E.; Gendebien, S.; Bertagnolio, S.; Lemort, V. Modeling and simulation of the domestic energy
use in Belgium following a bottom-up approach. In Proceedings of the CLIMA 2013 11th REHVA World
Congress & 8th International Conference on IAQVEC, Prague, Czech Republic, 16–19 June 2013. 28. Georges, E.; Quoilin, S.; Mathieu, S.; Lemort, V. Aggregation of flexible domestic heat pumps for the
provision of reserve in power systems. In Proceedings of the ECOS 2017 Conference, San Diego, CA, USA,
2–6 July 2017. 29. References 1. EU. Renewable Energy Progress Report; EU: Brussels, Belgium, 2019; 1. EU. Renewable Energy Progress Report; EU: Brussels, Belgium, 2019; 2. EU. The European Green Deal; EU: Brussels, Belgium, 2019 . EU. EU Reference Scenario 2016: Energy, Transport and GHG Emissions. TRENDS to 2050; EU: Brus
Belgium, 2016. 4. Paterakis, N.; Erdinç, O.; Catalão, J. An overview of Demand Response: Key-elements and international
experience. Renew. Sustain. Energy Rev. 2017, 69, 871–891. 5. Lynch, M.; Nolan, S.; Devine, M.; O’Malley, M. The impacts of demand response participation in capacity
markets. Appl. Energy 2019, 250, 444–451. 6. Dupont, B.; Dietrich, K.; De Jonghe, C.; Ramos, A.; Belmans, R. Impact of residential demand response on
power system operation: A Belgian case study. Appl. Energy 2014, 122, 1–10. 7. Pina, A.; Silva, C.; Ferrão, P. The impact of demand side management strategies in the penetration of
renewable electricity. Energy 2012, 41, 128–137. 8. Moura, P.S.; De Almeida, A.T. The role of demand-side management in the grid integration of wind power. Appl. Energy 2010, 87, 2581–2588. Energies 2020, 13, 6628 18 of 19 9. McPherson, M.; Stoll, B. Demand response for variable renewable energy integration: A proposed approach
and its impacts. Energy 2020, 197, 117205. 0. Gils, H.C. Assessment of the theoretical demand response potential in Europe. Energy 2014, 67, 1–18. 1. Bloess, A.; Schill, W.P.; Zerrahn, A. Power-to-heat for renewable energy integration: A review of technolo
modeling approaches, and flexibility potentials. Appl. Energy 2018, 212, 1611–1626. 12. Reynders, G.; Nuytten, T.; Saelens, D. Potential of structural thermal mass for demand-side management in
dwellings. Build. Environ. 2013, 64, 187–199. 13. Hedegaard, K.; Mathiesen, B.V.; Lund, H.; Heiselberg, P. Wind power integration using individual heat
pumps–analysis of different heat storage options. Energy 2012, 47, 284–293. 14. Kepplinger, P.; Huber, G.; Petrasch, J. Autonomous optimal control for demand side management with
resistive domestic hot water heaters using linear optimization. Energy Build. 2015, 100, 50–55. 15. Jacobson, M.; Delucchi, M.; Bauer, Z.A. 100% Clean and Renewable Wind, Water, and Sunlight All-Sector
Energy Roadmaps for 139 Countries of the World. Joule 2017, 1, 108–121. 16. Vivian, J.; Prataviera, E.; Cunsolo, F.; Pau, M. Demand Side Management of a pool of air source heat pumps
for space heating and domestic hot water production in a residential district. Energy Convers. Manag. 2020,
225, 113457. 17. Sperber, E.; Frey, U.; Bertsch, V. References Bolher, A.; Casari, R.; Fleury, E.; Marchio, D.; Millet, J. Méthode de Calcul des Consommations D’énergie des
Bâtiments Climatisés Consoclim; Ecole des Mines: Paris, France, 1999. 19 of 19 Energies 2020, 13, 6628 30. Quoilin, S.; Nijs, W.; Gonzalez, I.H.; Zucker, A.; Thiel, C. Evaluation of simplified flexibility evaluation tools
using a unit commitment model. In Proceedings of the 2015 12th International Conference on the European
Energy Market (EEM), Lisbon, Portugal, 19–22 May 2015; pp. 1–5. 31. Statbel. Building Stock; Statbel: Brussels, Belgium, 2020. Publisher’s Note: MDPI stays neutral with regard to jurisdictional claims in published maps and institutional
affiliations. © 2020 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access
article distributed under the terms and conditions of the Creative Commons Attribution
(CC BY) license (http://creativecommons.org/licenses/by/4.0/). 30. Quoilin, S.; Nijs, W.; Gonzalez, I.H.; Zucker, A.; Thiel, C. Evaluation of simplified flexibility evaluation tools
using a unit commitment model. In Proceedings of the 2015 12th International Conference on the European
Energy Market (EEM), Lisbon, Portugal, 19–22 May 2015; pp. 1–5. 30. Quoilin, S.; Nijs, W.; Gonzalez, I.H.; Zucker, A.; Thiel, C. Evaluation of simplified flexibility evaluation tools
using a unit commitment model. In Proceedings of the 2015 12th International Conference on the European
Energy Market (EEM), Lisbon, Portugal, 19–22 May 2015; pp. 1–5. Publisher’s Note: MDPI stays neutral with regard to jurisdictional claims in published maps and institutional
affiliations. © 2020 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access
article distributed under the terms and conditions of the Creative Commons Attribution
(CC BY) license (http://creativecommons.org/licenses/by/4.0/). © 2020 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access
article distributed under the terms and conditions of the Creative Commons Attribution
(CC BY) license (http://creativecommons.org/licenses/by/4.0/).
|
https://openalex.org/W2102919205
|
https://tbiomed.biomedcentral.com/counter/pdf/10.1186/1742-4682-3-27
|
English
| null |
Kinetic modeling of tricarboxylic acid cycle and glyoxylate bypass in Mycobacterium tuberculosis, and its application to assessment of drug targets
|
Theoretical biology and medical modelling
| 2,006
|
cc-by
| 10,450
|
Theoretical Biology and Medical Modelling 2006, 3:27
doi:10.1186/1742-4682-3-27 Theoretical Biology and Medical Modelling 2006, 3:27
doi:10.1186/1742-4682-3-27 This article is available from: http://www.tbiomed.com/content/3/1/27 © 2006 Singh and Ghosh; licensee BioMed Central Ltd. g
This is an Open Access article distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/2.0),
which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited. Theoretical Biology and Medical
Modelling BioMed Central Open Access Open Acc
Research
Kinetic modeling of tricarboxylic acid cycle and glyoxylate bypass in
Mycobacterium tuberculosis, and its application to assessment of
drug targets
Vivek Kumar Singh and Indira Ghosh* Address: Bioinformatics Centre, University of Pune, Pune-411007, India
Email: Vivek Kumar Singh - vivek@bioinfo.ernet.in; Indira Ghosh* - indira@bioinfo.ernet.in
* Corresponding author Email: Vivek Kumar Singh - vivek@bioinfo.ernet.in; Indira Ghosh* - indira@bioinfo.ernet.in
* Corresponding author Received: 03 April 2006
Accepted: 03 August 2006 Published: 03 August 2006 Published: 03 August 2006 Abstract Hence, interference with the
glyoxylate bypass is a potential approach to the design of
new drugs against persistent mycobacteria. This is consist-
ent with the suggestion that the regulation of M. tubercu-
losis metabolism in response to the environment of the
bacterium makes large contributions to its virulence [3]. At the branch point of the tricarboxylic acid (TCA) cycle
and glyoxylate bypass, isocitrate dehydrogenase (ICD),
involved in the TCA cycle, and ICL, involved in the glyox-
ylate bypass, compete for the same substrate, namely isoc-
itrate (ICIT). In Escherichia coli, flux at this branch point is
predominantly controlled through the reversible inactiva-
tion of ICD by phosphorylation, catalyzed by ICD-kinase
[4]. We have already identified the kinase in M. tuberculo-
sis, equivalent to ICD-kinase in E. coli, that is responsible
for reversible inactivation of ICD1 (Rv3339c) by phos-
phorylation [5]. Moreover, a method has been described
for inhibiting a metabolic pathway that is essential for the
viability of a microorganism by diverting the substrate to
a different metabolic pathway, and it has been suggested
that inhibiting ICD1-kinase could inhibit the flux through
the glyoxylate bypass in M. tuberculosis [5]. Since inhibi-
tion of ICD1-kinase would increase the amount of
dephosphorylated (active) ICD1, the flux through the gly-
oxylate bypass would be diminished. However, enzymes
are not isolated entities in living organisms but act as
components of systems, so the effect of modulation of any
enzyme activity on a metabolic flux depends on the prop-
erties of the other enzymes in the pathway concerned [6]. In this communication, we describe the construction of a
kinetic model of the TCA cycle and glyoxylate bypass in
M. tuberculosis, and we study the likely metabolic conse-
quences of inhibiting ICLs and ICD1-kinase. To the best
of our knowledge, this is the first attempt to model any
specific metabolic pathway in M. tuberculosis, and no
kinetic model is available for the TCA cycle and glyoxylate
bypass in this bacterium. Initially, we constructed a
kinetic model for the TCA cycle and glyoxylate bypass in
E. coli to validate the pathway modeling protocol used
and to test how well the metabolic profile correlates with
the gene expression profile while trying to predict the met-
abolic flux distribution using the gene expression data. The biochemical reactions considered for the models are
shown in figure 1 and the metabolites with known con-
centrations are listed in table 1. Abstract humans for centuries, and presently there is an urgent
need for new drugs to combat drug-resistant tuberculosis humans for centuries, and presently there is an urgent
need for new drugs to combat drug-resistant tuberculosis Background
Tuberculosis is an ancient disease that has plagued Background
Tuberculosis is an ancient disease that has plagued Page 1 of 11
(page number not for citation purposes) Page 1 of 11
(page number not for citation purposes) Theoretical Biology and Medical Modelling 2006, 3:27 http://www.tbiomed.com/content/3/1/27 an enzyme kills an organism, MCA can elucidate the
mechanism involved. an enzyme kills an organism, MCA can elucidate the
mechanism involved. an enzyme kills an organism, MCA can elucidate the
mechanism involved. and shorten the time of tuberculosis therapy. Tuberculosis
treatment is lengthy because of a population of persistent
bacilli that is not effectively eliminated by current drugs. Since modulation of target enzyme(s) activity is usually
aimed at altering the cell's metabolic profile, knowledge
of the metabolic profile is important for identifying the
target. Recent experiments have shown a positive correla-
tion between mRNA levels measured by DNA microarrays
and protein abundance in both E. coli [10] and yeast cells
[11,12], so the gene expression profile could be connected
to the metabolic profile via simulation of the pathway
under study. In E. coli, the in vivo kinetic parameters
required for estimating the metabolic profile of most
enzymes are available when the organism is grown using
glucose as the carbon source [13]. In contrast, when ace-
tate is used as the carbon source, the gene expression pro-
file of the TCA cycle and glyoxylate bypass enzymes
differed from that found with glucose [14]. The corre-
sponding metabolic flux distributions in central meta-
bolic pathways under both growth conditions are known
[15], so this seems an ideal system for testing the hypoth-
esis that the gene expression profile can be connected with
the metabolic profile via simulation of the pathway under
study. The persistent bacilli primarily use fatty acids as their car-
bon source [1]. This makes the glyoxylate bypass, consist-
ing of isocitrate lyase (ICL) and malate synthase (MS),
essential for the bacterium; in its absence there will be no
net formation of the intermediates required for synthesiz-
ing cellular materials. Inhibition of both ICL1 (prokaryo-
tic-like isoform) and ICL2 (eukaryotic-like isoform) has
been shown to block the growth of M. tuberculosis in mac-
rophages and in mice [2]. Abstract Background: Targeting persistent tubercule bacilli has become an important challenge in the
development of anti-tuberculous drugs. As the glyoxylate bypass is essential for persistent bacilli,
interference with it holds the potential for designing new antibacterial drugs. We have developed
kinetic models of the tricarboxylic acid cycle and glyoxylate bypass in Escherichia coli and
Mycobacterium tuberculosis, and studied the effects of inhibition of various enzymes in the M. tuberculosis model. Results: We used E. coli to validate the pathway-modeling protocol and showed that changes in
metabolic flux can be estimated from gene expression data. The M. tuberculosis model reproduced
the observation that deletion of one of the two isocitrate lyase genes has little effect on bacterial
growth in macrophages, but deletion of both genes leads to the elimination of the bacilli from the
lungs. It also substantiated the inhibition of isocitrate lyases by 3-nitropropionate. On the basis of
our simulation studies, we propose that: (i) fractional inactivation of both isocitrate dehydrogenase
1 and isocitrate dehydrogenase 2 is required for a flux through the glyoxylate bypass in persistent
mycobacteria; and (ii) increasing the amount of active isocitrate dehydrogenases can stop the flux
through the glyoxylate bypass, so the kinase that inactivates isocitrate dehydrogenase 1 and/or the
proposed inactivator of isocitrate dehydrogenase 2 is a potential target for drugs against persistent
mycobacteria. In addition, competitive inhibition of isocitrate lyases along with a reduction in the
inactivation of isocitrate dehydrogenases appears to be a feasible strategy for targeting persistent
mycobacteria. Conclusion: We used kinetic modeling of biochemical pathways to assess various potential anti-
tuberculous drug targets that interfere with the glyoxylate bypass flux, and indicated the type of
inhibition needed to eliminate the pathogen. The advantage of such an approach to the assessment
of drug targets is that it facilitates the study of systemic effect(s) of the modulation of the target
enzyme(s) in the cellular environment. Page 2 of 11
(page number not for citation purposes) Results and discussion cinate (SUC) for the continuation of the TCA cycle. This
conversion is brought about by ScAS. Model-2 does not
require ScAS because it converts αKG directly to SUC
using KGD and SSADH. The steady state fluxes computed
from the two models showed minor differences, but the
turnover of the TCA cycle and glyoxylate bypass was sim-
ilar in both models, indicating that M. tuberculosis can
manage without a functional KDH. Thus, this study illus-
trates that at the metabolic level, the absence of KDH
activity has no effect on the net flux through the TCA cycle
and glyoxylate bypass. cinate (SUC) for the continuation of the TCA cycle. This
conversion is brought about by ScAS. Model-2 does not
require ScAS because it converts αKG directly to SUC
using KGD and SSADH. The steady state fluxes computed
from the two models showed minor differences, but the
turnover of the TCA cycle and glyoxylate bypass was sim-
ilar in both models, indicating that M. tuberculosis can
manage without a functional KDH. Thus, this study illus-
trates that at the metabolic level, the absence of KDH
activity has no effect on the net flux through the TCA cycle
and glyoxylate bypass. Steady state fluxes in the E. coli model (table 2) were com-
pared to the experimental fluxes given by Zhao et al. [15];
the net fluxes were expressed in relative units. The unit
conversion is described in methods section. The steady
state fluxes calculated from the model accorded with the
experimental fluxes [15] (table 3), thus validating the pro-
tocol used. Since the maximal reaction rates (Vmax) of the enzymes
during growth on acetate were estimated using gene
expression data, it is possible to estimate the changes in
metabolic flux distribution due to changes in gene expres-
sion via simulation of the biochemical pathway under
study. This was also noted in the study of branched chain
amino acid biosynthesis in E. coli [19]. On the basis of the finding of Tian et al. [13], i.e. that KDH
activity is absent in M. tuberculosis, and of the observation
that there is little difference between the two models in
the turnover of the TCA cycle and glyoxylate bypass, M. tuberculosis model-2 was taken as the reference model in
the remaining parts of this study. The steady state fluxes in the M. Abstract In M. tuberculosis H37Rv
strain there are two isoforms of ICD [17], ICD1 (Rv3339c)
and ICD2 (Rv0066c), and two isoforms of ICL [17,18],
ICL1 (Rv0467) and ICL2 (Rv1915 and Rv1916). In addi-
tion, the inability of Nathan and co-workers to detect α-
ketoglutarate dehydrogenase (KDH) activity in M. tubercu-
losis [13] was taken into account while constructing the
model. M. tuberculosis model-1 represents a standard TCA
cycle and glyoxylate bypass with KDH present, while
model-2 lacks KDH activity. Our aim was to check the
metabolic consequences of the presence and absence of
KDH in this organism. Metabolic Control Analysis (MCA) is a theoretical frame-
work that relates the systemic properties of a metabolic
system to the properties of its components, in particular
the enzymes, in a quantitative manner [6]. Application of
MCA to the identification of potential drug targets is
exemplified by glycolysis in Trypanosoma brucei [7-9]. MCA also gives insight into the cellular effect(s) of inhibi-
tion of a particular enzyme. Eisenthal et al. [9] suggested
two basic metabolic methods for killing an organism:
decreasing the flux through an essential metabolic path-
way to a nonviable level, or increasing the concentration
of a metabolite to a toxic level. Therefore, if inhibition of Page 2 of 11
(page number not for citation purposes) Theoretical Biology and Medical Modelling 2006, 3:27 http://www.tbiomed.com/content/3/1/27 Table 1: Metabolites of the models with known concentrations (with references indicated in square brackets)
Escherichia coli
Mycobacterium tuberculosis
Metabolite
Concentration in
glucose condition (in
mM)
Concentration in
acetate condition (in
mM)
Metabolite
Concentration (in mM)
acetyl-CoA
0.5 [16]
0.5 [16]
succinate
2.464 (derived from Tian
et. al [13])
citrate
3 [16]
9 [16]
fumarate
0.08528 (derived from Tian
et. al [13])
isocitrate
0.018a [16]
0.15 [16]
malate
0.408 (derived from Tian
et. al [13])
succinate
0.6 [16]
6 [16]
oxaloacetate
0.0003 (assumed)
malate
1.8 [16]
5 [16]
CoA
0.0001 (assumed)
oxaloacetate
0.004b
0.0014 (assumed)
CoA
0.0001 (assumed)
0.0001 (assumed)
aIsocitrate concentration was inferred from a graph shown by Walsh et. al [16]. The value in the graph was 0.025 mM at 30 minutes after addition
of glucose to the medium, but it had a negative slope, so, a value of 0.018 mM was taken. bTaken as 2.4 times the concentration of oxaloacetate under growth on acetate because flux leading to the synthesis of oxaloacetate under growth
on glucose is 2.4 times of that under growth on acetate [15]. Abstract able 1: Metabolites of the models with known concentrations (with references indicated in square brackets aIsocitrate concentration was inferred from a graph shown by Walsh et. al [16]. The value in the graph was 0.025 mM at 30 minutes after addition
of glucose to the medium, but it had a negative slope, so, a value of 0.018 mM was taken. bTaken as 2.4 times the concentration of oxaloacetate under growth on acetate because flux leading to the synthesis of oxaloacetate under growth
on glucose is 2.4 times of that under growth on acetate [15]. Results and discussion tuberculosis model-1
(standard TCA cycle) and model-2 (absence of KDH activ-
ity) are shown in table 4. The fluxes in the two models of
the M. tuberculosis TCA cycle and glyoxylate bypass are
similar, with the following exceptions. (i) The entire flux
from α-ketoglutarate (αKG) towards the TCA cycle passes
through the α-ketoglutarate decarboxylase (KGD) and
succinic semialdehyde dehydrogenase (SSADH) steps in
model-2 (which has no other branch from αKG that con-
tinues in TCA cycle); in model-1, about 84% of the flux
from αKG passes through KDH and the remaining 16%
through KGD and SSADH, but the total flux from αKG
continuing in the TCA cycle is almost the same in both
models. (ii) Flux was observed through the succinyl-CoA
synthetase (ScAS) step in model-1 but was negligible in
model-2. This is expected because KDH converts αKG to
succinyl-CoA, and succinyl-CoA must be converted to suc- Page 3 of 11
(page number not for citation purposes) Inactivation of ICDs in M. tuberculosis model f
Inactivation of ICD1, which is brought about by ICD1-
kinase, leads to a change in the number of active ICD1
molecules. Since Vmax is a function of the amount of
enzyme, any change in the amount of enzyme will affect
the Vmax. Therefore, varying Vmax for ICD1 from 1% to
100% was used to monitor the effect of inactivation of
ICD1 by ICD1-kinase. Since there is no information about
any such kinase for ICD2, the activity value was kept at
100%. Plots of the sum of flux through ICD1 and ICD2
(JICD1 + JICD2) and the sum of flux through ICL1 and ICL2
(JICL1 + JICL2) against Vmax for the forward ICD1 reaction
(VfICD1) (figure 2A) showed that even at 99% inactivation
there was no perceptible flux through the glyoxylate
bypass. We then studied the effect of inactivation of ICD2
by a hypothetical inactivator, along with the inactivation Page 3 of 11
(page number not for citation purposes) Page 3 of 11
(page number not for citation purposes) Theoretical Biology and Medical Modelling 2006, 3:27 http://www.tbiomed.com/content/3/1/27 TCA cycle and glyoxylate bypass reactions considered in E. coli and M. tuberculosis models
Figure 1
TCA cycle and glyoxylate bypass reactions considered in E. coli and M. tuberculosis models. Reactions 1, 2, 3, 5, 8,
9, 10, 11, 12 and 13 were present in all the models; reaction 4 was present only in the E. coli model and M. tuberculosis model-
1, but absent from M. tuberculosis model-2; and reactions 6 and 7 were present in the M. tuberculosis models, but absent from E. coli model. 1, CS; 2, ACN; 3, ICD in E. coli model and ICD1 and ICD2 in M. tuberculosis models; 4, KDH; 5, ScAS; 6, KGD; 7,
SSADH; 8, SDH; 9, FUM; 10, MDH; 11, fraction of αKG utilized for precursor biosynthesis (SYN); 12, ICL in E. coli model and
ICL1 and ICL2 in M. tuberculosis models; 13, MS. glyoxylate
citrate
isocitrate
alpha-
ketoglutarate
succinyl-CoA
succinate
fumarate
oxaloacetate
precursor
succinic
semialdehyde
1
2
3
4
5
6
7
8
9
10
11
12
13
malate
acetyl-CoA
acetyl-CoA
CoA
CoA TCA cycle and glyoxylate bypass reactions considered in E. coli and M. tuberculosis models. Reactions 1, 2, 3, 5, 8,
9, 10, 11, 12 and 13 were present in all the models; reaction 4 was present only in the E. coli model and M. Inactivation of ICDs in M. tuberculosis model tuberculosis model-
1, but absent from M. tuberculosis model-2; and reactions 6 and 7 were present in the M. tuberculosis models, but absent from E. coli model. 1, CS; 2, ACN; 3, ICD in E. coli model and ICD1 and ICD2 in M. tuberculosis models; 4, KDH; 5, ScAS; 6, KGD; 7,
SSADH; 8, SDH; 9, FUM; 10, MDH; 11, fraction of αKG utilized for precursor biosynthesis (SYN); 12, ICL in E. coli model and
ICL1 and ICL2 in M. tuberculosis models; 13, MS. of ICD1. The plot of JICD1 + JICD2 and JICL1 + JICL2 against
Vmax for the ICD1 and ICD2 forward reactions (VfICD1
and VfICD2 respectively) (figure 2B) showed that the flux through the glyoxylate bypass (JICL1 + JICL2) starts to
increase after VfICD1 and VfICD2 have fallen to approxi-
mately 30% of the original values, and becomes equal to Page 4 of 11
(page number not for citation purposes) Theoretical Biology and Medical Modelling 2006, 3:27 http://www.tbiomed.com/content/3/1/27 Table 2: Steady state fluxes computed for E. coli model. Reaction step
Growth on glucose (mM/min)
Growth on acetate (mM/min)
CS
4.187
8.006
ACN
4.187
8.006
ICD
4.179
6.125
KDH
3.394
5.916
ScAS
3.394
5.916
SDH
3.401
7.798
FUM
3.401
7.798
MDH
3.409
9.679
SYN
0.786
0.209
ICL
0.008
1.882
MS
0.008
1.882 Table 2: Steady state fluxes computed for E. coli model. Deletion of genes encoding ICLs in M. tuberculosis model
McKinney and co-workers showed that deletion of either
of the genes icl1 or icl2 had little effect on mycobacterial
growth in macrophages or in mice [2]. In our model, dele-
tion of icl1 could be simulated by deleting the ICL1 reac-
tion. Plots of JICD1 + JICD2 and JICL2 as a function of VfICD1
and VfICD2 (figure 2C) showed that more than 90% inacti-
vation of both ICD1 and ICD2 is required to allow a per-
ceptible flux through the glyoxylate bypass in the absence
of ICL1. In contrast, when both ICLs were present, 70%
inactivation of both ICD1 and ICD2 sufficed to allow a
flux through the glyoxylate bypass (figure 2B). Simulating
icl2 gene deletion showed only a marginal difference in
the flux through the glyoxylate bypass or in JICD1 + JICD2
when plotted against VfICD1 and VfICD2 (figure 2D), com-
pared to the fluxes observed in the presence of both ICLs
(figure 2B). Inactivation of ICDs in M. tuberculosis model Thus, the model correctly simulates the exper-
imental observation that deletion of either of the two ICL
genes has little effect on the growth of mycobacteria in
macrophages and in mice [2]. It also shows that a flux of
approximately 26% through the glyoxylate bypass
remains in the absence of icl1, compared to the flux when
both ICLs are present (with VfICD1 and VfICD2 kept at 5% of JICD1 + JICD2 when VfICD1 and VfICD2 have fallen to about
3% of the original values. Thus, flux through the glyoxy-
late bypass was observed only when both ICD1 and ICD2
were more than 70% inactivated. Inactivation of ICD1 has
already been demonstrated experimentally [5], but no
such phosphorylation-induced inactivation of ICD2 has
been reported. The possibility of inactivation of ICD2
along with ICD1 in persistent mycobacteria, leading to an
up-regulation of flux through the glyoxylate bypass, is
suggested by our study. A novel protein might bring about
this inactivation, or the kinase that acts on ICD1 might
also act on ICD2. Since no differential expression of ICD1
and ICD2 has been reported in the literature, both the
ICDs were kept active in our study. Interestingly, the
model also suggests that if 30% or more of ICD1 and
ICD2 are in the active state, there will be no flux through
the glyoxylate bypass. Since the glyoxylate bypass is essen-
tial for persistent bacilli, they would perish under such
conditions. Inhibition of ICD1-kinase and/or the pro-
posed inactivator of ICD2 would increase the amount of
active ICD1 and/or ICD2 respectively, suggesting that this
is a potential target for the development of drugs against
persistent mycobacteria. Table 3: Comparison of the experimental fluxes to that computed from E. coli model. The reaction step SYN was not explicitly
mentioned by Zhao et al. [15], but was shown by a branch from αKG. Reaction step
Growth on glucose
(Experimental)
Growth on glucose
(Simulation)
Growth on acetate
(Experimental)
Growth on acetate
(Simulation)
CS
50
50
73.4
73.4
ACN
50
50
73.4
73.4
ICD
50
49.9
52.8
56.1
KDH
40.6
40.5
51.0
54.2
ScAS
40.6
40.5
51.0
54.2
SDH
40.6
40.6
71.6
71.5
FUM
40.6
40.6
71.6
71.5
MDH
40.6
40.7
86.3
88.7
SYN
9.4
9.4
1.8
1.9
ICL
0
0.1
20.6
17.2
MS
0
0.1
20.6
17.2 xperimental fluxes to that computed from E. coli model. Uncompetitive inhibition of ICLs Uncompetitive inhibition of ICLs The rate equations of the ICL1 (equation (2)) and ICL2
reactions were modified to account for uncompetitive
inhibition against isocitrate. The procedure used was sim-
ilar to that described for competitive inhibition. The plots
of JICD1 + JICD2 and JICL1 + JICL2 against (I/KI) showed that I/
KI ratios of about 35 (figure 4A) and 71 (figure 4B) respec-
tively were required to reduce JICL1 + JICL2 by 90%. The cor-
responding reductions in JICL1 + JICL2 by competitive
inhibition of ICL1 and ICL2 were 52.4% and 86.2%
respectively. Inactivation of ICDs in M. tuberculosis model The reaction step SYN was not explicitly
, but was shown by a branch from αKG. . coli model. The reaction step SYN was not explicitly able 3: Comparison of the experimental fluxes to that computed from E. coli model. The reaction step SYN
mentioned by Zhao et al. [15], but was shown by a branch from αKG. Page 5 of 11
(page number not for citation purposes) Theoretical Biology and Medical Modelling 2006, 3:27
http://www.tbiomed.com/content/3/1/27 Theoretical Biology and Medical Modelling 2006, 3:27 http://www.tbiomed.com/content/3/1/27 Table 4: Steady state fluxes computed for M. tuberculosis model-1 and model-2 (in persistent mycobacteria)
Reaction step
Fluxes in model-1 (mM/min)
Fluxes in model-2 (mM/min)
CS
0.988
0.988
ACN
0.988
0.988
ICD1
0.653
0.650
ICD2
0.331
0.333
KDH
0.797
-
ScAS
0.797
-5.65892 × 10-11
KGD
0.154
0.950
SSADH
0.154
0.950
SDH
0.955
0.955
FUM
0.955
0.955
MDH
0.959
0.959
SYN
0.034
0.034
ICL1
0.004
0.004
ICL2
0.000
0.001
MS
0.004
0.005 e 4: Steady state fluxes computed for M. tuberculosis model-1 and model-2 (in persistent mycobacteria) Uncompetitive inhibition of ICLs the original values). In the absence of icl2, the flux
through the glyoxylate bypass decreases only by 7.6%
compared to the flux in presence of both ICLs (with VfICD1
and VfICD2 kept at 5% of the original values). Such a reduc-
tion in flux due to the deletion of either of the two ICL
genes would be too small to lead to elimination of the
bacilli. Competitive inhibition of ICLs The rate equations of the ICL1 and ICL2 reactions were
modified to account for competitive inhibition, i.e. com-
petition against isocitrate, as shown in equation (1). The
ratio of inhibitor concentration to inhibitor constant (I/
KI) was assumed to be the same for both ICL1 and ICL2. Two simulations were performed, one with VfICD1 and
VfICD2 kept at 2.5%, the other at 5%, of the original values. The plots of JICD1 + JICD2 and JICL1 + JICL2 against (I/KI)
showed that I/KI ratios of about 477 (figure 3A) and 105
(figure 3B) respectively were required to reduce JICL1 + JICL2
by 90%. v
Vf
ICIT
K
Vr
SUC
K
GLY
K
ICIT
K
ICL
M ICIT
ICL
M SUC
M GLY
M ICIT
=
−
+
+
1
1
1
,
,
,
,
ICIT
K
I
K
SUC
K
GLY
K
ICIT
K
SUC
K
S
M ICIT
I
M SUC
M GLY
M ICIT
M SUC
,
,
,
,
,
+
+
+
+
UC
K
GLY
K
M SUC
M GLY
,
,
⎛
⎝
⎜
⎜
⎜
⎜⎜
⎞
⎠
⎟
⎟
⎟
⎟⎟
(
)
equation 2 In contrast to competitive inhibition of ICL1 and ICL2,
the efficiency of uncompetitive inhibition decreased with
an increase in VfICD1 and VfICD2 from 2.5% to 5% of the
original values. This is because an increase in the Vmax of
the ICDs leads to a decrease in isocitrate concentration,
and hence to a decrease in the enzyme-substrate complex
concentration. Because an uncompetitive inhibitor binds
only to the enzyme-substrate complex, a decrease in
enzyme-substrate complex concentration leads to a
decrease in inhibitor binding, resulting in less inhibition. In contrast to competitive inhibition of ICL1 and ICL2,
the efficiency of uncompetitive inhibition decreased with
an increase in VfICD1 and VfICD2 from 2.5% to 5% of the
original values. This is because an increase in the Vmax of
the ICDs leads to a decrease in isocitrate concentration,
and hence to a decrease in the enzyme-substrate complex
concentration. Because an uncompetitive inhibitor binds
only to the enzyme-substrate complex, a decrease in
enzyme-substrate complex concentration leads to a
decrease in inhibitor binding, resulting in less inhibition. Page 6 of 11
(page number not for citation purposes) Competitive inhibition of ICLs The effect of inhibitor is shown by varying the ratio of I/KI. 0
250
500
0
0.25
0.5
I / KI
Fluxes (mM / min)
0
250
500
0
0.5
1
I / KI
Fluxes (mM / min)
A
B ICD2, i.e. would indirectly cause an increase in the Vmax
of ICD1 and/or ICD2, thus indirectly improving the effi-
ciency of competitive inhibition of the ICLs by the availa-
ble isocitrate and reducing the competition between the
substrate isocitrate and inhibitor. The points to note in this
strategy are: (i) a competitive inhibitor of ICLs can serve
the purpose; and (ii) the percentage inhibition of the ICD-
kinase and/or proposed inactivator of ICD2 required here
would be less than required to increase the amount of
active ICD1 and/or ICD2 sufficiently to stop the flux
through the glyoxylate bypass. Mixed inhibition of ICLs
H
tt
t h
b
d
t
i
l t
th
i hibi
the simulation, driving the isocitrate towards the shunt
( l
l t
b
)
th
Th
lt
h
d th t
Effect on the flux through ICDs and ICLs with varying VfICD1
and VfICD2
Figure 2
Effect on the flux through ICDs and ICLs with varying
VfICD1 and VfICD2. Effects of varying (A) VfICD1 alone, (B)
both VfICD1 and VfICD2 simultaneously (abbreviated as VfICDs),
(C) VfICD1 and VfICD2 simultaneously (abbreviated as VfICDs)
with ICL1 reaction removed from the model to simulate
deletion of gene encoding ICL1, (D) VfICD1 and VfICD2 simulta-
neously (abbreviated as VfICDs), with ICL2 reaction removed
from the model to simulate deletion of gene encoding ICL2. Broken line represents the sum of flux through ICD1 and
ICD2, and solid line represents the sum of flux through ICL1
and ICL2. 0
50
100
0
0.5
1
% VfICD1
Fluxes
(mM / min)
0
50
100
0
0.5
1
% VfICDs
Fluxes
(mM / min)
0
50
100
0
0.5
1
% VfICDs
Fluxes
(mM / min)
0
50
100
0
0.5
1
% VfICDs
Fluxes
(mM / min)
A
B
C
D
Competitive inhibition of ICLs by an inhibitor with concen-
tration I and inhibitor constant KI
Figure 3
Competitive inhibition of ICLs by an inhibitor with
concentration I and inhibitor constant KI. Inhibition of
ICL1 and ICL2, with VfICD1 and VfICD2 both kept at (A) 2.5%
of the original values, (B) 5% of the original values. Competitiv
tration I an
Figure 3 p
yI
g
Competitive inhibition of ICLs by an inhibitor with
concentration I and inhibitor constant KI. Inhibition of
ICL1 and ICL2, with VfICD1 and VfICD2 both kept at (A) 2.5%
of the original values, (B) 5% of the original values. Broken
line represents the sum of flux through ICD1 and ICD2, and
solid line represents the sum of flux through ICL1 and ICL2. The effect of inhibitor is shown by varying the ratio of I/KI. Competitive inhibition of ICLs v
Vf
ICIT
K
Vr
SUC
K
GLY
K
ICIT
K
ICL
M ICIT
ICL
M SUC
M GLY
M ICIT
=
−
+
+
1
1
1
,
,
,
,
SUC
K
GLY
K
ICIT
K
SUC
K
SUC
K
GLY
K
M SUC
M GLY
M ICIT
M SUC
M SUC
M G
,
,
,
,
,
,
+
+
+
LY
I
I
K
+
⎛
⎝
⎜
⎜
⎜
⎜⎜
⎞
⎠
⎟
⎟
⎟
⎟⎟ v
Vf
ICIT
K
Vr
SUC
K
GLY
K
ICIT
K
ICL
M ICIT
ICL
M SUC
M GLY
M ICIT
=
−
+
+
1
1
1
,
,
,
,
SUC
K
GLY
K
ICIT
K
SUC
K
SUC
K
GLY
K
M SUC
M GLY
M ICIT
M SUC
M SUC
M G
,
,
,
,
,
,
+
+
+
LY
I
I
K
+
⎛
⎝
⎜
⎜
⎜
⎜⎜
⎞
⎠
⎟
⎟
⎟
⎟⎟
( )
equation 1 An increase was observed in the efficiency of competitive
inhibition of ICL1 and ICL2 with an increase in VfICD1 and
VfICD2 from 2.5% to 5% of the original values, because at
lower VfICD1 and VfICD2, inhibition of ICL1 and ICL2 leads
to an increase in isocitrate concentration, nullifying the
effect of competitive inhibition. The increase in efficiency of competitive inhibition with
an increase in the Vmax of the ICDs leads to an alternative
strategy for killing mycobacteria, i.e. by using a competi-
tive inhibitor of ICL1 and ICL2 along with inhibition of
ICD1-kinase and/or the proposed inactivator of ICD2. Inhibition of ICD1-kinase and/or proposed inactivator of
ICD2 would increase the amount of active ICD1 and/or Page 6 of 11
(page number not for citation purposes) Page 6 of 11
(page number not for citation purposes) Theoretical Biology and Medical Modelling 2006, 3:27 http://www.tbiomed.com/content/3/1/27 p
Competitive inhibition of ICLs by an inhibitor with concen-
tration I and inhibitor constant KI
Figure 3
Competitive inhibition of ICLs by an inhibitor with
concentration I and inhibitor constant KI. Inhibition of
ICL1 and ICL2, with VfICD1 and VfICD2 both kept at (A) 2.5%
of the original values, (B) 5% of the original values. Broken
line represents the sum of flux through ICD1 and ICD2, and
solid line represents the sum of flux through ICL1 and ICL2. Competitive inhibition of ICLs Broken
line represents the sum of flux through ICD1 and ICD2, and
solid line represents the sum of flux through ICL1 and ICL2. The effect of inhibitor is shown by varying the ratio of I/KI. 0
250
500
0
0.25
0.5
I / KI
Fluxes (mM / min)
0
250
500
0
0.5
1
I / KI
Fluxes (mM / min)
A
B 0
50
100
0
0.5
1
% VfICD1
Fluxes
(mM / min)
0
50
100
0
0.5
1
% VfICDs
Fluxes
(mM / min)
0
50
100
0
0.5
1
% VfICDs
Fluxes
(mM / min)
0
50
100
0
0.5
1
% VfICDs
Fluxes
(mM / min)
A
B
C
D
0
250
500
0
0.25
0.5
I / K
Fluxes (mM / min)
A 0
50
100
0
0.5
1
% VfICDs
Fluxes
(mM / min)
0
50
100
0
0.5
1
% VfICDs
Fluxes
(mM / min)
B
D 0
50
100
0
0.5
1
% VfICD1
Fluxes
(mM / min)
0
50
100
0
0.5
1
% VfICDs
Fluxes
(mM / min)
A
C B ICD1
0
50
100
0
0.5
1
% VfICDs
Fluxes
(mM / min)
C Effect on t
and VfICD2
Figure 2 Effect on the flux through ICDs and ICLs with varying VfICD1
and VfICD2
Figure 2
Effect on the flux through ICDs and ICLs with varying
VfICD1 and VfICD2. Effects of varying (A) VfICD1 alone, (B)
both VfICD1 and VfICD2 simultaneously (abbreviated as VfICDs),
(C) VfICD1 and VfICD2 simultaneously (abbreviated as VfICDs)
with ICL1 reaction removed from the model to simulate
deletion of gene encoding ICL1, (D) VfICD1 and VfICD2 simulta-
neously (abbreviated as VfICDs), with ICL2 reaction removed
from the model to simulate deletion of gene encoding ICL2. Broken line represents the sum of flux through ICD1 and
ICD2, and solid line represents the sum of flux through ICL1
and ICL2. g
y g
ICD1
ICD2
g
Effect on the flux through ICDs and ICLs with varying Effect on the flux through ICDs and ICLs with varying
VfICD1 and VfICD2. Competitive inhibition of ICLs Effects of varying (A) VfICD1 alone, (B)
both VfICD1 and VfICD2 simultaneously (abbreviated as VfICDs),
(C) VfICD1 and VfICD2 simultaneously (abbreviated as VfICDs)
with ICL1 reaction removed from the model to simulate
deletion of gene encoding ICL1, (D) VfICD1 and VfICD2 simulta-
neously (abbreviated as VfICDs), with ICL2 reaction removed
from the model to simulate deletion of gene encoding ICL2. Broken line represents the sum of flux through ICD1 and
ICD2, and solid line represents the sum of flux through ICL1
and ICL2. ICD2, i.e. would indirectly cause an increase in the Vmax
of ICD1 and/or ICD2, thus indirectly improving the effi-
ciency of competitive inhibition of the ICLs by the availa-
ble isocitrate and reducing the competition between the
substrate isocitrate and inhibitor. The points to note in this
strategy are: (i) a competitive inhibitor of ICLs can serve
the purpose; and (ii) the percentage inhibition of the ICD-
kinase and/or proposed inactivator of ICD2 required here
would be less than required to increase the amount of
active ICD1 and/or ICD2 sufficiently to stop the flux
through the glyoxylate bypass. Methods
h Considering that we focused on the TCA cycle and glyox-
ylate bypass only, and that the model was built with a
number of permissible assumptions, the results obtained
agree satisfactorily with the experimental data. The obser-
vation that inhibition of ICLs results in no marked
changes in the concentrations of any other metabolites in
the model (result not shown), but to a decrease in the flux
through glyoxylate bypass, indicates that the clearing of The steps in the construction of the kinetic model are
described below. Conclusion
h
d p
y
I
g
Uncompetitive inhibition of ICLs by an inhibitor with
concentration I and inhibitor constant KI. Inhibition of
ICL1 and ICL2, with VfICD1 and VfICD2 both kept at (A) 2.5%
of the original values and (B) 5% of the original values. Bro-
ken line represents the sum of flux through ICD1 and ICD2,
and solid line represents the sum of flux through ICL1 and
ICL2. The effect of inhibitor is shown by varying the ratio of
I/KI. This study constitutes a proof of concept: one can use
kinetic modeling of biochemical pathways to investigate
potential drug targets and to infer the type of inhibition
appropriate for eliminating the pathogen. The study high-
lights the difference between the inhibitor concentrations
required in vitro and in vivo to inhibit the glyoxylate bypass
pathway enzymes. The advantage of this approach to
assessing drug targets is that it facilitates the study of sys-
temic effect(s) of modulating the target enzyme(s) on the
pathway. The applicability of the study is certainly limited
by the approximations and assumptions made while con-
structing the models, but these should be overcome soon
because the required data are accumulating rapidly in this
post-genomic era. v
Vf
ICIT
K
Vr
SUC
K
GLY
K
ICIT
K
ICL
M ICIT
ICL
M SUC
M GLY
M ICIT
=
−
+
+
1
1
1
,
,
,
,
ICIT
K
I
K
SUC
K
I
K
GLY
K
GLY
K
I
K
M ICIT
I
M SUC
I
M GLY
M GLY
I
,
,
,
,
+
+
⎛
⎝⎜
⎞
⎠⎟+
+
+
+
⎛
⎝
⎜
⎜
⎜
⎜⎜
⎞
⎠
⎟
⎟
⎟
⎟⎟
ICIT
K
SUC
K
SUC
K
GLY
K
M ICIT
M SUC
M SUC
M GLY
,
,
,
,
equaation 3
( ) Mixed inhibition of ICLs Inhibition of
ICL1 and ICL2, with VfICD1 and VfICD2 both kept at (A) 2.5%
of the original values and (B) 5% of the original values. Bro-
ken line represents the sum of flux through ICD1 and ICD2,
and solid line represents the sum of flux through ICL1 and
ICL2. The effect of inhibitor is shown by varying the ratio of
I/KI. 0
250
500
0
0.25
0.5
I / KI
Fluxes (mM / min)
0
250
500
0
0.5
1
I / KI
Fluxes (mM / min)
A
B mycobacterial load from macrophages as observed by
McKinney and co-workers [2] can be correlated with a
decrease in the glyoxylate bypass flux, not with accumula-
tion of any toxic metabolite. Conclusion
This study constitutes a proof of concept: one can use
kinetic modeling of biochemical pathways to investigate
potential drug targets and to infer the type of inhibition
appropriate for eliminating the pathogen. The study high-
lights the difference between the inhibitor concentrations
required in vitro and in vivo to inhibit the glyoxylate bypass
pathway enzymes. The advantage of this approach to
assessing drug targets is that it facilitates the study of sys-
temic effect(s) of modulating the target enzyme(s) on the
pathway. The applicability of the study is certainly limited
by the approximations and assumptions made while con-
structing the models, but these should be overcome soon
because the required data are accumulating rapidly in this
post-genomic era. Methods
v
Vf
ICIT
K
Vr
SUC
K
GLY
K
ICIT
K
ICL
M ICIT
ICL
M SUC
M GLY
M ICIT
=
−
+
+
1
1
1
,
,
,
,
ICIT
K
I
K
SUC
K
I
K
GLY
K
GLY
K
I
K
M ICIT
I
M SUC
I
M GLY
M GLY
I
,
,
,
,
+
+
⎛
⎝⎜
⎞
⎠⎟+
+
+
+
⎛
⎝
⎜
⎜
⎜
⎜⎜
⎞
⎠
⎟
⎟
⎟
⎟⎟
ICIT
K
SUC
K
SUC
K
GLY
K
M ICIT
M SUC
M SUC
M GLY
,
,
,
,
equaation 3
( )
Simulation of the effect of inhibition of both ICL1 and ICL2
by 3-nitropropionate (3-NP)
Figure 5
Simulation of the effect of inhibition of both ICL1 and
ICL2 by 3-nitropropionate (3-NP). Broken line repre-
sents the sum of flux through ICD1 and ICD2, and solid line
represents the sum of flux through ICL1 and ICL2. Mixed inhibition of ICLs the simulation, driving the isocitrate towards the shunt
(glyoxylate bypass) pathway. The results showed that a
concentration of 0.38 mM 3-NP was required to reduce
the in vivo flux through glyoxylate bypass by 90%. An
almost 10-fold lower inhibitor concentration was
required for 50% inhibition of ICL1 in vitro compared to
the model (result not shown). A concentration of 0.1 mM,
which experimentally blocks the growth of mycobacteria
in macrophages [2], reduced the flux by 75.8%. It was also
observed that a concentration of 3 mM was required to
reduce the flux by 98.4%. Here, an attempt has been made to simulate the inhibi-
tion of ICLs by 3-nitropropionate (3-NP), a dual-specific
ICL inhibitor that is known to block the growth of myco-
bacteria in macrophages at a concentration of 0.1 mM [2]. 3-NP is competitive against succinate and uncompetitive
against either glyoxylate or isocitrate [20]. The ICL1 and
ICL2 rate equations were therefore modified to account
for mixed inhibition (rate equation for ICL1 is shown in
equation (3); 'I' denotes 3-NP concentration). A similar
equation was used for ICL2. The inhibitor constants (KI)
of 3-NP for ICL1 and ICL2 are 0.003 mM and 0.11 mM
respectively [18]. Using these KI values, simulations were
performed to study the effect of 3-NP concentration on
JICD1 + JICD2 and JICL1 + JICL2 in the model (figure 5). VfICD1
and VfICD2 were kept at 5% of the original values during Page 7 of 11
(page number not for citation purposes) Theoretical Biology and Medical Modelling 2006, 3:27 http://www.tbiomed.com/content/3/1/27 Theoretical Biology and Medical Modelling 2006, 3:27
htt
mycobacterial load fro
McKinney and co-wor
decrease in the glyoxyla
tion of any toxic metab
Conclusion
This study constitutes
kinetic modeling of bi
potential drug targets a
appropriate for elimina
lights the difference be
required in vitro and in
pathway enzymes. Th
Simulation of the effect o
by 3-nitropropionate (3-
Figure 5
Simulation of the effe
ICL2 by 3-nitropropio
sents the sum of flux thr
represents the sum of flu
VfICD2 both kept at 5% o
lation. 0
0
0.5
1
Fluxes (mM / min)
Uncompetitive inhibition of ICLs by an inhibitor with concen-
tration I and inhibitor constant KI
Figure 4
Uncompetitive inhibition of ICLs by an inhibitor with
concentration I and inhibitor constant KI. Mixed inhibition of ICLs VfICD1 and
VfICD2 both kept at 5% of the original values during the simu-
lation. 0
1
2
3
0
0.5
1
I (mM)
Fluxes (mM / min)
Uncompetitive inhibition of ICLs by an inhibitor with concen-
tration I and inhibitor constant KI
Figure 4
Uncompetitive inhibition of ICLs by an inhibitor with
concentration I and inhibitor constant KI. Inhibition of
ICL1 and ICL2, with VfICD1 and VfICD2 both kept at (A) 2.5%
of the original values and (B) 5% of the original values. Bro-
ken line represents the sum of flux through ICD1 and ICD2,
and solid line represents the sum of flux through ICL1 and
ICL2. The effect of inhibitor is shown by varying the ratio of
I/KI. 0
250
500
0
0.25
0.5
I / KI
Fluxes (mM / min)
0
250
500
0
0.5
1
I / KI
Fluxes (mM / min)
A
B Uncompetitive inhibition of ICLs by an inhibitor with concen-
tration I and inhibitor constant KI
Figure 4
Uncompetitive inhibition of ICLs by an inhibitor with
concentration I and inhibitor constant KI. Inhibition of
ICL1 and ICL2, with VfICD1 and VfICD2 both kept at (A) 2.5%
of the original values and (B) 5% of the original values. Bro-
ken line represents the sum of flux through ICD1 and ICD2,
and solid line represents the sum of flux through ICL1 and
ICL2. The effect of inhibitor is shown by varying the ratio of
I/KI. 0
250
500
0
0.25
0.5
I / KI
Fluxes (mM / min)
0
250
500
0
0.5
1
I / KI
Fluxes (mM / min)
A
B 0
1
2
3
0
0.5
1
I (mM)
Fluxes (mM / min) 3 Simulation of the effect of inhibition of both ICL1 and ICL2
by 3-nitropropionate (3-NP)
Figure 5
Simulation of the effect of inhibition of both ICL1 and
ICL2 by 3-nitropropionate (3-NP). Broken line repre-
sents the sum of flux through ICD1 and ICD2, and solid line
represents the sum of flux through ICL1 and ICL2. VfICD1 and
VfICD2 both kept at 5% of the original values during the simu-
lation. mycobacterial load from macrophages as observed by
McKinney and co-workers [2] can be correlated with a
decrease in the glyoxylate bypass flux, not with accumula-
tion of any toxic metabolite. Parameters of the models were identical. A reaction branching from α-ketoglutarate
(αKG = precursor; named SYN in the models) was added
to both the E. coli and M. tuberculosis models to account
for the fraction of αKG utilized for precursor biosynthesis
(as shown by Zhao et al. [15] in E. coli). A set of two reac-
tions catalyzed by α-ketoglutarate decarboxylase (KGD)
and succinic semialdehyde dehydrogenase (SSADH) that
together convert αKG to succinate (SUC) via succinic sem-
ialdehyde (SSA) was also included in the M. tuberculosis
model. The model also accounted for the presence of two
isoforms of ICD [17], ICD1 (Rv3339c) and ICD2
(Rv0066c), and two isoforms of ICL [17,18], ICL1
(Rv0467) and ICL2 (Rv1915 and Rv1916), in M. tubercu-
losis H37Rv strain. The requisite co-enzymes and co-fac-
tors were assumed to be present in large excess so their
effects on the reaction rates in the models were ignored. The reactions considered in the construction of the mod-
els are shown in figure 1. were identical. A reaction branching from α-ketoglutarate
(αKG = precursor; named SYN in the models) was added
to both the E. coli and M. tuberculosis models to account
for the fraction of αKG utilized for precursor biosynthesis
(as shown by Zhao et al. [15] in E. coli). A set of two reac-
tions catalyzed by α-ketoglutarate decarboxylase (KGD)
and succinic semialdehyde dehydrogenase (SSADH) that
together convert αKG to succinate (SUC) via succinic sem-
ialdehyde (SSA) was also included in the M. tuberculosis
model. The model also accounted for the presence of two
isoforms of ICD [17], ICD1 (Rv3339c) and ICD2
(Rv0066c), and two isoforms of ICL [17,18], ICL1
(Rv0467) and ICL2 (Rv1915 and Rv1916), in M. tubercu-
losis H37Rv strain. The requisite co-enzymes and co-fac-
tors were assumed to be present in large excess so their
effects on the reaction rates in the models were ignored. The reactions considered in the construction of the mod-
els are shown in figure 1. The kinetic parameters of the enzymes in the models (see
[additional file 1: Kinetic constants of the enzymes in E. coli model'] and [additional file 2: Kinetic constants of the
enzymes in M. tuberculosis model-1 and model-2]) were
either obtained from publicly available databases, namely
CyberCell Database (CCDB) [25] and BRENDA [26], or
extracted from the literature. Parameters of the models The maximal reaction rates
(Vmax) expressed in nmol/min/mg protein were con-
verted to mM/min by taking the intracellular volume of a
bacterial cell as 2 × 10-12 ml [27] and the total protein con-
tent as 3.2 × 10-10 mg [28]. We were interested in studying
the reactions of the pathway in the catabolic direction, i.e. the direction in which it usually works in the cell; so in
cases where the value of Vr was not available it was taken
as a fraction of Vf (after some trial and error, Vr = Vf/100). In cases where reverse reaction had been monitored and
Vr reported, Vf was taken as equal to Vr. Where a KM was
not available, usually for a reverse reaction, it was
assumed to be equal to 10 × KM of the substrate from
which that product was formed (by the same logic as used
for the Vr values). The metabolites acetyl-CoA, oxaloace-
tate and CoA were considered as boundary metabolites, so
their concentrations were fixed in the simulations. The
initial concentration of each variable metabolite was
taken as 2 × KM for the reaction for which that metabolite
is a substrate (except for those metabolites of which the
concentrations were known; see table 1). Recently, Nathan and co-workers failed to detect α-ketogl-
utarate dehydrogenase (KDH) activity in M. tuberculosis
[13]. They suggested that Rv1248c, annotated as encoding
SucA, the putative E1 component of KDH, encodes KGD
and produces SSA. SSA is then converted by SSADH to
SUC. This new finding was also incorporated into our
study by constructing another model for M. tuberculosis
(named M. tuberculosis model-2) in which the KDH reac-
tion was removed (see figure 1). In the E. coli model, the carbon flux through the pathway
was predicted under two growth conditions, viz. growth
on glucose and acetate as carbon sources. Most enzyme
kinetic parameters are available for E. coli grown on glu-
cose, but it is also necessary to estimate the enzyme kinetic
parameters for the acetate condition. The changes in E. coli
gene expression when growth shifts from glucose to ace-
tate were described by Oh et al. [14]. Assuming that the
change in mRNA level leads to a proportional change in
protein level (enzyme level in our study), there would be
a proportional change in the Vmax of that enzyme
(because Vmax is proportional to the amount of enzyme). Biochemical reactions in the pathway The biochemical reactions of the E. coli TCA cycle and gly-
oxylate bypass were obtained from EcoCyc [21], and those
of M. tuberculosis from MetaCyc [22]. These reactions for
the two organisms from the two different data sources Page 8 of 11
(page number not for citation purposes) Page 8 of 11
(page number not for citation purposes) Theoretical Biology and Medical Modelling 2006, 3:27 http://www.tbiomed.com/content/3/1/27 Parameters of the models Thus, using the Vmax values of enzymes under the glucose
condition and the fold change in gene expression of the
corresponding enzymes, the Vmax values under the ace-
tate condition were calculated. Reaction kinetics Michaelis-Menten equations for one substrate and two-
substrate reactions were used to describe the reaction
kinetics in the models. The reversible Michaelis-Menten
equation for two non-competing product-substrate cou-
ples is shown in equation (4) [23]: v
Vf S
K
S
K
Vr P
K
P
K
S
K
P
K
S
K
P
K
S
S
P
P
S
P
S
P
=
−
+
+
⎛
⎝
⎜
⎞
⎠
⎟
+
+
1
2
1
2
1
1
2
2
1
2
1
2
1
1
2
1
1
2
4
⎛
⎝
⎜
⎞
⎠
⎟
(
)
equation 4
(
)
equation where v = net rate of the reaction; Vf, Vr = maximal rates
of the forward and reverse reaction, respectively; S1, S2 =
concentrations of substrates S1 and S2 respectively; P1, P2 =
concentrations of products P1 and P2 respectively; KS1, KS2,
KP1, KP2 = Michaelis-Menten constants for S1, S2, P1 and P2
respectively. Authors' contributions VKS has contributed in developing the models, analysis
and interpretation of data, and writing the manuscript. IG
was involved in the overall design of this study, critical
analysis and interpretation of the data, and revision of the
draft of the manuscript. Computation Simulations were performed by writing scripts for Jarnac
2.14 [29]. First, steady states were calculated, then – start-
ing from the steady state solution for each model – a time-
dependent simulation was performed to test the stability
of the steady state. We checked that the program Gepasi
3.30 [30] generates the same results as Jarnac given the
same input, but we continued our work with Jarnac
because it offered us the flexibility of writing our own
scripts. Competing interests The author(s) declare that they have no competing inter-
ests. Abbreviations ICL, isocitrate lyase; ACN, aconitase; αKG, α-ketoglutar-
ate; CS, citrate synthase; FUM, fumarase; GLY, glyoxylate;
I, inhibitor concentration; ICD, isocitrate dehydrogenase;
ICIT, isocitrate; JICD1, flux through ICD1; JICD2, flux
through ICD2; JICL1, flux through ICL1; JICL2, flux through
ICL2; KDH, α-ketoglutarate dehydrogenase; KGD, α-
ketoglutarate decarboxylase; KI, inhibitor constant of
inhibitor I; MCA, Metabolic Control Analysis; MDH,
malate dehydrogenase; MS, malate synthase; 3-NP, 3-
nitropropionate; ScAS, succinyl-CoA synthetase; SDH,
succinate dehydrogenase; SSA, succinic semialdehyde;
SSADH, succinic semialdehyde dehydrogenase; SUC, suc-
cinate; TCA, tricarboxylic acid; Vf, maximal rate of the for-
ward reaction; VfICD1, Vmax of the reaction catalyzed by
ICD1 in the forward direction; VfICD2, Vmax of the reac-
tion catalyzed by ICD2 in the forward direction; Vmax,
maximal rate of an enzymatic reaction; Vr, maximal rate
of the reverse reaction. or Vmaxa = n × Vmaxg or Vmaxa = n × Vmaxg or Vmaxa = n × Vmaxg Thus, using the values of n and Vmaxg, Vmaxa values were
calculated and used as parameters for the model to simu-
late the condition of growth on acetate as the carbon
source. The rate of the SYN reaction was maintained at 0.188
times (for glucose condition) and 0.0341 times (for ace-
tate condition) the rate of the ICD reaction in the E. coli
model, as shown experimentally [15]. Owing to the una-
vailability of data for M. tuberculosis, the rate of the SYN
reaction was maintained at that under acetate conditions
in E. coli. The kinetic parameters for M. tuberculosis KDH
were also assumed to be same as for E. coli. As ICL activity
in persistent mycobacteria is 4 times that in the normal
condition [28], the concentration of the ICLs were taken
as 4 times those in normal conditions. Additional material Additional File 1
Kinetic constants of the enzymes in E. coli model. Additional file 1 con-
tains a table that enlist the kinetic constants of the enzymes in E. coli
model. Click here for file
[http://www.biomedcentral.com/content/supplementary/1742-
4682-3-27-S1.pdf] Additional File 1
Kinetic constants of the enzymes in E. coli model. Additional file 1 con-
tains a table that enlist the kinetic constants of the enzymes in E. coli
model. Click here for file
[http://www.biomedcentral.com/content/supplementary/1742-
4682-3-27-S1.pdf] The fluxes computed from the models were expressed in
mM/min. To compare the steady state fluxes of the E. coli
model with experimental findings [15], they were con-
verted to the units in which experimental fluxes were
expressed. The experimental fluxes were expressed relative
to (a) molar glucose uptake or (b) molar acetate uptake
rate depending on the carbon source. The following steps
were used to convert the units: flux through citrate syn-
thase during growth on glucose = 50; flux through citrate
synthase during growth on glucose in the model = 4.187
mM/min; hence, conversion factor x = (50)/(4.187 mM/
min). Using this conversion factor (x), all the fluxes com- Calculation of Vmax from gene expression data Let, the expression levels of a gene g1 under the acetate
and glucose conditions be g1a and g1g respectively. There-
fore, the fold change when growth shifts from glucose to
acetate is n = g1a/g1g. Taking account of the assumption
that a change in mRNA level leads to a proportional
change in protein level, The only reaction in which a different kinetic equation
was used was the reaction: ICIT = SUC + glyoxylate (GLY),
catalyzed by ICL. This is known to occur by an ordered
uni-bi mechanism [24] as described by Bakker et. al [7]. p1a/p1g = g1a/g1g = n equation (5) p1a/p1g = g1a/g1g = n equation (5) Page 9 of 11
(page number not for citation purposes) Page 9 of 11
(page number not for citation purposes) Theoretical Biology and Medical Modelling 2006, 3:27 http://www.tbiomed.com/content/3/1/27 where p1 is the amount of the protein encoded by g1 and
the subscripts 'a' and 'g' denote its level in acetate and glu-
cose respectively puted from the model were converted to the units in
which experimental fluxes were expressed. puted from the model were converted to the units in
which experimental fluxes were expressed. Example: flux through α-ketoglutarate dehydrogenase
(KDH) reaction step in the model = 3.394 mM/min =
(3.394 mM/min) × (x min/mM) = 40.5. Since Vmax = kcat × E (where kcat = turnover number, E =
amount of enzyme catalyzing the reaction) and kcat is a
constant, Vmax α E A similar conversion factor was calculated for growth on
acetate using flux through the citrate synthase step. Therefore, from equation (5), Vmaxa/Vmaxg = n (where Vmaxa, Vmaxg = Vmax of the enzyme in acetate
and glucose respectively) Additional File 1
Kinetic constants of the enzymes in E. coli model. Additional file 1 con-
tains a table that enlist the kinetic constants of the enzymes in E. coli
model.
Click here for file
[http://www.biomedcentral.com/content/supplementary/1742-
4682-3-27-S1.pdf] References 1. Bishai W: Lipid lunch for persistent pathogen. Nature 2000,
406:683-685. 1. Bishai W: Lipid lunch for persistent pathogen. Nature 2000,
406:683-685. 22. Krieger CJ, Zhang P, Mueller LA, Wang A, Paley S, Arnaud M, Pick J,
Rhee SY, Karp PD: MetaCyc: a multiorganism database of met-
abolic pathways and enzymes. Nucleic Acids Res 2004,
32:D438-D442. 2. Muñoz-Elías EJ, McKinney JD: Mycobacterium tuberculosis isoci-
trate lyases 1 and 2 are jointly required for in vivo growth and
virulence. Nat Med 2005, 11:638-644. 3. McAdam RA, Quan S, Smith DA, Bardarov S, Betts JC, Cook FC,
Hooker EU, Lewis AP, Woollard P, Everett MJ, Lukey PT, Bancroft GJ,
Jacobs WR Jr, Duncan K: Characterization of a Mycobacterium
tuberculosis H37Rv transposon library reveals insertions in
351 ORFs and mutants with altered virulence. Microbiology
2002, 148:2975-2986. 23. Segel IH: Enzyme Kinetics: Behavior and analysis of rapid equilibrium and
steady-state enzyme systems New York: Wiley Classics Library Edition,
John Wiley & Sons, Inc; 1993. J
y
24. Reinscheid DJ, Eikmanns BJ, Sahm H: Characterization of the iso-
citrate lyase gene from Cornebacterium glutamicum and bio-
chemical analysis of the enzyme. J Bacteriol 1994,
176:3474-3483. 4. LaPorte DC, Walsh K, Koshland DE Jr: The branch point effect
ultrasensitivity and subsensitivity to metabolic control. J Biol
Chem 1984, 259:14068-14075. 25. Sundararaj S, Guo A, Habibi-Nazhad B, Rouani M, Stothard P, Ellison
M, Wishart DS: The CyberCell Database (CCDB): a compre-
hensive, self-updating, relational database to coordinate and
facilitate in silico modeling of Escherichia coli. Nucleic Acids Res
2004, 32:D293-D295. 5. Balganesh TS, Datta S, Ghosh I: WO 2004/087943 A1. 2004. 6. Fell DA: Metabolic Control Analysis: a survey of its theoreti-
cal and experimental development. Biochem J 1992,
286:313-330. 26. Schomburg I, Chang A, Schomburg D: BRENDA, enzyme data
and metabolic information. Nucleic Acids Res 2002, 30:47-49. 26. Schomburg I, Chang A, Schomburg D: BRENDA, enzyme data
and metabolic information. Nucleic Acids Res 2002, 30:47-49. 27. Mengin-Lecreulx D, Flouret B, van Heijenoort J: Cytoplasmic steps 7. Bakker BM, Michels PAM, Opperdoes FR, Westerhoff HV: Glycoly-
sis in bloodstream form Trypanosoma brucei can be under-
stood in terms of the kinetics of the glycolytic enzymes. J Biol
Chem 1997, 272:3207-3215. and metabolic information. Nucleic Acids Res 2002, 30:47-49. 27. Mengin-Lecreulx D, Flouret B, van Heijenoort J: Cytoplasmic steps
of peptidoglycan synthesis in Escherichia coli. J Bacteriol 1982,
151:1109-1117. 8. Additional File 2 16. Walsh K, Koshland DE Jr: Branch point control by the phospho-
rylation state of isocitrate dehydrogenase a quantitative
examination of fluxes during a regulatory transition. J Biol
Chem 1985, 260:8430-8437. Kinetic constants of the enzymes in M. tuberculosis model-1 and
model-2. Additional file 2 contains a table that enlist kinetic constants of
the enzymes in M. tuberculosis models. Click here for file
[h
//
bi
d
l
/
/
l
/1742 Kinetic constants of the enzymes in M. tuberculosis model-1 and
model-2. Additional file 2 contains a table that enlist kinetic constants of
the enzymes in M. tuberculosis models. Click here for file 17. Cole ST, Brosch R, Parkhill J, Garnier T, Churcher C, Harris D, Gor-
don SV, Eiglmeier K, Gas S, Barry CE, Tekaia F, Badcock K, Basham
D, Brown D, Chillingworth T, Connor R, Davies R, Devlin K, Feltwell
T, Gentles S, Hamlin N, Holroyd S, Hornsby T, Jagels K, Krogh A,
Mclean J, Moule S, Murphy L, Oliver K, Osborne J, Quail MA, Rajan-
dream M-A, Rogers J, Rutter S, Seeger K, Skelton J, Squares R,
Squares S, Sulston JE, Taylor K, Whitehead S, Barrell BG: Decipher-
ing the biology of Mycobacterium tuberculosis from the com-
plete genome sequence. Nature 1998, 393:537-544. Acknowledgements 18. 18. Höner zu Bentrup K, Miczak A, Swenson DL, Russell DG: Charac-
terization of activity and expression of isocitrate lyase in
Mycobacterium avium and Mycobacterium tuberculosis. J Bacte-
riol 1999, 181:7161-7167. We thank Dr. S. Datta, AstraZeneca R&D, Bangalore, and Dr. S. Sinha,
Centre for Cellular and Molecular Biology, Hyderabad, for interesting dis-
cussions and help provided during this work. We are also grateful to Dr. V. Shankar, Institute of Bioinformatics and Biotechnology, University of Pune,
Pune for his inputs on the presentation and flow of the manuscript. V. K. Singh would like to thank Department of Biotechnology, Government of
India, for providing the Junior Research Fellowship. We are thankful to the
referees for painstakingly reading the manuscript and giving valuable sugges-
tions. 19. Yang CR, Shapiro BE, Hung SP, Mjolsness ED, Hatfield GW: A math-
ematical model for the branched chain amino acid biosyn-
thetic pathways of Escherichia coli K12. J Biol Chem 2005,
280:11224-11232. 20. Schloss JV, Cleland WW: Inhibition of isocitrate lyase by 3-
nitropropionate, a reaction-intermediate analogue. Biochem-
istry 1982, 21:4420-4427. 21. Keseler IM, Collado-Vides J, Gama-Castro S, Ingraham J, Paley S,
Paulsen IT, Peralta-Gil M, Karp PD: EcoCyc: a comprehensive
database resource for Escherichia coli. Nucleic Acids Res 2005,
33:D334-D337. 15.
Zhao J, Shimizu K: Metabolic flux analysis of Escherichia coli K12
grown on 13C-labeled acetate and glucose using GC-MS and Additional File 1 Page 10 of 11
(page number not for citation purposes) Page 10 of 11
(page number not for citation purposes) Theoretical Biology and Medical Modelling 2006, 3:27 http://www.tbiomed.com/content/3/1/27 powerful flux calculation method. J Biotechnol 2003,
101:101-117. Additional File 2
Kinetic constants of the enzymes in M. tuberculosis model-1 and
model-2. Additional file 2 contains a table that enlist kinetic constants of
the enzymes in M. tuberculosis models. Click here for file
[http://www.biomedcentral.com/content/supplementary/1742-
4682-3-27-S2.pdf] Additional File 2
Kinetic constants of the enzymes in M. tuberculosis model-1 and
model-2. Additional file 2 contains a table that enlist kinetic constants of
the enzymes in M. tuberculosis models. Click here for file
[http://www.biomedcentral.com/content/supplementary/1742-
4682-3-27-S2.pdf] References Bakker BM, Michels PAM, Opperdoes FR, Westerhoff HV: What
controls glycolysis in bloodstream form Trypanosoma brucei? J Biol Chem 1999, 274:14551-14559. 28. Wayne LG, Lin KY: Glyoxylate metabolism and adaptation of
Mycobacterium tuberculosis to survival under anaerobic con-
ditions. Infect Immun 1982, 37:1042-1049. 9. Eisenthal R, Cornish-Bowden A: Prospects for antiparasitic
drugs the case of Trypanosoma brucei, the causative agent of
African sleeping sickness. J Biol Chem 1998, 273:5500-5505. 29. Sauro HM, Hucka M, Finney A, Wellock C, Bolouri H, Doyle J, Kitano
H: Next generation simulation tools: The Systems Biology
Workbench and BioSPICE integration. OMICS 2003,
7:355-372. p
g
J
10. Arfin SM, Long AD, Ito ET, Tolleri L, Riehle MM, Paegle ES, Hatfield
GW: Global gene expression profiling in Escherichia coli K12
the effects of integration host factor. J Biol Chem 2000,
275:29672-29684. 30. Mendes P, Kell DB: Non-linear optimization of biochemical
pathways: applications to metabolic engineering and param-
eter estimation. Bioinformatics 1998, 14:869-883. Futcher B, Latter GI, Monardo P, McLaughlin CS, Garrels J 11. Futcher B, Latter GI, Monardo P, McLaughlin CS, Garrels JI: A sam-
pling of the yeast proteome. Mol Cell Biol 1999, 19:7357-7368. 12. Ideker T, Thorsson V, Ranish JA, Christmas R, Buhler J, Eng JK, Bum-
garner R, Goodlett DR, Aebersold R, Hood L: Integrated genomic
and proteomic analyses of a systematically perturbed meta-
bolic network. Science 2001, 292:929-934. 13. Tian J, Bryk R, Itoh M, Suematsu M, Nathan C: Variant tricarboxy-
lic acid cycle in Mycobacterium tuberculosis: Identification of
α-ketoglutarate decarboxylase. Proc Natl Acad Sci USA 2005,
102:10670-10675. 14. Oh MK, Rohlin L, Kao KC, Liao JC: Global expression profiling of
acetate-grown
Escherichia
coli. J
Biol
Chem
2002,
277:13175-13183. 15. Zhao J, Shimizu K: Metabolic flux analysis of Escherichia coli K12
grown on 13C-labeled acetate and glucose using GC-MS and Page 11 of 11
(page number not for citation purposes)
|
https://openalex.org/W4312797022
|
https://www.sworldjournal.com/index.php/swj/article/download/swj15-02-025/2882
|
Ukrainian
| null |
МЕТОДОЛОГІЧНІ АСПЕКТИ ОРГАНІЗАЦІЇ ПЕДАГОГІЧНОЇ ПРАКТИКИ МАГІСТРАНТІВ СПЕЦІАЛЬНОСТІ 014.09 - СО_ІНФОРМАТИКА
|
Mir nauki i innovacij
| 2,019
|
cc-by
| 1,811
|
Issue 15 / Part 2
le/view/swj15-02-025 Issue 15 / Part 2
le/view/swj15-02-025 SWorldJournal https://www.sworldjournal.com/index.php/swj/article/view/swj15-02-025
DOI: 10.30888/2663-5712.2022-15-02-025 https://www.sworldjournal.com/index ISSN 2663-5712 Д
METHODOLOGICAL ASPECTS OF THE ORGANIZATION OF THE
PEDAGOGICAL PRACTICE OF MASTER'S STUDENTS,
SPECIALTY 014.09 – SE_(INFORMATICS)
МЕТОДОЛОГІЧНІ АСПЕКТИ ОРГАНІЗАЦІЇ ПЕДАГОГІЧНОЇ ПРАКТИКИ
МАГІСТРАНТІВ СПЕЦІАЛЬНОСТІ 014.09 - СО_ІНФОРМАТИКА
Li
h
k V A / Лі
В А Litovchenko V.A. / Літовченко В.А. d.f.-m.s., prof. / д.фіз.-мат.н., проф. ORCID: https://orcid.org/0000-0001-9848-3167
Yu. Fed’kovych Chernivtsi National University, Ukraine,
Chernivtsi, Kotsubins’ky Str., 2, 58021
Чернівецький національний університет імені Юрія Федьковича, Україна,
Чернівці, вул. Коцюбинського, 2, 58012 Litovchenko V.A. / Літовченко В.А. d.f.-m.s., prof. / д.фіз.-мат.н., проф. ORCID: https://orcid.org/0000-0001-9848-3167
Yu. Fed’kovych Chernivtsi National University, Ukraine,
Chernivtsi, Kotsubins’ky Str., 2, 58021
Чернівецький національний університет імені Юрія Федьковича, Україна,
Чернівці, вул. Коцюбинського, 2, 58012 Анотація. Розглянуто важливі питання організації та проведення педагогічної
практики здобувачів вищої освіти другого (магістерського) рівня зі спеціальності 014.09
Середня освіта (Інформатика). При цьому, сформульовано мету, зміст і основні завдання
практики та запропоновано оптимальну програму її проходження. р
р
у р
р
у
р
Ключові слова: педагогічна практика, середня освіта, магістрант, вчитель
інформатики. www.sworldjournal.com Вступ Особливе місце в освітньому процесі займає виробнича практика, як засіб
закріплення теоретичних знань та опрацювання практичних навичок здобувачів
вищої освіти, вона є вагомою складовою системи навчання у вищому
навчальному закладі. Практика спрямована на закріплення теоретичних знань
студентів, отриманих ними за час навчання, набуття та удосконалення
практичних навичок і вмінь за відповідною спеціальністю. Для здобувачів вищої освіти другого (магістерського) рівня зі спеціаль-
ності 014.09 – Середня освіта (Інформатика) виробнича практика, зазвичай,
називається педагогічною. Ця практика покликана сприяти поглибленню та
закріпленню інтересу до професії вчителя інформатики, забезпечувати умови
для вивчення передового педагогічного досвіду, формувати у майбутнього
фахівця власний творчий підхід до педагогічної діяльності, а також, виховати в
нього потребу систематично оновлювати свої знання, вести наукові дослід-
ження та творчо застосовувати їх у практичній діяльності. У зв’язку з цим, при
плануванні та організації проведення цього виду практики, задля забезпечення
максимального навчального ефекту, необхідно вирішувати ряд важливих
питань, таких, як [1-3]: «де і коли проводити педагогічну практику», «який
обсяг кредитів відводити для неї», «які знання, вміння та навики повинні
набути практиканти», «які завдання мають виконувати практиканти для
досягнення необхідних компетентностей», «яка форма звітності» тощо. У даній роботі пропонується варіант відповідей на ці питання для
освітньо-професійної програми «Інформатика та математика», за якою
здійснюється підготовка здобувачів вищої освіти другого (магістерського) рівня
зі спеціальності 014.09 – Середня освіта (Інформатика) в Чернівецькому ISSN 2663-5712 www.sworldjournal.com 66 Issue 15 / Part 2 Issue 15 / Part 2 SWorldJournal національному університеті імені Юрія Федьковича (ЧНУ). При цьому,
враховується досвід організації та проведення педагогічної практики за ОПП1
галузі знань – 01 Освіта/Педагогіка [4, 5]. у
Методологія педагогічної практики У ЧНУ педагогічна практика здобувачів вищої освіти галузі знань – 01
«Освіта/Педагогіка», спеціальності 014.09 «Середня освіта (Інформатика)», є
обов’язковим компонентом з циклу загальної підготовки ОПП «Інформатика та
математика» з надання освітніх послуг для підготовки фахівців другого
(магістерського) рівня. Регламентується «Положенням про практику в ЧНУ». Загальний обсяг практики становить 15 кредитів ЄКТС2. Проходження цієї
практики заплановане в третьому семестрі навчання і триває 2 місяці. Практика
обов’язково відбувається на базі ЗЗСО3. Проходження практики є важливим
етапом у системі практичної підготовки здобувача освіти до майбутньої його
професійної діяльності. 2 ЄКТС – європейська кредитно трансферна система
3 3 ЗЗСО -загальні заклади середньої освіти 1 ОПП – освітньо-професійна програма
2 Вступ Мета практики – оволодіння здобувачами вищої освіти сучасними
методами, формами організації праці в галузі їхньої майбутньої професії,
формування у них на базі одержаних в університеті знань, професійних умінь і
навичок для прийняття самостійних рішень під час конкретної роботи в
реальних виробничих умовах, виховання потреби систематично оновлювати
свої знання, вести наукові дослідження та творчо застосовувати отримані
знання в практичній діяльності. Завдання педагогічної практики – закріпити та поглибити теоретичні
знання, отримані здобувачами вищої освіти у процесі вивчення теоретичних
дисциплін
професійної
підготовки,
сформувати
практичні
навички
зі
спеціальності, а також, збір фактичного матеріалу для виконання дипломних і
дослідних робіт. Зміст практики охоплює основні етапи функціонування ЗЗСО – бази
практики, включає в себе навчальну, виховну, методичну, науково-дослідну та
громадську роботу і складається з таких видів діяльності здобувача вищої
освіти: - ознайомлення з функціональними обов’язками керівництва закладу,
знайомство
з
матеріально-технічною
базою
навчального
закладу,
традиціями його колективу, відвідування уроків, виховних заходів; - вивчення методології ведення обліку психологічних особливостей у
навчальній та виховній роботі вчителя; - поглиблене вивчення навчальної діяльності вчителя інформатики; - освоєння методики проведення вчителями закладу інтегрованих уроків з
використанням інформатики; - вивчення передового досвіду та творчої діяльності провідних вчителів-
предметників закладу проходження практики, їх науково-методичного
доробку; - проведення навчальної та позакласної роботи з предмету спеціал ОПП – освітньо-професійна програма 2 ЄКТС – європейська кредитно трансферна система
3 3 ЗЗСО -загальні заклади середньої освіти www.sworldjournal.com ISSN 2663-5712 67 Issue 15 / Part 2 SWorldJournal Issue 15 / Part 2 - вивчення досвіду виховної роботи класного керівника; - вивчення планів роботи методичного об’єднання вчителів; - участь у роботі педагогічної ради, методичного об’єднання; - відвідування та аналіз навчальних занять і виховних заходів колег
практикантів; - організація в старших класах роботи з самовиховання. Програма педагогічної практики складається із наступних модулів. Модуль 1. «Організаційна діяльність майбутнього магістра інформатики». У ньому передбачається: Модуль 1. «Організаційна діяльність майбутнього магістра інформатики». Вступ У ньому передбачається: - участь практиканта в настановчій конференції; - знайомство з функціональними обов’язками керівництва, з - педагогічним колективом та внутрішнім розпорядком бази практик - знайомство зі специфікою оформлення та обладнанням навчальних
кабінетів і лабораторій; - аналіз навчальних планів учителів та виховних планів класних керівників
закладу; - формування індивідуального плану роботи на період проходження
практики, в якому передбачається навчальна, виховна, методична та
дослідницька робота; - своєчасне оформлення та подача керівнику практики звітної докум - своєчасне оформлення та подача керівнику практики звітної документації;
ас
і с
о ій о фере
ії (за
с і) ра - своєчасне оформлення та подача керівнику практики звітної документації;
- участь у підсумковій конференції (захисті) практики. своєчасне оформлення та подача керівнику практики звітної документації;
- участь у підсумковій конференції (захисті) практики. Модуль 2. «Педагогічна діяльність майбутнього магістра інформатики». Який полягає у: Модуль 2. «Педагогічна діяльність майбутнього магістра інформатики». Який полягає у: - виконанні практикантом обов’язків помічника класного керівника
учнівського класу; - ведення особових справ учасників освітнього процесу, обліку вікових та
індивідуальних
особливостей
учнів,
виявлення
особливостей
педагогічного спілкування у колективі; - робота з батьками і громадськістю, збір інформації для написання
психолого-педагогічної характеристики на учня та клас; - складання
письмового
аналізу
відвідуваних
комбінованих
занять
педагогічних працівників закладу; - вивчення передового досвіду та творчої діяльності провідних вчителів-
предметників бази практики, їх науково-методичного доробку; - вивчення планів роботи методичного об’єднання вчителів, класних
керівників, вихователів; участь у роботі педагогічної ради, методичного
об’єднання, семінарів класних керівників бази практики. дуль 3. «Професійна діяльність майбутнього магістра інформатики». Модуль 3. «Професійна діяльність майбутнього магістра інформатики». ISSN 2663-5712
4 ТЗН – технічні засоби навчання 4 ТЗН – технічні засоби навчання Вступ У ньому передбачено: - знайомство та аналіз шкільної програми й підручників з інформатики в
старших класах; - опрацювання наявної навчально-методичної літератури з предмету
«Інформатика»; ISSN 2663-5712 www.sworldjournal.com 68 Issue 15 / Part 2 SWorldJournal Issue 15 / Part 2 - поглиблене вивчення та аналіз навчальної діяльності вчителя інформатики
в старших класах: методика опитування, виклад нового матеріалу,
організація самостійної роботи з учнями на уроці, методи активізації
пізнавальної діяльності учнів, особливості використання ТЗН4 та
спеціалізованих програмних продуктів, ознайомлення з обладнанням
комп’ютерного класу, програмним забезпеченням; - допомога вчителю інформатики з розробкою навчальних планів,
підготовкою дидактичних і роздаткових матеріалів, наочних посібників,
планів занять предметного гуртка тощо; - розробка з допомогою викладача-методиста розгорнутих планів-конспектів
уроків з інформатики для старших класів; - добір змісту навчального матеріалу, методів, прийомів і засобів навчання
для кожного уроку з інформатики відповідно до його мети і завдань; ур
у
ф р
- проведення не менше 4 пробних уроки та 2 залікових уроки; - технічне і програмне обслуговування та підготовка комп’ютерного класу
для уроків з інформатики; - освоєння методики проведення вчителями бази практики інтегрованих
уроків з використанням інформатики; - відвідування уроків, факультативних занять за фахом колег-практикантів,
участь у їх обговоренні з метою вдосконалення педагогічної майстерності;
ї
б
ф
і ф - проведення позакласної роботи та факультативних занять з інформ Модуль 4. «Виховна діяльність майбутнього магістра інформатики». Він полягає в: Модуль 4. «Виховна діяльність майбутнього магістра інформатики». Він полягає в: - плануванні та проведенні виховної роботи з учнями закріпленого класу,
спрямованої на покращення їх успішності й дисципліни, патріотичного
виховання та виховання позитивних особистісних якостей кожного учня,
а також, роботи учнівського самоврядування; - проведенні виховних заходів, зокрема, один заліковий, у закріпленому
учнівському класі; - відвідування виховних заходів колег-практикантів та участь у їх
обговоренні. Під час проходження педагогічної практики студенти повинні виконувати і
дотримуватися прийнятих на базі практики правил охорони праці та безпеки
життєдіяльності з обов’язковим проходженням ними відповідних інструктажів. Проходження педагогічної практики завершується її захистом на кафедрі,
що забезпечує проведення практики. Результати підсумкового контролю зано-
сяться у відомість обліку успішності, залікову книжку здобувача вищої освіти. Проходження педагогічної практики завершується її захистом на кафедрі,
що забезпечує проведення практики. Результати підсумкового контролю зано-
сяться у відомість обліку успішності, залікову книжку здобувача вищої освіти. у
Висновки Висновки Педагогічний компонент практики сприяє формуванню особистості
майбутнього педагога з інформатики ЗЗСО, його готовності до безпосереднього
застосування теоретичних знань у повсякденній педагогічній діяльності,
засвоєння ним педагогічної майстерності, набуття, вивчення та аналіз
педагогічного досвіду. ISSN 2663-5712 www.sworldjournal.com 69 Issue 15 / Part 2 Issue 15 / Part 2 SWorldJournal Issue 15 / Part 2 Література: 1. Положення про проведення практики студентів вищих навчальних
закладів України, затвердженого наказом Міністерства освіти України від 8
квітня 1993 р. № 93. – ЗНД. – Київ, 1993. - №173. – 24 с. 1. Положення про проведення практики студентів вищих навчальних
закладів України, затвердженого наказом Міністерства освіти України від 8
квітня 1993 р. № 93. – ЗНД. – Київ, 1993. - №173. – 24 с. р
2. Гончаренко С.У. Педагогічні дослідження: методологічні поради
молодим науковцям / Семен Устимович Гончаренко. – К.; Вінниця: ДОВ
Вінниця, 2008. – 278 с. 2. Гончаренко С.У. Педагогічні дослідження: методологічні поради
молодим науковцям / Семен Устимович Гончаренко. – К.; Вінниця: ДОВ
Вінниця, 2008. – 278 с. 3. Методика викладання у вищій школі : навчальний посібник / О.В. Малихін, І. Г. Павленко, О.О. Лаврентьєва, Г.І. Матукова. – Київ: КНТ, 2014. –
262 с. 4. Сірант Н.П. Методичні рекомендації до проходження педагогічної
практики: навчально-методичний посібник / Нелі Сірант. –Львів, 2013 – 116 с. 4. Сірант Н.П. Методичні рекомендації до проходження педагогічної
практики: навчально-методичний посібник / Нелі Сірант. –Львів, 2013 – 116 с. 5. Методичні рекомендації до проходження педагогічної практики для
здобувачів вищої освіти другого (магістерського) рівня за освітньо-
професійною програмою «Педагогіка вищої школи» спеціальності 011 «Освітні,
педагогічні науки» денної форми навчання [Електронне навчання] / Кочубей
А.В., Шевчук Т.Є. – Рівне : НУВГП, 2019. – 44 с. Abstract. Considered important issues of organizing and conducting pedagogical practice of
students of higher education of the second (master's) level in the specialty 014.09 Secondary
education (Informatics). At the same time, the purpose, content and main tasks of the practice were
formulated and the optimal program for its completion was proposed. Key words: pedagogical practice, secondary education, master's student, computer science
teacher. Стаття відправлена: 25.09.2022 р. © Літовченко В. ISSN 2663-5712 ISSN 2663-5712 www.sworldjournal.com 70
|
https://openalex.org/W2171123113
|
https://europepmc.org/articles/pmc4246654?pdf=render
|
English
| null |
“No level up!â€: no effects of video game specialization and expertise on cognitive performance
|
Frontiers in psychology
| 2,014
|
cc-by
| 8,693
|
ORIGINAL RESEARCH ARTICLE
bli h d 28 N
b
2014 ORIGINAL RESEARCH ARTICLE
bli h d 28 N
b
2014 published: 28 November 2014
doi: 10.3389/fpsyg.2014.01337 1In this study, the term “video game” refers to any published computer or con-
sole video game from 1952 to the present day. The term “video game player” “No level up!”: no effects of video game specialization and
expertise on cognitive performance Fernand Gobet 1*, Stephen J. Johnston 2, Gabriella Ferrufino 3, Matthew Johnston 3, Michael B. Jones 3,
Antonia Molyneux 3, Argyrios Terzis 3 and Luke Weeden 3
1 Psychological Sciences, University of Liverpool, Liverpool, UK
2 Department of Psychology, Swansea University, Swansea, UK
3 Department of Psychology, Brunel University, Uxbridge, UK Fernand Gobet 1*, Stephen J. Johnston 2, Gabriella Ferrufino 3, Matthew Johnston 3, Michael B. Jones 3,
Antonia Molyneux 3, Argyrios Terzis 3 and Luke Weeden 3
1 Psychological Sciences, University of Liverpool, Liverpool, UK
2 Department of Psychology, Swansea University, Swansea, UK
3 Department of Psychology, Brunel University, Uxbridge, UK Previous research into the effects of action video gaming on cognition has suggested
that long term exposure to this type of game might lead to an enhancement of cognitive
skills that transfer to non-gaming cognitive tasks. However, these results have been
controversial. The aim of the current study was to test the presence of positive cognitive
transfer from action video games to two cognitive tasks. More specifically, this study
investigated the effects that participants’ expertise and genre specialization have on
cognitive improvements in one task unrelated to video gaming (a flanker task) and
one related task (change detection task with both control and genre-specific images). This study was unique in three ways. Firstly, it analyzed a continuum of expertise
levels, which has yet to be investigated in research into the cognitive benefits of
video gaming. Secondly, it explored genre-specific skill developments on these tasks
by comparing Action and Strategy video game players (VGPs). Thirdly, it used a very
tight experiment design, including the experimenter being blind to expertise level and
genre specialization of the participant. Ninety-two university students aged between 18
and 30 (M = 21.25) were recruited through opportunistic sampling and were grouped
by video game specialization and expertise level. While the results of the flanker task
were consistent with previous research (i.e., effect of congruence), there was no effect of
expertise, and the action gamers failed to outperform the strategy gamers. Additionally,
contrary to expectation, there was no interaction between genre specialization and image
type in the change detection task, again demonstrating no expertise effect. The lack
of effects for game specialization and expertise goes against previous research on the
positive effects of action video gaming on other cognitive tasks. INTRODUCTION for theories based on the notion that expertise in great part relies
on domain-specific perceptual knowledge [e.g., chunking theory
(Chase and Simon, 1973) and template theory (Gobet and Simon,
1996)]. While perceptual knowledge enables fluid behavior in the
original domain, it is of little use in other domains as it does
not match the new environment. Research on chess has provided
considerable support for this prediction. Chess players’ percep-
tual skills do not extend to visual memory for shapes (Waters
et al., 2002), nor do their planning capabilities transfer to the
Tower of London, a task measuring executive function and plan-
ning (Unterrainer et al., 2011). Moreover, contrary to widespread
belief, there is no robust empirical evidence that playing chess
improves scholastic abilities (Gobet and Campitelli, 2006). Transfer—the extent to which skills generalize—is an important
theoretical concept that has serious practical implications. In a
classic article, Thorndike and Woodworth (1901) propounded
their theory of “identical elements,” according to which transfer
from a first domain to a second domain is possible only when
the components of the skills required in each domain overlap. For example, a pianist can use their knowledge of music theory
to understand a violin concerto, and a mathematician will under-
stand the differential equations of an economics paper better than
a person without background in mathematics. But even in these
cases, transfer is far from perfect; for example, the pianist will not
be able play the violin concerto itself without extensive additional
training. One of the rare domains in which evidence of far transfer
has been found is playing action video games (e.g., Green et al.,
2009)1. Repeated playing of this kind of game has been reported Keywords: change detection task, expertise, flanker task, transfer, video game playing Edited by: Edited by:
David Zachary Hambrick, Michigan
State University, USA David Zachary Hambrick, Michigan
State University, USA
Reviewed by:
Brooke Noel Macnamara, Princeton
University, USA
Mark W. Becker, Michigan State
University, USA
Elizabeth Jane Meinz, Southern
Illinois University Edwardsville, USA
*Correspondence:
Fernand Gobet, Psychological
Sciences, University of Liverpool,
Bedford Street South, Liverpool
L69 7ZA, UK
e-mail: fgobet@liv.ac.uk Reviewed by:
Brooke Noel Macnamara, Princeton
University, USA
Mark W. Becker, Michigan State
University, USA
Elizabeth Jane Meinz Southern *Correspondence:
Fernand Gobet, Psychological
Sciences, University of Liverpool,
Bedford Street South, Liverpool
L69 7ZA, UK
e-mail: fgobet@liv.ac.uk FAR-TRANSFER Action VGPs showed faster reaction times to detect tar-
gets and a reduced effect of distractor interference, leading the
authors to conclude that the action video gamers had better top-
down attentional control, with the consequence that they spend
less time processing irrelevant distractors. Consistent with this
result, Hubert-Wallander et al. (2011) found that, compared with
non-action gamers, action gamers demonstrate superior visual
selective attention as measured in a visual search task, with the
greatest benefit occurring at the highest cognitive loads (largest
search arrays). Additional evidence comes from neuroimaging
where differences in brain activation support the idea that long-
term video game playing impacts on cortical functioning. For
example, using functional magnetic resonance imagery (fMRI),
Bavelier et al. (2012) compared a group of action and non-action
video gamers on a task involving locating a target stimulus under
conditions of increasing distractor load. In addition to overall
faster reaction times, compared to the non-video game playing
group, the VGPs showed little increase in the level of activation
in a network of fronto-parietal sites as distractor load increased. This fronto-parietal network has commonly been associated with
attentional processing (Ptak, 2012). These data were taken to
suggest that the VGPs were more efficient in their allocation of
attentional resources such that the cortical sites deploying atten-
tion were able to filter out the distracting information more
easily and therefore showed less load dependent increases in
cortical activity, supporting the behavioral finding of Chisholm
et al. (2010). The proposed superior attentional resource alloca-
tion of VGPs (e.g., Hubert-Wallander et al., 2011; Bavelier et al.,
2012), that may be at the heart of the observed enhancement
of stimulus processing and reduced distractor interference, has
now been observed in a number of experiments examining video
game expertise based improvements in spatial selective attention
(Green and Bavelier, 2003; Feng et al., 2007; Spence et al., 2009),
distractor inhibition (Chisholm et al., 2010; Hubert-Wallander
et al., 2011; Mishra et al., 2011; Bavelier et al., 2012), enhanced
image search (Dye et al., 2009) and target detection (Castel et al.,
2005; Dye et al., 2009). An experimental task that is homologous to the task require-
ments of many action video games is the change detection task. In the change detection task, participants are asked to moni-
tor a visual display for a small change that they indicate finding
via a keypress response. For example, Clark et al. FAR-TRANSFER In line with Thorndike and Woodworth’s (1901) hypothesis, most
theories of expertise predict that transfer from one domain to
another (far-transfer) will be difficult. This is particularly the case 1In this study, the term “video game” refers to any published computer or con-
sole video game from 1952 to the present day. The term “video game player” November 2014 | Volume 5 | Article 1337 | 1 www.frontiersin.org www.frontiersin.org Player 1 level up! Gobet et al. Flanker Compatibility Task (Eriksen and Eriksen, 1974; Bavelier
et al., 2012). The Flanker Task requires participants to ignore
salient laterally presented distractors while making responses to a
centrally presented target stimulus. Mishra et al. (2011) employed
a flanker task to examine whether there was any neuroelec-
trophysiological evidence of VGP showing enhanced distractor
inhibition. The results showed that, behaviourally, VGPs were bet-
ter able to ignore flanking items competing for attention with
a central stimulus than nVGPs and that this increase in behav-
ioral performance was associated with a greater P300 component
in the ERP. The P300 electrophysiological component has been
associated with perceptual discrimination and decision-making
(Picton, 1992; Mishra et al., 2011). These data, together, were
taken to support the hypothesis that VGP were better at filtering
out the distractor stimuli leading to improved perceptual deci-
sion making. Lavie (1995) also reported that through extensive
game play VGPs gain the ability to identify task-irrelevant flankers
before further processing stimuli. This indicates that VGPs pos-
sess an enhanced capability to logically filter information for
relevance before attempting to ignore distractors, rather than
trying to process everything at once as nVGPs do. to lead to improvements in perceptual and attentional processes
and to reduce reaction time in other tasks where one must be
both fast and accurate (e.g., Green et al., 2009; Bavelier et al.,
2012). One of the main advantages proposed to be the result of
habitual action video game playing is that of a more efficient
attentional system. For example, Chisholm et al. (2010) com-
pared action vs. non-action video game players (VGPs) on an
attentional capture task where participants searched for a target
that could appear in isolation or with a salient task-irrelevant
distractor. (shortened to VGP) refers to any individual who plays these games, and the
term “non-video game player” (shortened to nVGP) refers to any individual
who partakes in video game play for less than 1 h per week. FAR-TRANSFER Gobet et al. Boot et al. (2011) note that experts and novices have different
expectations about their performance, which is likely to affect
their behavior due to demand characteristics. They also observe
that playing video games might affect the kind of strategies that
are being used rather than basic perceptual or cognitive capacities. Finally, as some studies have failed to find differences between
VGPs and nVGPs, the literature might suffer from a file-drawer
problem. Kristjánsson (2013) note that, in many training stud-
ies, the control groups do not improve their performance on the
tasks of interest, as one would expect, based on the extensive lit-
erature on learning, given the test-retest methodology used. In
addition, the results might be affected by gender differences, as it
is difficult to find expert female VGPs. Both Boot et al. (2011) and
Kristjánsson (2013) note the necessity to carry out independent
replications. establishing whether differing levels of video gaming expertise
vary with performance on cognitive tasks. Thus, the first aim of
this study was to test the hypothesis that, as the level of exper-
tise increases, task accuracy increases, and reaction times become
faster. In addition, a number of studies compare VGPs who iden-
tify as “Action” players to nVGPs, but as yet there has been no
research into whether the skills demonstrated by action players
cross over into other genres, such as strategy games, or indeed
if each genre improves different skills. Data from a consumer
survey by the Entertainment Software Association found that
action and strategy games proved popular with both console and
computer VGPs, and so these two genres were chosen as a vari-
able to test the hypothesis (Entertainment Software Association,
2012). The second aim of the study was thus to test to what extent
different video-game genre specialization tap into different cog-
nitive abilities. Action games typically involve fast-paced game-
play, such as “Call of Duty: Modern Warfare 3” (Infinity Ward,
2011), which was the best selling action console video game in
2011 (Entertainment Software Association, 2012). It was pre-
dicted that the speed of gameplay will heighten action VGPs
speeded response times to stimuli other than those normally
responded to in a VG, as shown by Green and Bavelier (2003). Near transfer Research has also investigated whether transfer occurs between
sub-disciplines of the same field (near transfer). For example,
do physicians specializing in neurosurgery generalize their skills
when solving problems from pediatrics, or do chess players spe-
cializing in specific openings (i.e., the first moves of the game)
maintain their skill level when making decisions in board posi-
tions in which they are not specialized? Several studies have addressed this issue in medicine (Rikers
et al., 2002), political science (Chiesi et al., 1979), and the design
of experiments (Schunn and Anderson, 1999). The pattern of
results suggests that experts fall back on general heuristics when
they cannot use domain-specific knowledge. Emphasizing the
role of general problem-solving methods, these studies also high-
light the role of domain-specific patterns and methods, as clearly
some degree of expertise is lost when domain-specific methods
are replaced by domain-general one. While these studies com-
pared individuals of the same level of expertise, Bilali´c et al. (2009)
compared individuals of different skill levels. They took advantage
of several features of chess: chess skill is precisely and quantita-
tively measured by the Elo rating; chess players enjoy trying to
find the best move in a chess position; and chess players specialize
in different openings, which makes it relatively easy to find play-
ers who have the same strength (as measured by their Elo points)
but who have different specialized opening knowledge. Bilali´c et al. compared the performance, in both a memory and
problem solving task, of players who specialized in two different
chess openings. In addition to positions coming from these two
types of defense, they also used neutral positions (positions dif-
ficult to classify with respect to the opening they came from). The players were Candidate Masters, Masters, and International
Masters/Grandmasters. The results were dramatic. With only one
exception, all players obtained the best results with the positions
taken from the openings they specialized in. When confronted
with positions outside their domain of specialization, players per-
formed one standard deviation on average below the level shown
with positions taken from their domain of specialization. The final aim was to replicate the effect of action video-playing
on two tasks: a flanker task (Eriksen and Eriksen, 1974) and
change detection task (Clark et al., 2011). Near transfer In particular the flanker
task has shown a mixed pattern of results; a basic flanker task has
shown both no effect of expertise (Cain et al., 2012), and effects of
expertise only once an additional perceptual load has been added
(Green and Bavelier, 2003). In the case of Green and Bavelier
(2003) it was argued that the addition of a perceptual load pre-
vented flanker interference in the case of nVGP because, unlike
the VGP, there were fewer spare resources to process the distract-
ing flankers. However, it did appear in the original Green and
Bavelier (2003) that there was a small advantage for VGP com-
pared to nVGP at low loads. We therefore predict that, using a
basic “low-load” flanker task, there should be a smaller flanker
effect for VGP compared to nVGP and that this will increase as
VG expertise decreases. FAR-TRANSFER Strategy players, however, are predicted to possess a stronger
reliance on maintaining accuracy as a gained trait from long-
term play where accuracy over response time is key to success. This is because, typically, strategy games require the ability to
move and place items in carefully decided places and forma-
tions. While often these changes are in response to an in-game
target event and can result in swift determination and sequenc-
ing of new actions to fulfill a shifting long-term goal, there is
less emphasis on rapid direct responding to an appearing tar-
get. Games such as “StarCraft 2: Wings of Liberty” (Blizzard
Entertainment, 2010), the best-selling PC strategy game of 2011
(Entertainment Software Association, 2012), demonstrate the
need for this ability, particularly in games with a military basis. As a consequence, it was predicted that strategy players would
perform with significantly higher mean accuracy and action
players would perform with a significantly faster mean reaction
time. FAR-TRANSFER (2011) found
that VGPs display a superior ability to spot changes when pre-
sented with rapidly alternating sets of images. In this study, 35
participants were presented with an unedited/edited image cycle
switching at 4 Hz. The image cycle repeated until the participants
indicated via a mouse click that they had spotted the edited ele-
ment by clicking the image in the position they thought believed
contained the image edit. Video game players performed better
than nVGPs, replicating previous work on attentional improve-
ments in VGP (e.g., Green et al., 2009), and there were also
strategic changes in their search patterns. Compared with nVGP,
the VGP showed broader search strategies, further supporting the
view that VGPs develop top-down processing. It follows from the above arguments that, if video-game exper-
tise leads to the observed enhanced attentional and perceptual
processing, then it should be possible to train nVGP using video
games and observe an improvement in their cognitive function-
ing. Green and Bavelier (2003) recruited two groups of partici-
pants that had no history of video gaming; one group was then
trained on a fast-paced action game (Medal of Honor) and the
other on a slow-paced puzzle game (Tetris). After a period of 10 h
training (1 h a day over 10 days), compared with the Tetris group,
participants trained on Medal of Honor displayed better Useful
Field-of-View (Ball et al., 1988), that is they had an enhance-
ment in their ability to search for and identify cued targets. It was
also found that the Medal of Honor trainees showed a reduced
attentional blink (Raymond et al., 1992), i.e., a reduction in the
window of attentional “blindness” that occurs after detecting or
recognizing the first of two temporally close visual stimuli (Green
and Bavelier, 2003; Feng et al., 2007; Bailey et al., 2010). As detailed above, a number of reports using different atten-
tional tasks have suggested that VGPs have an improved ability to
“filter out” unnecessary or irrelevant stimuli partly through the
enhancement of attentional functioning (Chisholm et al., 2010;
Bavelier et al., 2012). One task that has been used successfully to
test VGPs proposed advantage at distractor filtering directly is the While intriguing, the research on the cognitive benefits of
video game playing has been criticized on several grounds. November 2014 | Volume 5 | Article 1337 | 2 Frontiers in Psychology | Cognition Player 1 level up! DESIGN This study was pseudo-experimental in nature, as the indepen-
dent variables were not directly manipulated. In both experi-
mental measures, the independent variables were skill (experts,
intermediates, novices, and controls) and specialization (action
vs. strategy). In some analyses, in order to allow direct compari-
son with the literature, we used skill with only two levels (VGPs
vs. nVGPs). ntermediates, novices, and controls) and specialization (action
s. strategy). In some analyses, in order to allow direct compari-
on with the literature, we used skill with only two levels (VGPs
s. nVGPs). In order to operationalise the study variables, criteria for each
etween-subject variable needed to be established. A question-
aire was given at the end of the study. In addition to stan-
ard questions such as asking age and gender, information was
btained on the participants gaming habits to allow for allocation
f each participant to the levels of the two independent variables
i.e., VGP or nVGP). The following three questions were asked. (i) “How many hours a week, on average, do you play video
games for? 0–1, 2–5, 6–10, 11–15, 16–20, 21+.” This ques-
tion was used to allocate participants to either the VGP or
nVGP group based on their hours of play. Participants who
answered “0–1” were allocated to the nVGP group and any
answers above were assigned to the VGP group. Sixty-two
VGPs and 19 nVGPs were identified. Table 1 | Participation allocation to the experimental groups (genre
specialization and expertise level). Genre specialization and expertise level
Frequency (N)
Action expert
12
Action intermediate
11
Action novice
13
Action control
12
Action total
48
Strategy expert
10
Strategy intermediate
13
Strategy novice
9
Strategy control
12
Strategy total
44
Total
92
rontiers in Psychology | Cognition
November 2014 | Volume 5 | Article 1337 | 4 Table 1 | Participation allocation to the experimental groups (genre
specialization and expertise level). Genre specialization and expertise level In order to operationalise the study variables, criteria for each
between-subject variable needed to be established. A question-
naire was given at the end of the study. In addition to stan-
dard questions such as asking age and gender, information was
obtained on the participants gaming habits to allow for allocation
of each participant to the levels of the two independent variables
(i.e., VGP or nVGP). The following three questions were asked. ETHICAL APPROVAL This study was granted ethical approval from the Brunel
University School of Social Sciences ethics board in accordance
with the British Psychological Society (BPS) guidelines. All par-
ticipants gave informed consent and were fully debriefed after the
study. PARTICIPANTS Ninety-two participants (56 male) aged between 18 and 30
(M = 21.25, SD = 2.07) were recruited by opportunistic sam-
pling through social networking sites and word of mouth. Most
of the participants had filled out the online questionnaire (Pilot
study). Each participant was offered a food reward for partici-
pating in the study, with a further cash reward incentive (£20) if
they achieved the best score on one of the two tasks out of all
participants. Table 1 presents the frequency of participants for each genre ×
expertise cell of the design. Table 1 presents the frequency of participants for each genre ×
expertise cell of the design. GENERAL PROCEDURE Upon arrival, one of the research team allocated the participant
a random number that corresponded to the participant’s entry in
the pilot database that contained information pertaining to their
game playing history and experience. They were then handed
over to a second experimenter who ran the experiment and who
was blind to the participant’s details and questionnaire scoring. This method ensured that both the participant and the second
researcher were unaware of the participants’ genre or expertise
allocation. Participants carried out the two experimental mea-
sures that formed the study in a random order. For each measure
(described below), the first screen to appear was a set of task
instructions (see below for details of each experiments instruc-
tion). Once the series of tasks was complete, participants com-
pleted the “General Information Sheet,” wherein they answered
questions such as age, genre specialization, average weekly hours
played etc. PILOT STUDY An online pilot, carried out several months before the main study,
asked participants (N = 115) to identify the last three action
and strategy video games they had played. The Call of Duty and
Assassin’s Creed video game series were identified as the most pop-
ular action video games, and the Starcraft and FIFA series were
found to be the most popular strategy video games. (iii) “Would you identify yourself predominantly as an action
or strategy video gamer?” This question was used to decide
each participant’s genre specialization. Thirty six partici-
pants identified themselves as predominantly action VGPs,
alongside 32 strategy VGPs. APPARATUS The experiment was run using the E-Prime software package
(Psychology Software Tools, Inc., 2008) on a Dell desktop com-
puter running Windows 7, with stimuli presented on a 15 inch
Lenovo LCD screen running at a resolution of 640 × 480 pixels at
a refresh rate of 60 Hz. Keyboard and mouse responses were col-
lected via a standard keyboard and mouse. Participants were sat
approximately 60 cm from the computer screen. November 2014 | Volume 5 | Article 1337 | 4 AIMS OF THE STUDY Many studies have investigated the differences between VGPs
and nVGPs but there is as yet, to our knowledge, no research November 2014 | Volume 5 | Article 1337 | 3 www.frontiersin.org Player 1 level up! Gobet et al. (ii) “On average, what percentage of the games that you
play do you complete? (By completed, we mean attain-
ing the highest in-game ‘level’ or ‘rank’ or completing
the game’s storyline. You don’t need to include optional
missions, achievements or DLC (Downloadable Content).”
Participants who answered “76–100%” to this were deemed
as Experts (n = 22), those who answered “51–75%” were
deemed Intermediates (n = 24) and those who answered
“26–50%” were deemed Novices (n = 22). Those who
answered “0–25%” were allocated to the Control group
(n = 24). (ii) “On average, what percentage of the games that you
play do you complete? (By completed, we mean attain-
ing the highest in-game ‘level’ or ‘rank’ or completing
the game’s storyline. You don’t need to include optional
missions, achievements or DLC (Downloadable Content).”
Participants who answered “76–100%” to this were deemed
as Experts (n = 22), those who answered “51–75%” were
deemed Intermediates (n = 24) and those who answered
“26–50%” were deemed Novices (n = 22). Those who
answered “0–25%” were allocated to the Control group
(n = 24). DATA ANALYSIS group of non-players, were presented with images from these
two games in addition to non-video game related (defined as
“neutral”) stimuli. Outliers were identified as either reaction times or correct
responses that were notably outside the general distribution. Boxplots for each data set were analyzed and any outliers SPSS
identified were removed. All reaction time analyses were per-
formed using correct only trials. Outliers were identified as either reaction times or correct
responses that were notably outside the general distribution. Boxplots for each data set were analyzed and any outliers SPSS
identified were removed. All reaction time analyses were per-
formed using correct only trials. The task itself was an adapted form of Clark et al. (2011). There were 13 trials in total, the first of which was a practice
trial and was not included in later analyses. Three repeated mea-
sures conditions were used: Call of Duty, StarCraft and Landscape
(Control) with each condition consisting of all those trials con-
taining the images derived from those games or scenes. There
were four images in each condition. The experiment comprised two measures, an Eriksen Flanker
task and a change detection task. Each of these measures is
described below. DESIGN (i) “How many hours a week, on average, do you play video
games for? 0–1, 2–5, 6–10, 11–15, 16–20, 21+.” This ques-
tion was used to allocate participants to either the VGP or
nVGP group based on their hours of play. Participants who
answered “0–1” were allocated to the nVGP group and any
answers above were assigned to the VGP group. Sixty-two
VGPs and 19 nVGPs were identified. Frontiers in Psychology | Cognition Player 1 level up! Gobet et al. DESIGN On detection of a change, the participant ceased the trial via
keypress (spacebar) and then reported both the location and
nature of the perceived change to the experimenter who remained
with the participant in the room during data collection. The par-
ticipant then pressed the spacebar again in order to trigger the
next trial presentation. The experiment was a mixed factorial design with the within sub-
jects factor being Congruence and between-subjects factors of
Expertise Level, Genre Affiliation and “Video Game Players vs. Non-Video Game Players.” The dependent variables in this task
were reaction time and percent correct accuracy. PROCEDURE This measure was a modified version of Eriksen and Eriksen’s
(1974) flanker task. Arrows were used instead of letters, similar
to other video gaming studies such as Cain et al. (2012). Participants fixated a centrally presented cross for 4000 ms prior
to the start of the change detection task image presentation. The
first, unedited (UE), image was then presented for 240 ms, fol-
lowed by a blank gray screen for 80 ms. A second image, identical
to the first, would then appear for a period of 240 ms before being
replaced by a blank gray screen for 80 ms. The process would then
repeat but with the edited (E) version (identical save for a change
in a single image feature) of the same image, i.e., the sequence
appeared as UE, UE, E, E, UE, UE, E, E... This cycle repeated
until the participant responded by pressing the spacebar on the
computer keyboard (see Figure 1). RESULTS Participants completed 24 congruent and 24 incongruent trials
(trial order randomized) in two blocks of equal number (i.e., 24
trials per block). On each trial the participant viewed a centrally
presented fixation cross for 500 ms that was replaced by either a
congruent or incongruent trial image. Participants viewed each
trial image and were asked to indicate in which direction the
central arrow using the arrow keys on the keyboard. Each trial
remained onscreen until the participant made a key press. The
next trial immediately followed. MATERIALS Congruent and incongruent stimuli were created prior to the start
of the experiment by combining arrow stimuli such that the cen-
tral arrow to which the participant responded was surrounded
by equally spaced, directionally congruent stimuli (e.g., < < <
or > > >) or directionally incongruent stimuli (e.g., < > <
or > < >). The flanker stimuli subtended 8◦of visual angle. MATERIALS Images were sourced from Google Image Search and were edited
using Adobe Photoshop CS5 (Adobe Systems, 2010). Based on the
pilot study, which provided information on the most commonly
played action and strategy games, images from Call of Duty and
StarCraft were chosen. As both games are part of a much larger
series of games, the most recent versions of each franchise were
used (StarCraft II: Wings of Liberty; Blizzard Entertainment, 2010)
and Call of Duty: Black Ops 2 (Treyarch, 2012). All trial images
were scaled such that they subtended 26◦of visual angle. FIGURE 1 | One cycle of a sample Change Detection trial showing a FIGURE 1 | One cycle of a sample Change Detection trial showing a
Landscape (Control) image. The sequence involved two presentations of
the unedited image prior to two presentations of an edited image. An
example of an unedited and associated edited image is shown in the
bottom right corner: The feature missing in the edited version of the image
is indicated by the red circle on the original image. FIGURE 1 | One cycle of a sample Change Detection trial showing a
Landscape (Control) image. The sequence involved two presentations of
the unedited image prior to two presentations of an edited image. An
example of an unedited and associated edited image is shown in the
bottom right corner: The feature missing in the edited version of the image
is indicated by the red circle on the original image. MEASURE 1: ERIKSEN FLANKER TASK Prior to analysis, one outlier was removed for failing to comply
with task instruction. Data were analyzed using a mixed ANOVA FIGURE 1 | One cycle of a sample Change Detection trial showing a
Landscape (Control) image. The sequence involved two presentations of
the unedited image prior to two presentations of an edited image. An
example of an unedited and associated edited image is shown in the
bottom right corner: The feature missing in the edited version of the image
is indicated by the red circle on the original image. www.frontiersin.org MEASURE 2: CHANGE DETECTION TASK After outliers were removed due to task non-compliance, 85 par-
ticipants remained from the original 92. A Mixed ANOVA was
carried out and outliers were controlled in an identical way to the
Flanker Task. Reaction time Analyses showed a main effect of congruence on reaction time,
F(1, 73) = 53.31, p < 0.001, η2
p = 0.42, f = 0.85; overall, a lower
mean reaction time was demonstrated in the Congruent (Same)
condition (M = 433.85, SD = 70.47) than the Incongruent
(Distractor) condition (M = 529.46, SD = 155.82). This effect
was not qualified by participant expertise, F(3, 73) = 0.42, p =
0.74, η2
p = 0.02, f = 0.14, or genre, F(1, 73) = 1.71, p = 0.59,
η2
p = 0.006, f = 0.08. No interaction was found between congru-
ence, expertise and genre, F(3, 73) = 1.71, p = 0.17, η2
p = 0.07,
f = 0.27. Analysis indicated a main effect of image type on reac-
tion time, F(2, 154) = 36.57, p < 0.001, n2
p = 0.32, f = 0.69. Response times were quicker in the Landscape condition (M =
9399 ms, SD = 5119 ms) than in the Call of Duty condition
(M = 16, 138 ms, SD = 7556 ms) and Starcraft condition (M =
20, 247 ms, SD = 10, 761 ms). This effect was not qualified by
participant expertise [F(6, 154) = 1.06, p = 0.39, n2
p = 0.04, f =
0.2] or genre [F(2, 154) = 0.57, p = 0.57, n2
p = 0.01, f = 0.1]. No
interaction was found between image type, expertise and genre
[F(6, 154) = 0.27, p = 0.95, n2
p = 0.01, f = 0.1]. p
We also analyzed the data by grouping the participants
into players and non-players. Mauchly’s Test indicated a vio-
lated assumption of sphericity, χ2
(2) = 21.57, p < 0.001, therefore
degrees of freedom (df ) were corrected using Greenhouse-Geisser
estimates of sphericity (ε = 0.81). Analysis indicated a main
effect of image type on reaction time, F(2, 166) = 25.06, p <
0.001, n2
p = 0.23, f = 0.55, that was not qualified by the “Video
Game Players vs. Non-Video Game Players” variable [F(2, 166) =
1.37, p = 0.26, n2
p = 0.02, f = 0.14]. Figure 3 shows a summary DESIGN As it used both expertise and specialization as independent vari-
ables, this experiment had the same design as described in Bilali´c
et al. (2009). Players of different skill levels and specialized with
the video games Call of Duty or StarCraft, as well as a control November 2014 | Volume 5 | Article 1337 | 5 www.frontiersin.org www.frontiersin.org Player 1 level up! Gobet et al. to determine the effect of specialization and skill (between-
subjects) with performance on congruent and incongruent trials
(within-subjects). With respect to accuracy, there was a main effect of con-
gruence on accuracy, F(1, 89) = 19.96, p < 0.001, η2
p = 0.18, f =
0.47, that was not qualified by the players’ status [F(1, 89) = 0.24,
p = 0.62, η2
p = 0.008, f = 0.09]. p
With respect to reaction time, both groups performed over-
all faster in the Congruent (Same) condition (M = 433.85, SD =
70.47) than the Incongruent (Distractor) condition (M = 529.46,
SD = 155.82). Analysis indicated a main effect of congruence on
reaction time, F(1, 79) = 32.23, p < 0.001, η2
p = 0.29, f = 0.64,
that was not qualified by the players’ status, F(1, 79) = 1.12, p =
0.29, η2
p = 0.01, f = 0.1. Accuracy Analyses indicated a main effect of congruence, F(1, 83) = 27.90,
p < 0.001, η2
p = 0.25, f = 0.58. The Congruent (Same) condi-
tion (M = 23.69, SD = 0.69) had a higher average score than
the Incongruent (Distractor) condition (M = 18.90, SD = 8.38). This effect was not qualified by participant expertise, F(3, 83) =
0.57, p = 0.64, η2
p = 0.02, f = 0.14, or genre, F(1, 83) = 2.55,
p = 0.11, η2
p = 0.03, f = 0.18. No interaction was found between
congruence, expertise and genre, F(3, 83) = 2.00, p = 0.12, η2
p =
0.07, f = 0.27. VGPs vs. nVGPs In order to attempt to replicate previous research using this task,
we also carried out analyses where the participants were allocated
to only two groups (VGPs and nVGPs). nVGPs were identified
as any participant who played, on average, less than 1 h of either
console or computer video games per week. Figure 2 shows a
summary of the reaction time and accuracy data for the flanker
task for VGPs and nVGPs. FIGURE 2 | Mean accuracy (left), reaction times (right) and the associated 95% confidence intervals for the VGP and nVGP from the Eriksen Flanker
Task. FIGURE 2 | Mean accuracy (left), reaction times (right) and the associated 95% confidence intervals for the VGP and nVGP from the Eriksen Flanker
Task. November 2014 | Volume 5 | Article 1337 | 6 Frontiers in Psychology | Cognition Player 1 level up! Gobet et al. FIGURE 3 | Mean reaction times along with the associated 95%
confidence intervals for VGP and nVGP from the Change Detection
Task. that playing this kind of game leads to substantial transfer, in par-
ticular with tasks engaging attentional processing. The aim of this
study was to replicate this phenomenon with two tasks that had
previously been used in the action video-game literature. In addi-
tion, the study aimed to use a finer measure of expertise than had
been done in the past, and to look at the extent to which skills
acquired in a specific VG genre (action or strategy) can be used in
a task using material linked to either of the two genres. g
g
In neither task were we able to find any effect of skill or a supe-
riority of the VGPs when compared to the nVGPs. Thus, our study
did not support the hypothesis of far transfer, in line with most
theories of expertise but in contradiction with previous VG stud-
ies. Our failure to replicate previous results cannot be ascribed to
a lack of power, as the number of participants (n = 92) was high
for this kind of study and our design, incorporating different lev-
els of skill, was in principle able to identify subtle skills effects that
cannot be found when only two groups are compared (VGPs vs. nVGPs). In addition, the results we obtained in each task were
consistent with the results normally obtained in these tasks. POWER SUMMARY One advantage of this study is the relatively large number of par-
ticipants who were involved. However, to ensure that the absence
of expertise effects was not due to a lack of statistical power, we
investigated the size of effects we could be expected to find. All
calculations are based on a power criterion of 0.8 and a 0.05 alpha
level. For Measure 1: Eriksen Flanker task, one could expect to
find significant differences of effect sizes for the interaction of
congruency and specialization of 0.31 and of 0.36 for the two-way
interaction of congruency with skill and for the three-way inter-
action of congruency with skill and specialization. For Measure 2:
Change Detection Task, effect size of 0.25 could be expected to be
identified for interactions of image type with specialization and
effect sizes of 0.30 for both the two-way interactions of special-
ization and skill with image type and the three-way interaction of
image type, specialization, and skill. We followed Boot et al.’s (2011) advice of pre-screening par-
ticipants long before the experiments per se, and of asking them
to fill in the questionnaire on video game activities at the end of
the study. Thus, our procedure minimized demand characteris-
tics. Together with the fact that other studies have failed to find
a VGP effect (e.g., Castel et al., 2005; Boot et al., 2008; Murphy
and Spencer, 2009), our results are consistent with the possibility
that such demands might have played a role in previous research
showing VGPs’ superiority. However, given that this conclusion
is based on null results, further research is needed to validate or
invalidate this hypothesis. In all cases, the observed effect sizes for game specialization
and skill in each of the measures was considerably smaller than the
minimum expected detectable effect size even given our relatively
large sample size. The power analyses also highlight that given our
sample size we could expect to detect small effect sizes. Most unexpected was perhaps the failure to find an expertise
effect in the change detection task. Near transfer did not occur
in spite of the fact that the material used came from the players’
domain of expertise (either action or strategy players) and that we
based the stimulus choice on the most popular games within each
genre, an expected level of increased familiarity. VGPs vs. nVGPs For
example, we found significantly faster reaction times and a greater
number of percent correct accurate trials for the congruent trials
compared to the non-congruent trials in the flanker task. FIGURE 3 | Mean reaction times along with the associated 95%
confidence intervals for VGP and nVGP from the Change Detection
Task. For the flanker task, despite strong congruency effects, the
absence of a significantly different interference effect for VGP vs. nVGP is not entirely unexpected, despite our predictions to the
contrary. We argued that based on previous work that we might
expect a small difference (Green and Bavelier, 2003), especially
given our sample size and a finer division of game expertise than
previously used. However, this was not the case and, although the
null hypothesis cannot be accepted, the absence of a significant
interaction of game expertise and congruency does add to a num-
ber of results showing that, at low loads there is little difference
between the performance on VGP and nVGP on a flanker task
(Green and Bavelier, 2003; Cain et al., 2012). of the reaction time data for the change detection task for VGPs
and nVGPs. POWER SUMMARY Why is it that
the patterns that the players had presumably acquired by playing
their favorite game did not enable them to find changes in the
stimuli more rapidly? One possible explanation is that the change
detection paradigm is unsuitable for detecting domain-specific
patterns used for unconscious pattern recognition. An examina-
tion of the mean reaction times shows that the average time spent www.frontiersin.org DISCUSSION One of the major conclusions of research into learning and exper-
tise is that transfer from one domain to another is rare and
difficult, and happens only when the two domains share com-
ponents that ask for the same cognitive skills. In recent years, a
series of experiments on action video game playing have found November 2014 | Volume 5 | Article 1337 | 7 www.frontiersin.org www.frontiersin.org Player 1 level up! Gobet et al. on task was very long (about 20 s on average in the Starcraft con-
dition) and thus is likely to engage more conscious mechanisms. This explanation gains further plausibility given that Gobet et al. (in preparation) did find an interaction between specialism and
expertise in a recognition task using a similar design as that used
here: action players specializing in the Call of Duty and rac-
ing players specializing in Gran Turismo performed better when
dealing with images from their own game. of visual search. Acta Psychol. 119, 217–230. doi: 10.1016/j.actpsy.2005. 02.004 of visual search. Acta Psychol. 119, 217–230. doi: 10.1016/j.actpsy.2005. 02.004 Chase, W. G., and Simon, H. A. (1973). Perception in chess. Cogn. Psychol. 4, 55–81. doi: 10.1016/0010-0285(73)90004-2 Chiesi, H. L., Spilich, G. J., and Voss, J. F. (1979). Acquisition of domain-related
information in relation to high and low domain knowledge. J. Verbal Learn. Verbal Behav. 18, 257–273. doi: 10.1016/S0022-5371(79)90146-4 Chisholm, J. D., Kingstone, A., Hickey, C., and Theeuwes, J. (2010). Reduced atten-
tional capture in action video game players. Attent. Percept. Psychophys. 72,
667–671. doi: 10.3758/APP.72.3.667 Our study was not without weaknesses. The measure of video-
game expertise and the allocation to a specific genre were based
on self-reports, and perhaps it would have been desirable (albeit
unpractical) to ask players to play segments of their favorite game
to estimate their level. With regard to obtaining a pure mea-
sure of specific video game genre benefits a study, such as this,
that concentrates on training effects in nVGP (e.g., Green and
Bavelier, 2003) may be in a better suited to detect subtle changes
in performance. A comparison of preferred gametype may be
expected to find subtle differences between players of different
games but only when those players partake of a single type of
game. Anecdotally, many game players are “poly-gamers” and will
play other game types in addition to their preferred category. DISCUSSION A
training study where nVGP gain experience playing only a single
genre of game would therefore be better placed to detect subtle
differences between expertise benefits of individual game types. A similar argument can be made to better examine any potential
gender differences. The sample here did not allow for a meaning-
ful investigation of potential gender differences and, for the same
reasons previously argued to account for poly-gamers, a training
study would be ideal. Finally, there were few trials in the change
detection task. Clark, K., Fleck, M. S., and Mitroff, S. R. (2011). Enhanced change detection per-
formance reveals improved strategy use in avid action video game players. Acta
Psychol. 136, 67–72. doi: 10.1016/j.actpsy.2010.10.003 Dye, M. W. G., Green, C. S., and Bavelier, D. (2009). The development of atten-
tion skills in action video game players. Neuropsychologia 47, 1780–1789. doi:
10.1016/j.neuropsychologia.2009.02.002 Entertainment
Software
Association. (2012). Essential
Facts
About
the
Computer
and
Video
Game
Industry. Available
online
at:
http://www.theesa.com/facts/pdfs/ESA_EF_2012.pdf
(Accessed
March
11,
2013) Eriksen, B. A., and Eriksen, C. W. (1974). Effects of noise letters upon identifica-
tion of a target letter in a nonsearch task. Percept. Psychophys. 16, 143–149. doi:
10.3758/BF03203267 Feng, J., Spence, I., and Pratt, J. (2007). Playing an action video game reduces
gender differences in spatial cognition. Psychol. Sci. 10, 850. doi: 10.1111/j.1467-
9280.2007.01990.x Gobet, F., and Campitelli, G. (2006). “Education and chess: a critical review,”
in Chess and Education: Selected Essays from the Koltanowski Conference, ed
T. Redman (Dallas, TX: Chess Program at the University of Texas at Dallas),
124–143. Gobet, F., and Campitelli, G. (2006). “Education and chess: a critical review,”
in Chess and Education: Selected Essays from the Koltanowski Conference, ed Gobet, F., de Voogt, A. J., and Retschitzki, J. (2004). Moves in Mind. Hove:
Psychology Press. Gobet, F., Patel, K., and Johnston, S., (in preparation). The Effects of Video Game
Specialisation on a Recognition Task. Gobet, F., and Simon, H. A. (1996). Templates in chess memory: a mechanism for
recalling several boards. Cogn. Psychol. 31, 1–40. doi: 10.1006/cogp.1996.0011 Green, C. S., and Bavelier, D. (2003). Action video game modifies visual selective
attention. Nature 423, 534. doi: 10.1038/nature01647 In the last year several professional associations and journals
have emphasized the need for more replications. However, in spite
of previous calls (e.g., Gobet et al., 2004), and partly due to the
difficulty of finding experts, research into expertise is rarely repli-
cated. REFERENCES Infinity Ward. (2011). Call of Duty: Modern Warfare 3 [Video Game]. Santa Monica,
CA: Activision. Bailey, K., West, R., and Anderson, C. A. (2010). A negative association between
video game experience and proactive cognitive control. Psychophysiology 47,
34–42. doi: 10.1111/j.1469-8986.2009.00925.x Kristjánsson, Á. (2013). The case for causal influences of action videogame
play upon vision and attention. Attent. Percept. Psychophys. 75, 667–672. doi:
10.3758/s13414-013-0427-z Ball, K., Beard, B., Roenker, D., Miller, R., and Ball, D. (1988). Visual search age and
practice. Invest. Ophthalmol. Vis. Sci. 29, 448. Lavie, N. (1995). Perceptual load as a necessary condition for selective atten-
tion. J. Exp. Psychol. Hum. Percept. Perform. 21, 451–468. doi: 10.1037/0096-
1523.21.3.451 Bavelier, D., Achtman, R. L., Mani, M., and Föcker, J. (2012). Neural bases of
selective attention in action video game players. Vision Res. 61, 132–143. doi:
10.1016/j.visres.2011.08.007 Mishra, J., Zinni, M., Bavelier, D., and Hillyard, S. A. (2011). Neural basis
of superior performance of action videogame players in an attention-
demanding task. J. Neurosci. 31, 992–998. doi: 10.1523/JNEUROSCI.4834-
10.2011 Bilali´c,
M.,
McLeod,
P.,
and
Gobet,
F. (2009). Specialization
effect
and its influence on memory and problem solving in expert chess
players. Cogn. Sci. 33,
1117–1143. doi:
10.1111/j.1551-6709.2009. 01030.x Murphy, K., and Spencer, A. (2009). Playing video games does not make for better
visual attention skills. J. Artic. Support Null Hypothesis 6, 1–20. Boot, W. R., Blakely, D. P., and Simons, D. J. (2011). Do action video games improve
perception and cognition? Front. Psychol. 2:226. doi: 10.3389/fpsyg.2011. 00226 Picton, T. W. (1992). The P300 wave of the human event-related poten-
tial. J. Clin. Neurophysiol. 9, 456–479. doi: 10.1097/00004691-199210000-
00002 Boot, W. R., Kramer, A. F., Simons, D. J., Fabiani, M., and Gratton, G. (2008). The
effects of video game playing on attention, memory, and executive control. Acta
Psychol. 129, 387–398. doi: 10.1016/j.actpsy.2008.09.005 Ptak, R. (2012). The frontoparietal attention network of the human brain: action,
saliency, and a priority map of the environment. Neuroscientist 18, 502–515. doi:
10.1177/1073858411409051 Cain, M. S., Landau, A. N., and Shimamura, A. P. (2012). Action video game experi-
ence reduces the cost of switching tasks. Atten. Percept. Psychophys. 74, 641–647. doi: 10.3758/s13414-012-0284-1 Raymond, J. E., Shapiro, K. L., and Arnell, K. M. (1992). Temporary sup-
pression of visual processing in an RSVP task—an attentional blink. J. Exp. Psychol. Hum. Percept. Perform. 18, 849–860. doi: 10.1037/0096-1523. 18.3.849 Castel, A. D., Pratt, J., and Drummond, E. (2005). DISCUSSION The current paper contributes to this effort of obtaining
more robust empirical data. Green, C. S., Li, R. J., and Bavelier, D. (2009). Perceptual learning during
action video game playing. Topics Cogn. Sci. 2, 202–216. doi: 10.1111/j.1756-
8765.2009.01054.x Hubert-Wallander, B., Sugarman, M., Bavelier, D., and Green, C. S. (2011). Changes in search rate but not in the dynamics of exogenous attention in
action videogame players. Attent. Percept. Psychophys. 73, 2399–2412. doi:
10.3758/s13414-011-0194-7 REFERENCES The effects of action video
game experience on the time course of inhibition of return and the efficiency November 2014 | Volume 5 | Article 1337 | 8 Frontiers in Psychology | Cognition Player 1 level up! Gobet et al. Rikers, R. M. J. P., Schmidt, H. G., Boshuizen, H. P. A., Linssen, G. C. M., Wesseling,
G., and Paas, F. G. W. C. (2002). The robustness of medical expertise: clinical
case processing by medical experts and subexperts. Am. J. Psychol. 115, 609–629. doi: 10.2307/1423529 Conflict of Interest Statement: The authors declare that the research was con-
ducted in the absence of any commercial or financial relationships that could be
construed as a potential conflict of interest. Received: 31 May 2014; accepted: 03 November 2014; published online: 28 November
2014. Schunn, C. D., and Anderson, J. R. (1999). The generality/specificity of exper-
tise in scientific reasoning. Cogn. Sci. 23, 337–370. doi: 10.1207/s15516709cog
2303_3 Citation: Gobet F, Johnston SJ, Ferrufino G, Johnston M, Jones MB, Molyneux A,
Terzis A and Weeden L (2014) “No level up!”: no effects of video game specialization Citation: Gobet F, Johnston SJ, Ferrufino G, Johnston M, Jones MB, Molyneux A,
Terzis A and Weeden L (2014) “No level up!”: no effects of video game specialization
and expertise on cognitive performance. Front. Psychol. 5:1337. doi: 10.3389/fpsyg. 2014.01337 Spence, I., Yu, J. J., Feng, J., and Marshman, J. (2009). Women match men when
learning a spatial skill. J. Exp. Psychol. Learn. Mem. Cogn. 35, 1097–1103. doi:
10.1037/a0015641 and expertise on cognitive performance. Front. Psychol. 5:1337. doi: 10.3389/fpsyg. 2014.01337 Thorndike, E. L., and Woodworth, R. S. (1901). The influence of improvement
in one mental function upon the efficiency of other functions. Psychol. Rev. 9,
374–382. This article was submitted to Cognition, a section of the journal Frontiers in
Psychology. Copyright © 2014 Gobet, Johnston, Ferrufino, Johnston, Jones, Molyneux, Terzis and
Weeden. This is an open-access article distributed under the terms of the Creative
Commons Attribution License (CC BY). The use, distribution or reproduction in other
forums is permitted, provided the original author(s) or licensor are credited and that
the original publication in this journal is cited, in accordance with accepted academic
practice. No use, distribution or reproduction is permitted which does not comply with
these terms. Unterrainer, J. M., Kaller, C. P., Leonhart, R., and Rahm, B. (2011). November 2014 | Volume 5 | Article 1337 | 9 REFERENCES Revising
superior planning performance in chess players: the impact of time restriction
and motivation aspects. Am. J. Psychol. 124, 213–225. doi: 10.5406/amer-
jpsyc.124.2.0213 Waters, A. J., Gobet, F., and Leyden, G. (2002). Visuo-spatial abilities in chess play-
ers. Br. J. Psychol. 30, 303–311. doi: 10.1348/000712602761381402 November 2014 | Volume 5 | Article 1337 | 9 www.frontiersin.org www.frontiersin.org
|
https://openalex.org/W2083280228
|
https://europepmc.org/articles/pmc2766383?pdf=render
|
English
| null |
The Self-Assessment Scale of Cognitive Complaints in Schizophrenia: A validation study in Tunisian population
|
BMC psychiatry
| 2,009
|
cc-by
| 5,806
|
BioMed Central BioMed Central BioMed Central The Self-Assessment Scale of Cognitive Complaints in
Schizophrenia: A validation study in Tunisian population
Ines Johnson*, Oussama Kebir, Olfa Ben Azouz, Lamia Dellagi,
Yasmine Rabah and Karim Tabbane Address: Research Unit "Cognitive dysfunctions in psychiatric diseases", Department of psychiatry "B", Razi Hospital. 24, rue des orangers. La
b Address: Research Unit "Cognitive dysfunctions in psychiatric diseases", Department of psychiatry "B", Razi Hospital. 24, rue des orangers. La
Manouba, Tunisia Email: Ines Johnson* - ines.johnson@yahoo.fr; Oussama Kebir - kebir_oussama@yahoo.fr; Olfa Ben Azouz - o.benazouz@yahoo.fr;
Lamia Dellagi - dellagilamia@yahoo.fr; Yasmine Rabah - yasmine_rabah@hotmail.com; Karim Tabbane - k_tabbane@yahoo.fr * Corresponding author Received: 22 April 2009
Accepted: 8 October 2009 This article is available from: http://www.biomedcentral.com/1471-244X/9/66 This article is available from: http://www.biomedcentr © 2009 Johnson et al; licensee BioMed Central Ltd. © 2009 Johnson et al; licensee BioMed Central Ltd. This is an Open Access article distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/2.0),
which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited. © 2009 Johnson et al; licensee BioMed Central Ltd. This is an Open Access article distributed under the terms of the Creative Commons Attribution License (http://creative
which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cit BMC Psychiatry Open Access Page 1 of 7
(page number not for citation purposes) Abstract Background: Despite a huge well-documented literature on cognitive deficits in schizophrenia,
little is known about the own perception of patients regarding their cognitive functioning. The
purpose of our study was to create a scale to collect subjective cognitive complaints of patients
suffering from schizophrenia with Tunisian Arabic dialect as mother tongue and to proceed to a
validation study of this scale. Methods: The authors constructed the Self-Assessment Scale of Cognitive Complaints in
Schizophrenia (SASCCS) based on a questionnaire covering five cognitive domains which are the
most frequently reported in the literature to be impaired in schizophrenia. The scale consisted of
21 likert-type questions dealing with memory, attention, executive functions, language and praxia. In a second time, the authors proceeded to the study of psychometric qualities of the scale among
105 patients suffering from schizophrenia spectrum disorders (based on DSM- IV criteria). Patients
were evaluated using the Positive and Negative Syndrome Scale (PANSS), the Global Assessment
Functioning Scale (GAF scale) and the Calgary Depression Scale (CDS). Results: The scale's reliability was proven to be good through Cronbach alpha coefficient equal to
0.85 and showing its good internal consistency. The intra-class correlation coefficient at 11 weeks
was equal to 0.77 suggesting a good stability over time. Principal component analysis with Oblimin
rotation was performed and yielded to six factors accounting for 58.28% of the total variance of
the scale. Conclusion: Given the good psychometric properties that have been revealed in this study, the
SASCCS seems to be reliable to measure schizophrenic patients' perception of their own cognitive
impairment. This kind of evaluation can't substitute for objective measures of cognitive
performances in schizophrenia. The purpose of such an evaluation is to permit to the patient to
express his own well-being and satisfaction of quality of life. Page 1 of 7
(page number not for citation purposes) Page 1 of 7
(page number not for citation purposes) http://www.biomedcentral.com/1471-244X/9/66 BMC Psychiatry 2009, 9:66 Description of the scale The authors constructed the Self-Assessment Scale of Cog-
nitive Complaints in Schizophrenia (SASCCS) based on a
questionnaire covering five cognitive domains which are
the most frequently reported in the literature to be
impaired in schizophrenia [6,26]. The scale consisted of
21 questions dealing with memory, attention, executive
functions, language and praxia. Memory was evaluated
through its components: working memory (item 1&2),
episodic memory (item 3 though 9) and semantic mem-
ory (item 10&11). Attention was investigated through its
components: distractibility (item 12), alertness (item13),
selective attention (item14), divided attention (item15)
and sustained attention (item16). Executive functions
were explored through their components: planning
(item17), organisation (item18) and flexibility (item19). Finally, language was examined through item 20 and
praxia through item 21. The scale was made to be as clear,
simple and easy to use by patients suffering from schizo-
phrenia. It was written in Tunisian Arabic dialect. 'See
additional file 1: Tunisian version of the SASCCS'. 'See
additional file 2: English version of the SASCSS'. Traditionally, the study of subjective symptoms of schizo-
phrenic patients has been limited to delusions and hallu-
cinations
[15]. Nowadays,
abnormal
subjective
experiences concerning fields other than delusions and
hallucinations are becoming more investigated since they
are believed to be important in understanding and treat-
ing schizophrenia [16,17]. From a historical point of
view, the first author who described a patient's subjective
experiences in schizophrenia was Huber [18,19]. This
German author introduced the term of "basic symptoms"
to designate the first symptoms of schizophrenia that con-
stitute the basis on which the others symptoms develop. These symptoms do not include behavioural abnormali-
ties or verbal impairments that can be assessed objectively
by clinicians. In fact, they are only reported by patients
that describe them as subjective experiences of deficits
including loss of energy, motor dysfunctions, abnormal
corporeal sensations, altered cognitive processes, difficul-
ties to feel emotions and vulnerability to stress [20]. The
basic symptoms were targeted by a multitude of scales
comprising the Bonn Scale [21], the Frankfurt Complaint
Scale [22], the Subjective Experience of Deficit Scale [23],
the Interview on Subjective Experience [16], the Subjec- Background tive Deficit Syndrome Scale [24] and the Ependorff Inven-
tory of Schizophrenia [25]. What is significant is that these
scales dealt with different aspects of subjective experiences
in schizophrenia including cognitive dysfunctions but
didn't focus specifically on the latter. Only one scale, the
SSTICS or Subjective Scale To Investigate Cognition in
Schizophrenia [14], assessed specifically the cognitive
subjective symptoms in schizophrenia. The psychometric
properties of this scale were evaluated within a popula-
tion of 114 French speaking schizophrenic patients. Vali-
dation study of the SSTICS was shown to be successful
proving that cognitive complaints in schizophrenia can be
reliably assessed. g
It is now well proven that schizophrenia is associated with
multiple cognitive deficits [1-3] which can be profound
and devastating [4]. Patients with chronic schizophrenia
demonstrate impairments that range between one and a
half to two standard deviations below healthy controls on
several key dimensions of cognition [5], especially verbal
memory, working memory, motor speed, attention, exec-
utive functions and verbal fluency [6]. These deficits are thought to be a core feature of schizo-
phrenia and not simply the result of the symptoms or the
current treatments of the illness [7,8]. Moreover, they
seem to have an impact on functional outcome [9] as they
are correlated with poor functional abilities including
skills acquisition, problem solving, and community living
[10-12]. Furthermore, neurocognitive deficits are believed
to be the single strongest correlate of real-world function-
ing [13]. To our knowledge, no similar instrument has been pub-
lished and validated in the Arabic language. Conse-
quently, the purpose of our study was to create a scale to
collect subjective cognitive complaints of patients suffer-
ing from schizophrenia whose mother tongue is Tunisian
Arabic. The number of publications on cognitive deficits in schiz-
ophrenia has grown vastly over the past two decades. At
the same time, an increasing number of sophisticated lab-
oratory tasks has been developed for a better assessment
of cognition [14]. However, little is known about how
patients suffering from schizophrenia perceive their own
cognition. Are they aware of their eventual cognitive
impairments? Do they realize that their social functioning
is highly influenced by these deteriorations? Do they com-
plain about their memory problems to their doctor and
do they demand specific treatments for them? Page 2 of 7
(page number not for citation purposes) Administration procedure The SASCCS is a self-rated questionnaire administered
during a structured interview in which the investigator
explains to the patient the way he should answer to the 21
Likert-type questions of the scale. The patient is asked to
read each of the items in which problems of memory or
concentration of daily life are presented and may have
been experienced by him self. He is then asked to estimate
the frequency of occurrence of such situations in his own
life. For that purpose, he must circle the number that best
corresponds to his experienced life. (4-very often; 3-often;
2-sometimes; 1-rarely; 0-never). The SASCCS total score is
calculated by adding each item score together. The more
the patient complained about cognitive impairments, the
higher was the scale's total score. Psychopathological assessment Psychopathological symptoms were evaluated using the
PANSS [29], the Calgary depression scale (CDS) [30] and
the Global Assessment Functioning scale (GAF scale) [27]. PANSS, CDS and EGF were administered by the same
trained psychiatrist for all participants. Mean scores on
these clinical scales were as follows: 52.84 (SD = 9.64) for
the PANSS total score, 1.35 [min = 0; max = 5] for the CDS
and 62.58 (SD = 13.88) for the GAF scale. Mean scores for
the PANSS subscales were as follows: 10.05 (SD = 2.5) for
the positive symptoms, 16.32 (SD = 4.49) for the negative
symptoms and 26.4 (SD = 5) for general psychopathol-
ogy. Mean score for the item G12 of the PANSS assessing
insight was 2.32 (SD = 1.15). The approximate time to completion was 15 minutes on
average. The questionnaire was administered at the outpa-
tient clinic. The same trained psychiatrist proceeded to the
administration of the scale among all participants. The
investigator should remain on site until the patient is
done with the questionnaire. He could provide explana-
tions to some questions or even examples to clarify the
meaning of items especially item 13, 15, 18 and 19. 'See
additional file 4: examples for items 13, 15, 18 and 19 of
the SASCCS'. Using the 5-factor model of the PANSS as identified by
Lindenmayer et al. [31], we calculated the cognitive factor
and the depression factor which had respectively a score
of 10.14 (SD = 2.49) and 5.93 (SD = 1.99). Statistical analysis We conducted an exploratory principal component anal-
ysis (PCA) on the correlation matrix of the 21 items of the
SASCCS. Several guidelines were used to select the
number of factors: the Kaiser criteria and the interpretabil-
ity of the factors. Oblimin rotation was then performed. (4) never undergone electroconvulsive therapy, (4) never undergone electroconvulsive therapy, Accordingly to this purpose, item 8 was modified in a way
to provide examples corresponding to both men and
women in their daily activities. The wording of items 10,
14 and 15 was reviewed in a way to be clarified. This pre-
test also served to harmonize the modalities of the scale's
administration and the instructions given each time to the
patients. (5) no evidence of organic brain pathology including cer-
ebral tumor, epilepsy, systemic disease, history of cranial
trauma, brain surgery (6) no history of substance abuse or dependence, and
consumption of psychoactive. Table 1 shows sociodemographic sample characteristics
and its psychiatric history. (2) no evidence of mental retardation, of the disease was 10.3 years (SD = 6.89). 'See additional
file 3: Table S1: demographic characteristics and psychia-
try history of pre-test sample'. of the disease was 10.3 years (SD = 6.89). 'See additional
file 3: Table S1: demographic characteristics and psychia-
try history of pre-test sample'. (3) being at the time of testing under unchanged medica-
tion dosage for the last 4 weeks. (3) being at the time of testing under unchanged medica-
tion dosage for the last 4 weeks. (1) have a minimum educational level of 5 years, Pre-test of experimental version The questionnaire was first administered to a reduced
sample of 38 patients (35 men, 3 women) meeting the
DSM-IV diagnostic criteria for schizophrenia (n = 35) or
schizoaffective disorder (n = 3) [27]. The aim of this pre-
liminary work was to collect comments from both
patients and investigators in order to better formulate the
items and furthermore, to add examples to the questions
that closely suit the patient's daily life. Mean age of the
patients was 34 ± 8.9 years and time elapsed since onset Page 2 of 7
(page number not for citation purposes) Page 2 of 7
(page number not for citation purposes) http://www.biomedcentral.com/1471-244X/9/66 BMC Psychiatry 2009, 9:66 BMC Psychiatry 2009, 9:66 (2) no evidence of mental retardation, Characteristics of the population The final version of the scale was then administered to
105 outpatients who met the DSM IV criteria for schizo-
phrenia (undifferentiated subtype, n = 47; paranoid sub-
type, n = 39; hebephrenic subtype, n = 6; residual subtype,
n = 3) or schizoaffective disorder (bipolar subtype, n = 8;
depressive subtype, n = 2). Patients were recruited from
three different outpatient clinics based in the Razi Hospi-
tal (La Manouba, Tunisia). They were carefully screened to
rule out an additional Axis I diagnosis or any disorder that
might alter brain functioning. They had to meet the fol-
lowing requirements: Construct validity and reliability were evaluated by calcu-
lating Cronbach's alpha coefficient and the average of cor-
relations between each item and the total score. Correlation analyses were performed using the Pearson
coefficient when data had normal distribution; elsewhere,
Spearman rank correlation was calculated. Statistical significance level was set at p = 0.01 (two-
tailed). Test-retest reliability Its was assessed within a subgroup of 39 patients exam-
ined by the same investigator at a mean interval of 80 days
(SD = 33). Intra-class correlation coefficient was equal to
0.77 (p = 0.00) suggesting a good stability over time. Table 2: Principal component analysis: Total variance explained Table 2: Principal component analysis: Total variance explained
Factor
% of variance
Cumulative %
1
26.55
26.55
2
7.68
34.24
3
6.77
41.01
4
6.55
47.56
5
5.65
53.22
6
5.06
58.28 Page 4 of 7
(page number not for citation purposes)
Table 2: Principal component analysis: Total variance explained
Factor
% of variance
Cumulative %
1
26.55
26.55
2
7.68
34.24
3
6.77
41.01
4
6.55
47.56
5
5.65
53.22
6
5.06
58.28 Characteristics of the population Page 3 of 7
(page number not for citation purposes) Page 3 of 7
(page number not for citation purposes) http://www.biomedcentral.com/1471-244X/9/66 BMC Psychiatry 2009, 9:66 Table 1: Demographic sample characteristics and psychiatric history
Variable
Age (years; mean, SD)
34
7
Gender (n)
Male
86
81.9%
Female
19
18.1%
Years of education (mean,SD)
9.7
3.1
Marital status (n)
Single
91
86.7%
Married
10
9.5%
Divorced
4
3.8%
Occupation (n)
Unemployed
60
57.2%
Working
40
38.1%
Studying
3
2.9%
Retired
2
1.9%
Duration of illness (years; mean, SD)
10.17
6.01
Number of hospitalisations (mean, min-max)
3
[0-22]
Total period of hospital stay (weeks; mean, min-max)
10.38
[0-60]
Neuroleptics (n)
Neuroleptics (n)
First generation
76
72.4%
Second generation
29
27.6%
Chlorpromazine equivalent of antipsychotic dosage (mean, SD)
482.5
322 Table 1: Demographic sample characteristics and psychiatric history
Variable
Age (years; mean, SD)
34
7
Gender (n)
Male
86
81.9%
Female
19
18.1%
Years of education (mean,SD)
9.7
3.1
Marital status (n)
Single
91
86.7%
Married
10
9.5%
Divorced
4
3.8%
Occupation (n)
Unemployed
60
57.2%
Working
40
38.1%
Studying
3
2.9%
Retired
2
1.9%
Duration of illness (years; mean, SD)
10.17
6.01
Number of hospitalisations (mean, min-max)
3
[0-22]
Total period of hospital stay (weeks; mean, min-max)
10.38
[0-60]
Neuroleptics (n)
Neuroleptics (n)
First generation
76
72.4%
Second generation
29
27.6%
Chlorpromazine equivalent of antipsychotic dosage (mean, SD)
482.5
322 Table 1: Demographic sample characteristics and psychiatric history According to these criteria, the factors extracted should
have an eigenvalue greater than 1, provided that the total
variance explained exceeded 50%. PCA with Oblimin
rotation yielded six factors with 58.2% explained variance
(Table 2). The eigenvalues of the first two factors were
5.57 and 1.61, respectively, and the corresponding vari-
ances were 26.55% and 7.68%. Statistical analyses were performed using SPSS software in
his 12th version. Statistical analyses were performed using SPSS software in
his 12th version. Ethics and Consent This research has been undergone in a psychiatric univer-
sity department in RAZI hospital. It has been approved by
the local ethic committee. Patients have signed a written
and informed consent. In order to evaluate cognition as conceptualized by sub-
jectivity, PCA with Oblimin rotation method [34] was
performed to see whether latent variables would emerge
and lead to a cognitive model different from the initial
theoretical one that have been the basis of our scale. After
carrying out an Oblimin Rotation, the items with a load-
ing higher then 0.50 were retained to be part of the sub-
jective cognitive factors (Table 3). Results The SASCCS global score mean was 24.98 (SD = 14.83;
min = 0, max = 109; median = 24). Correlations between psychopathological assessment and
scale's scores We examined whether correlations existed between scores
derived from the scale and positive, negative and disor- Internal consistency It was evaluated by calculating Cronbach's alpha coeffi-
cient [32] which was equal to 0.85 proving a good internal
consistency of the scale but furthermore, a satisfactory
reliability of its measure. Validity of internal structure We carried out a factor analysis using principal compo-
nent analysis as the extraction method. The decision-mak-
ing for factor extraction was based on Kaiser criteria [33]. http://www.biomedcentral.com/1471-244X/9/66 BMC Psychiatry 2009, 9:66 Table 3: Subjective cognitive domains of complaints
Subjective Factors
Items
Subjective domains
1
5, 12, 16, 20, 21
Distractibility
2
8, 9, 15, 18
Daily life
3
10, 11
Semantic memory
4
4, 6, 7
Disorder consciousness
5
1, 2
Working memory
6
3, 14, 17, 19
Executive skills Table 3: Subjective cognitive domains of complaints city might be imperfect as it refers to several overlapping
dimensions. These findings point to the complex representation of
schizophrenic patients of their own cognition. And even
though the latter does not correspond to the theoretical
construct of cognition, the scale remains reliable because
of both its good internal consistency and stability over
time. During this study, no other instrument evaluating cogni-
tive functions was administered simultaneously to our
population. Therefore, convergent validity was unneces-
sary. However, when reviewing the literature, no positive
correlation was found between objective and subjective
scores of cognition. Using the SSTICS, Prouteau et al. found that cognitive nature of subjective complaints did
not strictly match with that of impaired objective per-
formances [36]. Chan et al. assessed prospective memory
in patients with schizophrenia and did not find a correla-
tion between objective performances and subjective meas-
ures of this cognitive function [37]. ganisation factors derived from the factorial analyses stud-
ies of the PANSS [35]. We also considered the item G 12
assessing insight as well as the Calgary Depression Scale
total score. The SASCCS total score wasn't correlated to any of the
PANSS. A weak negative correlation between the SASCCS
total score and PANSS insight score was found (r = -.21)
but didn't reach the statistical significance (p = .03). Cal-
gary score was correlated with the SASCCS total score (r =
.33; p = .001). These results suggest that subjective evaluation of cogni-
tion could be an independent dimension from its objec-
tive assessment in patients with schizophrenia. The cognitive factor of the 5-factor model of the PANSS
wasn't correlated to the SASCCS total score or sub-scores. The depression factor was correlated to the SASCCS total
score (r = .20) although this correlation didn't reach the
statistical significance (p = .03). Validity of internal structure In fact, in our study, no correlation has been found
between the SASCCS scores and the PANSS cognitive fac-
tor which could also point to the fact that self-assessment
of cognition is a totally independent aspect from clinical
evaluation of the cognitive functions. Additional file 4 examples for items 13, 15, 18 and 19 of the SASCCS. these are the
examples that the investigator could provide to the patient when adminis-
tering the SASCCS to clarify the meaning of items 13, 15, 18 and 19. Click here for file examples for items 13, 15, 18 and 19 of the SASCCS. these are the
examples that the investigator could provide to the patient when adminis-
tering the SASCCS to clarify the meaning of items 13, 15, 18 and 19. Click here for file [http://www.biomedcentral.com/content/supplementary/1471-
244X-9-66-S4.DOC] References 1. Heinrichs RW, Zakzanis KK: Neurocognitive deficits in schizo-
phrenia: A quantitative review of the evidence. Neuropsychol
1998, 12:426-46. 1. Heinrichs RW, Zakzanis KK: Neurocognitive deficits in schizo-
phrenia: A quantitative review of the evidence. Neuropsychol
1998, 12:426-46. Authors' contributions IJ, OK and OBA led the study concept and design, data col-
lection, data analysis, and drafting of the manuscript. LD,
YR and KT participated in the pre-test of experimental ver-
sion of the scale. LD and YR participated in data collec-
tion. All authors read and approved the final manuscript. In our study, there was a positive correlation between SAS-
CCS total score and CDS score meaning that the more
depressive symptomatology is severe, the more the patient
reports cognitive troubles. Although it was not statistically
significant, we also did find a correlation between SAS-
CCS total score and the depression score of the 5-factor
model of the PANSS suggesting the influence that could
exert depression on self assessment of cognition by
emphasizing cognitive complaints when being more
depressed. List of abbreviations 2. Fioravanti M, Carlone O, Vitale B, Cinti ME, Clare L: A meta-anal-
ysis of cognitive deficits in adults with diagnosis of schizo-
phrenia. Neuropsychol Rev 2005, 15:73-95. SASCCS: Self-Assessment Scale of Cognitive Complaints
in Schizophrenia; PCA: Principal Component Analysis;
SSTICS: Subjective Scale To Investigate Cognition in
Schizophrenia. 2. Fioravanti M, Carlone O, Vitale B, Cinti ME, Clare L: A meta-anal-
ysis of cognitive deficits in adults with diagnosis of schizo-
phrenia. Neuropsychol Rev 2005, 15:73-95. p
p y
3. Keefe RSE, Fenton WS: How Should DSM-V Criteria for Schiz-
ophrenia Include Cognitive Impairment? Schizophr Bull 2007,
33:912-20. p
p y
3. Keefe RSE, Fenton WS: How Should DSM-V Criteria for Schiz-
ophrenia Include Cognitive Impairment? Schizophr Bull 2007,
33:912-20. 4. Green MF, Kern RS, Braff DL, Mintz J: Neurocognitive deficits and
functional outcome in schizophrenia: are measuring the «
right stuff?? ». Schizophr Bull 2000, 26:119-136. 4. Green MF, Kern RS, Braff DL, Mintz J: Neurocognitive deficits and
functional outcome in schizophrenia: are measuring the «
right stuff?? ». Schizophr Bull 2000, 26:119-136. Conclusion [http://www.biomedcentral.com/content/supplementary/1471-
244X-9-66-S3.DOC] We present here a self-assessment scale to evaluate cogni-
tive deficits as perceived by patients suffering from schiz-
ophrenia in domains of memory, attention and executive
functions. Given the good psychometric properties that
have been revealed in this study, the SASCCS seems to be
reliable to measure schizophrenic patients' perception of
their own cognitive impairment. This kind of evaluation
can not replace objective measures of cognitive perform-
ances in schizophrenia. Actually, the purpose of such an
evaluation is to allow the patient to express his own well-
being and satisfaction of quality of life. Furthermore, sub-
jective evaluation of cognitive functions could provide a
more complete picture of the cognitive profile of an indi-
vidual. Therefore, better therapeutic targets could be
adapted to his condition during cognitive rehabilitation
programs. Additional file 2 Lecardeur et al. found in their study using the SSTICS a
correlation between the scale total score and the PANSS
depression score. It could be suggested that subjective
complaints of cognitive deficits may influence a patient's
objective depressive state as rated by the clinician [41]. English version of the SACSS (not validated). this is the English version
of the SASCCS which is not validated. Click here for file
[http://www.biomedcentral.com/content/supplementary/1471-
244X-9-66-S2.DOC] English version of the SACSS (not validated). this is the English version
of the SASCCS which is not validated. Click here for file
[http://www.biomedcentral.com/content/supplementary/1471-
244X-9-66-S2.DOC] Considering the influence of depressive traits on subjec-
tive perception toward cognition, we recommend measur-
ing the patient's mood state when using the SASCCS. Additional file 3 Table S1: demographic characteristics and psychiatry history of pre-
test sample. this table describes the sociodemographic characteristics of
pre-test sample as well as its psychiatric history. Click here for file
[http://www.biomedcentral.com/content/supplementary/1471-
244X-9-66-S3.DOC] Table S1: demographic characteristics and psychiatry history of pre-
test sample. this table describes the sociodemographic characteristics of
pre-test sample as well as its psychiatric history. Click here for file Acknowledgements The authors would like to thank Marie-Chantal Bourdel for her help in sta-
tistical analyses. Discussion Page 5 of 7
(page number not for citation purposes) Page 5 of 7
(page number not for citation purposes) http://www.biomedcentral.com/1471-244X/9/66 BMC Psychiatry 2009, 9:66 BMC Psychiatry 2009, 9:66 Another important factor to be considered is depression
since a depressive state could be accompanied by cogni-
tive disturbances in several domains such memory and
attention [39,40]. Additional material Additional file 1
Tunisian version of the SACSS. this is the original version of the SAS-
CCS scale written in Tunisian Arabic. Click here for file
[http://www.biomedcentral.com/content/supplementary/1471-
244X-9-66-S1.DOC] Additional file 1
Tunisian version of the SACSS. this is the original version of the SAS-
CCS scale written in Tunisian Arabic. Click here for file
[http://www.biomedcentral.com/content/supplementary/1471-
244X-9-66-S1.DOC] Additional file 1
Tunisian version of the SACSS. this is the original version of the SAS-
CCS scale written in Tunisian Arabic. Click here for file
[http://www.biomedcentral.com/content/supplementary/1471-
244X-9-66-S1.DOC] Discussion The aim of this study was to construct and to validate a
scale to measure the subjectivity of patients with schizo-
phrenia regarding their cognition. The SASCCS, which
was easy to administer in less than 15 minutes, had good
reliability and stability over time. No cut-off has been
determined for this scale. In fact, the SASCCS total score is
used to estimate a patient's level of complaining. The correlation of insight with subjective perception of
cognition in schizophrenia is an aspect that deserves to be
considered and analyzed. In fact, awareness of one's own cognitive deficits could be
highly influenced by consciousness of one's whole condi-
tion as a mentally ill person. Schizophrenia is generally
accompanied by a lack of insight meaning an impaired
awareness of one's psychiatric condition and life situation
[38]. Therefore, low scoring at the SASCCS could be due
to a lack of insight. The composition of subjective cognitive domains as
derived from factor analysis was slightly different from
that of the initial theoretical model which has been the
basis of the scale's construction. Actually, the scale's items
have been distributed after PCA differently from the orig-
inal structure of the scale leading to a neo-construct of the
instrument. These differences were not surprising since
the questionnaire was based on the neuropsychological
theoretical conception of cognition whereas factor analy-
sis of the scale reflected the patient's own perception of his
cognition. Stip et al., using the Subjective Scale To Investi-
gate Cognition in Schizophrenia (SSTICS), have also
found a difference between the distribution of the items
in the initial model and in the neo-construct of their scale
[14]. It could be that the selected items did not exactly
measure what they were supposed to. Also, their specifi- In our study, a weak negative correlation between PANSS
insight score and SASCCS score has been found (r = -.21,
p = .03) but didn't reach the significance level set at 0.01. However, it should be noticed that our study included a
majority of subjects scoring no more than 4 on the PANSS
insight item. Only one patient had a score of 5. Since insight could influence one's subjective perception
of cognition, it is recommended to evaluate patient's
insight while using the SASCCS. Page 6 of 7
(page number not for citation purposes) (page number not for citation purposes) http://www.biomedcentral.com/1471-244X/9/66 BMC Psychiatry 2009, 9:66 6. Keefe RS, Goldberg TE, Harvey PD, Gold JM, Poe MP, Coughenour L:
The Brief Assessment of cognition in schizophrenia: reliabil-
ity, sensitivity and comparison with a standard neuroccogni-
tive battery. Schizophr Res 2004, 68:283-297. 32. Cronbach LJ: Coefficient alpha and the internal structure of
tests. Psychometrica 1951, 16:297-335. y
33. Jennrich RI, Sampson PF: Rotation for simple loadings. Psychomet-
rica 1966, 31:313-323. tive battery. Schizophr Res 2004, 68:283-297. 34. Kaiser HF: Image analysis. Madison, Wisconsin: Harris CW; 1963. 7. Green MF, Kern RS, Heaton RK: Longitudinal studies of cogni-
tion and functional outcome in schizophrenia: implications
for MATRICS. Schizophr Res 2004, 72:41-51. 35. Lepine JP, Piron JJ, Chapatot E: Factor analysis of the PANSS in
schizophrenic patients. In Psychiatry today: accomplishments and
promises Edited by: Stefanis CN, Soltados CR, Rabavilas AD. Amster-
dam: experta medica; 1989. 8. Saykin AJ, Shtasel DL, Gur RE, Kester DB, Mozley LH, Stafiniak P:
Neuropsychologial deficits in neuroleptic naïve patients with
first-episode schizophrenia. Arch Gen Psychiatry 1994,
51:124-131. 36. Prouteau A, Verdoux H, Briand C, Lesage A, Lalonde P, Nicole L,
Reinharz D, Stip E: Self-assessed cognitive dysfunction and
objective performance in out-patients with schizophrenia
participating in a rehabilitation program. Schizophr Res 2004,
69:85-91. 9. Green MF, Kern RS, Braff DL, Mintz J: Neurocognitive deficits and
functional outcome in schizophrenia: are we measuring the
« right stuff?? ». Schizophr Bull 2000, 26:119-136. g
p
10. Dickinson D, Bellack AS, Gold JM: Social/Communication Skills,
Cognition, and Vocational Functioning in Schizophrenia. Schizophr Bull 2007, 33:1213-20. g
p
kinson D, Bellack AS, Gold JM: Social/Communication Sk
i i
d V
i
l F
i
i
i
S hi
h 37. Chan RCK, Wang Y, Ma Z, Chan RC, Wang Y, Ma Z, Hong XH, Yuan
Y, Yu X, Li Z, Shum D, Gong QY: Objective measures of pro-
spective memory do not correlate with subjective com-
plaints in schizophrenia. Schizophr Res 2008, 103:229-239. J
,
Cognition, and Vocational Functioning in Schizophrenia. Schizophr Bull 2007, 33:1213-20. p
11. McClure MM, Bowie CR, Patterson TL, Heaton RK: Correlations of
functional capacity and neuropsychological performance in
older patients with schizophrenia: Evidence for specificity of
relationships? Schizophr Res 2007, 89:330-38. 38. Pini S, Cassano GB, Dell'Osso L, Amador XF: Insight into illness in
schizophrenia, schizoaffective disorder, and mood disorders
with psychotic features. Am J Psychaitry 2001, 158:122-125. p
p
12. http://www.biomedcentral.com/1471-244X/9/66 Cohen AS, Forbes CB, Mann MC, Blanchard JJ: Specific cognitive
deficits and differential domains of social functioning impair-
ment in schizophrenia. Schizophr Res 2006, 81:227-38. 39. Brébion G, Smith MJ, Amador X, Malaspina D, Gorman JM: Clinical
correlates of memory in schizophrenia: Differentail links
between depression, positive and negative symptoms, and
two types of memory impairment. Am J Psychiatry 1997,
154:1538-1543. p
p
13. Green MF: What are the functional consequences of neuro-
cognitive deficits in schizophrenia? Am J Psychiatry 1996,
153:321-330. 40. Holthausen EA, Wiersma D, Knegtering RH, Bosch RJ Van den: Psy-
chopathology and cognition in schizophrenic spectrum dis-
orders: the role of depressive symptoms. Schizophr Res 1999,
59:137-146. 14. Stip E, Caron J, Renaud S, Pampoulova T, Lecompte Y: Exploring
cognitive complaints in schizophrenia: The Subjective Scale
To Investigate cognition in Scizophrenia. Compr Psychiatry
2003, 44:331-340. 41. Lecardeur L, Briand C, Prouteau A, Lalonde P, Nicole L, Lesage A, Stip
E: Preserved awareness of their cognitive deficits in patients
with schizophrenia: Convergent validity of the SSTICS. Schiz-
ophr Res 2009, 107:303-306. 15. Peralta V, Cuesta MJ: Subjective experience in schizophrenia: A
critical review. Compr Psychiatry 1994, 35:198-204. p
y
y
16. Cutting J, Dunne F: Subjective experience of schizophrenia. Schizophr Bull 1989, 15:217-231. p
17. Strauss JS: Subjective experiences of schizophrenia: toward a
new dynamic psychiatry-II. Schizophr Bull 1989, 15:179-187. Pre-publication history p
y
The pre-publication history for this paper can be accessed
here: p
The pre-publication history for this paper can be accessed
here: y
p y
y
p
18. Huber G: Die coenesthetische Schizophrenie. Fortschr Neurol
Psychiatr 1957, 25:491-520. y
19. Huber G: Reine Defektsyndrome und Basisstadien endogener
Psychosen. Fortschr Neurol Psychiatr 1966, 34:409-426. http://www.biomedcentral.com/1471-244X/9/66/pre
pub http://www.biomedcentral.com/1471-244X/9/66/pre
pub 20. Koehler K, Saller H: Huber's Basic symptoms: Another
Approach to Negative Psychopathology in schizophrenia. Compr Psychiatry 1984, 25:174-182. p
y
y
21. Gross G, Huber G, Klosterkotter J, Linz M: Bonn Scale for Assass-
ment of Basic Symptoms. Berlin: Spinger Verlag; 1987. 22. Sullwold L: Frankfurter Beschwerde-Fragboden (FBF). In
Schizohrene Basisstorungen Berlin: Springer; 1986:1-36. g
p
g
23. Liddle PF, Barnes TR: The subjective experience of deficits in
schizophrenia. Compr Psychiatry 1988, 29:157-164. 24. Jaeger J, Bitter I, Czobor P, Volavka J: The measurement of sub-
jective experience in schizophrenia: the Subjective Deficit
syndrome scale. Compr Psychiatry 1990, 31:216-226. y
p
y
y
,
25. Mass R: Characteristic subjective experiences of schizophre-
nia. Schizophr Bull 2000, 26:921-931. p
26. Saykin AJ, Gur RC, Gur RE, Mozley PD, Mozley LH, Resnick SM, Kes-
ter DB, Stafiniak P: Neuropsychological function in schizophre-
nia. Selective impairment in memory and learning. Arch Gen
Psychiatry 1991, 48:618-624. y
y
27. American Psychiatric Association: Diagnostic and Statistical
Manual of Mental Disorders. 4th edition. Washington DC: Amer-
ican psychiatric Association; 1994. Pre-publication history Publish with BioMed Central and every
scientist can read your work free of charge
"BioMed Central will be the most significant development for
disseminating the results of biomedical research in our lifetime."
Sir Paul Nurse, Cancer Research UK
Your research papers will be:
available free of charge to the entire biomedical community
peer reviewed and published immediately upon acceptance
cited in PubMed and archived on PubMed Central
yours — you keep the copyright
Submit your manuscript here:
http://www.biomedcentral.com/info/publishing_adv.asp
BioMedcentral
Page 7 of 7
(page number not for citation purposes) Publish with BioMed Central and every
scientist can read your work free of charge
"BioMed Central will be the most significant development for
disseminating the results of biomedical research in our lifetime."
Sir Paul Nurse, Cancer Research UK
Your research papers will be:
available free of charge to the entire biomedical community
peer reviewed and published immediately upon acceptance
cited in PubMed and archived on PubMed Central
yours — you keep the copyright
Submit your manuscript here:
http://www.biomedcentral.com/info/publishing_adv.asp
BioMedcentral Publish with BioMed Central and every
scientist can read your work free of charge Competing interests p
g
The authors declare that they have no competing interests. The authors declare that they have no competing interests. g
p
5. Keefe RS: Should cognitive impairment be included in the
diagnostic criteria for schizophrenia? World Psychiatry 2008,
7:22-28. g
p
5. Keefe RS: Should cognitive impairment be included in the
diagnostic criteria for schizophrenia? World Psychiatry 2008,
7:22-28. Page 6 of 7
(page number not for citation purposes) http://www.biomedcentral.com/1471-244X/9/66 Publish with BioMed Central and every
scientist can read your work free of charge 28. Johnson I, Dhouib S, Bouaziz N, Ketata W, Dellagi L, Kebir O, Ben
Azouz O, Tabbane K: Evaluation de la perception subjective
des déficits cognitifs chez les patients atteints de schizo-
phrénie [Abstract]. L'Encéphale 2007, 32:52. p
[
]
p
29. Kay SR, Fiszbein A, Opler LA: The positive and negative syn-
drome scale (PANSS) for schizophrenia. Schizophr Bull 1987,
13:261-276. 30. Addington D, Addington J, Maticka-Tyndale E: Specificity of the
Calgary Depression Scale for schizophrenics. Schizophr Res
1994, 11:239-244. 31. Lindenmayer JP, Bernstein-Hyman R, Grochowski S: Five-factor
model of schizophrenia. Initial validation. J Nerv Ment Dis 1994,
182:631-38.
|
https://openalex.org/W3096766075
|
https://www.ocl-journal.org/articles/ocl/pdf/2020/01/ocl200037.pdf
|
English
| null |
Estimating crop model parameters for simulating soybean production in Iran conditions
|
Oilseeds and fats, crops and lipids/OCL. Oilseeds & fats crops and lipids
| 2,020
|
cc-by
| 6,970
|
Received 24 March 2020 – Accepted 2 October 2020 Abstract – Crop modelling has the potential to contribute to food security. In this study, to provide a simple
model for estimating the soybean potential yield and phenological stages in Iran, a simulation model
(SSM_iCrop2) was parameterized and tested. This model estimates the soybean phenological stages and
potential yield based on the weather data (minimum and maximum temperature, solar radiation and rainfall)
using the phenological models such as leaf area development, mass production and partitioning and soil
water balance. Regarding the model parametrization, the two maturities groups of 3 and 5 with the
temperature unit of 2000 and 2400 growth degrees day (GDD) were chosen. The model evaluation results
indicated that the soybean yield ranged between 1.9 and 4.8 with the average of 3.5 t.ha1, while the range of
simulated yield changes between 1.8 and 4.7 with the average of 3.7 t.ha1. Comparing the observed yield to
the simulated yield, values of r, CVand RMSE were obtained 0.84, 13%, 0.5 t.ha1 which indicates the high
accuracy of the model. All of these results indicated that the estimated model parameters are high accuracy
for use in the simulation of soybean yield at the country level. Keywords: parameterization / evaluation / food security / crop model / climate change Résumé – Estimation des paramètres d’un modèle de culture simulant la production de soja en Iran. La modélisation des cultures peut contribuer à la sécurité alimentaire. Dans cette étude, un modèle de
simulation (SSM_iCrop2) a été paramétré et testé afin d’estimer les stades phénologiques et le rendement
potentiel du soja en Iran. Ce modèle simple s’appuie sur une base de données météorologiques (température
minimale et maximale, rayonnement solaire et précipitations) pour estimer la phénologie, la surface foliaire,
la production et l’allocation de la biomasse, et le bilan hydrique du sol. Pour le paramétrage du modèle, seuls
les groupes de maturité III et V correspondant à des sommes de température de 2000 et 2400 degrés-jours
ont été retenus. Les résultats de l’évaluation du modèle ont indiqué que le rendement observé du soja se
situait entre 1,9 et 4,8 t.ha1, avec une moyenne de 3,5 t.ha1, tandis que la gamme des rendements simulés
variait de 1,8 à 4,7 t.ha1 avec une moyenne de 3,7 t.ha1. En comparant le rendement observé au rendement
simulé, on a obtenu des valeurs de r, CVet RMSE de 0,84, 13%, 0,5 t.ha1,respectivement, ce qui souligne la
grande précision du modèle. Received 24 March 2020 – Accepted 2 October 2020 Tous ces résultats indiquent que les paramètres estimés pour ce modèle sont
suffisamment précis pour être utilisés dans la simulation du rendement du soja au niveau national. Mots clés : paramétrage / évaluation / sécurité alimentaire / modèle de culture / changement climatique production in Iran are about 66 000 hectares and 151 000 tons,
respectively which does not meet the domestic needs, so
imported soybean meal to Iran amounted to 2.37 million tonnes
with the worth of $1.5 billion in 2012 (Ministry of Agriculture
Jihad,
2016). Also,
imported
soybean
oil
was
about
800 000 tonnes worth $960 million in 2013 (FAOSTAT,
2013). So far, several attempts have been made using field
experiments to better understand factors affecting crop yield ☆Contribution to the Topical Issue “Soybean / Soja”.
*Correspondence: V.dadrasi@gmail.com This is an Open Access article distributed under the terms of the Creative Commons Attribution License (https://creativecommons.org/licenses/by/4.0), which permits
unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited. OCL 2020, 27, 58
© A. Nehbandani et al., Hosted by EDP Sciences, 2020
https://doi.org/10.1051/ocl/2020057 OCL 2020, 27, 58
© A. Nehbandani et al., Hosted by EDP Sciences, 2020
https://doi.org/10.1051/ocl/2020057 OCL 2020, 27, 58
© A. Nehbandani et al., Hosted by EDP Sciences, 2020
https://doi.org/10.1051/ocl/2020057 Oil
d & f
C
OCL Available online at:
www.ocl-journal.org *Correspondence: V.dadrasi@gmail.com 1 This model can be downloaded from: “https://sites.google.com/site/
cropmodeling/-5-SSM_iCrop2”. 1 Introduction Due to
these constraints, crop models can be useful tools to study and
estimate the yield (Geerts and Raes, 2009) A mathematical
model is an equation or set of equations that describes the
behaviour of each system quantitatively (Soltani, 2009). To predict the crop growth, studies on phenology, mass
production and partitioning, leaf area development and soil
water balance are required (Dadrasi and Torabi, 2016). Precise
prediction of the crop phenology is one of the essential features
of the simulation models. The mass production and partition-
ing are largely regulated by the timing of developmental stages
in crop simulation models (Soltani, 2012). Simple models can
be more efficient in yield analysis and investigate the limiting
factors due to easy manufacturing, testing, applying, under-
standing and interpretation of results; In addition it needs
minimum inputs (Sinclair and Muchow, 1999). One of the
other benefits of modelling is the prediction of the food
production status in one area and making decisions based on
environmental changes. Models apply a variety of plant and
environmental parameters to simulate crop growth and they
should be calibrated and evaluated before being used (Hsiao
et al., 2009). In some models, the parameters related to plant
characteristics may have been calculated according to the
climate of a certain region, which is not usable in other regions
or may not have acceptable performance. Therefore, to predict
the crop growth and yield by the model, the compatibility of
the equations with the relationships between the different
processes of growth and yield and the climatic conditions of
the study area, access to the input parameters and the model
efficiency in predicting growth and yield should be considered
(Torabi et al., 2011). 1 Introduction Soybean (Glycine max) is one of the most important
oilseed crops cultivated in the world. Soybean crop area and A. Nehbandani et al.: OCL 2020, 27, 58 A. Nehbandani et al.: OCL 2020, 27, 58 Briefly, the model is tested using different values for a specific
parameter, then values are chosen that provide the closest
match to the observations of the major outputs, frequently final
yield. Also, several model parameters that are fixed within
different soybean cultivars (such as cardinal temperature) were
obtained from credible references. per unit area. Field experiments on the crop response to various
environmental conditions are laborious and costly. Due to
these constraints, crop models can be useful tools to study and
estimate the yield (Geerts and Raes, 2009) A mathematical
model is an equation or set of equations that describes the
behaviour of each system quantitatively (Soltani, 2009). To predict the crop growth, studies on phenology, mass
production and partitioning, leaf area development and soil
water balance are required (Dadrasi and Torabi, 2016). Precise
prediction of the crop phenology is one of the essential features
of the simulation models. The mass production and partition-
ing are largely regulated by the timing of developmental stages
in crop simulation models (Soltani, 2012). Simple models can
be more efficient in yield analysis and investigate the limiting
factors due to easy manufacturing, testing, applying, under-
standing and interpretation of results; In addition it needs
minimum inputs (Sinclair and Muchow, 1999). One of the
other benefits of modelling is the prediction of the food
production status in one area and making decisions based on
environmental changes. Models apply a variety of plant and
environmental parameters to simulate crop growth and they
should be calibrated and evaluated before being used (Hsiao
et al., 2009). In some models, the parameters related to plant
characteristics may have been calculated according to the
climate of a certain region, which is not usable in other regions
or may not have acceptable performance. Therefore, to predict
the crop growth and yield by the model, the compatibility of
the equations with the relationships between the different
processes of growth and yield and the climatic conditions of
the study area, access to the input parameters and the model
efficiency in predicting growth and yield should be considered
(Torabi et al., 2011). per unit area. Field experiments on the crop response to various
environmental conditions are laborious and costly. 2.2.1 Weather data Meteorological information of each experimental site
including minimum and maximum daily temperature, daily
precipitation, and solar radiation were obtained from the
nearest meteorological station. Outliers and missing data were
then estimated and restored using the WeatherMan program
(Hoogenboom et al., 2004). Therefore, the aim of this study was to determine the
SSM_iCrop2 parameters in simulating main cultivars soybean
growth and yield in Iran to provide a tool for analysing yield-
limiting factors, optimizing field management and identifying
the factors that influence the yield increase in certain
environmental conditions. 2.2 Crop model The model used in this study was SSM_iCrop21 (Soltani
and Sinclair, 2012; Soltani et al., 2020a). The model includes
daily phenology progress, leaf area development and senes-
cence, dry matter production, yield formation, and soil water
balance. Responses of crop processes to solar radiation,
temperature, water availability, and cultivar differences are
included in the model. Soil water sub-model accounts for soil
water additions from precipitation or irrigation, and increasing
rooting depth and water removal via deep drainage, run-off,
soil evaporation, and plant transpiration. The soil profile is
divided into two layers: one top layer of 15–20 cm thickness
and a second layer that includes the first layer and its depth
increases by root growth. Soil water balance of both layers is
calculated separately. The effect of water deficit and excess on
leaf area development and senescence, dry mass accumulation,
and phenological development are simulated. The model also
accounts for the effect of freezing temperatures on plant leaf
area that might take place in early spring sowings or winter
sowings. The model has been tested extensively for a wide
range of plant species and proved to be robust (Soltani et al.,
2020a). Some of the parameters required in the SSM_iCrop2 model
for soybean are presented in Table 2. Some used models for simulation of soybean yield and
phenological stages include EPIC (Williams and Watson,
1985), SoySim (Setiyono et al., 2010), CROPGRO-Soybean
(Boote et al., 1998) and APSIM (Keating et al., 2003). There is
another a group of plant models, called SSM_iCrop2 (Soltani
et al., 2020a). SSM_iCrop2 model simulated a large number of
plant species including orchard species and perennial forages
(Soltani et al., 2020a). 2.2.2 Soil data The required soil information included soil albedo index,
Drainage coefficient, soil water volume in the field capacity,
wilting point and saturation conditions. There is no local
digitized soil database for crop modeling in Iran, so the HC27
database (Koo and Dimes, 2013) was utilized. The resolution
of the soil database is also important. HC27 soil database used
in the current study has a resolution of 10-km which may
seem
coarse. Tests
using
SSM_iCrop2
for
crop
and
horticultural species indicated that using HC27 soil profiles
compared to actual, measured soil profiles resulted in similar
output for yield and the net amount of irrigation water
requirements
or
evapotranspiration
with
no
significant
difference with respect to mean, variance and distribution
(Nehbandani et al., 2020b). 2.1 Data used Location and year
Treatments
References
Parameterization
Gorgan, 2002
Sowing date, Cultivars
Zeinali et al., 2003
Gorgan, 2011
Cultivars, Sowing date
Gorzin et al., 2015
Ghaemshahr, 2010
Irrigation
Akbari Nodehi, 2012
Gharakhil, 2009
Cultivars, Plant density, Sowing date
Rameeh and Aghabozorgi, 2016
Ardabil, 2013
Seed inoculation
Sidi Sharifiand Khoramdel, 2013
Gorgan, 2002
Cultivars, Plant density
Zahleht Salmasi et al., 2004
Mahmod abad, 2015
Cultivars, Nitrogen fertilizer
Mahmodi and Zakipour, 2016
Gorgan, 2002
Cultivars, Plant density
Zahleht Salmasi et al., 2004
Mahmod abad, 2015
Cultivars, Nitrogen fertilizer
Mahmodi and Zakipour, 2016
Sari, 2011
Cultivars, Sowing date
Ghanbari Malayder et al., 2015
Evaluation
Gorgan, 2012
Crop density, Cultivars
Nehbandani, 2013
Babulsar, 2001
Zinc and Potassium fertilizer
Habibzadeh et al., 2003
Gorgan, 2006 and 2007
Cultivars, Plant density
Raeisi and Hezarjeribi, 2013
Gorgan, 2005
Plant density
Najafi, 2006
Gorgan, 2011 and 2012
Irrigation
Hosaini et al., 2016
Ardabil, 2011
Seed inoculation, Nitrogen fertilizer
Sidi Sharifiand Khoramdel, 2013
Ghaemshahr, 2010
Cultivars, Planting patern
Namdari and Mahmoodi, 2013
Moghan, 2007 and 2008
Cultivars, Plant density, Sowing date
Razmi, 2010
Sari, 2010
Cultivars
Fazeli et al., 2016
Gorgan, 2011
Cultivars, Plant density
Mosnei et al., 2015
Daland, 2014
Seed inoculation
Ghana, 2016
Ardabil, 2009
Cultivars, Sowing date
Mousavi and Chavoshi, 2013
Nekah, 2010
Irrigation
Akbari Nodehi, 2011
Bilesowar, 2011
Seed inoculation, Nitrogen fertilizer
Zendeh et al., 2016
Sari, 2002
Irrigation
Qajar Sepanlou and Behminar, 2004
Gorgan, 2011 and 2012
Irrigation
Faraji, 2016
Ardabil, 2013
Seed inoculation
Seyed Sharifi, 2015 Location and year References Zeinali et al., 2003
Gorzin et al., 2015
Akbari Nodehi, 2012
Rameeh and Aghabozorgi, 2016
Sidi Sharifiand Khoramdel, 2013
Zahleht Salmasi et al., 2004
Mahmodi and Zakipour, 2016
Zahleht Salmasi et al., 2004
Mahmodi and Zakipour, 2016
Ghanbari Malayder et al., 2015 Zeinali et al., 2003
Gorzin et al., 2015
Akbari Nodehi, 2012
Rameeh and Aghabozorgi, 2016
Sidi Sharifiand Khoramdel, 2013
Zahleht Salmasi et al., 2004
Mahmodi and Zakipour, 2016
Zahleht Salmasi et al., 2004
Mahmodi and Zakipour, 2016
Ghanbari Malayder et al., 2015 Zeinali et al., 2003
Gorzin et al., 2015
Akbari Nodehi, 2012
Rameeh and Aghabozorgi, 2016
Sidi Sharifiand Khoramdel, 2013
Zahleht Salmasi et al., 2004
Mahmodi and Zakipour, 2016
Zahleht Salmasi et al., 2004
Mahmodi and Zakipour, 2016
Ghanbari Malayder et al., 2015 Table 2. Required weather, soil and crop management input to run SSM_iCrop2. 2.1 Data used To parameterize and evaluate the model, in major areas in
terms of soybean cultivation, data of various studies on
soybean (treatments without any growth and development
limiting factors or environmental stresses) was applied
(Tab. 1). About 35% of published paper was applied to
parameterize the model (include 34 situations for days to
maturity and 26 situations for yield) and about 65% other was
utilized to evaluate it (include 49 situations for yield). Parameterization
of
SSM_iCrop2
is
straightforward
as
presented in Appendix I of Soltani and Sinclair (2012). Page 2 of 9 A. Nehbandani et al.: OCL 2020, 27, 58 Table 1. Experiments used for parameterization and evaluation of SSM_iCrop2. 2.1 Data used Input data
Abv
Unit
Weather data
Maximum daily temperature
TMAX
°C
Minimum daily temperature
TMIN
°C
Solar radiation
SRAD
MJ m2d1
Daily rainfall
RAIN
mm
Soil data
Soil albedo
SALB
–
Drainage factor
DRAINF
–
Volumetric soil water content at drained upper limit
IDUL
mm mm1
Volumetric soil water content at crop lower limit
ILL
mm mm1
Volumetric soil water content at saturation
ISAT
mm mm1
Curve number
CN
–
Soil depth
SOLDEP
mm
Crop management
Planting date
PDOY
day
Initial soil water at start of simulation
ISW
mm
Irrigation threshold level (for automatic irrigation)
IRGLVL
– Table 2. Required weather, soil and crop management input to run SSM_iCrop2. Page 3 of 9 A. Nehbandani et al.: OCL 2020, 27, 58 2.2.3 Crop management data Sexton et al. (2017) selected 14 parameters for this purpose. In
this study, 7 parameters were used for sensitivity analysis
(including tuHAR, LAIMX, KPAR, IRUE, TEC, Himax and
HImin). Variation range set at ±30% perturbation of the default
parameter values. The selection of perturbation range (±30%)
was based on Tan et al. (2017) and Noorhosseini et al. (2018). For sensitivity analysis, grain yield (t.ha1) and WTOP
(accumulated above-ground dry matter, t.ha1) were extracted
as model outputs. Box plot (created with SAS software) was
used to show changes in grain yield and WTOP under different
parameter variation ranges. For create box plots, SAS 9.4 was
used. The required crop management information included
sowing date, soil moisture content during simulation, irrigation
level. This data was obtained from the articles in Table 1. In this model, GDD was applied to determine the
difference between soybean maturity groups. For this purpose,
the phenological data of maturity groups 3 and 5 (major
maturity groups that are grown in Iran) were used (Tab. 1). The
GDD was calculated (based on soybean cardinal temperatures)
for each phenological step. 3.1 SSM_iCrop2 model parameterization To
parameterize
the
SSM_iCrop2
model,
common
soybean cultivars in Iran (groups 3 and 5) were considered. All the parameters indicated in Table 3 were necessary for
estimating soybean growth and production. Based on these
parameters, SSM_iCrop2 model could simulate the yield and
days to maturity. The results showed that the soybean observed
yields varied between 1.9 and 4.7 with an average of 3.5 t.ha1
and the range of simulated yield changes between 1.8 and 4.4
with an average of 3.9 t.ha1. The root mean square error
(RMSE) was 0.48 t.ha1 which is equivalent to 14% of both
the mean of simulated and observed yields, and the correlation
coefficient (r) was 0.63 (Fig. 1). The coefficient of variation in
field experiments is usually between 20 to 30% (Dadrasi and
Torabi, 2016). All data were within the range of ±20% of grain
yield which indicates the accurate estimation of the model
parameters. Regarding the phenological feature of days to
maturity, the model had a good estimation so that the observed
time intervals ranged from 104 to 154 days with an average of
127 days and the simulated time intervals ranged between 108
and 147 days with an average of 127 days (Fig. 2). For this
feature, RMSE, the coefficient of variation (CV) and r were
12 days, 9% and 0.60, respectively. Therefore, the results of
parametrization based on days to maturity are considered
optimal at the country level. RMSE ¼
ffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
1
n
X
n
i¼1
yi xi
ð
Þ2
s
;
ð1Þ
CV ¼
ffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
n P xi x
ð
Þ2
q
P xi
;
ð2Þ
r ¼
Xn
i¼1 xi x
ð
Þ yi y
ð
Þ
ffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
Xn
i¼1 xi x
ð
Þ2
q
ffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
Xn
i¼1 yi y
ð
Þ2
q
;
ð3Þ ð1Þ ð2Þ ð3Þ where xi is the observed values, yi is the simulated values, n is
the number of observations, x is the mean observed values
including days to maturity or grain yield in the independent
experiments, and y is the mean simulated values including
days to maturity or grain yield. 3 Results and discussion The statistical indices used for model evaluation were the
coefficient of variation (CV), root mean square error (RMSE)
and correlation coefficient (r). Also, the 1:1 line with 20%
discrepancy was used to show the amount of deviation of the
simulated versus the observed values. These statistical indices
were calculated as follows: 2.4 Sensitivity analysis Sensitivity analysis is the study of how the different input
variations of a mathematical model influence the variability of
its output (Monod et al., 2006). In this research, we used local
sensitivity analysis, which evaluates the local impact of the
variation in the input factors on a model response, focusing on
sensitivity in vicinity of a set of factor values. The evaluation is
conducted through gradients or partial derivatives of the output
functions at these factor values, while the values of other input
factors are kept constant. 3.2 Model evaluation To evaluate the model, the values of simulated yield was
compared to the observed values. For this purpose, a set of
experimental data was used (Tab. 1). For simulation, required
inputs for the model and the weather data of the areas in the
studied years were provided as a file to the model. The values
that were estimated in the parameterization section (Tab. 3)
were applied as parameters. The results showed that the model
had a very good prediction of the average yield. The evaluation
results of yield showed that the observed yield ranged between
1.9 to 4.8 t.ha1 with an average of 3.5 t.ha1 and simulated
yield changes between 1.8 to 4.5 t.ha1 and with an average of
3.7 t.ha1. Also, RMSE, r and CV were 0.46 t.ha1, 0.84 and
13% (Fig. 3), which indicated a high accuracy in the soybean
yield estimation in considered provinces. Therefore, this
model can be applied for different purposes. To get full coverage of parameter value space and
considering that the runtime for one execution of the
SSM_iCrop2 model is very short, we set the parameters for
the different variation ranges, separately. In the model, about
10% of the crop parameters are approved to be varied with
cultivar and environment (Bouman and van Laar, 2006; Tan
et al., 2017). In previous studies, Tan et al. (2017) selected
16 parameters for uncertainty and sensitivity analysis, while Page 4 of 9 A. Nehbandani et al.: OCL 2020, 27, 58 Table 3. SSM_iCrop2 parameter estimates for soybean in Iran. Name
Value
Reference Phenolog
Base temperature for development (°C)
TBD
7
(Soltani and Sinclair, 2012)
Lower optimum temperature for development (°C)
TP1D
27
(Soltani and Sinclair, 2012)
Upper optimum temperature for development (°C)
TP2D
34
(Soltani and Sinclair, 2012)
Ceiling temperature for development (°C)
TCD
45
(Soltani and Sinclair, 2012)
Temperature unit for emergence or beginning leaf growth (°C)
tuEMR
68–82
Parameterization
Temperature unit for beginning of seed or fruit growth (°C)
tuBSG
1200–1440
Parameterization
Temperature unit for termination of seed or fruit growth (°C)
tuTSG
1700–2088
Parameterization
Temperature unit for physiological maturity (end of dry
mass accumulation) (°C)
tuPM
1700–2088
Parameterization
Temperature unit for harvest or leaf fall (°C)
tuHAR
2000–2400
Parameterization
Leaf area development and senescence
Point #1 for normalized leaf area vs. normalized temperature unit (x1, y1)
x1, y1
(0.15, 0.05)
Parameterization
Point #2 for normalized leaf area vs. 3.2 Model evaluation Temperature unit for beginning leaf senescence (°C Leaf senescence rate coefficient Dry mass accumulation Base temperature for dry matter production (°C) Ceiling temperature for dry matter production (°C) Grain/fruit moisture content (%, dwb) * Used as maximum plant leaf area under optimal condition (PLAMX) product by plant density. 3.2 Model evaluation normalized temperature unit (x1, y1)*
x2, y2
(0.5, 0.95)
Parameterization
Maximum expected leaf area index*
LAIMX
2.5–4
Parameterization
Temperature unit for beginning leaf senescence (°C)
tuBLS
1200–1440
Parameterization
Leaf senescence rate coefficient
SRATE
1
Fixed for soybean
Low temperature / freezing threshold for leaf death (°C)
FrzTh
8
(Soltani and Sinclair, 2012)
Relative leaf death per each degree below low temperature/ freezing threshold
FrzLDR
0.01
(Soltani and Sinclair, 2012)
Heat threshold temperature for leaf senescence (°C)
HeatTH
37
(Soltani and Sinclair, 2012)
Relative increase in leaf senescence rate per each degree above
heat threshold (°C)
HtLDR
0.1
(Soltani and Sinclair, 2012)
Dry mass accumulation
Base temperature for dry matter production (°C)
TBRUE
10
(Soltani and Sinclair, 2012)
Lower optimum temperature for dry matter production (°C)
TP1RUE
20
(Soltani and Sinclair, 2012)
Upper optimum temperature for dry matter production (°C)
TP2RUE
30
(Soltani and Sinclair, 2012)
Ceiling temperature for dry matter production (°C)
TCRUE
40
(Soltani and Sinclair, 2012)
Extinction coefficient for photosyntheticaly active radiation
KPAR
0.65
(Soltani and Sinclair, 2012)
Radiation use efficiency under optimal growth conditions (g MJ1)
RUE
1.8
(Soltani and Sinclair, 2012)
Coefficient for response of RUE to CO2 concentration
C3C4
0.8
Fixed for soybean
Yield formation
Maximum harvest index/Liner increase in harvest index (g g1 d1)
HImax
0.34–0.40
Parameterization
Fraction of dry mass remobilizable from the vegetative tissue to the developing
seeds/fruits (g g1)
FRTRL
0.25
Parameterization
Grain/fruit moisture content (%, dwb)
MC
13
Fixed for soybean
Water relations
Temperature unit for beginning root growth (°C)
tuBRG
68–82
Parameterization
Temperature unit for termination root growth (°C)
tuTRG
1200–1440
Parameterization
Initial depth of roots at emergence or beginning leaf growth (mm)
iDEPORT
200
Fixed for soybean
Maximum effective depth of water extraction from soil (mm)
MEED
1000
Fixed for soybean
Transpiration efficiency coefficient (Pa)
TEC
4.5
(Soltani and Sinclair, 2012)
FTSW threshold when dry matter production starts to decline
WSSG
0.25
(Soltani and Sinclair, 2012)
FTSW threshold when leaf area development starts to decline
WSSL
0.31
(Soltani and Sinclair, 2012)
A coefficient that specifies acceleration or delaying in development in
response to water deficit
WSSD
0.4
(Soltani and Sinclair, 2012)
* Used as maximum plant leaf area under optimal condition (PLAMX) product by plant density. 3.3 Sensitivity analysis of model parameters The ±20% discrepancy
lines are indicated by dashed lines. Solid line is 1:1 line. HImax parameters caused differences in predicted yield by
model. Among these parameters, KPAR and HImax caused the
highest difference in the predicted total dry matter. Analysis of
the KPAR parameter showed that by increasing and decreasing
KPAR about 30%, the grain yield varied from 4.6 t.ha1
(constant and unchanged parameter) to 6.0 (increase in KPAR
value) and 3.2 t.ha1 (decrease in the amount of KPAR). Also,
Increasing and decreasing the HImax parameter value by 30%
changed the grain yield from 4.6 t.ha1 to 6.0 and 3.2 t.ha1,
respectively. The most sensitive parameters for the total dry
matter were tuHAR, LAImax, and KPAR. The increase and
decrease of 30% in the values of these three parameters
changed the total dry matter. In terms of grain yield sensitivity
to the change of parameters, different results were achieved,
Fig. 2. Simulated versus measured days to maturity by SSM_iCrop2
model based on data used in model parameterization. The ±20%
discrepancy lines are indicated by dashed lines. Solid line is 1:1 line. Fig. 4. The ranking of influential cultivar-specific parameters for
WTOP (Accumulated above-ground dry matter, t ha1) under
different parameter variation ranges included þ30% (White), 30%
(Grey) and Standard (Dotted line). Fig. 2. Simulated versus measured days to maturity by SSM_iCrop2
model based on data used in model parameterization. The ±20%
discrepancy lines are indicated by dashed lines. Solid line is 1:1 line. Fig. 4. The ranking of influential cultivar-specific parameters for
WTOP (Accumulated above-ground dry matter, t ha1) under
different parameter variation ranges included þ30% (White), 30%
(Grey) and Standard (Dotted line). Fig. 4. The ranking of influential cultivar-specific parameters for
WTOP (Accumulated above-ground dry matter, t ha1) under
different parameter variation ranges included þ30% (White), 30%
(Grey) and Standard (Dotted line). Fig. 2. Simulated versus measured days to maturity by SSM_iCrop2
model based on data used in model parameterization. The ±20%
discrepancy lines are indicated by dashed lines. Solid line is 1:1 line. value) and 3.2 t.ha1 (decrease in the amount of KPAR). Also,
Increasing and decreasing the HImax parameter value by 30%
changed the grain yield from 4.6 t.ha1 to 6.0 and 3.2 t.ha1,
respectively. The most sensitive parameters for the total dry
matter were tuHAR, LAImax, and KPAR. The increase and
decrease of 30% in the values of these three parameters
changed the total dry matter. 3.3 Sensitivity analysis of model parameters amount of KPAR caused a difference in the amount of
simulated dry matter, increasing and decreasing the amount of
KPAR by 30% changed the total dry matter from 11.1 t.ha1
(constant and unchanged parameter) to 14.4 (increase in KPAR
value) and 7.8 t.ha1 (decrease in KPAR value). Also, in terms
of grain yield (considering 13% moisture content at the harvest
time), changes in the values of tuHAR, LAIMX, KPAR and The comparison of the averages presented in Figures 4 and
5 showed that the change in the values of the tuHAR, LAIMX
and KPAR parameters in the SSM_iCrop2 model caused
a difference in the total dry matter predicted by the model
(Figs. 4 and 5). Among these parameters, the change in the Page 5 of 9 A. Nehbandani et al.: OCL 2020, 27, 58 HI
t
d diff
i
di t d
i ld b
value) and 3.2 t.ha1 (decrease in the amount of KPAR). Also,
Increasing and decreasing the HImax parameter value by 30%
h
d th
i
i ld f
4 6 t h 1 t
6 0
d 3 2 t h 1
Fig. 1. Simulated versus measured dry soybean yield by SSM_iCrop2
model based on data used in model parameterization. The ±20%
discrepancy lines are indicated by dashed lines. Solid line is 1:1 line. Fig. 2. Simulated versus measured days to maturity by SSM_iCrop2
model based on data used in model parameterization. The ±20%
discrepancy lines are indicated by dashed lines. Solid line is 1:1 line. Fig. 3. Simulated versus measured dry soybean yield by SSM_iCrop2
model based on data used in model evaluation. The ±20% discrepancy
lines are indicated by dashed lines. Solid line is 1:1 line. Fig. 4. The ranking of influential cultivar-specific parameters for
WTOP (Accumulated above-ground dry matter, t ha1) under
different parameter variation ranges included þ30% (White), 30%
(Grey) and Standard (Dotted line). Fig. 3. Simulated versus measured dry soybean yield by SSM_iCrop2
model based on data used in model evaluation. The ±20% discrepancy
lines are indicated by dashed lines. Solid line is 1:1 line. Fig. 1. Simulated versus measured dry soybean yield by SSM_iCrop2
model based on data used in model parameterization. The ±20%
discrepancy lines are indicated by dashed lines. Solid line is 1:1 line. Fig. 3. Simulated versus measured dry soybean yield by SSM_iCrop2
model based on data used in model evaluation. 3.3 Sensitivity analysis of model parameters In terms of grain yield sensitivity
to the change of parameters, different results were achieved, HImax parameters caused differences in predicted yield by
model. Among these parameters, KPAR and HImax caused the
highest difference in the predicted total dry matter. Analysis of
the KPAR parameter showed that by increasing and decreasing
KPAR about 30%, the grain yield varied from 4.6 t.ha1
(constant and unchanged parameter) to 6.0 (increase in KPAR Page 6 of 9 Page 6 of 9 A. Nehbandani et al.: OCL 2020, 27, 58 Fig. 5. The ranking of influential cultivar-specific parameters for
grain yield (t.ha1) under different parameter variation ranges
included þ30% (White), 30% (Grey) and Standard (Dotted line). Soltani et al. (2020b) estimated total plant production at
province and country levels by SSM_iCrop2 and a bottom-up
scaling protocol (GYGA). They provided a framework within
which assessing the possibility of increasing national plant
production via intensification, optimizing water allocation
across plant species at province and country levels by changing
the cropping pattern, and assessing and prioritizing possible
ways of adapting a country’s agriculture to limited land and
water resources and climate change. Alizadeh Dehkordi et al. (2020) evaluated potential yield of wheat by using the
SSM_iCrop2 model in the Northwest of Iran. Nehbandani
et al. (2020a) estimated soybean potential yield, amount of net
irrigation water, evapotranspiration, vapour-pressure deficit,
and soybean water productivity using the SSM_iCrop2 model
in Iran. Also, they investigated relationships between potential
yield and environmental factors (accumulated solar radiation,
rainfall, maximum temperature, and minimum temperature
during the soybean growing season). 4 Conclusion Fig. 5. The ranking of influential cultivar-specific parameters for
grain yield (t.ha1) under different parameter variation ranges
included þ30% (White), 30% (Grey) and Standard (Dotted line). Crop models are essential in undertaking large scale
estimation of crop production of diverse crop species, especially
in assessingfood availability and climate changeimpacts.In this
research,cropsimulationmodel(SSM_iCrop2)parameterswere
estimated and evaluated. The model requires limited, readily
available input information. The simulations account for plant
phenology, leaf area development and senescence, dry matter
accumulation, yield formation, and soil water balance in a daily
time step. Parameterization of this model is easy and straight-
forward. The results of this study showed that the SSM_iCrop2
modelprovidesreasonablepredictionofdevelopmentstagesand
yield for the soybean in Iran. The sensitivity analysis of the
parametervaluesshowedthatthemosteffectiveparameteronthe
total dry matter and grain yield are KPAR and HImax,
respectively. This model can help find the best management
plans to achieve the potential yield for different regions of the
country. so that, in addition to the parameters affecting the dry matter,
increasing and decreasing the total amount of HImax by 30%
caused the most significant difference on grain yield simulated
by the model. An interesting point about the grain yield was the
effect of the 30% change in the tuHAR parameter, which
appears to be due to the increase and decrease in the growth
period of the soybean. By 30% increase in the amount of
tuHAR, the vegetative growth period was increased and
reproductive stage (seed filling) starts in the undesirable
temperatures so yield decreased. A 30% decrease in the tuHAR
amount resulted in yield reduction due to the decrease in the
vegetation and reproductive growth periods which causes less
dry matter transfer to the seed. Our results indicated the SSM_iCrop2 model simulates
growth and yield with reasonable accuracy across a wide range
of environments in the Iran. For example, in the validation data
set, observed grain yields ranged from 1.9 to 4.8 t.ha1, sowing
dates were as April 8 to July 31, and cultivars differed in
maturity from MG 3 and MG 5. SSM_iCrop2 achieves this
robust simulation capability with only 9 parameters variable. Acknowledgements. This research was supported by the
Agricultural Research, Training and Promotion Organization
of country, which we announce our gratitude of them. References p
y
y
p
SSM_iCrop2 model has considerable advantages com-
pared with other models. Although SSM_iCrop2 requires a
maximum of 37 parameters, the actual relevant number of
parameters is about half of the total number (i.e., between 15
and 20 depending on plant species) because many parameters
are interconnected and some parameters are not important for
some species (Soltani et al., 2020a). Whereas, the APSIM and
DSSAT models needed 292 and 211 parameters to estimate
the potential yield and phenological stages, respectively. (Noorhosseini et al., 2018). Due to the number of required
parameters,
the
SSM_iCrop2
model
can
simulate
the
phenological stages of growth in a large area such as a country. Akbari Nodehi D. 2011. Evaluating soybean response to water stress
at different growth stages in Mazandaran province. Water Soil Res
Cons J 1: 53–62. Akbari Nodehi D. 2012. Effect of drought stress in different growth
stages on soybean yield and water use efficiency in Mazandaran. Akbari Nodehi D. 2012. Effect of drought stress in different growth
stages on soybean yield and water use efficiency in Mazandaran. J Sustain Agric Prod Sci 22(1): 13–23. J Sustain Agric Prod Sci 22(1): 13–23. Alizadeh Dehkordi P, Nehbandani A, Hassanpour-Bourkheili S,
Kamkar B. 2020. Yield gap analysis using remote sensing and
modelling approaches: Wheat in the northwest of Iran. IJPP 1–10. Boote KJ, Jones JW, Hoogenboom G. 1998. Simulation of crop
growth: CROPGRO Model. It is worth noting that, in the model can easily use an Excel
spreadsheet to provide input and produce output, and also it is
open source. Bouman M, van Laar HH. 2006. Description and evaluation of the
rice
growth
model
ORYZA2000
under
nitrogen-limited
conditions. Agric Syst 87: 249–273. The studies that have been done using SSM_iCrop2 model
include the following: The studies that have been done using SSM_iCrop2 model
include the following: Dadrasi A, Torabi B. 2016. Predict the growth and yield of corn in
Hamedan. Iran J Field Crop Sci 47: 595–610. Page 7 of 9 Page 7 of 9 A. Nehbandani et al.: OCL 2020, 27, 58 FAOSTAT. 2013. Data. [WWW Document], n.d. URL http://www. fao.org/faostat/en/#data/TP (Accessed 6/29/20). Namdari M, Mahmoodi S. 2013. Evaluation of grain yield and yield
components in intercropping of dwarf and tall cultivars of
soybean (Glycine max L.). Iran J Crop Sci 15(1): 1–11. Faraji A. 2016. Evaluation of soybean new genotypes in terms of
drought tolerance indices. Reprod Crops 8(18): 30–36. References Nehbandani A. 2013. Parameterization of SSM model for soybean
growth and yield. Master’s Thesis group Agriculture, Faculty of
Plant Production, Gorgan University of Agricultural Sciences and
Natural Resources. Fazeli F, NajafiZarini H, Arefrard M, Mirabadi A. 2016. Assessment
of relation of morphological traits with seed yield and their
diversity in M4 generation of soybean mutant lines [Glycine max
(L.) Merrill] through factor analysis. J Crop Breed 7: 47–55. errill] through factor analysis. J Crop Breed 7: 47–55. Nehbandani A, Soltani A, Rahemi Karizaki A, Dadrasid A,
Nourbakhsh F. 2020a. Determination of soybean yield gap and
potential production in Iran using modelling approach and GIS. JIA 19: 2–14. Ghana P. 2016. Evaluation the effect of phosphorus solubilizing
bacteria, inoculation of rhizobium and zinc solubility on soybean
yield and yield components of Katoul cultivar in Golestan. J Oil
Plants Prod 3(1): 25–35. Nehbandani A, Soltani A, Taghdisi Naghab R, et al. 2020b. Assessing
HC27 Soil Database for Modeling Plant Production. Int J Plant
Prod accepted. Ghanbari Malayder A, Janbaz Ghobadi Gh, Dastan S, Shahidifar A. 2015. Evaluation of changes in the length of day and growth
phenology of soybean cultivars in different planting dates in Sari. Agric Res 7(1): 41–53. Noorhosseini SA, Soltani A, Ajamnoroozi H. 2018. Simulating
peanut (Arachis hypogaea L.) growth and yield with the use of the
Simple Simulation Model (SSM). Comput Electron Agric 145:
63–75. Geerts S, Raes D. 2009. Deficit irrigation as an on-farm strategy to
maximize crop water productivity in dry areas. Agric Water
Manag 96: 1275–1284. Qajar Sepanlou M, Behminar MA. 2004. Irrigation effect in different
stages of growth on yield, water use efficiency and harvest index
of soybean cultivars in Mazandaran. Agric Nat Res Caspian Sea 2
(2): 79–88. Gorzin M, GhaderiFar F, Zainali E, Razavi SE. 2015. Effect of the
length of various developmental periods on soybean yield and
yield components. Elec J Crop Prod 8(1): 21–41. Habibzadeh F, Amini A, Miernia Kh. 2003. Evaluation the effect of
different potassium value and zinc on the yield and yield
components of soybean in Mazandaran province. J Pazhu Saz 60:
18–24. Raeisi S, Hezarjeribi A. 2013. Evaluation of soybean cultivars in
different planting arrangements and densities in Golestan
Province of Iran. Seed Plant Prod J 29(2): 215–233. Rameeh V, Aghabozorgi M. 2016. Effect of different seed rates and
sowing dates on yield components, yield and qualitative traits of
soybean (Glycine max (L.) Merr.) variety Sari. References J Oil Plants Prod 3
(1): 15–24. Hoogenboom G, Jones JW, Wilkens PW, et al. 2004. Decision support
system for agrotechnology transfer version 4.0. Honolulu, HI:
Univ. Hawaii. Hosaini T, Khoshravesh M, Ziatabar Ahmadi M, Ghadami Firozabadi
A. 2016. Evaluation soybean yield using AquaCrop model under
salinity and water difinency management. J Water Res Agric 30
(3): 361–372. Razmi N. 2010. Effect of sowing date and plant density on some
agronomic characteristics, grain yield and its components in
soybean genotypes in Moghan region. Seed Plant Prod J 26(4):
403–418. Hsiao TC, Heng L, Steduto P, Rojas-Lara B, Raes D, Fereres E. 2009. AquaCrop – The FAO crop model to simulate yield response to
water: III. Parameterization and testing for maize. Agron J 101:
448–459. Setiyono TD, Cassman KG, Specht JE, et al. 2010. Simulation of
soybean growth and yield in near-optimal growth conditions. Field Crop Res 119: 161–174. Sexton J, Everingham L, Inman-Bambera G. 2017. A global
sensitivity analysis of cultivar trait parameters in a sugarcane
growth model for contrasting production environments in
Queensland, Australia. Eur J Agron 88: 96–105. Keating BA, Carberry PS, Hammer GL, et al. 2003. An overview of
APSIM, a model designed for farming systems simulation. Eur J
Agron 18: 267–288. Koo J, Dimes J. 2013. HC27 Generic Soil Profile Database. Harvard
Dataverse Ver. 4. Seyed SharifiR. 2015. The Effect of Zinc Application and Biological
Fertilizers on Seed, Function and Some Soybean Growth
Characteristics. Agric Prod 17(1): 109–130. Mahmodi M, Zakipour V. 2016. The effect of nitrogen on yield and
yield components of soybean in the city of Mahmudabad. Agron
Res Edge Desert 13(1): 73–81. Sidi SharifiA, Khoramdel S. 2013. Effects of nano-zinc oxide and
seed inoculation by plant growth promoting rhizobacteria (pgpr)
on yield, yield components and grain filling period of soybean
(Glycine max l.). IRN J Field Crop Res 13: 738–753. g
( )
Ministry of Agriculture Jihad. 2016. Agricultural Statistics. [WWW
Document], n.d. URL https://www.maj.ir/ (Accessed 6/29/20). Sinclair TR, Muchow RC. 1999. Radiation use efficiency. In:
Advances in Agronomy. Elsevier, pp. 215–265. Monod H, Naud C, Makowski D. Uncertainty and sensitivity analysis
for crop models. In: Wallach D, Makowski D, Jones JW, eds. Working with dynamic crop models: Evaluation, analysis,
parameterization, and applications. Amsterdam (The Netherlands):
Elsevier, 2006. Soltani A. 2009. Mathematical modeling of the crop. Publication
Mashhad University Jihad, 175 p. Soltani A. 2012. Modeling physiology of crop development, growth
and yield. Cite this article as: Nehbandani A, Soltani A, Nourbakhsh F, Dadrasi A. 2020. Estimating crop model parameters for simulating soybean
production in Iran conditions. OCL 27: 58. References CABi. Mosnei H, Ajmnorouzi H, Rangriz Z, Zarez Zadeh Z. 2015. The effect
of planting pattern on phenological stages, yield and yield
components of soybeans promising in Gorgan. Plant Physio 7
(20): 43–58. Soltani A, Sinclair TR. 2012. Identifying plant traits to increase
chickpea yield in water-limited environments. Field Crop Res
133:186–196. Mousavi S, Chavoshi S. 2013. Evaluation of phenological traits and
some important cultivars of soybean cultivars under different
planting dates in Ardabil. Iran J Agric Sci 7(4): 371–382. Soltani A, Alimagham SM, Nehbandani A, et al. 2020a. SSM_i-
Crop2: A simple model for diverse crop species over large areas. Agric Syst 182: 102855. NajafiR. 2006. Effect of seed power on soybean growth and yield. Master’s Thesis group Agriculture, Faculty of Plant Production,
Gorgan University of Agricultural Sciences and NaturalResources. Soltani A, Alimagham M, Nehbandani A, et al. 2020b. Modeling
plant production at country level as affected by availability and
productivity of land and water. Agric Syst 183: 102859. Page 8 of 9 Page 8 of 9 A. Nehbandani et al.: OCL 2020, 27, 58 filling of three soybean varieties in different densities. Agric Sci 1
(4): 141–152. Tan J, Cui Y, Luo Y. 2017. Assessment of uncertainty and sensitivity
analyses for ORYZA model under different ranges of parameter
variation. Eur J Agron 91: 54–62. Zeinali E, Ghadirifar F, Soltani A, Kashiri H. 2003. Effect of planting
date on yield and yield components of three soybean cultivars in
Gorgan. Iran Crop Res 1: 81–92. g
Torabi B, Soltani A, Galeshi S, Zeinali E. 2011. Assessment of yield
gap due to nitrogen management in wheat. Aust J Crop Sci 5: 879. g p
g
g
p
Williams SL, Watson N. 1985. Perceived danger and perceived self-
efficacy as cognitive determinants of acrophobic behavior. Behav
Ther 16: 136–146. Zendeh A, Ansari MH, Khorshidi MB. 2016. Effect of nitrogen
fertilizer and Pseudomonas putida inoculation on yield and yield
component of soybean as second crop in Moghan region. Agron J
12(2): 35–42. Zahleht Salmasi S, Mehaghani R, Ghasemi Golazani K, Alari H,
Raeisi S. 2004 Yield evaluation, speed and durability of seed Page 9 of 9
|
https://openalex.org/W2576276446
|
https://www.duo.uio.no/bitstream/10852/64630/2/econometrics-05-00006.pdf
|
English
| null |
Between Institutions and Global Forces: Norwegian Wage Formation Since Industrialisation
|
Econometrics
| 2,017
|
cc-by
| 33,469
|
Ragnar Nymoen 1,2 Department of Economics, University of Oslo, POB 1095 0317 Oslo, Norway; ragnar.nymoen@econ.uio.no
Tel.: +47-22-855-148 2
Centre for Wage Formation at Economic Analysis, POB 0650, Oslo, Norway †
Paper presented at the workshop Macroeconomics and Policy Making, arranged in honour of Asbjørn Rødseth,
18 May, 2016, by the Department of Economics, University of Oslo. Thanks to Olav Bjerkholt for comments,
and for showing me the article written by Frisch about “rational wage policies”, and the correspondence
with Haavelmo that it led to. Discussions at the Workshop in econometrics at Statistics Norway, 21 October
2016, were also very useful, thanks to the participants. Thanks also to Jan Morten Dyrstad, David F. Hendry ,
Steinar Holden, Tord S. Krogh and Mikkel Myhre Walbækken for important comments and suggestions. Finally, thanks to the editors and to two anonymous referees for their comments, both critical and
constructive. The numerical results in this paper were obtained by the use of OxMetrics 7/PcGive 14 and
Eviews 9.5. †
Paper presented at the workshop Macroeconomics and Policy Making, arranged in honour of Asbjørn Rødseth,
18 May, 2016, by the Department of Economics, University of Oslo. Thanks to Olav Bjerkholt for comments,
and for showing me the article written by Frisch about “rational wage policies”, and the correspondence
with Haavelmo that it led to. Discussions at the Workshop in econometrics at Statistics Norway, 21 October
2016, were also very useful, thanks to the participants. Thanks also to Jan Morten Dyrstad, David F. Hendry ,
Steinar Holden, Tord S. Krogh and Mikkel Myhre Walbækken for important comments and suggestions. Finally, thanks to the editors and to two anonymous referees for their comments, both critical and
constructive. The numerical results in this paper were obtained by the use of OxMetrics 7/PcGive 14 and
Eviews 9.5. Academic Editors: Gilles Dufrénot, Fredj Jawadi and Alexander Mihailov Received: 31 August 2016; Accepted: 13 December 2016; Published: 12 January 2017 Abstract: This paper reviews the development of labour market institutions in Norway, shows
how labour market regulation has been related to the macroeconomic development, and presents
dynamic econometric models of nominal and real wages. Single equation and multi-equation
models are reported. The econometric modelling uses a new data set with historical time series of
wages and prices, unemployment and labour productivity. Impulse indicator saturation is used
to achieve robust estimation of focus parameters, and the breaks are interpreted in the light of the
historical overview. Ragnar Nymoen 1,2 A relatively high degree of constancy of the key parameters of the wage setting
equation is documented, over a considerably longer historical time period than earlier studies have
done. The evidence is consistent with the view that the evolving system of collective labour market
regulation over long periods has delivered a certain necessary level of coordination of wage and price
setting. Nevertheless, there is also evidence that global forces have been at work for a long time, in a
way that links real wages to productivity trends in the same way as in countries with very different
institutions and macroeconomic development. Keywords: wage formation; economic history of Norway; structural breaks; labour market regulation;
econometric models of inflation JEL Classification: C22; C31; E23; E24; E31; J38; J50; J51; N14; N34; O52 In the days of Manchester liberalism, the wage contract was a matter between the individual
worker and employer. Any “wage policies”, either by the society or by the organizations,
were non-existent. Luckily, this has been changed. Ragnar Frisch (Arbeiderbladet 30 August 1945) [1]. econometrics econometrics econometrics 1
Ragnar Frisch was professor at the University of Oslo from 1931 to 1965, a founder of The Econometric Society and was
awarded the first Nobel Prize in economics in 1969. Frisch seems to have been deeply influenced by observing at close range
the impact of deflationary policies in Norway in the 1920s and by the Great Depression in the 1930s. His scientific work was
motivated by the need for social improvements as much as an intellectual interest, Bjerkholt (2014) [2] (p. 299) and Bjerkholt
and Qin (2011) [3] (p. 11). All quotes from the newspaper article (and the correspondence that it led to) have been translated
by the author. 1. Introduction The newspaper article by professor Ragnar Frisch, where the quotation is taken from, continued
with the observation that one of the (“lucky”) things that had happened was that “wage policies” The newspaper article by professor Ragnar Frisch, where the quotation is taken from, continued
with the observation that one of the (“lucky”) things that had happened was that “wage policies” www.mdpi.com/journal/econometrics www.mdpi.com/journal/econometrics Econometrics 2017, 5, 6; doi:10.3390/econometrics5010006 2 of 54 Econometrics 2017, 5, 6 had come to take a central place in economic policy thinking and practice, alongside monetary and
fiscal policy 1. Frisch put this down to the increased political importance of redistribution and “social
justice”. However, even more importantly, he cited the fact that the general wage level is one of the
main factors that determine the activity level in a capitalist industrialised economy, both as a cost factor
for producers, and as a main determinant of aggregate demand in the economy. In that way, Frisch
wrote, the general wage level had become a central variable in the “most important of all economic
processes” 2. p
Frisch’s main motive for writing the newspaper article may have been to present some ideas about
what he called a “rational wage policy” that would make it possible to reconcile full employment and
a certain stability of the price level. In this Frisch was not alone. For example, American and British
economists commented on the challenges and dilemmas that the western economies would face during
the post war period, as inflation and international competitiveness replaced mass unemployment as
the main problems for macroeconomic policy makers. Frisch was clearly looking for a conceptualization, and an operationalization, of a wage norm
for the Norwegian economy as a whole. Yet he did not give attention to the important developments
towards practical collective labour market regulation that had taken place in the early decades of
the 20th century. It was those developments, which we review in Section 2.5, not Frisch’s theoretical
formulations, that came to provide the operational definition of the wage norm, which became a
mainstay of the system of wage formation during the whole post-war period. Nevertheless, there is nothing in this that would have reduced the relevance of Frisch’s timely
identification of wage setting as one of the most important economic processes of the modern market
based economy. y
2
Frisch’s newspaper article is interesting also because it led to a correspondence with his pre-war assistant and colleague
Trygve Haavelmo, who was still in USA, where he had been exiled during the war. Haavelmo opened by saying that he was
“very interested in the problems” that Frisch had analysed in the article, and then went on to present a detailed note with
comments (and improvements). Frisch, who wanted Haavelmo to come back to the University of Oslo, wrote back with
enthusiasm and said that it was of the “greatest importance” that Haavelmo committed himself intellectually to this “all
important field”, the theory of rational wage setting.
3
See Olstad (2009) [4], in particular chapter 5, and the concluding chapter. p
y
g
g
3
See Olstad (2009) [4], in particular chapter 5, and the concluding chapter. 1. Introduction This has been confirmed time and time again, not only by the (ebbing and flowing)
stream of academic offerings in the field, but even more by the many political involvements and
initiatives that have been launched over the years, some of them ill-fated, others more successful. The continued relevance of wage setting is also the motivation of this paper, where I attempt to
give econometric treatment to the formation of the general wage level in Norway over a period of
115 years. Since industrialisation started very late in the 1800s in Norway, the sample period therefore
covers the epoch with an industrialised economy. In turn, because the organization among workers
and firms happened in tandem with the growth of modern electricity based heavy industry, it meant
that one premiss for collective labour market regulation was a reality already early in our sample
period. Inevitably, another force established itself at it same time, and that was the economic laws of
international product markets. Early in the century, nationwide collective agreements were struck in important manufacturing
sectors for the first time 3. The idea that industrial peace, not strife, was possible as a sort of normal
situation in the labour market, seems to have motivated both union leaders and industrialists quite
early in our data period. That did not mean that the days of industrial unrest were over though. The 1920s, and part of the 1930s, saw years when the number of working days lost in strikes and 3 of 54 Econometrics 2017, 5, 6 lockouts were extremely high. However, these years can hardly be counted as normal in any sense of
the word. The difficulties for the economy, and the strains on the political system, were enormous. Underlying the early development was a recognition among union leaders that international
product markets and capitalist principles had certain consequences. For example: private owners’
right to organize and lead work, and that the firm’s profitability represented the basis for wage claims. As long as there was enough protection from unwanted competition in the labour market (which could
undermine the ability to organise), competition and international trade in the product markets brought
many benefits for trade union members. In this perspective, the system of wage setting was formed, and has been adapted under
the influence of two strong forces: Domestic labour market institutions and global market forces. 1. Introduction In order to avoid Frisch’s Manchester liberalism, collective bargaining had to become institutionalised. However, in order to become stable, labour market regulation in turn needed to be compatible with
private ownership and with competitive product markets. This balancing act, between liberalised
product markets and labour markets with enough protection from unwanted competition to sustain
a system of collective bargaining, not only defined the development of wage setting in Norway. It was unavoidable and ubiquitous in western economies for most of the 20th century 4. In Norway,
the balancing act is still going on, centered around the consequences of liberalised labour market
immigration from Europe during and after the financial crisis. From the perspective of the modeller of wage setting, there is an interesting side effect of this
duality, namely that the forces of institutions and of markets have been present in the data over a
long historical period. Hence, there may be an element of continuity in the data generation process,
despite the variability and huge changes in individual years, and this motivates empirical modelling
over the long sample. This study therefore starts with a review, in Section 2, of a century (plus) of
economic history, focusing on labour market regulation and institutions. Section 3 recounts briefly the economic theory of wages, and I then specify a theoretical macro
economic model that can be a relevant framework for empirical modelling. It is shown that the
specified Incomplete Competition Model (ICM) has dynamic solutions that are qualitatively similar
to the historical development of, for example, nominal and real wages, but also unemployment and
the real exchange rate. The theoretical framework also contains the wage and price Phillips Curve
Model (PCM) as a special case. These models represent different theoretical possibilities of wage-price
coordination, especially when there is a target of near full unemployment, see Kolsrud and Nymoen
(2014, 2015) [6,7]. The framework is therefore relevant for our empirical modelling project, which
however requires identification and representation of structural breaks to achieve robust estimation of
focus parameters, parameter constancy and identification, see Hendry (2017) [8]. Section 4 documents empirical coefficient constancy, and invariance, of the nominal wage
setting equation over a considerably longer historical time period than earlier studies have done. The relationship is not (merely) a statistical relationship, it has a clear interpretation as an economic
relationship showing that collective bargaining has been the main factor in wage formation. 4
Britain is an interesting case for historical comparison. At the time when the industrial unrest of the late 1960s was more of a
nuisance than an unmanageable problem, Labour party minister Barbara Castle became frustrated by the unions’ lack of
understanding that living in a liberalised market economy had consequences for how the unions could act. She mothered
the ill fated White Paper In place of Strife in 1969, which in retrospect defined a turning point in the history of labour market
regulation in Britain, see Sandbrook (2006) [5] (pp. 709–710). 5
The number is for 1903 which is the earliest year with data, cf. Appendix A. 2. A Century Plus of Labour Market Change The sample period for our empirical econometric analysis begins at the start of the 20th century
and ends in 2015. However, the Norwegian economy of 1900 had a pre-history in the 1800s, which is
relevant for understanding the development that took place in the first decade of our main sample
period. I therefore first comment on some important trends of the last part of the 19th century,
see Section 2.1. I then turn to the development after 1900, which was marked by the end of mass
emigration and of the underemployment that it reflected, see Section 2.2. The other parts of the
chapter, Sections 2.3–2.6, contain a presentation of the time series for unemployment and productivity,
and discuss the development of labour market regulation and institutions. The dual, sometimes
conflicting, developments towards product market deregulation and globalization, and a strong trait
of collective regulation of the labour market, is a main theme. That section also gives the backdrop for
the empirical modelling, including the assessment of the relevance of the economic theory of wages,
see Section 3, which of course is central in the econometric models in Section 4. 1. Introduction The econometric modelling results for wage formation, both single- and multiple-equation,
also clarify methodological issues, and they have policy implications. They invalidate the still common
way of obtaining estimates of a natural rate of unemployment by inverting the wage Phillips curve. Instead, equilibrium unemployment is defined by a multi-equation model of wages, domestic and
foreign prices, productivity and the unemployment rate. The results therefore entail that empirical 4 of 54 Econometrics 2017, 5, 6 relevance suffers if the steady state of a medium term macro model is taken as exogenous, as the
dynamic stochastic general equilibrium models (DSGEs) do. Another implication is that unions and firm owners’ organizations, through, for example,
coordinated wage formation, can aid economic policy by making sure that inflation, labour market
disputes, low mobility and low productivity growth do not become obstacles for the attainment of
other important policy targets like full employment. However, the social partners cannot determine
unemployment, or secure full employment, as some economists suggested would be their right role
in a regime where the central bank supposedly takes care of the nominal path (wages and prices) of
the economy, see, e.g., Norges Bank (2002) [9] (pp. 28,29), Isachsen (2008) [10] (p. 8). To this they do
not have the instruments. The results of this investigation support that a much more concerted policy
adjustment is required, and that this was well understood during the post-war period, but also that
even this may not be enough, if the economy is hit by external shocks, or has to correct imbalances that
have been allowed to build up over time. 2.1. The Norwegian Economy at the Start of the 20th Century The last 25–30 years of the 19th century nevertheless marked a time-shift. In 1900 the real wage
was 67 percent higher than in 1870, which was a much better performance than earlier in the century. GDP per capita had grown by 38 percent over the same period 6. 6
Fixed 1990 International Geary-Khamis dollars, Maddison project database:
http://www.ggdc.net/maddison/
maddison-project/data.htm. 7
Norway left the Gold standard in 1914, even though it was a neutral country with no war efforts to finance,
cf. Lie (2012) [15] (pp. 32,33). Norges Bank’s monetary expansion during the first two years of the war was related to
international trade flows. However, during the two last war years another part of the monetary expansion was result of
credit to domestic borowers, Værholm and Øksendal (2010) [16].
8
There was more than 25 000 immigrants in total in 1917 while only 2500 left Norway Søbye (2014) [14] (Tables 4 and 15) 2.1. The Norwegian Economy at the Start of the 20th Century At the start of the 20th century, Norway was barely an industrialised country. In 1900, almost
half of the employment was in agriculture, forestry and fisheries. In comparison, only 11% worked in
the primary sectors of the earlier industrialised UK economy, see Skoglund (2013) [11] and Lindsay
(2000) [12]. Twenty-four percent of the employment was in manufacturing and other secondary
industries, while it was 54% in the UK. Long periods of the previous century had been marked by
low economic growth, stagnation in real wage growth, and by mass emigration, to North America
in particular. As the graph in Figure 1 shows, emigration included three big waves with peaks in emigration
rates above 1 percent of the population size. In Europe, only Ireland had higher emigration rates. Emigration was selective within cohorts, with many leaving who would have contributed to the
Norwegian economy if given a chance, Bævre et al. (2001) [13]. There can be little doubt that
the emigration to North America during the second half of the 1800s reflected a high degree of
underemployment in Norway. It is also easy to imagine that the Norwegian unemployment rate early
in the 20th century would have been considerably larger than 4.5 percent, if emigration had not been
possible 5. If we use ratios for the period 1903–1910, when we have data of emigration, employment
and unemployment, the implied unemployment “without emigration” in 1900 becomes between 7%
and 9% . Of course, this is only a crude interval, but it is nevertheless a reminder that the scale of 5 of 54 Econometrics 2017, 5, 6 emigration was large enough to affect the balance between supply and demand in the domestic labour
market, see Søbye (2014) [14] (p. 107). Emigration in percent of population
1850
1860
1870
1880
1890
1900
1910
1920
1930
1940
0.25
0.50
0.75
1.00
1.25
1.50
Years with net immigration
Emigration in percent of population
Figure 1. Emigration in percent of the Norwegian population. Source: Søbye (2014) [14] (Table 15). Figure 1. Emigration in percent of the Norwegian population. Source: Søbye (2014) [14] (Table 15). The last 25–30 years of the 19th century nevertheless marked a time-shift. In 1900 the real wage
was 67 percent higher than in 1870, which was a much better performance than earlier in the century. GDP per capita had grown by 38 percent over the same period 6. (
) [
]
8
There was more than 25,000 immigrants in total in 1917, while only 2500 left Norway, Søbye (2014) [14] (Tables 4 and 15). 6
Fixed 1990 International Geary-Khamis dollars, Maddison project database:
http://www.ggdc.net/maddison/
maddison-project/data.htm.
7
Norway left the Gold standard in 1914, even though it was a neutral country with no war efforts to finance,
cf. Lie (2012) [15] (pp. 32,33). Norges Bank’s monetary expansion during the first two years of the war was related to
international trade flows. However, during the two last war years another part of the monetary expansion was result of
credit to domestic borowers, Værholm and Øksendal (2010) [16].
8
There was more than 25,000 immigrants in total in 1917, while only 2500 left Norway, Søbye (2014) [14] (Tables 4 and 15). 2.3. Unemployment and Productivity The boom during WW-I continued until 1919, but was stopped short in 1920 by a deflationary
economic policy in the western European economies, notably the UK and Sweden, Norway’s most
important trading partners. This policy was monitored by the central banks in order to decrease
prices, and thereby increase the value of their currencies back to par gold values. Norway followed
suit, and one of the deflationary consequences was the increase in the rate of unemployment seen in
Figure 2. The gold-parity target was reached in 1928, but Norway, again following Bank of England’s
example, left the gold standard in September 1931. Monetary policy came off the deflationary track
that had been followed (with only a few stops) since 1920. Lowered central bank interest rates may
have contributed to higher economic activity. Other, more indirect effects of the policy change may
also have been important. The international value of the krone was lowered, which made it easier
to successfully compete for market shares in the export market, and in the domestic market against
imported goods. Fiscal policy, after a while along “Keynesian lines”, may have contributed to the fall
in unemployment. However, the main impression is that budget discipline was given priority, also
during the years with Labour party rule in the 1930s, see Grytten (2008) [17]. Although the historians still debate the causes, the depression in the 1930s was less severe in
Norway (and Sweden), than in many other western european countries, and the USA. In Norway, the
1920s was a tougher decade than the late 1930s. As noted, Grytten (2008) [17] finds it noteworthy that
unemployment did not fall more during the 1930s, but pointed to the increased labour supply noted
due to (re-)immigration as an important explanatory factor of relatively high unemployment rates,
see Figure 2. 1910
1920
1930
1940
1950
1960
1970
1980
1990
2000
2010
1
2
3
4
5
6
7
8
9
10
11
Par gold value policy
1920
Gold standard
1928
Percent
Leave gold standard
1931
Banking crisis
Financial
consolidation
WW-I
WW-II
Figure 2. The rate of unemployment cf. Appendix A, together with empirical breaks in mean and
labels for major events. Percent Figure 2. The rate of unemployment cf. Appendix A, together with empirical breaks in mean and
labels for major events. 2.2. The End of Mass Emigration It is well known that emigrants often returned to Norway, and it is realistic to think that the
propensity to re-immigrate increased during the 1920s, not to speak of in the 1930s, during the Great
Depression in USA. In addition, many workers from Sweden came to several large construction and
building works in Norway early in the 1900s. A boom in building and construction was a consequence
of the first world war, and of a very accommodating monetary policy 7. Also the government sector
contributed to the demand for construction workers in these years, with 1919 and 1920 as particularly
buoyant years, see Søbye (2014) [14] (p. 94). According to Søbye (2014) [14] (pp. 92–97), 1915 and 1916 (when gross emigration was in any case
very low) may have been the first years with net immigration, and definitively 1917, when more than
thirteen thousand immigrants were added to the Norwegian labour force 8. High labour immigration
continued until 1920. In 1921 labour immigration dropped sharply, and in 1922 or 1923 the situation
had no doubt changed back to positive net emigration from Norway. However, in the period 1931
to 1940, net immigration again contributed to population growth in Norway. For example, Grytten
(2008) [17] argues that the tens of thousands who returned to Scandinavia after the start of the Great
Depression in USA, meant a significant increase in labour supply. 6 of 54 Econometrics 2017, 5, 6 2.3. Unemployment and Productivity The unemployment rates for the Nazi occupation years have been constructed by utilizing an
empirical post-war relationship between employment growth and unemployment, as explained in
the data appendix. The result is a series which shows an unemployment rate during occupation that
was lower than in any year between 1921 and 1939. Unemployment may have been even lower, as
the historians argue, see Hodne and Grytten (2002) [18] and Bjørnhaug and Halvorsen (2009) [19]
(p. 124). In any case, mass unemployment was a thing of the past already in the first war years. With 7 of 54 Econometrics 2017, 5, 6 the exception of 1983–1984, unemployment stayed below 3 percent right until the housing price crash
and the banking crisis in 1990–1991 9. Productivity is one of the main determinants of the trend growth in real wages, and ultimately
also of living standards. Conversely, the trend in productivity can be conditioned by the system of
wage formation. In particular during epochs of full employment, collective wage setting may “free”
more labour to move to the more efficient production units, than a local and individual wage setting
will do, see Barth et al. (2014) [22]. Hence, labour market regulation with collective agreements needs
not be an impediment to productivity growth. On the contrary, it can be a productivity increasing
factor, since it makes a larger share of the employment work with the latest and best technology, Barth
and Moene (2015) [23]. Labour productivity is also shaped by many other factors than organization of the labour
market. For a country at the technological frontier, productivity improvement depends on innovations,
education and institutions, and these dimensions are interdependent as well. Over long historical
periods, any one country is however likely to find itself lagging in development and adaptation of
new technologies. Although, at first thought such countries must surely catch-up relatively easily, the
evidence shows that this does not always happen. One explanation may again be that institutions are
also important for technology adaptation and copying, see Bergeaud et al. (2015 ) [24]. Figure 3 plots LP-growth together with the rate of unemployment for comparison. Three epochs
are evident. First, productivity growth was very volatile until WW-II. Second, productivity growth
was both high and relatively stable, from the beginning of the re-building period until the second
half of the 1970s. The 1980s and 1990s were characterised by lower productivity growth. 9
The Norwegian economy was affected by the Korean war of 1951–1952, but mainly in the form of a spurt of imported
inflation, SSB (1965) [20] (pp. 385–392). The increase in unemployment in 1957 and 1958 may have been jointly caused by
weak development in export markets, and unintended deflationary effects of fiscal policy, SSB (1965) [20] (pp. 406–408).
The increase in unemployment in 1983 and 1984 had background in the weak development of the international economy and
structural problems in Norway. It was first tackled by expansionary fiscal policy, which however was switched-off, primarily
because the projected international recovery did not materialize. The large balance of payment deficits in the period with
expansionary policy gave rise to concern about a possible loss of “economic scope for maneuvre”, SSB (1985) [21] (p. 97). 2.4. Labour Market Regulation Hydroelectric power,
and new electrotechnical and electrochemical industries led to
industrialisation of Norway at the start of the 20th century. These and other new large scale industries
that had developed during the 1880s, were organised in ways that regulated competition. As a
result, the 1900s started with a movement away from free trade and market liberalism in some
important product markets 10. Hence a wider acceptance of the legitimacy of protection against
unwanted competition was “in the air”, and this may have favoured changes in the regulation of
the labour market, where collective agreements took over from individualised work contracts as the
main principle. The late industrialisation of Norway may have been a blessing, since society escaped
the fractures that decades of socially harsh “Victorian” liberalism would have created. Not that the
conservative paternalism of 19th century Norwegian capitalism represented any less of an impediment
for the individual worker and his family, as the very high emigration rates also were evidence of. And of course, the growth of trade unions and the acceptance of collective bargaining did not
happen without conflict. The Norwegian trade union confederation (LO) was formed in 1899, and the
first decades was marked by struggles to limit competition for jobs and to push for higher wages,
Olstad (2009) [4] (p. 89). As Figure 4 shows, years when working hours lost in strikes and lockouts took
a substantial share of total hours were much more common before WW-II than after. In particular 1921,
1924 and 1931 were years with serious conflicts. Still, the 2.3 percent lost in the worst year, which was
in 1931, may have been less than the percentage lost due to sick absence from work, as indicated by the
graph showing work absenteeism in the 1970s. Apart from the strike-free West-Germany and Japan,
industrial unrest returned in some scale to western economies in the 1960s and 1970s 11. To some
extent the Scandinavian countries were also affected. However, strikes tend perhaps to loom higher in
the public consciousness than in the actual figures, as Figure 4 also indicates. Labour market reforms have typically started from below, and have later been supported
(or extended) by law. One reason why this has been a regular pattern is the limited reach of a
collective agreement, Evju (2014) [27]. See Lie (2012) [15] (pp. 15–23).
11
The increased number of labour disputes, also illegal (“wild cat”) strikes, in the 1960s and 1970s is often seen as a “British
disease”. However, in the 1960s and 1970s United Kingdom finished a mere seventh and sixth in a league table of working
days lost per 1000 workers, with Canada, Italy, Australia, United States and Ireland all recording more strikes in those two
decades. Even during the 1980s, United Kingdom finished third, behind Canada and Australia, cf. Wrigley (2002) [25]
(Table 4.4), Sandbrook (2011) [26] (p. 98). 2.3. Unemployment and Productivity By and
large, this performance after the WW-II is not very different from other western European countries,
cf. Bergeaud et al. (2015) [24]. Unemployment rate
LP growth
1900
1910
1920
1930
1940
1950
1960
1970
1980
1990
2000
2010
-2.5
0.0
2.5
5.0
7.5
10.0
Percent
Unemployment rate
LP growth
Figure 3. Labour productivity (LP), measured as GDP in fixed prices in Mainland-Norway per hour
worked in Mainland Norway, and the rate of unemployment. Source: Appendix A. LP is a centered
moving average using one period lead and one period lag, the raw series is shown in the data appendix. Figure 3. Labour productivity (LP), measured as GDP in fixed prices in Mainland-Norway per hour
worked in Mainland Norway, and the rate of unemployment. Source: Appendix A. LP is a centered
moving average using one period lead and one period lag, the raw series is shown in the data appendix. 8 of 54 Econometrics 2017, 5, 6 As we will discuss below, collective bargaining became the main principle already before the
occupation. Although the system may have been at its strongest in the 1950s and 1960s, it is still in
place today. Hence, it does not seem that a break in labour market regulation can explain the secular
drop in LP growth towards the end of the sample. Neither is there a simple and stable correlation
between LP growth and unemployment. The graphs show examples of positive correlation (1920s and
1930s), high LP growth together with constant and full employment (1946–1975), and a few examples
of negative correlation as well. 10
See Lie (2012) [15] (pp. 15–23). 2.4. Labour Market Regulation It it only binding for the parties that have signed the agreement:
The union’s members and firm(s) that have negotiated with the union. An important early collective agreement was the iron worker settlement of 1907. In addition to
settling important issues about economics and principles between two strong parties, that agreement
showed, by example, that much could be achieved by trade unions that respected firms’ right to
manage, was positive to technological progress and which allowed for wage differentiation according
to individual qualifications and working hours, Olstad (2009) [4] (p. 89). Many of these principles later
became associated with the so called Norwegian model of labour market organization. 9 of 54 9 of 54 Econometrics 2017, 5, 6 Work absenteeism
Industrial unrest
1900
1910
1920
1930
1940
1950
1960
1970
1980
1990
2000
2010
0.5
1.0
1.5
2.0
2.5
3.0
3.5
4.0
Percent
Work absenteeism
Industrial unrest
Figure 4. Hours lost in industrial unrest in percent of hours worked and the absenteeism percent. Source: Appendix A. Figure 4. Hours lost in industrial unrest in percent of hours worked and the absenteeism percent. Source: Appendix A. There can be little doubt that the bargaining position of Norwegian manufacturing workers
was weaker at the start of the 20th century than later in our period. The organization percentage
for workers (union density) may have been below 10 percent in 1900, Olstad (2009) [4]. However, it
increased year by year, and reached 50 percent at the end of the 1930s. The number and coverage
of collective agreements also increased both before WW-I, and in the interwar years, in spite of the
difficult economic situation in that period, Olstad (2009) [4] (pp. 436,437). In an econometric paper,
Bårdsen and Klovland (2010) [28] presented evidence showing that wages responded to changes in
firms’ profitability during the Great Depression, which is a typical characteristic of wage formation
with mutual bargaining power 12. Leiserson (1959) [29] is an example of an early “onlooker’s” impression of Norwegian labour
market regulation. Leiserson emphasised the importance of the Master agreement between the two
confederations LO and NAF in 1935 as a turning point: away from strife and towards a capacity for
coordinated, concerted adjustments in several key areas. Olstad (2009) [4] (p. 419).
14
Reiersen (2015) [30] is an interesting analysis of the mental re-orientation during the 1920s and 1930, which probably was
needed on both sides in order to break the deadlock marked by strife and industrial unrest. Reiersen’s view is that the main
step was to move from a situation of distrust, and hence conflict as the main strategy on both sides, to a situation with
sufficient trust so that the mutual strategy became one of compromise and cooperation. 12
The data used by Bårdsen and Klovland is a panel data set of individual firms.
13
Ol t d (2009) [4] (
419) 12
The data used by Bårdsen and Klovland is a panel data set of individual firm
13 y
13
Olstad (2009) [4] (p. 419). 15
As the interesting spat between Sæther and Eriksen (2014) [32] and Bjerkholt (2014) [2] shows, the “economic planning” of
Norwegian post-war economy may have been misunderstood by some commentaries, or wrongly presented, as directives.
The plans set out in the annual National Budget were expectations and intentions, not directives, Bjerkholt (2014) [2] (p. 301).
The outcome depended on the international development in particular, as well as on economic control measures. In the
early reconstruction years, the monitoring of the economy took place at a detailed level, and so did the use of control
measures. However, the approach was more a reflection of pragmatism and an unorthodox view on policy instruments,
than a principal position against product market liberalisation and consumer sovereignty SSB (1965) [20] (pp. 369,370).
16
The TCC had its origin in two important reports from 1966 about the system of wage and income formation, which we refer
to in Section 2.5, see Longva (1994) [33]. 2.4. Labour Market Regulation Olstad (2009) [4] writes in his book about LO from 1989 to 1935, that the Master agreement of
1935 saved the labour movement from a possibly destructive confrontation with both firm owners and
the government 13. When the process away from strife started around 1930, LO did not participate
in the government’s “industrial peace commission” out of strength. It was in a defensive position. On the other hand, the factory owners and the employer confederation NAF had experienced that
an ambition to dictate wage setting was illusionary. At a critical point in 1934, when the government
had already taken controversial labour laws through parliament, it took a step back and accepted to
replace those laws by self-enforced rules by the unions, about secret ballots in particular. In that way,
the unions ended up setting up rules for their own behaviour that the employers’ confederation and
the government had already accepted, Olstad (2009) [4] (p. 419) 14. 10 of 54 Econometrics 2017, 5, 6 There is also a political side of this development. The Labour party moved away from
Moscow-communism during the 1920s, and the labour government that was elected in 1937,
was basically committed to the idea that the working-class could benefit from living in a society with
liberalised product markets and private ownership to productive capital in those markets, but with
collective bargaining in the labour markets. On the other hand, it was seen as almost an prerequisite
that a viable system of labour market regulation had to be in place before the Labour party could take
responsibility for national economic policy, Olstad (2009) [4] (p. 419). Hence, the historical process
may have been characterized by positive feed-back between institutions in the labour market and
in the political sphere. In the words of Barth and Moene (2015) [31], “institutions were beginning
to reciprocate”. This development continued after WW-II, when the ambitious combination of macroeconomic
planning, political democracy and free collective bargaining was noted by American economists and
political scientists, see Bjerkholt (2014) [2] 15. In particular, free collective wage bargaining continued
as the main principle, SSB (1965) [20] (p. 370). Legislation and institutions were introduced to bolster up the wage-setting system, with the
aim to reduce probability of conflicts, and to increase the degree of coordination in wage-setting. The legislation that regulates labour disputes, and a separate Labour Court, dates back to 1915. 2.4. Labour Market Regulation The Technical Calculation Committee (TCC) was established in 1967 by a tripartite agreement,
and is vested with elaborating a common understanding about recent wage developments and
about the forecast for cost of living, and other parameters of relevance for the upcoming agreement
revisions 16. The state mediator has had a strong position, and the period of validity of agreements has
become coordinated (two years). A machinery for interest dispute resolution was built up quite early. The “peace obligation” in
disputes of rights (in practice everything that is regulated by collective agreement), goes back to the
Master agreement of 1935. There has been a relatively low threshold for the use of compulsory
arbitration. For example, when the petroleum sector was built up, arbitration was often used
to settle wage disputes in that sector. The phasing-in of a super-profitable industry in the small
open Norwegian economy was going to be challenging under any circumstances. Completely free
collective bargaining in petroleum could have destabilised the nominal path of the economy, or
at least undermined the competitiveness of non-petroleum based industry. Dyrstad (2016) [34]
provides evidence indicating that government intervention was effective in establishing an element of
co-ordination in the “oil-sector”, before the wider consequences for wage formation became too large
to be reversed. Like in many other countries with collective bargaining, there have been epochs with (different
versions) of incomes policies, as well as a few examples of completely centralised wage setting (by law,
as in 1988). Free collective bargaining has in periods no doubt been regarded as a major problem
as well. In particular, like in many other countries, in the inflation decades of the 1970s and 1980s. In 1973, a proposal about replacing free wage bargaining by a Price and Income Policy Council almost
became government policy, but the largest union confederation LO made a U-turn, Lie and Venneslan
(2010) [35] (pp. 200–202), Bergh (2009) [36] (p. 122). 11 of 54 Econometrics 2017, 5, 6 As noted, union membership was low early in the 20th century, but increased through the
1920s and 1930s, and union strength became a factor in the evolution of the collective labour market
regulation that continued in the postwar Norwegian economy. As Table 1 shows, the unionisation rate
(“union density”), may have peaked around 1990, and the overall impression is one of stability. 2.4. Labour Market Regulation In a
comparison with other western countries, the Norwegian unionisation rate has not been particularly
high, Stokke et al. (2013) [37] (Chapter 2.3.1 and p. 81) 17. But because there has been a secular decline
in the union density of many countries, Norway’s stability at 52%–53% places the country higher in
the league table in 2013 than it would have done in 1980 for example. We have less data about the degree of organisation on the employer side of the bargain, but Table 1
indicates increasing organisation tendencies among firms. The numbers for the firm side are for
the private business sector (and for the number of employees, not firms, to make them comparable
with union density numbers). If government administration is included, the organisation density
becomes 75%. Table 1. Organisation densities in Norway in selected years. (Source: Stokke et al. (2013) [37], Nergaard
(2014) [38]). Unionization Rate
Employer Organization
1948
50%
1972
51%
1990
57%
50 %
2005
53%
60 %
2013
52%
65 % Table 1. Organisation densities in Norway in selected years. (Source: Stokke et al. (2013) [37], Nergaard
(2014) [38]). Table 1. Organisation densities in Norway in selected years. (Source: Stokke et al. (2013) [37], Nergaard
(2014) [38]). The power and political influence of the main union confederation (LO) has varied over the
period, and so too has the role of the main employers association (NHO (which used to be NAF)). As pointed out by Soskice (1990) [39], the analysis of collective wage formation may become too
narrowly focused on the worker organisations, Their counterparts on the employer side are usually
not passive on-lookers to the developments in labour market organisations, but contribute actively out
of organisational and political strength. Above, when we discussed the 1935 Master agreement, we
noted that the leaders of the NAF opted for a compromise, when another employer strategy would
have meant a more direct conflict with the weakened trade unions. There are other examples of
the importance of the power (or weakness) of employer organisation, and if the 1935 compromise
came out of burgeoning power, the unfortunate lock-out in 1986 may have been a nadir for collective
employer behaviour. More generally, the secular trend in the strength of employer organization may be one of the
main determinants of how systems of wage setting have evolved. 17
In the period 1980–2010, Denmark, Finland and Sweden had higher union densities well above 70% for most of the time. 2.4. Labour Market Regulation In some epochs, with perhaps
Western-Germany as prime example, the strong employer organisations were arguably more
instrumental to the system of pattern bargaining than the unions, see Soskice (1990) [39] and
Ruoff (2016) [40]. Bargaining coverage denotes the proportion of wage earners to whom a collective agreement
signed by a union or worker representative and the employer or employers’ association applies. Table 2
indicates that the coverage rate in Norway is somewhat higher than the unionisation degree, but not
by a large margin. It is higher in manufacturing and other goods producing sectors, than in service
production. However, this reflects the same difference in unionisation. 12 of 54 Econometrics 2017, 5, 6 Table 2. Coverage rates in Norway in selected years. (Source: Nergaard (2014) [38] (Table 2.5)). Table 2. Coverage rates in Norway in selected years. (Source: Nergaard (2014) [38] (Table 2.5)). Private Sector
Production of Goods
Service
1998
63%
71%
58%
2004
60%
63%
58%
2005
59%
64%
56%
2008
59%
65%
55%
2013
58%
62%
56% Private Sector
Production of Goods
Service In comparison with other western countries, the Norwegian bargaining coverage would take a
place at the bottom half of that league table, Stokke et al. (2013) [37] (pp. 81–92). The reason is that
there are formal extension mechanisms in many countries. Hence, in the balance between collective
bargaining and the use of law in labour market regulation, the weight is much more on the legal pillar
in countries like Austria, Belgium, France and even Finland and Sweden, than it has been in Norway. In sum, the postwar Norwegian system can be characterized as a voluntarily system for regulation
of wage compensation and working conditions. The parties have little direct support in the legislation
when it comes to extending their agreement to other wage contracts, see Evju (2014) [27]. Hence, we
can draw a distinction between formal bargaining coverage, as measured in Table 2, and the effective
bargaining coverage that results when employers without membership in a confederation nevertheless
offer their workers compensation in line with the relevant collective agreement. It is not unrealistic to
believe that voluntary extension of collective agreements has been a feature of actual labour market
regulation for long periods, in particular in the post WW-II epoch. 2.4. Labour Market Regulation For example, in a situation with
“excess demand” for labour, it can be rational for employers to remove the wage compensation issue
from the competition interface, to avoid cost increasing bidding rounds for employers 18. But it is also possible to imagine that a system of voluntary extension of collective agreement
can be unstable, and that there are tipping points in the organisations rates. If those lines are crossed,
both the effective and formal bargaining coverage can decline sharply 19. A relatively new element in
the labor market regulation in Norway is the The General Application Act (of Collective Agreements),
of June 1993. Although it was far from a semi-automatic extension mechanism, and considering that it
targeted social dumping, the act was contested by organizations on both sides of the bargain at the
time. It has become more in use after 2007 and 2009, see Evju (2014) [27,41], possibly as a response to
real-life problems of maintaining collective bargaining as a regulator of labour markets with many EU
labour immigrants. 19
An example of the relevance of this point is found in the contentious issues in shipyards’ regulation and the
“STX-case”, where EFTA court advisory has collided with the Norwegian High Court domestic conception of public
policy, see Evju (2014) [27]. 8
Again there may be an interesting parallel to Germany, where employer organisations were instrumental in operating
system of pattern bargaining, Soskice (1990) [39] (pp. 43–46). 2.5. Coordination Therefore, the source of the domestic cost-push inflation could not be the wage
settlements in the export and import competing part of the economy. That problem instead resided in
the sectors where there was little foreign competition in the product market. In those markets, pressure
for higher wages could be compensated by price increases. It was easy to foresee that a process of
mutual wage and price increases which started in the “sheltered” sector of the economy, would over
time feed into wage growth in the competing sectors as well. With near full employment, claims for
wage compensation, could become near impossible to withstand. Hence, there was a fundamental horizontal co-ordination problem in wage and price setting. In Norway, the solution became to grant the wage settlement in the manufacturing sector a special role
as wage-leader (or wage-norm setter, or front runner), and sweetening the pill for the wage earner
in the following sectors by reminding them that if they are loyal to the system, they can on average
expect to get the value of a much higher productivity growth than they could count on if they break
out of the system. As noted above, the wage-leader system has performed variably over the decades, with the
the late 1970s and 1980s as possible low-marks, see, e.g., Skånland (1981) [46], Llewellyn (1994) [47]. It clearly relies on strong confederate unions, and it seems to have adapted to the increase in such
organizations. LO was alone, and dominant until the start of the 1970s, but now there are five. The fragmentation of organisations at the employee side may have increased the importance of
the Technical Calculation Committee, (TCC). As noted above, the organizations’ participation in TCC
means that the expectations about cost-of-living increase become synchronised before the negotiations
about wage adjustment start each year. A returning point of concern has been wage drift, which denotes the part of the total wage
change which is not due to the agreement between the confederate organisations. Wage drift arises
mainly from the local wage agreements in the manufacturing sector, not in the wage following
sectors. As a result, the actual wage growth in the wage-leading industry can end up considerably
above the wage-norm. Wage drift has been so large in some epochs (the 1980s in particular) that
it could potentially have undermined the system. 2.5. Coordination As the postwar period unfolded, with de facto full employment, and with a commitment to free
collective bargaining, the management of the economy in many western countries centered around the
trade balances, exchange rate policies and “the inflation problem”. Inflation was not popular among
union leaders and members, Bergh (2009) [36] (p. 118). For the policy makers, it represented a problem
for the attainment of important goals, not an instrument towards attainment of those goals. Contrary
to the academic Phillips curve myth that emerged between 1975 and 1977, there are almost no evidence
of Phillips curve inflationism in Britain, as Forder (2014) [42] shows convincingly. In Norway, the central role of wage formation in the inflation process was clearest
conceptualized in the so called “main-course model”, or the Norwegian model of inflation as it
was dubbed in 1977 when one of the intellectual fathers of the model finally published a paper 13 of 54 Econometrics 2017, 5, 6 in English, Aukrust (1977) [43]. The main-course model was the outcome of two reports that an expert
group of Norwegian economists (Aukrust, Holte and Stoltz) published as background material for the
wage and agricultural price negotiations in 1966. The second report, contained the long-run model that
we refer to as the main-course model, see Aukrust (1977) [43]. The connotation is navigation over long
distances, not the dinner table. It is also referred to as the “front runner model”, or “leader model”,
since the collective agreement in the internationally competing manufacturing sector represents the
wage-norm that other sectors in the economy follow. The premise is that wage growth must be
adjusted to a level which over time is capable of sustaining the competitiveness of import and export
competing industries. p
g
In that historical epoch, there were similar developments in, for example, Sweden, see
Edgren et al. (1969) [44], and the Netherlands. This model became the framework for both medium
term forecasting and normative judgements about “sustainable” centrally negotiated wage growth
in Norway and Sweden 20. A key-point in the analysis was that in the export and import competing
sectors of the economy, considerations about the required return to capital served as an automatic
stabilizer of nominal wage cost growth. Over time it was one of the corrective mechanism that would
make the wage cost level fluctuate around a main-course growth path defined by the value of average
labour productivity. 20
On the role of the main-course model in Norwegian economic planning, see Bjerkholt (1998) [45]. 2.6. Development of Working Time and Wages Beside wage compensation and health hazards at work, working time is the main variable that
needs regulation in the labour market. Unlike wage agreements, which to a very limited degree have
been law regulated in Norway, working time reforms have usually started with collective agreements
before it has been extended to all wage contracts by law. Figure 5 shows the development of the length
of the working week, and the number of working days, relative to 1900. Working days
Length of working week
1900
1910
1920
1930
1940
1950
1960
1970
1980
1990
2000
2010
0.65
0.70
0.75
0.80
0.85
0.90
0.95
1.00
Working days
Length of working week
Figure 5. Working days in the year, and the length of the working week. 1900 = 1. Source: Appendix A. Working days
Length of working week
Working days
Length of working week Figure 5. Working days in the year, and the length of the working week. 1900 = 1. Source: Appendix A. At the start of the century, the number of working days was 300 and the length of the working
week for regular day time work was 60 h. By the end of WW-I, weekly hours has been reduced by
20 percent (to 48 h) but then it stayed unchanged until 1959. The last reduction in normal hours
came in 1987, and was the result of the wage settlement in 1986 (which also involved a somewhat
bizarre lockout, since the economy was in a boom in that year). Hour reductions have usually
been compensated, so that annual earning are intended to be unchanged, Nymoen (1989) [50]. This
happened in 1987, but also in 1976 (40 h) even though many firms struggled with the consequences of
stagflation internationally, the industrial structure needed an overhaul, and cost-push inflation was
already a recognized problem, Bergh (2009) [36] (p. 135). The more gradual reduction in the number of workdays, from 300 in year 1900, reflects the
increasing length of annual holidays. A major reduction came in the short time span between 1965
(280 days) and 1969 (231 days). Again this was an effect of extension of agreements about a fourth
holiday week, but the main part of the reduction was due to the introduction of the five-day working
week. Clearly, the reduction in annual working days has been compensated. 2.5. Coordination However, as analysed by Holden (1989) [48],
since there is no right to strike or lock-out during local negotiations, a bargaining model implies
that wage drift would not completely undo the outcomes of the settlements at the confederate level. Holden reported empirical evidence that supported the theoretical conclusions, and hence there
may be a structural explanation for why wage drift has not perverted the system. Nevertheless, the 14 of 54 Econometrics 2017, 5, 6 worrying about wage-drift has never disappeared. For example, if fragmentation of organisations
means weaker ability to contain firm level wage increases, especially for the higher paid white-collar
workers, the unions of the wage-followers might loose patience, and horizontal coordination will
suffer. In 2013, an official report where the organisations participated, reinforced the extension of the
wage-norm: it should also regulate the wage negotiations for white-collar workers in manufacturing,
cf. NOU (2013) [49]. 2.6. Development of Working Time and Wages Hence, labour productivity
per hour worked needed to be increased, either before or, more realistically, as a response to increased
holiday length and shorter working week. Figure 6 shows wage and price growth. Inflation was low (some years negative) at the start
of the last century, which was an international phenomenon. That was soon over, and wages in 15 of 54 15 of 54 Econometrics 2017, 5, 6 particular grew steadily until the war, in part as a result of construction and building activity and low
unemployment as noted above. During WW-I, prices first shot up, and wages followed quickly. The high inflation rates during WW-I were extraordinary, in particular when we take into account
that Norway did not participate in the war. Great Britain in comparison recorded inflation rates around
25%. Lie (2012) [15] describes what seems to have been a near meltdown of the monetary system
during the war, and in the first years after it ended. The deflationary par policy that began in 1920 was
a reaction. It may have been needed to restore confidence in the system, but the real economic costs of
the chosen policy became huge as the par policy period dragged on. There were 12 years with nominal
wage reduction between 1920 and 1934. Only in 1924 and 1925 did nominal wages grow. Wage growth
CPI inflation
1900
1910
1920
1930
1940
1950
1960
1970
1980
1990
2000
2010
-20
-10
0
10
20
30
40
Master
agreement
1935
WW-I
WW-II
Par-policy
Devaluation
1949
VAT
1970
Devaluation
1986
Wage-law
1988
Banking crisis
Inflation targeting
Percent
Wage growth
CPI inflation
Figure 6. Annual growth in annual wage and in consumer price index (CPI). Source: Appendix A. Percent Figure 6. Annual growth in annual wage and in consumer price index (CPI). Source: Appendix A. After WW-II, and after the effect of the 1949 devaluation was over, inflation was stable and
relatively low until the early 1970s. During the 1940s and 1950s, rationing and direct price control
was used to contain what was clearly understood as a situation with “excess demand” at the time,
see SSB (1965) [20]. But gradually, price formation in the product markets was normalised, and as
noted above, there was in principle free collective wage bargaining during the whole period. As also
noted, the 1970s were marked by gradually increasing inflation, in Norway as elsewhere in western
Europe. 2.6. Development of Working Time and Wages Early in the 1970s, North Sea oil production was still nowhere large enough to shelter the
country from the price increases that followed after the international oil crisis. The 1980s were even
more problematic with a string of self-inflicted unemployment in 1983–1984, “technical devaluations”,
a collapse in coordination of wage formation, and the mentioned lock-out in 1986, followed by a
relatively large devaluation. The decade ended with a collapse of the housing market, a huge banking
crisis and finally a big rise in the rate of unemployment. The consequences of the housing price
crisis were also felt long into the 1990s, as financial consolidation in the household sector depressed
private consumption. In a small open economy, inflation is always associated with foreign inflation, and with the rate of
change in the international value of the domestic currency (rate of depreciation). In Figure 7, this is
brought out by the graph for the annual change in the Norwegian import price index. Import price
growth was clearly leading inflation at the start of WW-I. Even though it is thinkable that the domestic
deflation contributed to the appreciation of the currency during the 1920s, it is more realistic that
the causation was mainly the other way, from par-policy to domestic deflation. During the 1930s,
the effects of the Great Depression on foreign currency denoted imports must also have played a role. 16 of 54 16 of 54 Econometrics 2017, 5, 6 Finally, towards the end of the sample, the secular reduction in import price growth seems to have
weighted down the nominal path of the Norwegian economy, CPI inflation in particular. This can in
part be due to the increased value of the krone, a consequence of the high oil price level of the period. However, the increasing role of China-produced commodities in the world economy, also depressed
the prices of many imported goods to Norway. Wage growth
Import price growth
CPI inflation
1900
1910
1920
1930
1940
1950
1960
1970
1980
1990
2000
2010
-20
-10
0
10
20
30
40
50
60
Percent
Wage growth
Import price growth
CPI inflation
Figure 7. Annual change in the import price index, together with annual wage change and consumer
price inflation. Source: Appendix A. Percent Figure 7. Annual change in the import price index, together with annual wage change and consumer
price inflation. Source: Appendix A. 2.6. Development of Working Time and Wages 17 of 54 Econometrics 2017, 5, 6 nometrics 2017, 5, 6
17 of
1900
1910
1920
1930
1940
1950
1960
1970
1980
1990
2000
2010
0
20
40
60
80
100
Percent
Figure 8. 15 year real wage growth percentages from 1900 to 2015, for example, the height of the first
bar shows that real wage growth from 1885 to 1900 was 25 percent, and the last bar shows that growth
in the real wage from 2000 to 2015 was close to 40 percent. Source: Appendix A. Figure 8. 15 year real wage growth percentages from 1900 to 2015, for example, the height of the first
bar shows that real wage growth from 1885 to 1900 was 25 percent, and the last bar shows that growth
in the real wage from 2000 to 2015 was close to 40 percent. Source: Appendix A. The relationship between labour productivity and the general wage level is nearly always close to
the centre of discussion about wage formation. Often, and in particular in periods of practically full
employment, the question is how to avoid that the growth in real wage costs, i.e., the producer real
wage, does not exceed the growth in labour productivity, which could make the share of labour become
so high that it harms necessary investments in the import and export competing sector. As mentioned
above, the system with the manufacturing sector acting as the wage leading and norm setting sector,
can be seen as an operation that solves that issue. In many other countries, the focus is on another, related but nevertheless different, relationship,
namely between the consumer real wage and labour productivity. There is evidence, across a number
of countries, of consumer real wages falling short of productivity over the the last decades of our
sample period, Haldane (2015) [52]. For example, in the US, this has been apparent since the 1970s,
and in the UK since the 1990s. In an econometric paper, Bårdsen and Nymoen (2009) [53] modelled the
US case by showing empirically that the trend in the wage level was weakly linked to the productivity
level, and more strongly linked to a reference wage determined by the probability of getting a job
elsewhere and the cost of living. 2.6. Development of Working Time and Wages As a result, growth in the consumer real wage was quite high, and also stable, during the first
15–16 years of the new millennium, as Figure 6 shows. For example, from 1997 to 2012 the consumer
real wage increased by 46%. In 2015, the 15 year growth rate had been reduced somewhat, to 38%,
but was still high compared to other countries. The source of the strong recent real wage performance
is still debated. A plausible argument is that by the end of the 1990s, Mainland-Norway had become
integrated with the international petroleum industry. Hence, even though as noted above, the wage
level in that sector has had a limited direct influence on the general wage level in Norway, the indirect
effect nevertheless became quite large when the oil price and oil investments started to grow after the
financial crisis. In a way, the super profitable petroleum sector had come to influence the wage norm
trough the back-door, see Anundsen (2016) [51]. The bar chart in Figure 8 shows all the 15-year growth rates from 1900 to 2015. The heights of the
bars show the growth rates. The steady increase in the growth rates after the low-point at the start
of the 1990s is easy to see. At the start for the graph, the impression is not so much that real wage
growth was absent, but that it was relatively uneven. When the economy came out of the deflation
years, real wage growth was very weak. Real wage growth was actually more positive during the
deflation itself (the CPI index was more reduced than the yearly wage). However, that did not help
the real economy much, since the increasing weight of debt depressed aggregate demand for product
and labour. Figure 8 also shows that the highest growth rates occurred in the late 1950s, not so surprising
given how the economy developed during the reconstruction years. In 1956, the consumer real wage
had doubled compared to the real wage of the first year of the Nazi occupation. The 15-year real-wage
growth rates continued to be very high during the 1960s. Even the 1970s bad reputation for real wage
eroding inflation seems a little exaggerated when we look at this graph: The real wage growth rates
did not dip below 50 percent before 1980. 2.6. Development of Working Time and Wages Another reason why the consumer real wage can drift away from productivity, perhaps most
relevant for small open economies, is that by definition, the relative price of imports drives a wedge
between the producer real wage and the consumer real wage. Hence if there are secular changes in the
relative price of imports, a gap can open up between the consumer real wage and labour productivity,
even if the producer real wage still tracks productivity. In Norway, this effect, due to fortunate terms of
trade development, may have pushed the consumer real wage above the productivity trend at the start
of the new millennium. In any case there are no traits in Figure 9 of anything like the Anglo-American
experience where labour has not shared in the fruits of the recent productivity growth, at least not
so far. onometrics 2017, 5, 6
18 of 5
log consumer real wage
log labour productivity
1900
1910
1920
1930
1940
1950
1960
1970
1980
1990
2000
2010
3.0
3.5
4.0
4.5
5.0
5.5
6.0
log productivity and real wage
log consumer real wage
log labour productivity
Figure 9. The consumer real wage and labour productivity, logarithmic scale. The real wage graph has
been shifted down for easier comparison with the productivity graph. Source: Appendix A. 18 of 54 Econometrics 2017, 5, 6 Figure 9. The consumer real wage and labour productivity, logarithmic scale. The real wage graph has
been shifted down for easier comparison with the productivity graph. Source: Appendix A. 3. Modelling: Theory Given the importance of wage setting in the national economy, and the rise of the economics
profession as economic policy makers and advisors, one would perhaps expect that that the system of
wage setting that evolved in Norway had a strong foundation in economic analysis. However, that
does not seem to have been the case. The Phillips curve for example, is hardly given a mention, even
by the time Aukrust published his article about the Norwegian model of inflation from 1977. On the other hand, the problem of cost-push inflation and the acute need for horizontal
coordination in a situation with full employment, was well understood, in Norway and elsewhere. In particular, and in the same manner as in international literature surveyed convincingly by
Forder (2014) [42], Aukrust listed unemployment well down on the list of wage-determining factors,
allocating it some importance when the rate was low and falling, but not when it was moderately high
and increasing. What this reflected, again typical of its time, is a certain exogeneity proposition about the wage
level, but not about the wage level as unconditionally fixed. Rather, it was a proposition about
exogeneity of nominal wages and prices to the level of (un)employment, see Forder (2014) [42]
(Chapter 1.3). Hence the view was twofold: First that unemployment could vary quite a lot (at least above a
certain low level defined by friction) without any very noticeable effects on wages and prices in macro. This was later known as the “L-shaped” price (and wage) curve. Second, that wages could increase
(or be reduced) a good deal without any simultaneous or preceding change in unemployment taking
place. In turn, this conceptualization opened for a clear understanding of cost-push elements and
possibility of wage and price spirals. 3.1. Theory of Wages and the Development of Wage Modelling The exogeneity proposition that was characteristic of applied macro in the 1950s and 1960s was
therefore not a sign of lack of thinking about wage formation. On the contrary, it was an expression of
how far economics had come in finding a relevant theoretical perspective on wage setting. There was in any case a clearer recognition that there was an indeterminacy in the economic
theory of wages in the 1940s and 1950s than it is today. In the article from which the Frisch-quotation Econometrics 2017, 5, 6 19 of 54 is taken from, he lamented the “blatant hole in the science of economics”. However, he also expressed
good hope that the matter would be brought in order if sufficient funding was given to a proper
research program in the field, which he clearly was happy to trust Haavelmo with (but it did not come
to that as we have heard). Frisch was not alone. On this, Forder (2014) [42] (Chapter 1.4), cites Samuelson (1951) [54] (p. 312)
and Hicks (1955) [55] (p. 390) and other leading theorists. The economic theory of supply and demand
could set some limits to what wages can be set, but within those limits closer determination requires
that other relationships are introduced. That was in the 1950s. However, the indeterminacy of wages
from theory also characterizes the now standard Diamond-Mortensen-Pissarides (DMP) search and
matching model. In the DMP model, the wage is usually determined in a Nash bargaining game. But is the wage
logically equal to the Nash solution given the assumptions of the DMP model? As Hall (2005) [56]
pointed out, any wage in the bargaining set is in principle consistent with private efficiency on the part
of both the firm and the worker. In that sense, the equilibrium wage rate is only set-identified. He then
went on to analyze a solution where the real wage is fixed, which however is only one possibility of
what in the DMP-literature is referred to as wage stickiness. Following Hall (2005) [56], several papers
have incorporated rigid wage setting in search models. For instance, Gertler and Trigari (2009) [57]
present a DMP model where the frequency of wage bargaining is constrained by a Calvo (1983) [58]
style lottery, leading to sticky wages. 3.1. Theory of Wages and the Development of Wage Modelling Blanchard and Galí (2010) [59] combine a reduced form search
model with real wage rigidity with a New Keynesian model to study how this impacts monetary
policy. Krogh (2016) [60] generalizes the Hall-approach to a small open economy model where there is
a non-trivial distinction between the consumer real wage and the producer real wage. Another theoretical approach, where utility functions of trade unions and firms are used to derive
a Nash-bargaining wage level, is also incomplete, since the wage equations that follow directly from
this theory are static, while time obviously plays a fundamental role in actual wage formation, see e.g.,
Nickell and Andrews (1983) [61] for an early contribution. Additional theoretical arguments have to
be added in order to bridge the gap from static theory to dynamic equations that can be confronted
with the data, see e.g., Nickell (1985) [62]. Below, I use a version of this approach, but I interpret the
static equations as long-run equilibrium equations, not as equations that pin down wages to the actual
amount of kroner paid per hour or year. As Forder (2014) [42] noted, with reference to Usher (2012) [63], understanding bargaining
requires an assessment of not only self-interest among workers and firms, but also of compromise. Compromise, is then a real world phenomenon, not just re-labelling of self-interest, and social, political
and institutional forces are among the fundamental determinants of decisions. In this view, even a full
analysis of economic rational behaviour leads to an indeterminacy of wages, and other considerations
(“non rational” or non economic) had to be introduced to resolve it. Again, there is nothing anti-theoretical about this view, it is just a realistic view that used to be
widely accepted, as in Samuelson’s textbook (3 ed. 1955, p. 547): [64] [wage formation]...depends on psychology, politics, and thousands of other intangible
factors. As far as the economist is concerned, the final outcome is indeterminate—almost
as indeterminate as the haggling between two millionaires over the price paid for a rare
oil painting. [wage formation]...depends on psychology, politics, and thousands of other intangible
factors. As far as the economist is concerned, the final outcome is indeterminate—almost
as indeterminate as the haggling between two millionaires over the price paid for a rare
oil painting. 3.1. Theory of Wages and the Development of Wage Modelling At the same time as we find it challenging to determine wages theoretically, we also observe that actual
wage bargains are struck year after year, and that they are rationalized by considerations of profits,
actual and required (to attract investments), cost of living and relative wages (fairness). These observed
regularities, that were documented early by for example Dunlop (1944) [65], give reason to believe
that wage formation can be subject to econometric treatment. This is also what have motivated much of the econometric literature that may have started with
Phillips’ 1958 paper [66], but which soon lost contact with it. Perhaps this happened because Phillips’ 20 of 54 Econometrics 2017, 5, 6 research question was very clear: His view was that over a long data sample, the relationship that
determined the change in money wages was determined by supply and demand, as captured by the
rate of unemployment, institutional factors did not go into it. In this he was clearly in opposition to the
wage theory of his day, which claimed that many institutional and psychological factors mattered, that
the “wage equation” was L-shaped and that the attainment of full employment with price stability
was possible. None of these views or claims were correct if Phillips was right. But, as Forder concluded his assessment, Phillips was wrong 21. Already, Lipsey (1960) [67]
had noted that his estimated Phillips curves were different in different periods. Afterwards, the
econometric time series modelling of wages seems to have been split in two main currents. The
first is the augmented Phillips curves, where price changes are brought into the model, and where
there is a distinction between a downward sloping short-run relationship between wage change and
unemployment and a vertical long-run Phillips curve. The natural rate of unemployment, and the
accelerationist Phillips curve are central concepts in this class of models, (Bårdsen et al. (2005) [68]
(Chapter 4)). The second branch of the econometric literature, is possible to interpret as a continuation of
the L-shaped theory of the 1950s. In these econometric models, it is possible but far from certain
that price growth can be stabilised at any going rate of unemployment. Hence, we recognise the
exogenitey theory of wages with respect to unemployment. 3.1. Theory of Wages and the Development of Wage Modelling Mathematically, this is just an implication
of modelling wage and price levels as generated by two stochastic difference equations, and making
that system subject to rank reduction (at the zero (long-run) frequency). As long as that rank is not zero,
price stabilization is vindicated as a theoretical property. It was the British econometrician Denis Sargan
who formulated this class of models, and they were first known as error-correction models, see Sargan
(1964, 1971, 1980) [69–71]. David Hendry has later suggested that they are referred to as equilibrium
correction models, ECMs, since they are modelling wages and prices as adjusting towards dynamic
equilibrium relationships, which in turn are interpretable as cointegration relationships. In terms of economic interpretation, equilibrium correction models sit well with the idea of a
wage-norm that serves as an attractor to actual wages. They are relevant models to consider when we
attempt to model Norwegian wages, since the wage-norm is a central variable of the system of wage
setting. Another reason why wage-price ECMs are a good starting point for modelling, is that they can
be formulated in such a way that the standard Phillips curve specification become encompassed by
the ECM system. In that way, it also gives the econometric framework for testing the Phillips curve,
and restrictions that are associated with it, such as vertical long-run Phillips curve restrictions and
natural rate of unemployment restrictions. 21
Forder (2014) [42] (p. 31). 22
Hatton (1988) [72] (p. 84) concluded along the same lines for the British labour market: It did not change from “one where
wages were set by atomistic competition to one in which the process was entirely institutional. Institutional wage setting
was well established before 1914, though it became increasingly centralized until the late 1960s”. 3.3. A Dynamic Model for Trends in Wage-and-Price Setting The basic nominal variables in the model we formulate are: yearly wage w, domestic producer
price q, domestic consumer price p, and import price pi in domestic currency. The average labour
productivity a and the unemployment rate u are real variables. All variables are in logarithmic scale,
primarily to facilitate relationships that are linear in the parameters. The following presentation follows
Kolsrud and Nymoen [6,74] closely, but the framework has been developed in stages and applied to
different data sets, see e.g., Bårdsen et al. (2005) [68], Bårdsen and Nymoen (2003) [75], Bårdsen and
Nymoen (2009) [76], Akram and Nymoen (2009) [77], Bårdsen et al. (2012) [78]. I present the model briefly, with focus on the main parameters of interest that I will try to estimate
later in the paper. We then look at how the theoretical model behave when we put in numbers for the
parameters and solve the model by dynamic simulation, from a given starting point. We specify the
theoretical model with one large break in the mean of the rate of unemployment. The question is then
whether the break has a lasting effect on wages? And, in levels or in growth rate form? The answers
the theory gives to these questions, can aid the interpretation of the empirical model’s results. 3.2. Change and Continuity “Change” is the single word that summarised the economy and labour market history that we
reviewed above. However, there are also elements of continuity which may be used to motivate
econometric modelling of wages over this long period. The first is that wage formation has been “free”
for the whole period. Although unions were weak at the start of the period, that may have changed
quite rapidly during the two first decades. Any ambitions on the employer side to dictate nominal
wage compensation (Frisch’s “Manchester liberalism”) may have become frustrated quite early in
the century. The employer side’s forceful attempt to re-take lost ground during the 1920s, was also
changed to a more cooperative line in labour market policy issues in the early 1930s. As we have seen,
there were similar development on the workers’ side, as unions set up internal rules that the firm 21 of 54 Econometrics 2017, 5, 6 owners and the government could accept, and which also strengthened the leaders of the unions and
the confederate level 22. Hence, it seems a worthwhile project to represent econometrically the main factors behind nominal
wage changes in a system of collective bargaining: Firms profitability, cost of living (current level and
outlook), and demand and supply conditions on the labour market. Another “continuity trait” is that price formation in the product markets has been up to the firms,
in competition with the price of imports, and strongly influenced by the exchange rate. A model
broadly along the lines of monopolistic competition is therefore a relevant conceptual framework. It captures that firms attempt to secure required profits by adjusting prices relative to unit labour
cost, and to adjust production to sales opportunities which in turn depends on both market size
(foreign and domestic demand), and on price competitiveness. In the same manner as for nominal
wages, there have been periods where free price formation did not apply: The occupation years
and the first years after the war with rationing and a strong focus on price control. Price-freeze
periods were not unknown during the 1960 and 1970s either, then as a part of incomes policies,
see Bowitz and Cappelen (2001) [73]. However, in the longer historical perspective, they represented
short departures from the main principle of free price setting and collective bargaining. 3.3.1. Nominal and Real Trends We begin by defining the exogenous trend variables of the model. By trends we mean, for now,
stochastic trends, or unit-root non-stationarity. Deterministic trends with breaks will be discussed
later. Stochastic trends are represented by integrated variables, I(d), where d denotes the order of
integration. In our analysis d will be 1 (unit root) or 0 (stationarity). There are two exogenous I(1) variables in the model: one nominal trend and one real trend. The nominal trend is the price of imports pi in domestic currency. We write the equation as a
random-walk with a positive drift: pit = gpi + pit−1 + εpit, gpi > 0 and εpit ∼N(0, σ2
pi). (1) (1) The drift parameter gpi represents underlying foreign inflation. The disturbance term εpit may include
international price shocks or a stationary nominal foreign currency exchange rate (normalised to
zero mean). Econometrics 2017, 5, 6 22 of 54 Since pit is defined as the sum of the logarithms of a price index in foreign currency and an
effective nominal exchange rate index, we could refine the model by specifying regime dependent
variance and drift parameters. However, it is plausible to assume that pit ∼I(1) across floating and
fixed exchange rate regimes, because the I(1) foreign price component will dominate. pit is one of three price variables in the model. The two others are the producer price index, qt,
which will be modelled from the firm-side of the economy. The third is the logarithm of the consumer
price index, pt. It is defined by: pt = φqt + (1 −φ)pit,
(2) (2) where the parameter 0 < φ < 1 measures the share of imports in total consumption. (2) is only a
stylized “consumer price index equation”, but it allows us to make the important distinction between
the consumer real-wage and the producer real-wage. The above discussion of historical labour productivity growth, shows that to be able to claim any
realism at all, we should include a real trend in average labour productivity at: at = ga + at−1 + εat, ga > 0 and εat ∼N(0, σ2
a ). (3) (3) We model domestic wage and price dynamics as conditional on pit and at. Equations (1) and (3) will
therefore imply that qt, the (log of the) price level on domestic products, and wt, the (log of) wage
compensation per hour, will both be non-stationary. 3.3.1. Nominal and Real Trends In the case where the domestic wage and price
setting system is dynamically stable, qt and wt will be integrated of order one, I(1), in a common
notation. If the domestic wage-price spiral is unstable, domestic wages and prices become “more
non-stationary” than their foreign counterparts, theoretically they may become I(2), or there may be
wage and price bubbles (“due to” explosive characteristics roots). We model domestic wage and price dynamics as conditional on pit and at. Equations (1) and (3) will
therefore imply that qt, the (log of the) price level on domestic products, and wt, the (log of) wage
compensation per hour, will both be non-stationary. In the case where the domestic wage and price
setting system is dynamically stable, qt and wt will be integrated of order one, I(1), in a common
notation. If the domestic wage-price spiral is unstable, domestic wages and prices become “more
non-stationary” than their foreign counterparts, theoretically they may become I(2), or there may be
wage and price bubbles (“due to” explosive characteristics roots). 23
We do not introduce explicit notation for firms’ expected wage, because with zero mean I(0) expectation errors, it will
not have any implications for co-integration (or not) between the variables. However, it is understood that wt in (4) is an
expected variable, and that p and q in (5) likewise denote expected prices in this context. 24
For the coefficients ψwq, ψqw and ψwp, ψqpi, the non-negative signs are standard in economic models. Negative values of
θw and θq, can give rise to explosive dynamics in wages and prices (hyperinflation), which is different from the low to
moderately high inflation scenario that we have in mind for this paper. 3.3.2. Wage-Price Spiral We next define two theoretical (latent) real wage variables: The targeted producer real wage from
the point of view of the firms, rw f
t , and the planned or expected bargained producer real wage, rwb
t . They are given by the following two equations: rw f
t = wt −q f
t = −mq + at + ϑ ut, ϑ ≥0
(4)
rwb
t ≡wb
t −qt = mw + ω (pt −qt) + ι at −ϖ ut,
(5) (4) (5) with ι > 0, 0 ≤ω ≤1, ϖ ≥0, see Nymoen and Rødseth (2003) [79]. with ι > 0, 0 ≤ω ≤1, ϖ ≥0, see Nymoen and Rødseth (2003) [79]. Equations (4) and (5) can, if we want, be drawn as a “price-curve” and a “wage curve” respectively
in a diagram with real wage along the vertical axis and ut along the horizontal axis. In the case of
ϑ = 0, the “price curve” becomes horizontal. If (pt −qt) changes, the “wage-curve” shifts if ω > 0. Care must be taken: The intersection of the curves does not represent the determination of the rate
of unemployment, in the NAIRU meaning. Formally, this is an under-determined model, with more
variables than equations. f In the price-setting Equation (4), q f
t denotes the price level set by the firm on basis of expected
nominal marginal labour costs wt −at 23. In (5), wb
t denotes the planned bargained nominal wage,
again given expectations of the two price levels. A reasonable assumption is that the elasticity ι with respect to productivity is close to unity,
as in Nymoen and Rødseth (2003) [79]. The standard assumption about the sign of the coefficient for
unemployment, ϖ, is that it is non-negative, hence −ϖ < 0, as already indicated. The coefficient ω is Econometrics 2017, 5, 6 23 of 54 called the wedge coefficient since it is multiplied by (p −q)t which is the wedge between consumer and
producer real wages (we abstract from tax rates). The wedge coefficient is assumed to be non-negative,
ω ≥0, see Rødseth (2000) [80] (Chapter 8.5). called the wedge coefficient since it is multiplied by (p −q)t which is the wedge between consumer and
producer real wages (we abstract from tax rates). The wedge coefficient is assumed to be non-negative,
ω ≥0, see Rødseth (2000) [80] (Chapter 8.5). 3.3.2. Wage-Price Spiral f Both rw f
t and rwb
t are I(1) variables by construction and can be co-integrated with the producer
real wage rwt. With that in mind we define two variables ecm f
t and ecmb
t ecm f
t ≡rwt −rw f
t = q f
t −q = wt −qt −at −ϑ ut + mq
(6)
ecmb
t ≡rwt −rwb
t = wt −wb
t = wt −qt −ι at −ω(1 −φ)ret + ϖ ut −mw,
(7) (6) (7) where we have used (2) to write the wedge variable as (p −q)t = (1 −φ)(pi −q)t, and then defining the real exchange rate ret ≡pit −qt. The point of this step will soon become clear,
since the complete model specification can be solved in terms of ret and one other real variable called
the wage-share. We may note, in passing, that if cointegration does hold, both ecm f
t and ecmb
t are I(0)
variables. However, we do not impose that from the outset. For short-run dynamics, the wage-price spiral, we use the simultaneous equation model: ∆qt = cq + ψqw ∆wt + ψqpi ∆pit −ς ut−1 + θqecm f
t−1 + εqt,
(8)
∆wt = cw + ψwq ∆qt + ψwp ∆pt −ϕ ut−1 −θwecmb
t−1 + εwt,
(9) (8)
(9) (8) (8) (9) where ∆is the difference operator, ∆qt ≡qt −qt−1, and where ψqw, ψqpi, ψwq, ψwp, ς, ϕ, θq, θw ≥0,
εqt ∼N(0, σ2
q ), εwt ∼N(0, σ2
w) 24. At first it may seem cumbersome that ut−1 is included in each equation, since it is already included
in ecm f
t and ecmb
t . However, this is done to extend the framework to the case where cointegration fails
in the sense that ecm f
t and/or ecmb
t may be I(1), instead of I(0). For example, ecmb
t ∼I(1) =⇒θw = 0,
but ϕ > 0 is still a logical possibility as long as ut ∼I(0), which is seen to give a wage Phillips-curve
model, (wage PCM). Conversely, ecmb
t ∼I(0) =⇒θw > 0 since an equilibrium correction model (ECM)
is implied by the Granger-Engle (1987) [81] representation theorem. However, in that case, ϕ = 0 is the
only logically consistent possibility. 3.3.2. Wage-Price Spiral y
g
y
p
y
Substituting for the two ecms in (6) and (7), we get: Substituting for the two ecms in (6) and (7), we get: the two ecms in (6) and (7), we get: ∆qt = (cq + θq mq) + ψqw ∆wt + ψqpi ∆pit −µq ut−1
+ θq (wt−1 −qt−1 −at−1) + εq,t,
(10)
∆wt = (cw + θw mw) + ψwq ∆qt + ψwp ∆pt −µw ut−1
−θw (wt−1 −qt−1 −ι at−1) + θw ω (pt−1 −qt−1) + εw,t,
(11) (10) + θq (wt−1 −qt−1 −at−1) + εq,t,
(10)
∆wt = (cw + θw mw) + ψwq ∆qt + ψwp ∆pt −µw ut−1
−θw (wt−1 −qt−1 −ι at−1) + θw ω (pt−1 −qt−1) + εw,t,
(11) (11) where the notation µq = θq ϑ + ς and µw = θw ϖ + ϕ are used to nest PCM and ECM specifications in
the same formulation, as just noted. Although ECM is an econometric concept, it also has an economic connotation: Since the
wage-curve is associated with bargaining models, θw > 0 is a relevant operational definition of
an Incomplete Competition Model, ICM. The point was expressed early by Sargan (1971) [70] (p. 52),
who saw “error correction” variables as natural to use in a model of bilateral monopoly, and stating 24 of 54 Econometrics 2017, 5, 6 in a later paper that “clearly both sides in a wage bargaining procedure are concerned with the real
wage”, see Sargan (1980) [71] (p. 98) 25. As pointed out above, the understanding that nominal wage setting was a result of a
genuine bargaining process with power relationships, was quite common in the 1960s and 1970s. Another econometrician, Trygve Haavelmo, presented the same line of thought in some lectures on
inflation in the mid 1970s, see Anundsen et al. (2012) [83]. The label “Conflict models” were put on
these models at the time, but a better name might have been “Compromise models”. Haavelmo in
particular, was clear that if the model solution was stable, the equilibrium real wage represented a
compromise; like a “quasi peace”. p
q
p
The relevance of incomplete competition in the product market is mainly to draw the distinction
between perfect competition and monopolistic competition. 3.3.2. Wage-Price Spiral In the labour market, the label incomplete
competition is even more to the point, since limitation of unwanted competition is a prerequisite
for well functioning labour market regulation based on collective agreements. The emphasis is on
unwanted because it is clear that unless there is some consensus about how much competition is wanted,
or “right”, in the labour market, a system of collective bargaining will probably be undermined by
industrial unrest and strife. As we have seen above, the Norwegian system of wage fixing did not reach
a real stability before the two parties gave up the ambition of unilateral control over wage changes. The main alternative to the cointegrated ICM is the wage Phillips curve, defined by θw = 0,
which is relevant if supply and demand conditions, after all, are the only really wage determining
factors, see Bårdsen et al. (2005) [68] (Chapter 3–6). In order to distinguish between ICM and PCM
specifications we therefore have: ICM: θw, θq > 0 and ς = ϕ = 0 =⇒µw = θwϖ and µq = θq ϑ
PCM: θw = θq = 0 and ς = ϕ > 0 =⇒µw = ϕ and µq = ς Note that the ICM is consistent with a horizontal “price curve”, since ϑ = 0 =⇒µq = 0, but not θq = 0. Note that the ICM is consistent with a horizontal “price curve”, since ϑ = 0 =⇒µq = 0, but not θq = 0. 25
Later theoretical derivations in the literature, using the Nash-solution, agree that θw > 0 is implied by collective bargaining,
but also find ω = 0 to be equally theoretically important, and even purge the compensation for cost-of-living increases from
the short-run dynamics, see Forslund et al. (2008) [82].
f
f y
26
An alternative is to define a “productivity corrected producer real wage”: rw f c
t
= rw f
t −ι = 1at 3.3.3. VAR Formulation Using the differenced version of the definitional (2): Using the differenced version of the definitional (2): ∆pt = φ ∆qt + (1 −φ) ∆pit
(12) (12) in combination with (10) and (11), the model can be re-formulated as a (open) VAR model in the two
variables ret and wst = wt −qt −at, the logarithm of the wage-share. If the parameter restriction
ι = 1 on at−1 (in wage-setting) is imposed, the VAR for ret and wst becomes independent of the labour
productivity level, but depends on the growth rate, ∆at. As long as at is specified as a random walk
with drift that we have specified, this step is mainly to save notation 26. This conditional VAR is found in the two first rows of (13), while the third row of the VAR (13)
contains a simple dynamic relationship between ut, the rate of unemployment in period t and the
lagged ret.
ret
wst
ut
yt
=
l
−k
n
λ
κ
−η
−ρ
0
α
R
ret−1
wst−1
ut−1
yt−1
+
e
0
−d
ξ
−1
δ
0
0
cu
P
∆pit
∆at
1
xt
+
ϵre,t
ϵws,t
ϵu,t
ϵt
(13) (13) 25
Later theoretical derivations in the literature, using the Nash-solution, agree that θw > 0 is implied by collective bargaining,
but also find ω = 0 to be equally theoretically important, and even purge the compensation for cost-of-living increases from
the short-run dynamics, see Forslund et al. (2008) [82]. 26
An alternative is to define a “productivity corrected producer real wage”: rw f c
t
= rw f
t −ι = 1at 25
Later theoretical derivations in the literature, using the Nash-solution, agree that θw > 0 is implied by collective bargaining,
but also find ω = 0 to be equally theoretically important, and even purge the compensation for cost-of-living increases from
the short-run dynamics, see Forslund et al. (2008) [82]. f c
f 25 of 54 Econometrics 2017, 5, 6 It represents a stylized dynamic aggregate demand relationship. Since increased ret means improved
competitiveness (real depreciation), it is reasonable that the parameter ρ in the third row is non-negative,
ρ ≧0. 3.3.3. VAR Formulation For the autoregressive parameter we set 0 < α < 1, which implies persistence of unemployment,
but within the limits of stationarity. The disturbance ϵu,t contains all other variables that might
affect ut. For the ICM-case, the R and P coefficients associated with ret can be shown to be: l = 1 −θw ω ψqw (1 −φ)/χ,
k = (θq −θwψqw)/χ,
n = (µq + µw ψqw)/χ,
e = 1 −(ψqpi + ψqw ψwp (1 −φ))/χ,
= 0 if dynamic homogeneity
d = (mq θq + cq + (mw θw + cw) ψqw)/χ, where the denominator is: χ = 1 −ψqw(φψwp + ψwq) > 0. For wst the coefficients in R and P, in the
case of ICM, are: where the denominator is: χ = 1 −ψqw(φψwp + ψwq) > 0. For wst the coefficients in R and P, in the
case of ICM, are: λ = θw ω (1 −ψqw)(1 −φ)/χ,
κ = 1 −(θw (1 −ψqw) + θq (1 −ψwq −φ ψwp))/χ,
η = (µw (1 −ψqw) −µq (1 −ψwq −φ ψwp))/χ,
ξ = (ψwp (1 −ψqw)(1 −φ) −ψqpi (1 −ψwq −φ ψwp))/χ,
= 0 if dynamic homogeneity
δ = ((mw θw + cw)(1 −ψqw) −(mq θq + cq)(1 −ψwq −φ ψwp))/χ. λ = θw ω (1 −ψqw)(1 −φ)/χ,
κ = 1 −(θw (1 −ψqw) + θq (1 −ψwq −φ ψwp))/χ,
η = (µw (1 −ψqw) −µq (1 −ψwq −φ ψwp))/χ,
ξ = (ψwp (1 −ψqw)(1 −φ) −ψqpi (1 −ψwq −φ ψwp))/χ,
= 0 if
δ = ((mw θw + cw)(1 −ψqw) −(mq θq + cq)(1 −ψwq −φ ψwp))/χ. These coefficients are non-negative for reasonable values of the structural coefficients. The exception is
δ which can be both positive and negative, see Kolsrud and Nymoen (2014) [6]. These coefficients are non-negative for reasonable values of the structural coefficients. The exception is
δ which can be both positive and negative, see Kolsrud and Nymoen (2014) [6]. 3.3.5. Numerical Simulation of the Theory Model To round off the theoretical section, we simulate the theory model for a typical ICM calibration. This will demonstrate dynamic properties of the system, when it is dynamically stable. It is easiest to calibrate the structural form. In the system of equations below, (14) correspond to
(10), and (15) to (11) and so on: ∆qt = cc
q + 0.6 ∆wt + 0.2 ∆pit + 0.12 (w −q −a)t−1 + εq,t,
(14)
∆wt = cc
q + 0.15 ∆qt + 0.7 ∆pt −0.15 × 0.35 ut−1
(15)
−0.15(w −q −1 a)t−1 + 0.15 × 0.8 (pt−1 −qt−1) + εw,t,
ut = cc
q + 0.25ut−1 −0.25(pi −q)t−1 + εu,t
(16)
∆pt = 0.5 ∆qt + 0.5∆pit
(17)
pit = pit−1 + 0.04 + εu,t
(18)
at = at−1 + 0.02 + εa,t
(19) (14)
(15) (14) (
)
(15) (15) (16)
(17)
(18)
(19) (19) Note that θw has been set to 0.15 and θq to 0.12 in order to capture the essence of ICM, and given that,
we have set ς = ϕ = 0. The numbers chosen for the remaining intercepts cc
q and cc
w only influence the
means (if they exists) of the simulations. cc
u is calibrated with a large structural break in period 150, to
simulate a permanent change in the mean of u. In the simulations, the artificial time series of the observable economic variables qt, wt, ut, pt, pit
and at are generated by using the VAR representation of the complete system (14)–(19). We use the
computer generated data to estimate the simultaneous equations model with the use of FIML (a one-off
Monte Carlo exercise of estimation of the correctly specified econometric model). We then use dynamic
simulation of the estimated structural model and plot the solutions paths of the endogenous variables. The first row of Figure 10 shows the graphs of the rates of change in the wage and in the consumer
price index in the simulation. Wage and price growth are stable at the start of the period. The reduction
of nominal growth rates that appear in period 151 is caused by the structural break in the rate of
unemployment in period 150. Since that shift is permanent, the rate of unemployment does not return
to the initial low level, as shown in the graph for unemployment in the second row in the figure. 3.3.4. Conditions for Global Asymptotic Stability corrective mechanism may be deflationary (or inflationary) economic policies (but with the unavoidable
uncertainty about right timing and strength). 3.3.5. Numerical Simulation of the Theory Model 3.3.4. Conditions for Global Asymptotic Stability The condition is that the three eigenvalues of R have modulus inside the unit-circle. Even though
the theory model is kept simple and transparent, Kolsrud and Nymoen show that the general analytic
expressions for the eigenvalues of R are too large and complex to be of much help. However, simulation
of a numerical version of the theoretical model gives insight, as shown below. One thing that can be established from the general expression though, is that the dynamic
homogeneity restrictions: ψqw + ψqpi = 1
ψwq + ψwp = 1 ψwq + ψwp = 1 have no direct implication for dynamic stability. However if the system is stable, dynamic homogeneity
implies that the steady states of ret, wst and ut do not depend on the nominal growth rate. In the PCM case, θw = θq = 0, R simplifies to have no direct implication for dynamic stability. However if the system is stable, dynamic homogeneity
implies that the steady states of ret, wst and ut do not depend on the nominal growth rate. In the PCM case, θw = θq = 0, R simplifies to RPCM=
1
0
n
0
1
−η
−ρ
0
α
and this matrix has one unit root. This shows that the PCM version of the model is less “inherently
stable” than the ECM. Intuitively, if the price and wage equations become disconnected from the
profitability and real wage levels, other variables in the system must take over their role in the
stabilization. In our theory model, this can be achieved by changing the specification so that
ut adjusts to real-wages directly, and not only to the real exchange rate. In the real world, the and this matrix has one unit root. This shows that the PCM version of the model is less “inherently
stable” than the ECM. Intuitively, if the price and wage equations become disconnected from the
profitability and real wage levels, other variables in the system must take over their role in the
stabilization. In our theory model, this can be achieved by changing the specification so that
ut adjusts to real-wages directly, and not only to the real exchange rate. In the real world, the Econometrics 2017, 5, 6 26 of 54 corrective mechanism may be deflationary (or inflationary) economic policies (but with the unavoidable
uncertainty about right timing and strength). 3.3.5. Numerical Simulation of the Theory Model -.04
-.02
.00
.02
.04
.06
.08
.10
130
140
150
160
170
180
190
200
CPI growth rate
U
l
t
t -.04
.00
.04
.08
.12
130
140
150
160
170
180
190
200
Wage growth rate Wage growth rate Consumer real wage Consumer real wage Unemployment rate Unemployment rate 10.0
10.4
10.8
11.2
11.6
130
140
150
160
170
180
190
200
g 1.6
2.0
2.4
2.8
3.2
3.6
130
140
150
160
170
180
190
200 Figure 10. Solution paths for endogenous variables shown in graphs with dashed lines, together with
“actuals” (the computer generated time series) and 95 percent uncertainty intervals. The first solution
period is period 125 and the final period of the dynamic simulation is number 200. -11.6
-11.4
-11.2
-11.0
-10.8
-10.6
-10.4
130
140
150
160
170
180
190
200
Relative import price (re)
-2.60
-2.56
-2.52
-2.48
-2.44
-2.40
130
140
150
160
170
180
190
200
Wage share (ws)
10.0
10.4
10.8
11.2
11.6
130
140
150
160
170
180
190
200
Consumer real wage
1.6
2.0
2.4
2.8
3.2
3.6
130
140
150
160
170
180
190
200
Unemployment rate
Figure 11. Solution paths for endogenous variables shown in graphs with dashed lines, together with
“actuals” (the computer generated time series) and 95 percent uncertainty intervals. The first solution
period is period 125 and the final period of the dynamic simulation is number 200. -11.6
-11.4
-11.2
-11.0
-10.8
-10.6
-10.4
130
140
150
160
170
180
190
200
Relative import price (re)
Consumer real wage -2.60
-2.56
-2.52
-2.48
-2.44
-2.40
130
140
150
160
170
180
190
200
Wage share (ws)
Unemployment rate Wage share (ws) Relative import price (re) Consumer real wage Consumer real wage Unemployment rate Unemployment rate 1.6
2.0
2.4
2.8
3.2
3.6
130
140
150
160
170
180
190
200 10.0
10.4
10.8
11.2
11.6
130
140
150
160
170
180
190
200 Figure 11. Solution paths for endogenous variables shown in graphs with dashed lines, together with
“actuals” (the computer generated time series) and 95 percent uncertainty intervals. The first solution
period is period 125 and the final period of the dynamic simulation is number 200. 3.3.5. Numerical Simulation of the Theory Model However, the reductions in wage and price inflation are temporary, which is a typical trait of ICM,
in contrast to the accelerationist PCM. The consumer real wage, shown in the panel to the left in the
second row, is also affected by the increase in unemployment. This is because, in the theory model, it is
nominal wage growth which is directly hit by the increased unemployment. CPI inflation is reduced,
but as a reaction to the moderation in wage growth. Hence, the growth in the real consumer wage is
practically brought to a halt by the rise in unemployment, before it gradually “finds its way back” to
the secular productivity driven trend growth of the calibrated model. As the graph shows. Figure 11 shows the same graphs as Figure 10 in the second row, but in the first row, we have
the relative import price (the real-exchange rate re) and the wage-share. In the chosen calibration,
and abstracting from the break in ut, both these variables are stationary, as indicated by the solutions
for the period leading up to the increase in unemployment. After the break, the relative price of
imports is permanently increased towards a new equilibrium level (a permanent real depreciation). The wage share, even though it is markedly reduced as a consequence of the step-increase in ut, slowly
equilibrium corrects back to the pre-break mean. 27 of 54 Econometrics 2017, 5, 6 -.04
.00
.04
.08
.12
130
140
150
160
170
180
190
200
Wage growth rate
-.04
-.02
.00
.02
.04
.06
.08
.10
130
140
150
160
170
180
190
200
CPI growth rate
10.0
10.4
10.8
11.2
11.6
130
140
150
160
170
180
190
200
Consumer real wage
1.6
2.0
2.4
2.8
3.2
3.6
130
140
150
160
170
180
190
200
Unemployment rate
Figure 10. Solution paths for endogenous variables shown in graphs with dashed lines, together with
“actuals” (the computer generated time series) and 95 percent uncertainty intervals. The first solution
period is period 125 and the final period of the dynamic simulation is number 200. 3.3.5. Numerical Simulation of the Theory Model Hence, the economic interpretation of the dynamics becomes distinctly Keynesian: Even if the
unions have some (or a lot) of influence over the money-wage bargains they do not really control
the dynamics and the long run level of the real wage, Keynes (1936) [84] (p. 12). Because, in
that case we would expect the wage share to be lowered by the step-up in the unemployment
rate. In terms of econometrics, this property amount to endogenous co-breaking, Clements and
Hendry (1999) [85] (Chapter 9): there are “permanent large-shifts” in the real exchange rate and in
unemployment, but there is no break in the unconditional mean of the equally endogenous wage-share,
even if the nominal wage in particular is directly affected by unemployment. Econometrics 2017, 5, 6 28 of 54 Finally, we note that the nominal growth rates are stable (and the same) on each side of the break
in the unemployment mean. Clearly, in this model, a unique steady-state rate of unemployment does
not follow from constant wage and price inflation rates, as it will do in a natural rate PCM. Hence, the simulations demonstrate that the theoretical model has several properties that makes it
relevant for empirical analysis of long historical time series. For example, it can explain why post-war
unemployment could stay low and stable for decades without causing a wage explosion. The answer
does not lie primarily in the L-shaped wage and price curves, but in the parametrization that allows
wage growth to equilibrium correct to the lagged wage-share and profitability, at any going rate of
unemployment, not just a “natural rate” of unemployment. However, it should be remembered that by looking at only one calibration we may underestimate
the complex dynamics that the framework can generate. For example, Kolsrud and Nymoen (2014) [6]
show by analysis and simulations that the interplay between parameters can give rise to very different
dynamics, some with cycles, other with more smooth stabilization after a shock. This is also echoing
Sargan (1980) [71] (p. 108), who noted how “critically dependent” the dynamics of his wage and price
model were on the estimated coefficients in the wage and price equations. 4. Empirical Models In this section I discuss empirical models of wct (annual wage cost), qt (GDP Mainland-Norway
deflator), pt (consumer price index) and ut (unemployment rate). All these time series are in
logs. The same is the case for the two remaining variables at (labour productivity) and pit (import
price index). I first present models that are conditional on the at and pit. Treating at and pit as non-modelled
variables, is consistent with the assumptions of the theoretical model. However, it is of course
interesting to get empirical results also for closed systems, where at and pit are endogenous. We
present results for the extended (or “closed”) multi-equation model at the end of the section. 4.1. Robust Estimation with Structural Breaks The theory model was formulated with constant parameters, with one exception, the location
shift in the unemployment series. As the descriptive section showed, the actual time series contain
several non-stationarities and breaks. An empirical model must represent also those aspects of the
data, if only to avoid serious residual mis-specification that would pull the rug from under inference
about parameter significance. When a model contains indicator variables for intermittent breaks,
the parameters that are in the focus of the research purpose may be more robustly estimated with
those indicators included in the model, than without them. Finally, unless all indicator variables
appear in all equations of the model, there may a pattern implying that the parameters of one part
of the model are invariant to the breaks in another part of the model. Hence, there may be a partial
structure in the model, in the form of invariance of (at least) a sub-set of focus parameters, to some of
the structural breaks. For example, consider estimates of the parameters in the VAR recursion matrix R, and derived
parameters like the eigenvalues which are important for the question of dynamic stability. In the
case of no-breaks and no unit-root stationary, OLS, equation-by-equation, also gives conditional FIML
estimators, with well known asymptotic properties. However, if there are location shifts, inference
based on the OLS estimators are unreliable, because of non-normal disturbances for instance. In that
case, OLS with impulse indicator saturation, which we can refer to as OLS-IIS, is a more robust
estimator, Johansen and Nielsen (2009) [86]. One way of testing the null hypothesis of no outliers, or location shifts, is by impulse-indicator
saturation (IIS). We then create a complete set of impulse indicator variables,
n
1{j=t}
o
= 1 when j = t
and 0 otherwise for j = 1, . . . , T, and add T (in our case 113) indicators and estimate a “saturated” VAR
on the 1904–2015 sample. The deadlock created by creating more parameters than observations, is Econometrics 2017, 5, 6 29 of 54 elegantly resolved by the feasible split-sample IIS algorithm, see Hendry et al. (2008) [87], Johansen
and Nielsen (2009) [86], and implemented in Autometrics by Doornik (2009) [88]. elegantly resolved by the feasible split-sample IIS algorithm, see Hendry et al. (2008) [87], Johansen
and Nielsen (2009) [86], and implemented in Autometrics by Doornik (2009) [88]. 27
Since at is included as unrestricted, R has four columns, against three in the theoretical section. The VAR includes an
unrestricted constant and a restricted trend. The H0 of no cointegration is rejected at the 1 percent level using the trace-test
and the critical values in Table 13 in Doornik (2003) [92]. Also the hypothesis of one cointegrating vector against the
alternative of two can be formally rejected. Accepting two long-run relationships, the joint hypothesis that the trend has
zero coefficients in both co-integration relationships cannot be rejected at the 1 percent level. 4.1. Robust Estimation with Structural Breaks IIS has correct null retention frequency in constant conditional models for a chosen nominal
test size. The alternative retention-frequency function has appropriate probabilities of retaining
location shifts. Step-dummies are relevant for persistent breaks in location parameters. In Autometrics
the extension of IIS to step-indicator saturation (SIS) by adding a complete set of step-indicators is
integrated with the other algorithms. As a description of the data generating process of wages in Norway over a long period of time,
an empirical model with indicators or step-dummies is incomplete, as the causes of the shifts in model
parameters are left unexplained. A more satisfactory model would include the reasons for the shifts. In the empirical modelling Section 4.3, we investigate whether a theory of non-linear response of
wage adjustments to inflation can lead to removal of dummies that are found to be of importance in a
linear-in-parameters wage equation model. However, knowledge of the factors responsible for parameter instability is not essential in order
to study the impact of the resulting breaks on the properties of the model. For example, we will be
very interested in whether our empirical models have “extended constancy”: When an indicator is
added, the model is enlarged by a deterministic variable which is a mere proxy for factors that are part
of the true data generating process. However, a crucial index of the constancy of the model is that all
the previous parameters retain their original values, if the extension is done with a proxy for the true
economic variables. This is an essential feature of partial structure in an empirical model, as well as of
co-breaking, Nymoen (2002) [89], Clements and Hendry (1999) [85] (p. 260). Clearly, it is also possible to achieve progress over time, since shifts that are represented by
dummies at one point in time (by one researcher) can later be replaced by variables that represent the
underlying factors, thus creating constancy by model extension, see Eitrheim et al. (2002) [90]. 4.2. Reduced form VAR and Long-Run Wage Equation As a first approach, and to illustrate IIS for our data, I attempt a direct estimation of the VAR where
the model of nominal wage and price adjustment have been boiled down to the three real variables:
The producer real wage , the real exchange rate (relative price of imports) and the unemployment
rate. In order to test the restriction that implies the wage-share formulation (ι = 1 in the formal model
above), the level of labour productivity is included as a conditioning variable. For the sample 1904–2015, Autometrics (with overall significance level 0.001) returns 13 impulse
indicator variables: {I:1911, I:1915, I:1916,I:1918 ,I:1919, I:1921, I:1924, I:1941, I:1945, I:1948, I:1951,
I:1960, I:1970}. Most of them are from the volatile and unsettled years early in the sample. An intended
consequence is to remove some of the non-normality that would otherwise be apparent in the residuals. On the other hand, with a strict significance level, the chance is large that some large residuals during
the last decades of the sample pass undetected because they are dominated by the even larger ones
early in the sample. The largest eigenvalue (which is also a squared correlation coefficient) of bRIIS −I, is 0.56, and
the trace statistic used to test the hypothesis of no cointegration is significant, based on the critical
values in Doornik (1998, 2003) [91,92].27 If we accept cointegration, bRIIS −I can be written as bα bβ′
where the columns of bβ contains the cointegration vectors. Because the long-run price equation has 30 of 54 Econometrics 2017, 5, 6 Econometrics 2017, 5, 6 Econometrics 2017, 5, 6 been substituted in the derivation of the VAR, only the long-run wage-relationship can be identified by
this analysis. However, it is interesting that the first column of bβ, namely been substituted in the derivation of the VAR, only the long-run wage-relationship can be identified by
this analysis. However, it is interesting that the first column of bβ, namely wc −q
−1
re
0
u
−0.07
a
0.94 is interpretable as an estimated long-run wage-equation (5), with ˆκ ≈1, and ˆϖ ≈−0.1. and with
a no-wedge restriction imposed (ω = 0), which is necessary for identification relative to a second
cointegration relationship. is interpretable as an estimated long-run wage-equation (5), with ˆκ ≈1, and ˆϖ ≈−0.1. and with
a no-wedge restriction imposed (ω = 0), which is necessary for identification relative to a second
cointegration relationship. 4.3. Single-Equation Modelling As noted above there is a wage-modelling literature in Norway that has estimated wage equations
on post WW-II data. Several of these modelling results have resulted in models with steady-state
relationships with coefficients that look like the cointegration parameters just estimated. Hence,
it seems worth trying to formulate an empirical single-equation model for wage cost changes, ∆wct,
and with (wc −q −a)t−1 and ut−1 as explanatory variables. I first present a conditional model, and use Autometrics (again with significance level 0.001) to
obtain IIS-estimates of the parameters. The resulting model is found in the OLS column of Table 3. This is an equilibrium wage equation which corresponds to the theoretical Equation (11), the main
difference is the extra lags ∆pt−1 and ∆wct−1 and the variable taxt representing the change in the
payroll tax-rate (tax = ln(1+payroll tax rate)). The estimated version of the steady-state Equation (4)
based on the IIS-OLS results becomes rwb = ˆc + 0 (pt −qt) + 1 at −0.125
(0.04), ut
(20) (20) where ˆc represents an estimated intercept. Figure 12 shows the thirteen indicator variables visualised,
using the estimated coefficients as weights, by the bar denoted wbreakt, together with the line graph
for ∆wct. There are five indicators between 1905 and 1925, three between 1941 and 1947, and the last
group is representing effects of the incomes policies of the 1970s and 1980s. ∆wc fitted
wbreak
1910
1920
1930
1940
1950
1960
1970
1980
1990
2000
2010
-0.2
-0.1
0.0
0.1
0.2
0.3
∆wc fitted
wbreak
Figure 12. Fitted values from the estimated wage equation together with sequence of breaks. Figure 12. Fitted values from the estimated wage equation together with sequence of breaks. 31 of 54 Econometrics 2017, 5, 6 Clearly, if we look back at the theory model, the OLS estimators contain a simultaneity bias unless
the short run dynamics are restricted in a way that gives a recursive structure. The table therefore
also includes a column with generalised instrumental variables estimation (GIVE), treating ∆pt and
∆ut as endogenous, and using the instruments listed at the bottom of the table. The Sargan IV-test is
insignificant and supports the validity of the instruments, see Sargan (1964) [69] and Davidson and
MacKinnon (2004) [93] (Chapter 8.6). 4.3. Single-Equation Modelling Given this result, it follows that
the accelerationist wage PCM is even more firmly rejected 29. Re-writing the estimated equation with real wage change (∆wct −∆pt) on the left hand side gives,
for the OLS estimates: (21) ∆wct −∆pt = −0.47
(0.03)∆2pt −0.15
(0.03)∆pt−1 + 0.11
(0.04)(∆wct−1 −∆pt−1) + .... (21) Since this is a re-parametrization that only affects the wage and price change coefficients, the rest of
the model specification is unaffected. (21) gives some intuition for why the main difference between
the OLS and IV results in Table 3 is that the coefficient of ∆pt is reduced and has a higher standard
error: Finding strong instruments for price acceleration ∆2pt is not easy, and weak instruments drive
up the residual standard error, and may induce notable finite sample bias in IV estimates. In terms of economic interpretation, the re-parametrization in (21) is also useful, since it isolates
the dynamic inhomogeneity parameter of the wage equation as −0.15. Although this coefficient is
quite small numerically, it is large enough for the null hypothesis of dynamic homogeneity to be
rejected by a solid margin (the t-value is −4.8) 30. Another 1-for-1 re-parametrization is: ∆wct −∆pt = −0.15
(0.03)∆pt −0.32
(0.04)∆2pt + 0.11
(0.04)(∆wct−1 −∆pt−1) + ...,
(22) (22) where both the inflation and the acceleration terms enter contemporaneously. Castle and Hendry
(2014) [96] reports a model for UK wages estimated on data for the long historical period 1864–2004,
which has several traits in common with our empirical wage equation 31. Their left hand side variable
is the change in the real wage, like in our re-parameterised equation. Castle and Hendry included a
lagged wage-share (wc −q −a), the unemployment percentage, and indicator variables. They also
included inflation and price acceleration (lagged) in their real wage growth equation. An interesting feature of the UK-wage equation in Castle and Hendry (2014) [96] is the inclusion
of non-linear terms in the unemployment percentage and in inflation. Building on their earlier work,
(in Castle and Hendry (2009) [97]), they present a theory of a non-linear, U-shaped, mapping from
inflation to real wage change: workers become more attentive when price inflation rises, and act to
prevent further erosion of their real wages, whereas employers may find it possible to cut nominal
wages when prices fall. 4.3. Single-Equation Modelling The differences between the two estimated wage equations are not large, and both sets of
estimates are similar to those that have been estimated on data from the 1960s and later, for
example, Nymoen (1989) [50] and Johansen (1995) [94]. Even though most of those models were
for manufacturing wages, the basic structure is the same as we see in Table 3. In particular we get a
healthy estimate of the coefficient of the rise in cost-of-living, and profitability plays a role in the form
of the lagged w −q −a term. Finally, the unemployment variable is significant, and the wage-curve is
L-shaped, even if more convex functional forms could have fitted that point better. Table 3. Empirical versions of the wage equation in (11) in the theory model. ˆσ denotes the
estimated residual standard error. The standard test for residual mis-specification (autocorrelation,
heteroskedasticity and non-normality) are insignificant in the two models reported. Table 3. Empirical versions of the wage equation in (11) in the theory model. ˆσ denotes the
estimated residual standard error. The standard test for residual mis-specification (autocorrelation,
heteroskedasticity and non-normality) are insignificant in the two models reported. OLS
GIVE
∆pt
0.53
0.44
(0.03)
(0.11)
∆ut
−0.01
−0.05
(0.005)
(0.03)
∆at
0.25
0.17
(0.03)
(0.074)
∆pt−1
0.21
0.27
(0.04)
(0.04)
∆tax
0.89
0.94
(0.26)
(0.26)
∆wct−1
0.11
0.15
(0.04)
(0.04)
(wc −q −a)t−1
−0.08
−0.06
(0.02)
(0.03)
ut−1
−0.01
−0.01
(0.002)
(0.003)
Constant
0.56
0.47
(0.14)
(0.19)
wbreakt
0.98
0.96
(0.06)
(0.08)
Sample: 1905–2015
ˆσ100
1.38
1.72
Sargan IV-test
χ2(3) = 2.56
∆at−1,∆qt−1, ∆qt−1, ∆ut−1,∆nht OLS
GIVE
∆pt
0.53
0.44
(0.03)
(0.11)
∆ut
−0.01
−0.05
(0.005)
(0.03)
∆at
0.25
0.17
(0.03)
(0.074)
∆pt−1
0.21
0.27
(0.04)
(0.04)
∆tax
0.89
0.94
(0.26)
(0.26)
∆wct−1
0.11
0.15
(0.04)
(0.04)
(wc −q −a)t−1
−0.08
−0.06
(0.02)
(0.03)
ut−1
−0.01
−0.01
(0.002)
(0.003)
Constant
0.56
0.47
(0.14)
(0.19)
wbreakt
0.98
0.96
(0.06)
(0.08)
Sample: 1905–2015
ˆσ100
1.38
1.72
Sargan IV-test
χ2(3) = 2.56
∆at−1,∆qt−1, ∆qt−1, ∆ut−1,∆nht Econometrics 2017, 5, 6 32 of 54 As Sargan was early to point out, finding that the lagged real wage is significant when added to a
standard wage Phillips curve, is enough to reject models that exclude that variable, see Ericsson et al. (2001) [95]. In Table 3 the null hypothesis of θw = 0 (which would imply a wage Phillips curve) is
rejected at the 5% significance level for both estimation methods 28. χ ( )
[
]
29
That is, the joint hypothesis of θw = 0 and the coefficients of the three nominal growth rates summing to one.
30
C
di
ψ
ψ
1 i
h
h
i
l
d l 30
Corresponding to ψwq + ψwp = 1 in the theoretical model.
31 28
In the IV column the Likelihood-ratio test statistic for the restriction θw = 0 becomes χ2(1) = 5.59[0.018] with p-value i
brackets. The OLS results are: χ2(1) = 14.31[0.0002].
29
That is the joint hypothesis of θ
= 0 and the coefficients of the three nominal growth rates summing to one 28
In the IV column the Likelihood-ratio test statistic for the restriction θw = 0 becomes χ2(1) = 5.59[0.018] with p-value i
brackets. The OLS results are: χ2(1) = 14.31[0.0002].
29
That is, the joint hypothesis of θw = 0 and the coefficients of the three nominal growth rates summing to one.
30
Corresponding to ψwq + ψwp = 1 in the theoretical model. p
g
ψ q
31
Their Equation (21). ,
j
yp
w
g
g
30
Corresponding to ψwq + ψwp = 1 in the theoretical model.
31 9
That is, the joint hypothesis of θw = 0 and the coefficients of the three nominal growth rates summing to one.
0
Corresponding to ψwq + ψwp = 1 in the theoretical model.
1 χ ( )
[
]
29
That is, the joint hypothesis of θw = 0 and the coefficient 4.3. Single-Equation Modelling Such behaviour can induce unstable wage-price inflation, whereas the
wage-price adjustments in the theory model above implied stable steady-states for both ∆wct and ∆pt,
and is relevant to test for our data as well. Castle and Hendry (2014) [96] showed that the U-shaped function just mentioned can be
approximated by the polynomial in the inflation rate: α1∆pt + α2(∆pt)3 + α3(∆pt)4. 33 of 54 Econometrics 2017, 5, 6 Of course, the linear part of this approximation is already present in the model, but the two non-linear
terms are not. One possibility is that the indicator dummies are substitutes for them in periods with
deflation or very high inflation. To test that possibility, (∆pt)3 and (∆pt)4 were included as regressors before Autometrics was
used to select over those two terms as well as over the full set of indicator variables again. In the
resulting model, (∆pt)4 was retained with a negative coefficient, but not (∆pt)3. The indicator variables
for 1915, 1922 and 1925 were removed, indicating that the wage increases in this year may have been
driven by non-linear responses. Nevertheless, six indicators in the period from 1904 to 1945 were
retained, so IIS does not “substitute” for included non-linear terms when they matter. In the following, the linear specification in Table 3 will be grafted into a system of wage-price
dynamics. There may be other nonlinearities lurking in the data but we leave the development of
non-linear wage equations, and their integration in a system, to later work. 4.4. Multiple-Equation Modelling The estimated models in Table 3 are empirical versions of the theoretical wage Equation (11),
and which was calibrated in (15), but it clearly needs to be embedded in a multi-equation model in
order to account for the many plausible dependencies between the variables wct, qt, pt and ut. 32
As long as the covariance matrix of the disturbances (which are omitted in the notation for simplicity) is unrestricted. 4.4.2. Reduced Form Estimates Reduced form estimation results for the vector (∆wct, ∆qt, ∆pt, ∆ut) are shown in Table 4. In this
system, which can be referred to as a contegrated VAR with exogenous variables, also called VAR-EX,
we allow one lag in each endogenous variable, but we only include the contemporaneous growth
rates of productivity and import prices. ∆pmt and ∆at. In addition, there are four composite indicator
variables: wbreakt, and three new ones for ∆qt, ∆pt and ∆ut. Finally, I have included constants,
the change in the payroll tax rate (∆tax) and the lagged levels variables consistent with (25) and (26). All in all this gives 17 parameters in each line of the VAR-EX (68 in all) to be estimated from the 111
observations in the sample from 1905 to 2015. Table 4. Reduced form estimates (cointegrated VAR-EX). Sample 1905–2015. ∆wt
∆qt
∆pt
∆ut
ˆσ100 †
1.48
2.04
1.56
0.22
w-p ecm ‡ χ2(2)
5.5 *
30.8 **
17.8 *
1.1
Breaks χ2(4)
207 **
254 **
340 **
34 **
Vector mis-spesification tests:
FAR(33, 307)
1.4
χ2
N(8)
11.5
FHet(128, 301)
1.8 **
† except for ∆utwhere ˆσ2 is reported; ‡ Test of exclusion of (w −a −q)t−1and (q −0.7(w −q) −0.3pi)t−1. * and ** indicate significance levels of 10 % and 5 % respectively. Table 4. Reduced form estimates (cointegrated VAR-EX). Sample 1905–2015. † except for ∆utwhere ˆσ2 is reported; ‡ Test of exclusion of (w −a −q)t−1and (q −0.7(w −q) −0.3pi)t−1. * and ** indicate significance levels of 10 % and 5 % respectively. We note that the residual percentage standard deviation for ∆wt, 1.48, is a little higher than in the
single equation analysis, which is natural since we no longer condition on ∆pt and ∆ut. For the other
nominal variables we can note that the unexplained variation is larger for the GDP deflator than for
the consumer price index, which seems reasonable. The second row in the table contains test statistics for the joint significance of the two equilibrium
correction terms (w −a −q)t−1 and (q −0.7(w −q) −0.3pi)t−1. They are jointly significant in all
equations, with the exception of the change in the unemployment rate, implying that a Phillips curve
wage-price system is rejected by this test. We also give the test for joint significance of the four, in the
outset, “variable specific” composite indicator variables. 4.4.1. Identificaton of Wage and Price The wage equation specification with ω = 0, is helpful for identification of the long-run
(cointegration) relationships. Formally, setting q f
t −q = 0 in (6) and wt −wb
t = 0 in (7) give the
theoretical long-run equations as two simultaneous equations (23)
(24) (23)
(24) w −q = a + ϑ u −mq
(23)
w −q = −ιa −ω(1 −φ)re + ϖ u −mw
(24) (24) rank and order condition therefore applies, Hsiao (1997) [98] 32. and rank and order condition therefore applies, Hsiao (1997) [98] 32. Formulating the original theory model with ω = 0 (no wedge) and ϑ = 0 (horizontal price curve),
does not identify the wage equation. But if we extend the theory model to allow the price mark up mq
to depend on the import price (pi), then ω = ϑ = 0 imply identification by the order condition. In this
section I estimate an empirical model consistent with the extended theory model. A consequence of
this approach is that we have more difficulty arguing theoretically that we have identified a producer
price equation, as different from a consumer price equation. That issue is however secondary to the
benefit of having strong identification of a wage equation, given our focus on wage setting over a long
historical period. p
Hence, in the following, the framework for the long-run price and wage setting equations become w = q + at −ϑ u + mw
(25)
q = κ(w −a) + (1 −κ)pi + m
′
q
(26) (25) (25)
(26) (26) where we provide new estimates of ϑ from the multiple-equation modelling, and where we fix the
parameter κ at 0.7, based on empirical ratios from the last four decades. where we provide new estimates of ϑ from the multiple-equation modelling, and where we fix the
parameter κ at 0.7, based on empirical ratios from the last four decades. 32
As long as the covariance matrix of the disturbances (which are omitted in the notation for simplicity) is unrestricted. 34 of 54 Econometrics 2017, 5, 6 34 of 54 4.4.2. Reduced Form Estimates The result of the test gives clear indication
that these variables are needed for constancy of the other parameters of the model, at least until a more
more substantial extension of the model can be achieved. The last part of the table contains the values of vector versions of tests for autoregressive residual
autocorrelation, non-normality and heteroskedasticity due to squares of regressors, Doornik and
Hendry (2013) [99] (Chapter 11). Even if the third test rejects, indicating remaining heteroskedastcity
(for example omitted non-linear terms in the system), the insignificance of the tests of autocorrelation
and non-normality is encouraging, given the obvious non-normality that characterizes the time series. 4.4.3. FIML Estimates of Structural Model 4.4.3. FIML Estimates of Structural Model Econometrics 2017, 5, 6 35 of 54
1
0
0.03
−0.5
−0.36
1
0
0
0
0
1
−0.98
0
−0.33
0
1
∆c
wct
∆bqt
∆but
∆bpt
=
0.24
0
0.23
0.13
0
−0.36
0.34
0.1
0
0
0
−0.98
0
0
0.3
0
0.16
0.32
0
0
∆at
∆pit
∆pt−1
∆wct−1
∆ut−1
+
−0.08
−0.01
0
0
0
0
0
−0.14
0
0
0
−0.37
0
−0.25
−0.26
0.04
0
0
−0.02
0.03
(wc −a −q)t−1
ut−1
qt−1 −0.7(w −a)t−1 −0.3pit−1
(w −p −a)t−1
(pi −p)t−1
+
0.91
0
0
0
0
0.96
0
0
0.95
1.1
wbreakt
qbreakt
ubreakt
pbreakt
(27)
LR test of over-identifying restrictions:
χ2(37) = 57.645[0.02]
1
0
0.03
−0.5
−0.36
1
0
0
0
0
1
−0.98
0
−0.33
0
1
∆c
wct
∆bqt
∆but
∆bpt
=
0.24
0
0.23
0.13
0
−0.36
0.34
0.1
0
0
0
−0.98
0
0
0.3
0
0.16
0.32
0
0
∆at
∆pit
∆pt−1
∆wct−1
∆ut−1
1
0
0.03
−0.5
−0.36
1
0
0
0
0
1
−0.98
0
−0.33
0
1
∆c
wct
∆bqt
∆but
∆bpt
=
0.24
0
0.23
0.13
0
−0.36
0.34
0.1
0
0
0
−0.98
0
0
0.3
0
0.16
0.32
0
0
∆at
∆pit
∆pt−1
∆wct−1
∆ut−1
+
−0.08
−0.01
0
0
0
0
0
−0.14
0
0
0
−0.37
0
−0.25
−0.26
0.04
0
0
−0.02
0.03
(wc −a −q)t−1
ut−1
qt−1 −0.7(w −a)t−1 −0.3pit−1
(w −p −a)t−1
(pi −p)t−1
+
0.91
0
0
0
0
0.96
0
0
0.95
1.1
wbreakt
qbreakt
ubreakt
pbreakt
(27)
LR test of over-identifying restrictions:
χ2(37) = 57.645[0.02] (27) χ2(37) = 57.645[0.02] Given that the VAR-EX is accepted as not seriously misspecified, the most important specification
test of the model in (27) is the log-likelihood test of the over-identifying restriction, shown at the
bottom of the display. 4.4.3. FIML Estimates of Structural Model An empirical model of the VAR-EX is reported, in compact form, in display (27). The estimation
method is Full Information Maximum Likelihood (FIML). To save space, the constant terms and
the ∆tax variable, which only enters the wage equation, have been omitted. For the same reason,
standard errors of the coefficient estimates are not shown. However, all individual “t-tests” would be
significantly different from zero at the usual significance level. 4.4.3. FIML Estimates of Structural Model y
Finally, for the change in the log of the consumer price index, we get: Finally, for the change in the log of the consumer price index, we get: −0.33∆qt + ∆pt = 0.16∆pit + 0.32∆pt−1 + 0.04(wc −a −q)t−1
−0.02 (w −p −a)t−1 + 0.03 (pi −p)t−1 + 1.11pbreak
(31) (31) which, as noted, is less well identified than the other equations (being more confluent), but it can play
a role as a link we need in order to represent the important distinction between the consumer real
wage and the producer real wage. Table 5 shows the empirical residual standard errors for the wage and price equations of the FIML
estimated model, together with the test for the Phillips-curve restrictions, which again are statistically
significant at very low level of significance. Table 5. Estimated residual standard errors, and tests of θw = 0 and θq = 0 in the structural wage and
price equations. Table 5. Estimated residual standard errors, and tests of θw = 0 and θq = 0 in the structural wage and
price equations. ∆wt
∆qt
ˆσ2100
1.43
1.90
W-P ECM ‡ χ2(1)
16.1 **
33.2 **
‡ Test for excluding (w −a −q)t−1in ∆wct and (q −0.7(w −q) + 0.3pi)t−1in ∆qt. ** indicate significance
levels of 10% and 5% respectively 4.4.3. FIML Estimates of Structural Model If the p-value of this test is found to be acceptable, the interpretation is that the
structural model, with 37 over-identifying restrictions, is explaining “as much” of the variation in the
data as the reduced form VAR-EX. As we see, the p-value is relatively low, and the test rejects at the 5%
and 2.5% levels. However, on the other hand, since the VAR EX is fitting the data closely, a p-value of
0.02 may be a price worth paying for an interpretable and parsimonious model of the system. y
p
p y
g
p
p
y
By multiplying out in (27) we get the FIML-estimated version of the wage equation: ∆wct + 0.03∆ut −0.5∆pt = 0.24∆at + 0.23∆pt−1 + 0.13∆wct−1
−0.08(wc −a −q)t−1 −0.01ut−1 + 0.91wbreak,
(28) (28) which is close to the single equation modelling results, the IV version in particular. From the second
row of the matrices and vectors we get, for the change in the GDP deflator: −0.36∆wct + ∆qt = −0.36∆at + 0.34∆pit
−0.14 [qt−1 −0.7 {(w −a) −0.3pi}]t−1 + 0.91qbreak,
(29) (29) showing that the change in unit labour cost, (∆wc −∆a)t has estimated elasticity 0.36, and the change
in import price gets an estimated elasticity of 0.34. Hence, the estimated short run elasticities are
smaller for wages, and higher for import prices, than the corresponding long-run elasticities, without
any inconsistency arising as a consequence of this. The estimated equation for ∆ut is ∆ut −0.98∆pt = −0.98∆pit + 0.3∆ut−1 −0.37ut−1
−0.25 (w −p −a)t−1 −0.26 (pi −p)t−1 + 0.95ubreak,
(30) (30) which shows that high CPI-inflation relative to the price change of imports (worsening of price
competitiveness) is increasing unemployment according to this empirical model. The lagged level
variables also has interpretable signs: A high real exchange rate is pushing unemployment up. Econometrics 2017, 5, 6 36 of 54 However, a high consumer real wage relative to productivity is predicting lower unemployment
in this model. Taken together, these estimation results indicate that it is the balance between the
aggregate demand determining consumer real wage, productivity, and cost-competitiveness which is
important for keeping unemployment stable (and low). It is not too much to say that maintenance of
the right balance has at times been extremely difficult during the post war period, in Norway like in
many other countries. 4.5. Simulation and Closing of the Model In order to solve the model by dynamic simulation, the structure in (27) has been supplemented
by identities that link the changes (∆wct, ∆qt, ∆pt, ∆ut) to the corresponding lagged level variables. The dynamic simulations for a chosen set of variables are shown in Figure 13. The solutions are
conditional on the initial values of all endogenous variables in 1904 and 1903, and the actual values of
(∆pit, ∆at, wbreakt, qbreakt, ubreakt, pbreakt) from 1905 to 2015. The first row in the figure contains panels with the solutions for ∆wct and ∆qt together with the
actuals. The second row shows ∆pt to the left and then the relative import price (pi −p)t. The third
row contains the level of unemployment in the first column and the consumer real wage in the second
column. Among the level variables, the trend in the consumer real wage is well captured by the model. The two other level variables are without a clear deterministic (or “global”) trend, but there is more or
less continuous fall in the relative import price for most of the post WW-II epoch. Figure 14 plots the solutions for wage growth and the unemployment rate together with their
“within-equation” indicator variables. The indicators in the unemployment equation take the form
of a step-indicator function and implies shifts in the mean of unemployment: First upward, starting
with the beginning of par-policy deflation in 1920, and continuing to the end of the occupation by the
nazis, and then downward, during the period from 1950 to 1982. Wage growth seems, on average, to
be higher in periods when there are breaks that pull unemployment down, than in other epochs. This
may mean that wage inflation after WW-II has been in part a consequence of the “new normality” of
full employment. However, this is not, by itself, evidence that the unions were taking advance of the
situation by increasing the wage mark-up. To conclude in that way, we must find empirically that
there are non-constancies in the wage equation in the same period that there are breaks in the equation
for ut. I return to this question in Section 4.6. 4.5. Simulation and Closing of the Model 37 of 54 37 of 54 Econometrics 2017, 5, 6 ∆wct
Simulated
1920
1940
1960
1980
2000
2020
0.00
0.25
∆wct
Simulated
∆qt
Simulated
1920
1940
1960
1980
2000
2020
0.00
0.25
∆qt
Simulated
∆pt
Simulated
1920
1940
1960
1980
2000
2020
0.0
0.2
0.4
∆pt
Simulated
(pi−p)t
Simulated
1920
1940
1960
1980
2000
2020
0.0
0.5
1.0
(pi−p)t
Simulated
ut
Simulated
1920
1940
1960
1980
2000
2020
0
1
2
ut
Simulated
(w−p)t
Simulated
1920
1940
1960
1980
2000
2020
11
12
13
(w−p)t
Simulated
Figure 13. Dynamic simulation of the FIML estimated model in (27), conditional on observed time
series of at and pit. Figure 13. Dynamic simulation of the FIML estimated model in (27), conditional on observed time
series of at and pit. ∆wc simulated
wbreak
1910
1920
1930
1940
1950
1960
1970
1980
1990
2000
2010
-0.2
0.0
0.2
∆wc simulated
wbreak
u simulated
ubreak
1910
1920
1930
1940
1950
1960
1970
1980
1990
2000
2010
0
1
2
u simulated
ubreak
Figure 14. The solutions for wage growth (∆wc) and unemployment (u) shown together with their
(within equation) indicator variables. ∆wc simulated
wbreak
1910
1920
1930
1940
1950
1960
1970
1980
1990
2000
2010
-0.2
0.0
0.2
∆wc simulated
wbreak u simulated
ubreak
1910
1920
1930
1940
1950
1960
1970
1980
1990
2000
2010
0
1
2
u simulated
ubreak Figure 14. The solutions for wage growth (∆wc) and unemployment (u) shown together with their
(within equation) indicator variables. All multiple-equation results so far have been conditional on the observed data for labour
productivity at and import prices pit. In order to endogenise all the random variables in the model,
the model was extended by specification of marginal equations for ∆at and ∆pit. Compared to the
corresponding parameter estimates in the model that conditioned on at and pit, the FIML estimates of
the extended multiple-equation model changed very little. 38 of 54 Econometrics 2017, 5, 6 Figure 15 shows the dynamic simulation results for this model. The solutions shown in
the simulations are now only conditioned by the initialization years (1903 and 1904) and by the
development of the composite indicator variables over the solution period 1905–2015. Unlike the first
dynamic simulation, the values for labour productivity and import prices are solved by simulation,
together with all the other endogenous variables of the model to include productivity and import price
as endogenous. 4.5. Simulation and Closing of the Model ∆wc
Simulated
1920
1940
1960
1980
2000
2020
0.00
0.25
∆wc
Simulated
∆q
Simulated
1920
1940
1960
1980
2000
2020
0.00
0.25
∆q
Simulated
∆a
Simulated
1920
1940
1960
1980
2000
2020
-0.2
0.0
0.2
∆a
Simulated
∆pi
Simulated
1920
1940
1960
1980
2000
2020
-0.25
0.00
0.25
0.50
∆pi
Simulated
pi−p
Simulated
1920
1940
1960
1980
2000
2020
0.0
0.5
1.0
pi−p
Simulated
u
Simulated
1920
1940
1960
1980
2000
2020
0
1
2
u
Simulated
Figure 15. Dynamic simulation of the Full Information Maximum Likelihood (FIML) estimated model
in (27), extended by equations for ∆at and ∆pit, and actuals. Simulation starts in 1905. Figure 15. Dynamic simulation of the Full Information Maximum Likelihood (FIML) estimated model
in (27), extended by equations for ∆at and ∆pit, and actuals. Simulation starts in 1905. The graphs with the solutions for the two new endogenous variables are shown in the second
row of the figure. Since labour productivity acts as a fundamental trend in the whole system, it is
important that the model gets the rate of change correct, on average. Otherwise the solutions paths for
the real wage for example, will substantially overshoot or undershoot the historical graph. The graph
for ∆at shows that the model solution tracks this variable quite well. Thus, as a direct comparison also
shows, the simulation results for the focus variables, in Figure 15 represented by ∆wct, ∆qt, (pi −p)t
and ut, are not much affected by the extension of the model. The secular trend in Norwegian consumer real wage growth (see Figure 8) is shown again in
Figure 16, but now together with the simulated 15-year change from the multiple-equation model with
endogenous productivity and import price index. The degree of co-movement between actual and
model simulated real wage development is quite high. Hence, the empirical model is capable of explaining secular real wage growth in very different
historical epochs: The par policy driven deflation of the 1920s, the 1960s and 1970s with both high
annual wage increases and high inflation, and the last epoch with low and falling inflation but
comparatively generous wage increases. 39 of 54 Econometrics 2017, 5, 6 Actuals
Simulated
1900
1910
1920
1930
1940
1950
1960
1970
1980
1990
2000
2010
0
20
40
60
80
100
Actuals
Simulated
Figure 16. 4.5. Simulation and Closing of the Model 15 year real wage growth percentages, actuals and simulated from FIML estimated
multiple-equation model, with at and pit as additional endogenous variables in the model. Figure 16. 15 year real wage growth percentages, actuals and simulated from FIML estimated
multiple-equation model, with at and pit as additional endogenous variables in the model. 4.6. Constancy and Invariance of the Coefficients of the Wage Equation, Lucas Citique 4.6. Constancy and Invariance of the Coefficients of the Wage Equation, Lucas Citique It is more plausible that coefficients Econometrics 2017, 5, 6 Econometrics 2017, 5, 6 40 of 54 40 of 54 and eventual bias-parameters are both constant. The combination of constant conditional models
and structural breaks in the marginal models for the regressors is the hallmark of super-exogeneity,
Engle et al. (1983) [101]. Figure 17 shows graphs for recursive estimation of coefficients and stability tests. In order to
simplify the specification, I have utilised that the estimated coefficients of ∆ut and ut−1 in Table 3 are
so close in magnitude that they can be substituted by ut. This also allows 1904 to be the start of the
sample period. The panels in the first three rows show recursive coefficient estimates ±2 estimated
standard errors. The start of the graphs shows estimation results for the 1904–1925 sample, and the
end of graphs are the full sample estimates. The graphs in the three panels in the fourth row shows
the sequence of 1-step residuals (i.e., yt −xt ˆβt in a common notation) bordered by 0 ± 2ˆσt (to the
left); the sequence of 1-step ahead Chow tests (middle) and the sequence of break-point tests (right),
Doornik and Hendry (2009) [102] (pp. 273,274), and Nielsen and Whitby (2015) [103] for an important
development 33. The Chow-tests are scaled by their one-off 1% percent significance levels. 1950
2000
0.0
0.1
0.2
∆wct−1 ±2se
∆at ±2se
1950
2000
0.2
0.4
∆at ±2se
∆taxt ±2se
1950
2000
-5
0
5
∆taxt ±2se
1950
2000
0.5
0.7
∆pt ±2se
∆pt−1 ±2se
1950
2000
0.2
0.4
∆pt−1 ±2se
Constant ±2se
1950
2000
0
1
Constant ±2se
(wc−a−q)t−1 ±2se
1950
2000
-0.1
0.0
0.1
(wc−a−q)t−1 ±2se
wbreakt ±2se
1950
2000
1.00
1.25
wbreakt ±2se
ut ±2se
1950
2000
-0.025
0.000
0.025
ut ±2se
1-step residuals ±2 ^σ
1950
2000
-0.025
0.000
0.025
0.050
1-step residuals ±2 ^σ
1-step Chow tests
1%
1950
2000
0.5
1.0
1-step Chow tests
1%
Break point Chow test
1%
1950
2000
0.25
0.50
0.75
1.00
Break point Chow test
1%
Figure 17. Recursive estimation results (OLS) for a conditional wage equation (1925–2015). Figure 17. Recursive estimation results (OLS) for a conditional wage equation (1925–2015). 4.6. Constancy and Invariance of the Coefficients of the Wage Equation, Lucas Citique Clearly, the sets of indicator variables are important both in the cointegrated VAR and in the
structural model estimated by FIML. It is of interest for the economic interpretation whether the
indicator variables are “correctly placed” in the structural model or not. For example, it strengthens
the interpretation of the wage equation if the “outside” indicators, from the other parts of the system,
are insignificant if they are added to the wage model that already includes the “within-equation” break
indicator wbreak. We recall that the within wage equation breaks reflect three epochs: (i) Individual years with
strife, price inflation and extreme difficulty during Norges Bank’s par-policy; (ii) Nazi-occupation
and liberation; and (iii) Incomes policies of the late 1970s and 1980s. Constancy of parameters after
this kind of model extension is an example of empirical co-breaking in econometric relationships,
see Hendry (2000) [100]. But we test first whether the outside-wage equation break indicators are insignificant when added
to the wage model. The results for the IV estimation in Table 3 (the results are similar for the FIML
estimation) become: ubreak:
χ2(1) = 0.26[0.61]
ubreak and qbreak:
χ2(2) = 1.17[0.56]
ubreak and qbreak and pbreak:
χ2(3) = 2.58[0.46] ubreak:
χ2(1) = 0.26[0.61]
ubreak and qbreak:
χ2(2) = 1.17[0.56]
ubreak and qbreak and pbreak:
χ2(3) = 2.58[0.46] which directly support the interpretation that, for example, a location shift can affect the rate of
unemployment, without that “forcing” a coefficient change in the wage equation. which directly support the interpretation that, for example, a location shift can affect the rate of
unemployment, without that “forcing” a coefficient change in the wage equation. The degree of constancy of coefficients can be taken one step further by re-considering the OLS
estimation of the wage equation. The OLS estimators are affected by simultaneity bias, but that does
not necessarily imply that OLS estimated wage equation coefficients are non-constant over the sample
period. It may be that the bias components are constant. Hence, it is interesting to investigate the
empirical stability of the OLS estimated wage equation. If we find empirically that the estimates are constant, the reason may be that structural parameters
and corresponding bias components change in exactly such a way that the conditional model for
wage growth has constant parameters. that does not seem likely. 33
The first test (marking the start of the graph) tests the stability of the relationship over the period from 1926 to 2015.
The second uses data to 1926 for estimation, and tests stability over the 1927–2015 period. Hence the number of out-of-sample
periods are decreasing as we move from left to right along this graph. 4.6. Constancy and Invariance of the Coefficients of the Wage Equation, Lucas Citique Although the uncertainty of the coefficient estimates appear to be quite large early in
the sample period shown in the graphs, the coefficients are relatively stable and increasingly
statistically significant when the last years of the 1940s are included in the estimation sample. The constancy is conditional on the indicators that are collected in the wbreak variable, which
following Clements and Hendry (1999) [85] (Chapter 9) is the essence of co-breaking by extension
of the information set. An important point is of course that the structural changes represented by
wbreak are not one-for-one one with the changes in the “outside” break-indicators ubreak, qbreak and
pbreak, since this would mean that there is no information in (no test strength) in adding these variables
to the conditional wage equation. However, already Figure 14 showed that there are considerably new
information in ubreak, relative to wbreak. 41 of 54 Econometrics 2017, 5, 6 On this basis it is of interest to test the conditional wage equation for super exogeneity, Engle and
Hendry (1993) [104]. The results are shown below, with very clear indication that super exogeneity
cannot be rejected: ubreak:
F(1, 99) = 0.62[0.43]
ubreak and qbreak:
F(2, 99) = 1.16[0.31]
ubreak and qbreak and pbreak:
F(3, 99) = 1.32[0.27] One implication of the empirical presence of super exogeneity is that it immunizes the conditional
wage equation from the Lucas (1976) [105] critique, see Hendry (1988) [106], Favero and Hendry
(1992) [107]. For example, consider the period after WW-II when the rate of unemployment was low
and stable, in part because of policy decisions. A change in policy, aimed to counteract the effects
of a shock that would otherwise have increased unemployment, could then take place without an
accompanying structural break in the coefficients of the wage equation. Which is opposite to the
implications of the Lucas critique. The relatively high degree of coefficient constancy also has implications for wage PCM based
estimates of “natural rate of unemployment” estimates. For example, the increase in non-accelerating
wage rate of unemployment (NAWRU indicator) during the 1980s and 1990s, has been interpreted
by the OECD as evidence of structural breaks in wage formation, see Elmeskov and MacFarland
(1993) [108], Elmeskov (1994) [109]. However, a stable empirical wage equation contradicts increased
NAWRU estimates. 34
Holden and Nymoen (2002) [110] concluded on the basis of results for Denmark, Finland, Norway and Sweden, for samples
that started in the early 1960s and ended in 1994 4.6. Constancy and Invariance of the Coefficients of the Wage Equation, Lucas Citique Table 6. Split sample (1935) and full sample estimates (OLS) of the wage equation (with ut replacing
the two terms ∆ut and ut−1). The standard test for residual mis-specification (autocorrelation,
heteroscedasticity and non-normality) are insignificant in all three samples. Table 6. Split sample (1935) and full sample estimates (OLS) of the wage equation (with ut replacing
the two terms ∆ut and ut−1). The standard test for residual mis-specification (autocorrelation,
heteroscedasticity and non-normality) are insignificant in all three samples. 1904–1934
1935–2015
1904–2015
∆pt
0.55
0.46
0.53
(0.06)
(0.06)
(0.03)
∆at
0.26
0.21
0.26
(0.08)
(0.06)
(0.03)
∆pt−1
0.24
0.19
0.21
(0.08)
(0.07)
(0.04)
∆tax
0.78
1.01
0.89
(2.67)
(0.21)
(0.20)
∆wct−1
0.07
0.21
0.12
(0.06)
(0.07)
(0.03)
(wc −q −a)t−1
−0.06
−0.06
−0.08
(0.05)
(0.03)
(0.02)
ut
−0.01
−0.01
−0.01
(0.008)
(0.002)
(0.002)
Constant
0.43
0.42
0.52
(0.30)
(0.27)
(0.14)
wbreakt
1.00
0.91
0.97
(0.11)
(0.10)
(0.06)
ˆσ100
1.63
1.32
1.37
Chow test
F(20, 92) = 0.24[0.99] 4.6. Constancy and Invariance of the Coefficients of the Wage Equation, Lucas Citique In the same manner as Holden and Nymoen (2002) [110] we therefore conclude
that the instability of the NAWRU estimates is an artefact of a misspecified underlying wage equation,
and not due to instability in wage setting itself 34. Although the constancy of the model is quite convincing over the post-war period, the instability
and lack of significance of some of the variables is open to interpretation. It can be a result of few
observations initially, but on the other hand it can also be a result of partial non-constancy, for example
before and after the Master agreement of 1935, which were discussed with reference economic historical
accounts in Section 2.4 Table 6 reports the estimation results for two sub-samples, 1904–1934 and 1935–2015. The full
sample estimates are also reported for reference. With the exception of the coefficient of the lagged wage change, and of the payroll tax rate, the
point estimates are not very different in the two sub-samples. The main difference is the estimated
model coefficient standard errors, which are a good deal higher on the first sub-sample. The two-sample Chow test reported at the bottom of the table does not reject that hypothesis of
equality of coefficients of the two wage equations. Formally though, the validity of that classic test
depends on constancy of the disturbance standard deviation, which may not hold since the model is
fitting the data better on the second sub-sample than on the first sub-sample. The econometric results nevertheless appear consistent with the historical interpretation of
Olstad (2009) [4], who argued that the Master agreement of 1935 was less of a definitive watershed
than earlier accounts have presented it. The 1935 agreement, although it may nearly have capsized on the rough seas of the 1920
deflation policy, radicalization, and comeback by conservative forces (both business and political),
was prepared trough a long history of labour contracts, negotiations and bargaining. Moreover, in its 42 of 54 42 of 54 Econometrics 2017, 5, 6 turn, it contributed to the understanding that between workers and firm owners; there were not only
conflicting, but also shared interests. With the positive attitudes towards collective action that prevailed after WW-II, the 1935–1940
system represented the natural default to return to, when normal labour market relationships were to
be restored. Frisch’s statement in August 1945, that the days of “Manchester liberalism” were over,
was therefore not even a gamble. 5. Summary and Discussion 5. Summary and Discussion The evolution of the system of labour market regulation has been important for the development
of the Norwegian macro economy since the industrialisation. Early in the 20th century, the idea formed
among trade union leaders and modern industrialists that a system based on collective bargaining
could be functional in a (small open) capitalist economy. As long as there was enough protection
from unwanted competition in the labour market, competition and international trade in the product
markets might bring benefits for all. In order to become stable, labour market regulation in its
turn needed to become compatible with private ownership. Thus a system based on the mutual
recognition of rights, and on genuine compromises, not strife, can be traced right back to the start of
our data sample. However, as with all real world processes, the evolution of the system of wage setting has not
been smooth and deterministic. There have been several crises down the decades, most notably during
the critical years for politics and the economy between the two world wars. But also in the 1970s,
when the system began demonstrating a self-destructive loss of coordinating ability. As noted above, there are both possibilities and potentially unsolvable challenges for econometric
modelling in such a situation. On the one hand, if the formative forces of institutions and markets
have been present in the data over a long historical period, there is a good chance that they can be
represented by an empirical econometric model. On the other hand, if the characteristics of the system 43 of 54 Econometrics 2017, 5, 6 have changed too much, or too often, there may be little to learn from an empirical econometric
modelling project, and it may be just as well to leave the field to economic historians. However, we can only try. The modelling strategy adopted in this paper has been to start from a
theoretical model that contains important collective bargaining parameters, and allow for different
forms of non-stationarities in the estimation of the empirical model. Conditional on the results for
those non-stationarities, we can investigate the constancy and invariance of the main parameters of
interest, for example, of those parameters that separate collective bargain from more individualistic
wage regulation. 5. Summary and Discussion g
g
Specifically, the results in Section 4.3 show that an econometric wage equation of the same type
that has been specified on data from the 1960s and 1970s, represents a congruent empirical model also
on the long historical data set. In this model, the target nominal wage adjusts to the value of labour
productivity, which makes it consistent with collective bargaining, but also to changes in cost of living
and to the level of the unemployment rate. The results show that the model’s derivative coefficients
have been relatively constant despite the structural breaks elsewhere in the economy. In particular,
during the long period of practically speaking full employment after WW-II, there is no evidence of
wage-setting breaking down in an inflationary direction. Invariance to such “outside-breaks” appears
to be a feasible form of coefficient constancy and invariance. the wage model is not invariant to all
historical shocks, and we include a set of indicators in the model, which we refer to as “within-equation
breaks”, and conditional on those indicators, the remaining coefficients in the wage equation (of cost
of living, productivity and unemployment) are empirically constant. Although the comparative perspective has not been central in this paper, we have made some
comparison with modelling results on historical UK data, cf. Castle and Hendry [96,97]. There are close
similarities in the time series, especially unemployment, but also wage and price inflation, albeit with
even more extreme fluctuations in Norway than in the UK. This concordance in timing occurs despite
the very different institutional structures and 20th century histories. Global forces seem to have
operated for a long time. The extension of the empirical modelling from single-equation to multiple-equations is important
in order to model the true nature of real wage dynamics, which is affected by both nominal wage
adjustments (collective bargaining is the regulation principle), and nominal price adjustments (decided
by firms). The empirical multiple-equation model that we formulated is of the Incomplete Competition
Model (ICM) type, and can explain why bargaining power (in the way we have parameterised it) is
important for the real wage to be able to track productivity growth. Updates of the model, and similar
models estimated on quarterly data for a shorter sample period, can be used to judge whether real
wages will continue to track the productivity trend, or not. 5. Summary and Discussion This could be a useful aid if one wants to
avoid to be taken by surprise by a “decoupling” of the real wage trend from productivity, in the way
that happened in the USA (1970s) and UK (1980s). pp
The econometric model thus supports historical accounts of collective bargaining as an
important and formative force of labour market organization and macro economic performance since
industrialisation. The “social order” based on free collective was further developed after the second
world war. The clarification of the “Norwegian model”, with a negotiated wage-norm rooted in the
profitability of the manufacturing industry, and with clear roles for wage-leaders and wage-followers,
was a significant step in the attempt to avoid that (cost-push) inflation became an obstacle for the
attainment of important economic policy targets. Nevertheless, as the post-war economy developed, the wear and tear on the system of free
collective bargaining started to emerge. In the early 1970s, a reform in the direction of corporatism
was shrugged-off in the last minute. The economic problems and the ideological currents of 1980s did
not bode well for the system of collective bargaining, as is was seen as part of the post-war system of
economic policies that had outlived itself. However, again history turned a card, coming up with a
large macro economic and banking crisis, and the unions playing a significant part in the rekindling of
collective rationality and of “solidarity” bargaining. Econometrics 2017, 5, 6 44 of 54 44 of 54 Formation of expectations about prices are always relevant for wage formation. The econometric
results are inconsistent with rational expectations, as a model of how expectations are formed. They are
more consistent with wage setters being able to make use of a relatively good forecast on consumer
price developments. Incidentally, but not coincidentally, this has been one of the roles of the Technical
Calculation Committee for the last fifty years. A system for labour market regulation that has survived for more than a century can be regarded
as a quite advanced product of civilisation. Coordinating capacity is the key contribution of the
system. However, even as our model implies, coordination is quite complex: it is true that the system
of wage-setting can contribute to a balanced nominal path for the economy. In that way it aids the
attainment of other important policy goals. 5. Summary and Discussion However, not even a well functioning wage-setting system
can resolve an unemployment problem, which has been caused by lack of aggregated demand, for
example. Concerted policy adjustment is required in a modern open economy, and even that may
not be enough to avoid unemployment consequences of adverse foreign shocks, or of corrections to
internal imbalances that have been allowed to build up over time. Norwegian policy makers, industrialists and trade union leaders today face low oil price and
reduced oil investments, and have to worry about the possible destabilisation of collective labour
market regulation in the era of globalization and internationalization of labour markets. However, as
the history shows, it is not obvious that today’s difficulties are larger than those that earlier generations
have taken on, and solved. Conflicts of Interest: The author declares no conflict of interest. Appendix A. Data Definitions and Sources The data set for this project is obtained from a number of input data files (with sources given
below) in Oxmetrics format, and by running an Eviews program (prg) file. The data series are listed
here in the same order as they are generated in the prg-file. This appendix documents version 1.0 of
the database (Norwagehistory-1.0). Wages Level For the period 1900–1969 we use the time series for the nominal and real annual wage in
Norges Bank’s database Historical Monetary Statistics for Norway, Norwegian wages classified by industry
(http://www.norges-bank.no/en/Statistics/Historical-monetary-statistics/), Table_total in the linked
xlsx file p2c5_7.xlsx, cf. Grytten (2007) [111], WNB. For the period 1970–2015 the wage series, WNR,
is calculated from Statistics Norway-Statistikkbanken as the wage per whole time equivalent in the
national accounts. Figure A1 shows the close correspondence between the two time series, which
means that it is unproblematic to combine them in the annual wage rate W that we use in the paper. Consumer Price Index 1900–1969: For consistency with the nominal annual wage (WNB) and real annual wage (RWNB),
CPI for this period is calculated as: CPINB = WNB/RWNB. For the second sub-sample, 1970–2015,
the source is Statistics Norway-Statistikkbanken, Tabell: 08184: Konsumprisindeks, historisk serie. CPIKPI. Both CPINB and CPIKPI have 1998 as base year. For the spliced series, CPI, the base year is changed to 2005, which is the base year for fixed price
kroner variables in the data set. Consequently, 2005 = 1 for CPI and the other price indices in the
data set. 45 of 54 45 of 54 Econometrics 2017, 5, 6 onometrics 2017, 5, 6
45 of 5
WNB
WNR
1900
1910
1920
1930
1940
1950
1960
1970
1980
1990
2000
2010
100000
200000
300000
400000
500000
Kroner
WNB
WNR
Figure A1. The annual wage rate from Norges Bank’s Historical Monetary Data, WNB and from the
annual national accounts, WNR. conometrics 2017, 5, 6
45 of 5
WNB
WNR
1900
1910
1920
1930
1940
1950
1960
1970
1980
1990
2000
2010
100000
200000
300000
400000
500000
Kroner
WNB
WNR
Figure A1. The annual wage rate from Norges Bank’s Historical Monetary Data, WNB and from the
annual national accounts, WNR. Figure A1. The annual wage rate from Norges Bank’s Historical Monetary Data, WNB and from the
annual national accounts, WNR. Price Indices for Imports and GDP 1900–1939 and 1949–1969:
Norges Bank’s database Historical Monetary Statistics for
Norway,
The
gross
domestic
product
of
Norway
(http://www.norges-bank.no/en/Statistics/ 1900–1939 and 1949–1969:
Norges Bank’s database Historical Monetary Statistics for
Norway,
The
gross
domestic
product
of
Norway
(http://www.norges-bank.no/en/Statistics/ Grytten (2004) [112]. 1940–1945: For the occupation years we assume that import prices grow at the same rate as the
consumer price index. 1970–2015:
The
implicit
deflator
of
total
imports
from
Statistics
Norway-Statistikk-
banken-National accounts. consumer price index. 1970–2015:
The
implicit
deflator
of
total
imports
from
Statistics
Norway-Statistikk-
banken-National accounts. GDP Fixed Prices 1900–1969: Statistics Norway Historical National Accounts, http://www.ssb.no/a/histstat/aarbok/
ht-0901-bnp.html. 1979–2015: Statistics Norway-Statistikkbanken-National accounts. The unit is million kroner in fixed 2005 prices and is in market values. ht 0901 bnp.html. 1979–2015: Statistics Norway-Statistikkbanken-National accounts. The unit is million kroner in fixed 2005 prices and is in market values. The unit is million kroner in fixed 2005 prices and is in market values. The unit is million kroner in fixed 2005 prices and is in market values. Price Index, Mainland Norway GDP 1900–1969: The GDP price index is used. 1970–2015: The implicit deflator of Mainland Norway GDP (in basic values) from Statistics
Norway-Statistikkbanken-National accounts. 1970–2015: The implicit deflator of Mainland Norway GDP (in basic values) from Statistics
Norway-Statistikkbanken-National accounts. GDP Fixed Prices per Capita The GDP series multiplied by 1,000,000 and divided by the size of the Norwegian population. The source of the population data is Søbye (2014) [14], and https://www.ssb.no/a/folketellinger/
tabeller/3-13.html for updated data. 46 of 54 46 of 54 Econometrics 2017, 5, 6 Man-Years for Mainland Norway 1900–1969: The man-years data for Norway in the historical national accounts (cf. Hansen and
Skoglund (2005, 2008, 2009) [113–115]) was adjusted by using the employment data for whaling and 1900–1969: The man-years data for Norway in the historical national accounts (cf. Hansen and
Skoglund (2005, 2008, 2009) [113–115]) was adjusted by using the employment data for whaling and
international transportation in the historical national accounts international transportation in the historical national accounts. 1970–2015: Statistics Norway-Statistikkbanken-National accounts. The unit is thousand man-years. international transportation in the historical national accounts. 1970–2015: Statistics Norway-Statistikkbanken-National accounts. The unit is thousand man-years. 1970–2015: Statistics Norway-Statistikkbanken-National accounts. Man-Hours for Mainland Norway 1900–1969: The man-hours data for Norway in the historical national accounts (cf. Hansen and 00–1969: The man-hours data for Norway in the historical national accounts (cf. Hansen and
nd (2005 2008 2009) [113–115]) was adjusted by using the employment data for whaling and 1900–1969: The man-hours data for Norway in the historical national accounts (cf. Hansen and
Skoglund (2005, 2008, 2009) [113–115]) was adjusted by using the employment data for whaling and
international transportation in the historical national accounts. 1900–1969: The man-hours data for Norway in the historical national accounts (cf. Hansen and
Skoglund (2005, 2008, 2009) [113–115]) was adjusted by using the employment data for whaling and
international transportation in the historical national accounts. international transportation in the historical national accounts. 1970–2015: Statistics Norway-Statistikkbanken-National accounts. The unit is million hours. 1970–2015: Statistics Norway-Statistikkbanken-National accounts. The unit is million hours GDP Fixed Prices for Mainland Norway 1900–1969: The total GDP data was adjusted with the same historical employment data that was
d t
l
l t
f
h
f
M i l
d N 1900–1969: The total GDP data was adjusted with the same historical employment data that was
used to calculate a measure of man hours for Mainland Norway. 1900–1969: The total GDP data was adjusted with the same historical employment data that was
used to calculate a measure of man hours for Mainland Norway. used to calculate a measure of man hours for Mainland Norway. 1979–2015: Statistics Norway-Statistikkbanken-National accounts used to calculate a measure of man hours for Mainland Norway. 1979–2015: Statistics Norway-Statistikkbanken-National accounts 1979–2015: Statistics Norway-Statistikkbanken-National accounts y
The unit is million kroner in fixed 2005 prices and is in market values. GDP per Man Hour for Mainland Norway The series was obtained by dividing the GDP numbers for Mainland Norway with the
corresponding man-hours data. The unit is kroner in fixed 2005 prices per man-hour. The unit is kroner in fixed 2005 prices per man-hour. For labour productivity we can compare our series with Bergeaud et al. (2015) [24] who include
Norway in their data set of labour productivity (GDP relative to labour (total employment × working
time) , cf. Figure A2. For the period 1970–2015, we give a measure of labour productivity in Mainland
Norway, so it is natural that there are systematic differences. For the 1900–1969 period we are on
thinner ground, since our measure is basically an adjusted series of total GDP and man-hours, as
explained above. Bergeaud et al. (2015) [24] cite Maddison (2001) [116] as their source. The main differences in the graphs are for 1919, 1945 and 1946. For 1945 and 1946, the Hansen
and Skogstad data that we use, represent a careful revision of the employment and working hours data
in the historical national accounts, and therefore our measure is likely give a more correct picture of
those years. For 1919, both series show a marked increase in productivity, but in our series it is twice as
large increase as the BCL series. Although 28% increase is extremely high, it is the consequence of two
aspects of the Hansen and Skogstad data. First, Norwegian GDP increased sharply in the first year after
the end of WW-I, by 18% in our calculation. Second, in 1919 there was also a large reduction in normal
working time. For example, the length of the working week was reduced form 54 to 48 h for day-time
work. This change contributed significantly to a reduction in man-hours worked by 10% in 1919. In particular the drop in labour input may have been underestimated by Bergeaud et al. (2015) [24],
if they interpolated between the annual working time per head of population in 1913 and 1950, for
example, (cf. Maddison (2001) [116] (Table E-10)). If we removed completely the effect of the reduction
in working-time, we would still have a higher number than Bergeaud et al. (2015) [24], so there is a
difference also in the GDP numbers for the year 1919. GDP per Man Hour for Mainland Norway 47 of 54 Econometrics 2017, 5, 6 Labour productivity growth
BCL labour productivity growth
1900
1910
1920
1930
1940
1950
1960
1970
1980
1990
2000
2010
-5
0
5
10
15
20
25
30
Percent
Labour productivity growth
BCL labour productivity growth
Figure A2. Labour productivity (GDP per man hour, Mainland-Norway) and productivity growth in
Bergeaud, Cette and Lecat , BCL. Labour productivity growth
BCL labour productivity growth
Labour productivity growth
BCL labour productivity growth Figure A2. Labour productivity (GDP per man hour, Mainland-Norway) and productivity growth in
Bergeaud, Cette and Lecat , BCL. Payroll-Tax Rate 1900–1969: We used the information given in Hansen and Skoglund (2005, 2008, 2009) [113–115
ut the development in the wedge between wages and wage costs. 1970–2015: Statistics Norway-Statistikkbanken, national accounts. Wage costs in million kroner
ded by wages in million kroner. GDP per Man-Year for Mainland-Norway The series was obtained by dividing the GDP numbers for Mainland-Norway with the
corresponding man-years data. p
g
y
The unit is 1000 kroner in fixed 2005 prices per man-year. The unit is 1000 kroner in fixed 2005 prices per man-year. GDP per Man Hour (Wage Earners) for Mainland Norway For 1970–2015 we use data from Statistikkbanken for both GDP and man-hours, in the same
manner as for the series where GDP for Mainland Norway is divided by total man hours. For 1949–1969,
we use the data for man hours by wage earners in Hansen and Skoglund (2005) [113]. For 1900–1948,
we have adjusted total man hours by a factor based on the employment shares in different sectors of
the economy, using data from SSB (1926) [117]. The man hour data for wage earners for this period is
uncertain. The unit is kroner in fixed 2005 prices per man-hour. 35
https://www.nav.no/no/NAV+og+samfunn/Statistikk/Arbeidssokere+og+stillinger+-+statistikk/Historisk+statistikk. Unemployment Rate NAV has published a historical data for registered unemployed in percent of the labour force
is available since 1948 35. This represents the longest consistent time series for the Norwegian
unemployment rate. We use this series for the period 1948 to 2015, it is denoted UNAV in Figure A3. https://www.nav.no/no/NAV+og+samfunn/Statistikk/Arbeidssokere+og+stillinger+-+statistikk/Historisk+statistikk. Econometrics 2017, 5, 6 48 of 54 For the period with 1940–1945 (Nazi occupation) and 1946–1947 we have made use of the
following equation For the period with 1940–1945 (Nazi occupation) and 1946–1947 we have made use of the
following equation Ut = 0.38 + 0.88Ut−1 −33∆ln(EMPt), t = 1940, ..., 1948
(A1) (A1) where the parameters have been estimated by regressing UNAV on the growth rate in total employment
(EMP), obtained from Hansen and Skoglund (2009) [115] 36. The estimation sample is 1949–1959. To start the forward recursion we used U1939 = UGrytten
1939
where UGrytten
1939
denotes the unemployment
rate for 1939 reported in Grytten (2008) [17] 37. The resulting seven year long series is denoted UWW in
Figure A3. s have been estimated by regressing UNAV on the growth rate in total employment
om Hansen and Skoglund (2009) [115] 36. The estimation sample is 1949–1959. UTU
UWW
UGrytten
UNAV
1900
1910
1920
1930
1940
1950
1960
1970
1980
1990
2000
2010
1
2
3
4
5
6
7
8
9
10
11
UTU
UWW
UGrytten
UNAV
Figure A3. The rate of unemployment (in percent) with the fours sub-samples and sources indicated. UTU
UWW
UGrytten
UNAV
UTU
UWW
UGrytten
UNAV Figure A3. The rate of unemployment (in percent) with the fours sub-samples and sources indicated. For the interwar period 1920–1939 we use the series from Grytten (2008) [17] (Table 3) just
mentioned, and for 1903–1919 we use the unemployment rate of trade union members (in percent
of union membership), this series is denoted UTU in the graph 38. This source has the weakness that
it only covers part of the labour force, namely insured trade unionists in the era where trade unions
administered unemployment insurance, and manufacturing and construction which is most sensitive
to business cycles. However, for the 1903–1910 period, the tendency to reduced unemployment shown
by the graph, and at a relatively low level, indicates that a bias in direction of overestimation of
unemployment might not be a problem, for this period 39. 36
Column labelled “Sysselsatte personer” in “Vedlegg 2 Table, page 35–37. ,
,
p
y
(
) [
]
38
The data are taken from SSB (1948) [119] (Table 187) and SSB (1958) [120] (Table 162).
39 9
Galenson and Zellner (1957) [121], who included Norway in their comparison of unemployment data, used this source
longer period of data. p
y
38
The data are taken from SSB (1948) [119] (Table 187) and SSB (1958) [120] (Table 162). y
p
gg
p g
mbers are revised, not much, compared to Grytten (1995) [118 Number of Working Days and Length of the Working Week For the period 1900–1965, Hansen and Skoglund (2005, 2008, 2009) [113–115]. For the period
1966–2015 we have counted working days from the calendar and have substracted the length of the
annual holidays. The length of the working week is measured by the number of hours in the main
collective agreement (between LO and NHO (earlier NAF)) for full-time day-workers. This series is
more or less identical to the legislative length of the working week (“normal hours”). 40
“Norge hører til de land som har hatt bort imot permanent full sysselsetting i nesten hele etterkrigstiden, og der
konjunkturelle bølgebevegelser produksjon og sysselsetting har vært lite merkbare ” SSB (1965) [20] (p. 352). Unemployment Rate The sources show that the 1920 observation
of the trade union series is 2.3%, which of course is larger than Grytten’s number for that year (1.7%),
but not by a very wide margin. On the whole, there are reasons to believe that the trade union numbers
may be realistic for the first years of the sample period. As mentioned, for the 1920–1939 sub-sample, we use the series due to Grytten, and which gives a
considerably lower level of unemployment than the records from the trade unions (with the noted
exception of 1920). 49 of 54 49 of 54 Econometrics 2017, 5, 6 As also noted, the UWW series for the 2-WW series is based on the estimated relationship between
unemployment rates and employment growth in (A1), using ten years of post-war data and with
the 1933 observation of Grytten’s series as the initial values. UWW indicates that unemployment fell
moderately during the occupation, which may be an underestimation according to some economic
historians. For example, both Hodne and Grytten (2002) [18] and Bjørnhaug and Halvorsen (2009) [19]
(p. 124) conclude that the unemployment of the 1930s disappeared during the two first years of the
war. The marked drop in the UWW series in 1946 and 1947 is on the other hand quite plausible. In
the calculation it is driven by the increase in employed persons in the data by Hansen and Skoglund. It is consistent with the view that near full employment became a reality when normality returned
to Norwegian labour markets after the long summer of 1945.When The Central Bureau of Statistics
(today’s Statistics Norway) in 1965 took stock of the first 20 years after the war, it was plainly
stated that “Norway is among those countries that has enjoyed practically full employment for in
almost the whole post-war period” 40. Moreover, the unemployment rates of the period 1947–1950
were the lowest Norway had experienced after the war, and probably also in all pre-war years, cf. SSB (1965) [20] (pp. 93,94). In that perspective the 2.6% unemployment rate for 1947 may be on the
high side, compared to the first observation of the UNAV series, which is 1.0% for 1948. Working Hours Lost in Labour Conflicts Data on number of working days (million hours) from 1903 to 1930 are taken from
Hansen and Skoglund (2009) [115] (Tables 2, 3 and 20). For 1931–1969 the source is SSB (1969) [122]
(Table 60), for 1971–1975 SSB (1978) [123] (Table 50), and for 1976-1991 SSB (1994) [124] (Table 9.18). For 1992–2015 the source is Statistikkbanken (Table 03629). The series on hours (in million) lost due to
labour conflicts has been constructed by adjusting the number of lost days by normal hours per week
and the number of working days in the year. The assumption we made is that as a rule, wage earners
keep their jobs when the conflict is over. g
.
Olstad, F. Med Knyttet Neve. LOs Historie 1899–1935; Pax Forlag: Oslo, Norway, 2009. 3.
Bjerkholt, O.; Qin, D. Teaching Economics as a Science: The Yale Letcures of Ragnar Frisch; Routledge: Milt
Abington, UK, 2011. .
Sandbrook, D. White Heat. A Historry of Britain in the Swinging Sixties; Abacus: London, UK, 2006. References Norsk økonomisk Politikk etter 1905; Universitetsforlaget: Oslo, Norway, 2012. no/befolkning/artikler og publikasjoner/folkemengdens bevegelse 1735 2014 (accessed on 12 April 2016). 15. Lie, E. Norsk økonomisk Politikk etter 1905; Universitetsforlaget: Oslo, Norway, 2012. g
g p
j
g
g
15. Lie, E. Norsk økonomisk Politikk etter 1905; Universitetsforlaget: Oslo, Norway, 2012. 16. Værholm, M.; Øksendal, L.F. Letting the Anchor Go: Monetary Policy in Neutral Norway During World War I;
Norges Banks Bicentenary Project; Working Paper 28; Norges Bank: Oslo, Norway, 2010. 17. Grytten, O.H. Why Was the Great Depression not so Great in the Nordic Countries? Economic Pol
Unemployment. J. Eur. Econ. Hist. 2008, 37, 369–393, 395–403. 18. Hodne, F.; Grytten, O.H. Norsk økonomi i det 20. århundre; Fagbokforlaget: Bergen, Norway, 2002. 19. Bjørnhaug, I.; Halvorsen, T. Medlemsmakt og Samfunnsansvar. LOs Historie 1935–1969; Pax Forlag: Oslo,
Norway, 2009. 20. Statistisk Sentralbyrå (SSB). Norges økonomi etter krigen. [The Norwegian Post-War Economy];
No. 12
in Samfunnsøkonomiske Studier; Statistisk Sentralbyrå [Central Bureau of Statistics Norway]: Oslo,
Norway, 1965. 21. Statistisk Sentralbyrå (SSB). Økonomisk utsyn over året 1984. [Economic Survey 1984];
NOS. Statistisk
Sentralbyrå: Oslo, Norway, 1985. 22. Barth, E.; Moene, K.O.; Willumsen, F. The Scandinavian Model—An Interpretation. J. Publ. Econ. 2014, 117,
60–72. 23. Barth, E.; Moene, K.O. Innovasjon, Kunnskap og Omstilling—Reell Versus Ideel Konkurranse. Notat Skrevet til Produktivitetskommisjonen, Produktivitetskommisjonen. 2015. Available online:
http://produktivitetskommisjonen.no/notater/ (accessed on 7 February 2016). 4. Bergeaud, A.; Cette, G.; Lecat, R. Productivity Trends in Advanced Countries between 1890 and 2
Rev. Income Wealth 2015, doi:10.1111/roiw.12185. 25. Wrigley, C. British Trade Unions Since 1933; Cambridge University Press: Cambridge, MA, USA, 2002. 26. Sandbrook, D. State of Emergency. The Way We Were: Britain 1970–1974; Penguin Books: London, UK 27. Evju, S. Norway. In Viking, Laval and Beyond; Freedland, M., Prassl, D., Eds.; Hart Publishing: Oxford, UK;
Portland, OR, USA, 2014; Chapter 12, pp. 211–228. 28. Bårdsen, G.J.A.D.; Klovland, J.T. Wage Formation and Bargaining Power During the Great Depression. Scand. J. Econ. 2010, 112, 211–233. 29. Leiserson, M. Wages and Economic Control in Norway: 1945–1957; Harvard University Press: Cambridge, MA,
USA, 1959. 30. Reiersen, J. Department of Business and Management, Buskerud and Vestfold University College. From
C
fli
C
i
N
i
L b
M
k
I
i
i
i
Th M ki
U
bli h d
k 2015 30. Reiersen, J. Department of Business and Management, Buskerud and Vestfold University College. From
Conflict to Cooperation. References 1. Frisch, R. Rasjonell lønnspolitikk [Rational Wage Policy]. Arbeiderbladet, 30 and 31 August 1945 2. Bjerkholt, O. Ragnar Frisch and the Postwar Nowegian Economy: A Critical Comment on Sæther and
Eriksen. Econ J. Watch 2014, 11, 297–312. 2. Bjerkholt, O. Ragnar Frisch and the Postwar Nowegian Economy: A Critical Comment on Sæther and
Eriksen. Econ J. Watch 2014, 11, 297–312. 3. Bjerkholt, O.; Qin, D. Teaching Economics as a Science: The Yale Letcures of Ragnar Frisch; Routledge: Milton Park,
Abington, UK, 2011. 3. Bjerkholt, O.; Qin, D. Teaching Economics as a Science: The Yale Letcures of Ragnar Frisch; Routledge: Milton Park,
Abington, UK, 2011. g
4. Olstad, F. Med Knyttet Neve. LOs Historie 1899–1935; Pax Forlag: Oslo, Norway, 2009. 5. Sandbrook, D. White Heat. A Historry of Britain in the Swinging Sixties; Abacus: Londo 6. Kolsrud, D.; Nymoen, R. Macroeconomic Stability or Cycles? The Role of the Wage-Price Spiral. Aust. Econ. Pap. 2014, 53, 41–68. 6. Kolsrud, D.; Nymoen, R. Macroeconomic Stability or Cycles? The Role of the Wage-Price Spiral. Aust. Econ. Pap. 2014, 53, 41–68. p
7. Kolsrud, D.; Nymoen, R. Heuristic vs formal dynamics of the wage- and price-curve model of equilibrium
unemployment J. Econ. Stud. 2015, 42, 186–206. 7. Kolsrud, D.; Nymoen, R. Heuristic vs formal dynamics of the wage- and price-curve model of equilibrium
unemployment J. Econ. Stud. 2015, 42, 186–206. 50 of 54 Econometrics 2017, 5, 6 50 of 54 8. Hendry, D.F. Deciding Between Alternative Approaches in Macroeconomics. Int. J. Forecast. 2017,
forthcoming. 9. Norges Bank. Inflation Report; Technol Report 3; Norges Bank: Oslo, Norway, 2002. 10. Isachsen, A.J. Jarle Bergo: A professional monetary policymaker steps down. Economic Bulletin, pp. 4–9. Norges Bank. Available online: http://static.norges-bank.no/pages/66487/Jarle_Bergo_Economic_Bulletin_
1_2008.pdf?v=3/17/201415043PM&ft=.pdf (accessed on 10 November 2016). 11. Skoglund, T. Fra jordbruk til tjenester. Økonomiske Anal. 2013, 2013, 53–55. 12. Lindsay, C. A Century of Labour Market Change. Labour Market Trends 2003, March, 133–144. 3. Bævre, K.; Riis, C.; Thonstad, T. Norwegian Cohort Emigration. J. Polpulation Econ. 2001, 14, 473–489. g
g
p
14. Søbye, E. Folkemengdens Bevegelse 1735–2014; Oktober Forlag, Oslo, 2014. Available online: https://www.ssb. no/befolkning/artikler-og-publikasjoner/folkemengdens-bevegelse-1735-2014 (accessed on 12 April 2016). g
g
p
14. Søbye, E. Folkemengdens Bevegelse 1735–2014; Oktober Forlag, Oslo, 2014. Available online: https://www.ssb
/b f lk i
/
ikl
blik
j
/f lk
d
b
l
1735 2014 (
d
12 A
il 2016) no/befolkning/artikler-og-publikasjoner/folkemengdens-bevegelse-1735-2014 (accessed on 12 April 2016). 15. Lie, E. References Interaction Between Model Builders and Policy Makers in the Norwegian Tr
Econ. Model. 1998, 15, 317–339. 46. Skånland, H. Inntektspolitikkens dilemma–kan det løses?; Cappelen Forlag: Oslo, Norway,1981. 47. Llewellyn, J. Can a Small Open Economy Attain Full Employment in the 1990s? In Stabilitet og langsiktighet. Festskrift til Hermod Skånland; Aschehoug: Oslo, Norway, 1994; pp. 313–334. 48. Holden, S. Wage Drift and Bargainig: Evidence from Norway. Economica 1989, 56, 419–432. 49. NOU. Lønnsdannelsen og Utfordringer for Norsk økonomi. Norges Offentlige Utredninger 2013:13,
Departementenes Servicesenter, Informasjonsforvaltningen, Oslo. Official Nowegian Reports (NOUs),
2013. Available online: https://www.regjeringen.no/no/dokumenter/nou-2013-13/id747181/ (accessed
on 15 March 2016). 50. Nymoen, R. Wages and the Length of the Working Day. A Empirical Test Based on Norwegian Quarterly
Manufacturing Data. Scand. J. Econ. 1989, 91, 599–612. 51. Anundsen, A. Oljepris og Lønnsutvikling i K-Sektor; Norges Bank: Oslo, Norway, 29 April 2016. 52. Haldane, A.G. Labour’s Share;
Speech, Bank of England. Speech Given by Andrew G. Haldane,
Chief Economist, Bank of England, Trades Union Congress, London, 12 November 2015. Available online:
http://www.bankofengland.co.uk/publications/Documents/speeches/2015/speech864.pdf (accessed on
6 May 2016). 53. Bårdsen, G.; Nymoen, R. U.S. Natural Rate Dynamics Reconsidered. In The Methodology and Practise of
Econometrics. A Festschrift in Honour of David F. Hendry; Castle, J., Shephard, N., Eds.; Oxford University Press:
Oxford, UK, 2009; Chapter 16, pp. 389–414. 54. Samuelson, P.A. Economic Theory and Wages. In The Impact of the Union; Wright, D., Ed.; Harcourt Brace &
Company: New York, NY, USA, 1951; pp. 312–343. 55. Hicks, J. Economic Foundations of Wages Policy. Econ. J. 1955, 65, 388–404. 56. Hall, R.E. Employment Fluctuations with Equilibrium Wage Stickiness. Am. Econ. Re 56. Hall, R.E. Employment Fluctuations with Equilibrium Wage Stickiness. Am. Econ. Rev. 2005, 95, 50–65. 57. Gertler, M.; Trigari, A. Unemployment Fluctuations with Staggered Nash Wage Bargaining. J. Polit. Econ. 2009 177 38 86 56. Hall, R.E. Employment Fluctuations with Equilibrium Wage Stickiness. Am. Econ. Rev. 2005, 95, 50–65. 57
Gertler M ; Trigari A Unemployment Fluctuations with Staggered Nash Wage Bargaining J Polit Econ 56. Hall, R.E. Employment Fluctuations with Equilibrium Wage Stickiness. Am. Econ. Rev. 2005, 95, 50–65. ,
p
y
q
g
,
,
7. Gertler, M.; Trigari, A. Unemployment Fluctuations with Staggered Nash Wage Bargaining. J. Polit. E
2009, 177, 38–86. 58. Calvo, G. Staggered Prices in a Utility Maximizing Framework. J. Monet. Econ. 1983, 12, 383–398. 59. Blanchard, O.; Galí, J. References Norwegian Labor Market Institutions in The Making. Unpublished work, 2015 , J
p
g
,
y
g
Conflict to Cooperation. Norwegian Labor Market Institutions in The Making. Unpublished work, 2015 31. Barth, E.; Moene, K.O. When Institutions Reciprocate. In European Social Models from Crisis to Crisis;
Dølvik, J.E., Martin, A., Eds.; Oxford University Press: Oxford, UK, 2015; Chapter 10, pp. 306–324. 32. Sæther, A.; Eriksen, I.E. Ragnar Frisch and the Postwar Norwegian Economy. Econ J. Watch 2014, 11 32. Sæther, A.; Eriksen, I.E. Ragnar Frisch and the Postwar Norwegian Economy. Econ J. Watch 2014, 11, 46–80. 33. Longva, S. Beregningsutvalg og lønnsoppgjør. In Stabilitet og langsiktighet. Festskrift til Hermod Skånland;
Aschehoug: Oslo, Norway, 1994; pp. 335–346. 33. Longva, S. Beregningsutvalg og lønnsoppgjør. In Stabilitet og langsiktighet. Festskrift til Hermod Skånland;
Aschehoug: Oslo, Norway, 1994; pp. 335–346. 34. Dyrstad, J.M. Resource Curse Avoidance: Government Intervention and Wage Formation in the Norwegian
Petroleum Sector. Oxf. Econ. Pap. 2016, forthcoming. 35. Lie, E.; Venneslan, C. Over Evne. Finansdepartementet 1965–1992; Pax Forlag: Oslo, Norway, 35. Lie, E.; Venneslan, C. Over Evne. Finansdepartementet 1965–1992; Pax Forlag: Oslo, Norway, 2010. 36. Bergh, T. LOs Historie. Kollektiv Fornuft.Bind 3 1969–2009; Pax Forlag: Oslo, Norway, 2009. 37. Stokke, T.A.; Nergaard, S.; Evju, S. Det kollektive Arbeidslivet, 2nd ed.;
Universitetsforlaget: Oslo,
Norway, 2013. 37. Stokke, T.A.; Nergaard, S.; Evju, S. Det kollektive Arbeidslivet, 2nd ed.;
Universitetsforlaget: Oslo,
Norway, 2013. 51 of 54 51 of 54 Econometrics 2017, 5, 6 38. Nergaard, K. Organisasjonsgrader, Tariffavtaledekning og Arbeidskonfliket 2013; Fafo-notat 2014-14,
FAFO, 2014. Available online: http://www.fafo.no/index.php/nb/zoo-publikasjoner/fafo-notater/item/
organisasjonsgrader-tariffavtaledekning-og-arbeidskonflikter-2013 (accessed on 10 October 2016). 39. Soskice, D. Wage Determination: The Changing Role of Institutions in Advanced Industrialized Co
Oxf. Rev. Econ. Policy 1990, 6, 36–61. 40. Ruoff, B. Labor Market Developments in Germany: Tales of Decency and Stability; Working Paper 39;
Global Labour University: Berlin, Germany, 2016. y
y
41. Evju, S. Norway. Supreme Court. Rt 2013 p. 258. Int. Labour Law Rep. 2014, 33, 3–29. 42. Forder, J. Macroeconomics and the Phillips Curve Myth; Oxford Studies in the History of Economics; Oxford
University Press: Oxford, UK, 2014. 43. Aukrust, O. Inflation in the Open Economy. A Norwegian Model. In World Wide Inflation. Theory and Recent
Experience; Klein, L.B., Sâlant, W.S., Eds.; Brookings: Washington, DC, USA, 1977. 44. Edgren, G.; Faxén, K.O.; Odher, C.E. Wages, Growth and Distribution of Income. Swed. J. Econ. 1969, 71,
133–160. 45. Bjerkholt, O. References Labor Markets and Monetary Policy: A New Keynesian Model with Unemployment. Am. Econ. J. Macroecon. 2010, 2, 1–30. 60. Krogh, T.S. Real wage rigidity and the unemployment volatility puzzle in small open economies. Oxf. Econ. Pap. 2016, 68, 131–151. 61. Nickell, S.J.; Andrews, M. Unions, Real-Wages and Employment in Britain 1951–79. Oxf. Econ. Pap. 1983,
35, 183–206. 62. Nickell, S. Error Correction, Partial Adjustment: and all that:: An expostionary Note. Oxf. Bull. Econ. Stat. 1985, 47, 119–129. 63. Usher, D. Bargaining Unexplained. Publ. Choice 2012, 151, 23–41. Usher, D. Bargaining Unexplained. Publ. Choice 201 64. Samuelson, P.A. Economics;
Oxford Studies in the History of Economics; Oxford University Press:
McCraw-Hill, Cambridge, MA, USA, 1955. 52 of 54 52 of 54 Econometrics 2017, 5, 6 65. Dunlop, J. Wage Determination under Trade Unions; Reprints of Economic Classic, 1966; Augustus M. Kelley
Publishers: New York, NY, USA, 1944. 66. Phillips, A.W. The Relationship Between Unemployment and the Rate of Unemployment and the Rate of
Change of Money Wage Rates in the U.K., 1861–1957. Economica 1958, 25, 283–299. 67. Lipsey, R.G. The Relationship Between Unemployment and the Rate of Change in Money Wages in the
United Kingdom 1862–1957: A Further Analysis. Economica 1960, 27, 1–31. 68. Bårdsen, G.; Eitrheim, Ø.; Jansen, E.S.; Nymoen, R. The Econometrics of Macroeconomic Modelling;
Oxford University Press: Oxford, UK, 2005. 69. Sargan, J.D. Wages and Prices in the United Kingdom: A Study of Econometric Methodology. In Econometric
Analysis for National Economic Planning; Hart, P.E., Mills, G., Whitaker, J.K., Eds.; Butterworth Co.: London,
UK, 1964; pp. 25–63. 70. Sargan, J.D. A Study of Wages and Prices in the U.K. 1949–1968. In The Current Inflation; Johnson, H.G.,
Nobay, A.R., Eds.; MacMillan: London, UK, 1971; Chapter 4, pp. 52–71. 71. Sargan, J.D. A Model of Wage-Price Inflation. Rev. Econ. Stud. 1980, 47, 113–135. 72. Hatton, T. Institutional Change and Wage Rigidity in the U.K. 1880–1985. Oxf. Rev. Econ. Policy
74–86. 73. Bowitz, E.; Cappelen, Å. Modelling Incomes Policies: Some Norwegian experiences 1973–1993. Econ. M
2001, 18, 349–379. 74. Kolsrud, D.; Nymoen, R. Unemployment and the Open Economy Wage-Price Spiral. J. Econ. Stud. 1998, 25,
450–467. 75. Bårdsen, G.; Nymoen, R. Testing Steady-State Implications for the NAIRU. Rev. Econ. Stat. 2003, 85,
1070–1075. 76. Bårdsen, G.; Nymoen, R. Macroeconometric Modelling for Policy. In Palgrave Handbook of Econometrics
Volume 2; Mills, T., Patterson, K., Eds.; Palgrave Mac-Millan: Basingstoke, UK, 2009; Chapter 17, pp. 851–916. References 77. Akram, Q.F.; Nymoen, R. Model Selection for Monetary Policy Analysis—How Important is Empirical
Validity? Oxf. Bull. Econ. Stat. 2009, 71, 35–68. 78. Bårdsen, G.; Reijer, A.; Jonasson, P.; Nymoen, R. MOSES: Model for studying the economy of Sweden. Econ. Model. 2012, 29, 2566–2582. 79. Nymoen, R.; Rødseth, A. Explaining Unemployment: Some Lessons from Nordic Wage Formation. Labour Econ. 2003, 10, 1–29. 80. Rødseth, A. Open Economy Macroeconomics; Cambridge University Press: Cambridge, MA, USA, 20 81. Engle, R.F.; Granger, C.W.J. Co-integration and Error Correction: Representation, Estimation and Testing. Econometrica 1987, 55, 251–276. 82. Forslund, A.; Gottfries, N.; Westermark, A. Prices, Productivity and Wage Bargaining in Open Economies. Scand. J. Econ. 2008, 110, 169–195. 83. Anundsen, A.K.; Krogh, T.S.; Nymoen, R.; Vislie, J. The macroeconomics of Trygve Haavelmo. Nord. J. Polit. Econ. 2012, 37, 2. 84. Keynes, J.M. The General Theory of Employment, Interest and Money, 1973rd ed.; Volume VII of The Collected
Writings of John Meynard Keynes; MacMillan: London, UK, 1936. 85. Clements, M.P.; Hendry, D.F. Forecasting Non-Stationary Economic Time Series; The MIT Press: Cambridge,
MA, USA, 1999. 86. Johansen, S.; Nielsen , B. Analysis of the Indicator Saturation Estimator as a Robust Regression Estimator. In The Methodology and Practise of Econometrics; Castle, J.L., Shephard, N., Eds.; Oxford University Press:
Oxford, UK, 2009. 87. Hendry, D.F.; Johansen, S.; Santos, C. Automatic Selection of Indicators in a Fully Saturated Regression. Comput. Stat. 2008, 23, 317–335, 337–339. 88. Doornik, J.A. Autometrics. In The Methodology and Practice of Econometrics; Castle, J., Shephard, N., Eds.;
Oxford University Press: Oxford, UK, 2009; Chapter 8, pp. 88–121. 89. Nymoen, R. Faulty Watch Towers: Structural Models in Norwegian Monetary Policy Analysis; Technol Report;
Department of Economics, University of Oslo: Oslo, Norway, 2002. 89. Nymoen, R. Faulty Watch Towers: Structural Models in Norwegian Monet
Department of Economics, University of Oslo: Oslo, Norway, 2002. 90. Eitrheim, Ø.; Jansen, E.S.; Nymoen, R. Progress from Forecast Failure: The Norwegian Consumption
Function. Econom. J. 2002, 5, 40–64. 53 of 54 53 of 54 Econometrics 2017, 5, 6 91. Doornik, J. Approximations to the Asymptotic Distribution of cointegration Tests. J. Econ. Surv. 1998, 12,
573–593. 92. Doornik, J. Asymptotic Tables for Cointegration Tests Based on the Gamma-Distribution;
Technol Report;
Nuffield College, University of Oxford: Oxford, UK, 2003. ffield College, University of Oxford: Oxford, UK, 2003 3. Davidson, R.; MacKinnon, J.G. References Econometric Theory and Methods; Oxford University Press: Oxford, UK, 2
4
J h
K N
i
W
C
O f B ll E
St t 1995 57 229 247 93. Davidson, R.; MacKinnon, J.G. Econometric Theory and Methods; Oxford University Press: Oxford, UK, 2004. 94. Johansen, K. Norwegian Wage Curves. Oxf. Bull. Econ. Stat. 1995, 57, 229–247. 93. Davidson, R.; MacKinnon, J.G. Econometric Theory and Methods; Oxford University Press: Oxford, UK, 2004. 94. Johansen, K. Norwegian Wage Curves. Oxf. Bull. Econ. Stat. 1995, 57, 229–247. 94. Johansen, K. Norwegian Wage Curves. Oxf. Bull. Econ. Stat. 1995, 57, 229–247. 95. Ericsson, N.R.; Maasoumi, E.; Mizon, G. A Retrospective on J. Denis Sargan and His Contributions to
Econometrics. Econom. Rev. 2001, 20, 132–158. 96. Castle, J.L.; Hendry, D.F. Semi-Automatic Non-Linear Model Selection. In Essays in Non-Linear Time Series
Econometrics; Haldrup, N., Saikkonen, P., Eds.; Oxford Univeristy Press: Oxford, UK, 2014; pp. 163–197. 96. Castle, J.L.; Hendry, D.F. Semi Automatic Non Linear Model Selection. In Essays in Non Linear Time Series
Econometrics; Haldrup, N., Saikkonen, P., Eds.; Oxford Univeristy Press: Oxford, UK, 2014; pp. 163–197. 98. Hsiao, C. Cointegration and Dynamic Simultaneous Equations Model. Econometrica 1997, 65, 647–670. 99. Doornik, J.A.; Hendry, D.F. Modelling Dynamic Systems PcGive 14. Volume 2; Timberlake Consultants:
London, UK, 2013. 00. Hendry, D.F. On Detectable and Non-detectable Structural Change. Struct. Chang. Econ. Dyn. 2000
45–65. 101. Engle, R.F.; Hendry, D.F.; Richard, J.-F. Exogeneity. Econometrica 1983, 51, 277–304. 102. Doornik, J.A.; Hendry, D.F. Empirical Econometric Modelling PcGive 13. Volume 1; Timberlake Consultants:
London, UK, 2009. 103. Nielsen, B.; Whitby, A. A Joint Chow Test for Structural Instability. Econometrics 2015, 3, 156–186. 104. Engle, R.F.; Hendry, D.F. Testing Super Exogeneity and Invariance in Regression Models. J. Econom. 1993, 56,
119–139. 105. Lucas, R.E., Jr. Econometric Policy Evaluation: A Critique. Carnegie-Rochester Conf. Ser. Public Policy 1976, 1,
19–46. 106. Hendry, D.F. The Encompassing Implications of Feedback verus Feedforward Mechanisms in Econometrics. Oxf. Econ. Pap. 1988, 40, 132–149. 107. Favero, C.; Hendry, D.F. Testing the Lucas Critique: A Review. Econom. Rev. 1992, 11, 265–306. 108. Elmeskov, J.; MacFarland, M. Unemployment Persistence. OECD Econ.Stud. 1993, 21, 59–88. 109. Elmeskov, J. Nordic Unemployment in a European Perspective. Swed. Econ. Policy Rev. 1994, 1, 27– 109. Elmeskov, J. Nordic Unemployment in a European Perspective. Swed. Econ. Policy Rev. 1994, 1, 27–70. c⃝2017 by the author; licensee MDPI, Basel, Switzerland. This article is an open access
article distributed under the terms and conditions of the Creative Commons Attribution
(CC-BY) license (http://creativecommons.org/licenses/by/4.0/). References 110
Holden S ; Nymoen R Measuring Structural Unemployment: NAWRU Estimates in the Nordic Countries Holden, S.; Nymoen, R. Measuring Structural Unemployment: NAWRU Estimates in the Nordic Countries
Scand. J. Econ. 2002, 104, 87–104. 111. Grytten, O.H. Norwegian wages 1726–2006 classified by industry. In Historical Monetary Statistics for
Norway-Part II; Volmue 38 of Occasional Papers; Eitrheim, J.K.Ø., Qvigstad, J., Eds.; Norges Bank: Oslo,
Norway, 2007; Chapter 6, 343–383. 112. Grytten, O.H. The gross domestic product for Norway 1830–2003. In Historical Monetary Statistics for Norway
1830–2003; Volume 35 of Occasional Papers; Eitrheim, J.K.Ø., Qvigstad, J., Eds.; Norges Bank: Oslo, Norway,
2004; Chapter 6, pp. 241–288. 113. Hansen, S.; Skoglund, T. Sysselsetting og Lønn i Historisk Nasjonalregnskap.Beregninger for 1949–1969;
Notater 2009/38; Statistisk Sentralbyr: Oslo, Norway, 2005. 114. Hansen, S.; Skoglund, T. Sysselsetting og Lønn i Historisk Nasjonalregnskap.Beregninger for 1949–1969;
Notater 2008/54; Statistisk Sentralbyr: Oslo, Norway, 2008. 115. Hansen, S.; Skoglund, T. Sysselsetting og Lønn i Historisk Nasjonalregnskap.Beregninger for 1900–1930;
Notater 2009/38; Statistisk Sentralbyr, Oslo, Norway, 2009. 116. Maddison, A. The World Economy. a Millenian Perspective; OECD Publishing: Paris, France, 2001 117. Statistisk Sentralbyrå (SSB). Statistiske Oversikter 1926;
Det Statistiske Centralbyrå. Available online:
http://www.ssb.no/a/histstat/hs1926.pdf (accessed on 12 April 2016). 118. Grytten, O.H. The Scale of Norwegian Interwar Unemployment in International perspective. Scand. Econ. Hist. Rev. 1995, 43, 226–250. 119. Statistisk Sentralbyrå (SSB). Statistical Surveys 1948; Volume Oslo of NOS; The Central Bureau of Statistics
of Norway: Oslo, Norway, 1948. 120. Statistisk Sentralbyrå (SSB). Statistical Surveys 1958; Volume Oslo of NOS; The Central Bureau of Statistics
of Norway: Oslo, Norway, 1958. 54 of 54 Econometrics 2017, 5, 6 54 of 54 121. Galenson, W.; Zellner, A. International Comparison of Unemployment Rates. In The Measurement and
Behavior of Unemployment; NBER: Cambridge, MA, USA, 1957; pp. 439–584. 122. Statistisk Sentralbyrå (SSB). Historisk Statistikk 1968; Statistisk Sentralbyrå: Oslo, Norway, 1969. 123. Statistisk Sentralbyrå (SSB). Historisk Statistikk 1978; Statistisk Sentralbyrå: Oslo, Norway, 1978. 124. Statistisk Sentralbyrå (SSB). Historisk Statistikk 1994; Statistisk Sentralbyrå: Oslo, Norway, 1994. c⃝2017 by the author; licensee MDPI, Basel, Switzerland. This article is an open access
article distributed under the terms and conditions of the Creative Commons Attribution
(CC-BY) license (http://creativecommons.org/licenses/by/4.0/).
|
https://openalex.org/W3181550827
|
https://escholarship.org/content/qt1ds9g1j6/qt1ds9g1j6.pdf?t=rwwup6
|
English
| null |
Bias, precision and timeliness of historical (background) rate comparison methods for vaccine safety monitoring: an empirical multi-database analysis
|
medRxiv (Cold Spring Harbor Laboratory)
| 2,021
|
cc-by
| 6,994
|
Title Bias, Precision and Timeliness of Historical (Background) Rate Comparison Methods for
Vaccine Safety Monitoring: An Empirical Multi-Database Analysis Bias, Precision and Timeliness of Historical (Background) Rate Comparison Methods for
Vaccine Safety Monitoring: An Empirical Multi-Database Analysis Permalink
https://escholarship.org/uc/item/1ds9g1j6 Authors
Li, Xintong
Lai, Lana YH
Ostropolets, Anna
et al. Publication Date
2021
DOI
10.3389/fphar.2021.773875
Peer reviewed UCLA
UCLA Previously Published Works UCLA
UCLA Previously Published Works Edited by:
Emanuel Raschi,
University of Bologna, Italy 1Centre for Statistics in Medicine, NDORMS, University of Oxford, Oxford, United Kingdom, 2School of Medical Sciences,
University of Manchester, Manchester, United Kingdom, 3Department of Biomedical Informatics, Columbia University, New York,
NY, United States, 4Department of Biostatistics, University of California, Los Angeles, California, United States, 5Real-World
Evidence, Trial Form Support, Barcelona, Spain, 6College of Pharmacy, Riyadh Elm University, Riyadh, Saudi Arabia, 7Institut
Universitari D’Investigació en Atenció Primària Jordi Gol (IDIAPJGol), Barcelona, Spain, 8O’Brien Institute for Public Health, Faculty
of Medicine, University of Calgary, Calgary, AB, Canada, 9Nuffield Department of Clinical Neurosciences, University of Oxford,
Oxford, United Kingdom, 10Quality Use of Medicines and Pharmacy Research Centre, Clinical and Health Sciences, University of
South Australia, SA, Australia, 11Observational Health Data Sciences and Informatics, New York, NY, United States,
12Observational Health Data Analytics, Janssen R&D, Titusville, NJ, United States, 13Medical Informatics Services, NewYork-
Presbyterian Hospital, NewYork, NY, United States, 14Health Data Sciences, Medical Informatics, Erasmus Medical Center
University, Rotterdam, Netherlands Reviewed by:
Rosa Gini,
ARS Toscana, Italy
Bo Ram Yang,
Chungnam National University, South
Korea Reviewed by:
Rosa Gini,
ARS Toscana, Italy
Bo Ram Yang,
Chungnam National University, South
Korea *Correspondence:
Martijn J. Schuemie
schuemie@ohdsi.org *Correspondence:
Martijn J. Schuemie
schuemie@ohdsi.org Specialty section:
This article was submitted to
Pharmacoepidemiology,
a section of the journal
Frontiers in Pharmacology
Received: 10 September 2021
Accepted: 02 November 2021
Published: 24 November 2021 Specialty section:
This article was submitted to
Pharmacoepidemiology,
a section of the journal
Frontiers in Pharmacology
Received: 10 September 2021
Accepted: 02 November 2021
Published: 24 November 2021 Using real-world data and past vaccination data, we conducted a large-scale experiment
to quantify bias, precision and timeliness of different study designs to estimate historical
background (expected) compared to post-vaccination (observed) rates of safety events for
several vaccines. We used negative (not causally related) and positive control outcomes. The latter were synthetically generated true safety signals with incident rate ratios ranging
from 1.5 to 4. Observed vs. expected analysis using within-database historical
background rates is a sensitive but unspecific method for the identification of potential
vaccine safety signals. Despite good discrimination, most analyses showed a tendency to
overestimate risks, with 20%-100% type 1 error, but low (0% to 20%) type 2 error in the
large databases included in our study. Edited by:
Emanuel Raschi,
University of Bologna, Italy Efforts to improve the comparability of background
and post-vaccine rates, including age-sex adjustment and anchoring background rates
around a visit, reduced type 1 error and improved precision but residual systematic error
persisted. Additionally, empirical calibration dramatically reduced type 1 to nominal but
came at the cost of increasing type 2 error. Keywords: incidence rate, vaccine safety, real world data, empirical - comparison, background rate Powered by the California Digital Library
University of California eScholarship.org ORIGINAL RESEARCH
published: 24 November 2021
doi: 10.3389/fphar.2021.773875 ORIGINAL RESEARCH
published: 24 November 2021
doi: 10.3389/fphar.2021.773875 Xintong Li 1, Lana YH Lai 2, Anna Ostropolets 3, Faaizah Arshad 4, Eng Hooi Tan 1,
Paula Casajust 5, Thamir M. Alshammari 6, Talita Duarte-Salles 7, Evan P. Minty 8,
Carlos Areia 9, Nicole Pratt 10, Patrick B. Ryan 11,12, George Hripcsak 3,13, Marc A. Suchard 4,11,
Martijn J. Schuemie 4,11,12* and Daniel Prieto-Alhambra 1,14 Edited by:
Emanuel Raschi,
University of Bologna, Italy Edited by:
Emanuel Raschi,
University of Bologna, Italy INTRODUCTION As regulators across the world evaluate the first signals of post-
marketing safety potentially associated with coronavirus disease
2019 (COVID-19) vaccines, they rely on the use of historical
comparisons with so-called “background rates” for the events of
interest to identify outcomes appearing more often than expected
following vaccination. However, a literature gap remains on the
reliability of these methods, their associated error(s), and the impact
of potential strategies to mitigate them. We therefore aimed to study
the bias, precision, and timeliness associated with the use of historical
comparisons between post-vaccine and background rates for the
identification of safety signals. We tested strategies for background
rate estimation (unadjusted, age-sex adjusted, and anchored around Citation: Li X, Lai LYH, Ostropolets A, Arshad F,
Tan EH, Casajust P, Alshammari TM,
Duarte-Salles T, Minty EP, Areia C,
Pratt N, Ryan PB, Hripcsak G,
Suchard MA, Schuemie MJ and
Prieto-Alhambra D (2021) Bias,
Precision and Timeliness of Historical
(Background) Rate Comparison
Methods for Vaccine Safety
Monitoring: An Empirical Multi-
Database Analysis. Front. Pharmacol. 12:773875. doi: 10.3389/fphar.2021.773875 Li X, Lai LYH, Ostropolets A, Arshad F,
Tan EH, Casajust P, Alshammari TM,
Duarte-Salles T, Minty EP, Areia C,
Pratt N, Ryan PB, Hripcsak G,
Suchard MA, Schuemie MJ and
Prieto-Alhambra D (2021) Bias,
Precision and Timeliness of Historical
(Background) Rate Comparison
Methods for Vaccine Safety
Monitoring: An Empirical Multi-
Database Analysis. Front. Pharmacol. 12:773875. doi: 10.3389/fphar.2021.773875 November 2021 | Volume 12 | Article 773875 1 Frontiers in Pharmacology | www.frontiersin.org Historical Comparisons for Vaccine Safety Li et al. FIRST PAGE a healthcare visit), and studied the impact of empirical calibration on
type 1 and type 2 error. Key Messages Historical rate comparison has been suggested for use in several
vaccine safety guidelines, including the European Network of
Centres
of
Pharmacoepidemiology
and
Pharmacovigilance
(ENCePP), Council for International Organizations of Medical
Sciences
(CIOMS),
and
Good
Pharmacovigilance
Practices
(GVP). It has also been applied extensively in various clinical
domains,
including
the
Center
for
Disease
Control
and
Prevention (CDC)’s Vaccine Safety Datalink (VSD) project,
which used background rates to detect safety signals for the
human papillomavirus vaccine (HPV) (Gee et al., 2011), adult
tetanus-diphtheria-acellular pertussis (Tdap) vaccine (Yih et al.,
2009), and a broad range of paediatric vaccines (Lieu et al., 2007;
Yih et al., 2011). Historical data were used in Australia to detect
signals for the rotavirus vaccine (Buttery et al., 2011), and in Europe
to detect signals for the influenza A H1N1 vaccine (Black et al., 2009;
Wijnans et al., 2013; Barker and Snape, 2014). While this study
design is widely implemented, there is high variability in the specifics
of methods used to calculate historical rates, including selection of
target populations, time-at-risk windows, observation time and
study settings. • Within-database background rate comparison is a sensitive
but unspecific method to identify vaccine safety signals. The
method is positively biased, with low ( ≤20%) type 2 error,
and 20–100% of negative control outcomes were incorrectly
identified as safety signals due to type 1 error. • Within-database background rate comparison is a sensitive
but unspecific method to identify vaccine safety signals. The
method is positively biased, with low ( ≤20%) type 2 error,
and 20–100% of negative control outcomes were incorrectly
identified as safety signals due to type 1 error. • Age-sex
adjustment
and
anchoring
background
rate
estimates around a healthcare visit are useful strategies to
reduce false positives, with little impact on type 2 error. • Sufficient sensitivity was reached for the identification of
safety signals by month 1-2 for vaccines with quick uptake
(e.g., seasonal influenza), but much later (up to month 9) for
vaccines with slower uptake (e.g., varicella-zoster or
papillomavirus). • Empirical calibration using negative control outcomes
reduces type 1 error to nominal at the cost of increasing
type 2 error. Background
f h One of the most common study designs in vaccine safety surveillance
is the use of a cohort study with a historical comparison as a
benchmark. This design allows the observed incidence of adverse
events of the studied vaccine following immunization (AEFI) to be
compared with the expected incidence of AEFI projected based on
historical data (Belongia et al., 2010). Alleged strengths include
greater statistical power to detect rare AEFIs, as well as improved
timeliness in detecting potential safety signals by leveraging
retrospective data for analysis. There are, however, also caveats
with this study design (Mesfin et al., 2019). Firstly, the historical
population must be similar to the vaccinated cohort to obtain
comparable estimates of baseline risk. Secondly, the design is
subject to various temporal confounders such as seasonality,
changing trends in the detection of AEFIs, and variation in
diagnostic or coding criteria over time. Thirdly, the design is
highly dependent on an accurate estimation of background
incidence rates of the AEFIs for comparison. Observed
vs. expected
analysis
using
within-database
historical
background
rates
is
a
sensitive
but
unspecific
method for the identification of potential vaccine safety
signals. Despite good discrimination, most analyses showed a
tendency to overestimate risks, with 20–100% type 1 error, but
low (0–20%) type 2 error in the large databases included in our
study. Efforts to improve the comparability of background and
post-vaccine rates, including age-sex adjustment and anchoring
background rates around a visit, reduced type 1 error and
improved precision but residual systematic error persisted. Additionally, empirical calibration dramatically reduced type 1
to nominal but came at the cost of increasing type 2 error. Standfirst Using real-world data and past vaccination data, we conducted a
large-scale experiment to quantify bias, precision and timeliness
of different study designs to estimate historical background
(expected) compared to post-vaccination (observed) rates of
safety events for several vaccines. We used negative (not
causally related) and positive control outcomes. The latter
were synthetically generated true safety signals with incident
rate ratios ranging from 1.5 to 4. Frontiers in Pharmacology | www.frontiersin.org Exposures We used retrospective data to study the following vaccines within
the corresponding study periods: 1) H1N1 vaccination (Sept 2009
to May 2010), 2) different types of seasonal flu vaccination (Sept
2017 to May 2018), 3) varicella-zoster vaccination (Jan 2018 to
Dec 2018), and 4) HPV 9-valent recombinant vaccine (Jan 2018
to Dec 2018). Vaccines were captured as drug exposure in the
common data mode. Specific CVX codes and RxNorm codes,
follow-up periods, and cohort construction details are available in
– Supplementary Appendix Table S2. Post-vaccination rates
were obtained for the period of 1–9 months for H1N1 and
seasonal flu, and 1 to 12 for varicella-zoster and HPV
vaccines. Background (historical) rates were obtained from the
general population, for the same range of months 1 year
preceding each of these vaccines (Unadjusted). To minimise
confounding, three additional variations of background rates
were estimated: 1) age-sex adjusted rates; 2) visit-anchored
rates; and 3) visit and age-sex adjusted rates. In the first,
background rates were stratified by age (10-years bands) and
sex. In the second option, background rates were estimated using
the time-at-risk following a random outpatient visit (visit-
anchored). The third combined the two above to account for
differences in socio-demographics and for the impact of
anchoring
(similar
to
anchoring
post-vaccination
in
the
exposed group). We used retrospective data to study the following vaccines within
the corresponding study periods: 1) H1N1 vaccination (Sept 2009
to May 2010), 2) different types of seasonal flu vaccination (Sept
2017 to May 2018), 3) varicella-zoster vaccination (Jan 2018 to
Dec 2018), and 4) HPV 9-valent recombinant vaccine (Jan 2018
to Dec 2018). Vaccines were captured as drug exposure in the
common data mode. Specific CVX codes and RxNorm codes,
follow-up periods, and cohort construction details are available in There have been suggestions on how to mitigate some of the
differences between the historical and observed populations,
including stratifying by age, gender, geographical or calendar
time (Gee et al., 2011; Yih et al., 2011). While these approaches
may reduce some differences, the distribution of the observed
population is rarely known unless the study uses the spontaneous
case’s demographic characteristics (of which the cases may be
identified through the adverse event spontaneous system) as a
proxy of the demographic characteristics of the observed
population. This could potentially lead to a bias due to the
estimation misclassification in each stratum based on the
reporting rate (i.e., high vs. Outcomes We therefore aimed to study the bias, precision, and timeliness
associated with the use of historical comparisons between post-
vaccine and background rates for the identification of safety
signals. We evaluated strategies for estimating background
rates and the effect of empirical calibration on type 1 and type
2 error using real-world outcomes presumed to be unrelated to
vaccines (negative control outcomes) as well as imputed positive
controls (outcomes simulated to be caused by the vaccines). We employed negative control outcomes as a benchmark to
estimate bias (Schuemie et al., 2016; Schuemie et al., 2020). Negative controls are outcomes with no plausible causal
association with any of the vaccines. As such, negative control
outcomes should not be identified as a signal by a safety
surveillance method, and any departure from a null effect is
therefore suggestive of bias due to type 1 error. A list of negative
control outcomes was pre-specified for all four vaccine groups. To
identify negative control outcomes that match the severity and
prevalence of suspected vaccine adverse effects, a candidate list of
negative controls was generated based on similarity of prevalence
and percent of diagnoses that were recorded in an inpatient
setting (as a proxy for severity). Three clinical experts manually
reviewed the list, which led to a final list of 93 negative control Exposures low reporting rates). follow-up periods, and cohort construction details are available in
– Supplementary Appendix Table S2. Post-vaccination rates
were obtained for the period of 1–9 months for H1N1 and
seasonal flu, and 1 to 12 for varicella-zoster and HPV
vaccines. Background (historical) rates were obtained from the
general population, for the same range of months 1 year
preceding each of these vaccines (Unadjusted). To minimise
confounding, three additional variations of background rates
were estimated: 1) age-sex adjusted rates; 2) visit-anchored
rates; and 3) visit and age-sex adjusted rates. In the first,
background rates were stratified by age (10-years bands) and
sex. In the second option, background rates were estimated using
the time-at-risk following a random outpatient visit (visit-
anchored). The third combined the two above to account for
differences in socio-demographics and for the impact of
anchoring
(similar
to
anchoring
post-vaccination
in
the
exposed group). Large databases that link medical outcomes with vaccine
exposure data provide a means of assessing signals identified,
as well as estimates of a true incidence of clinical events after
vaccination. However, these systems can be affected by relatively
small denominators (given the rarity of the event) of vaccinated
subjects, and a time lag in the availability of data. Very rare events
or outcomes affecting a subset of the population might still be
under-powered to assess a safety concern even when the data
reflect the experience of millions of individuals (Black et al.,
2009). Heterogeneity in background rates across databases and
age-sex
strata
may
also
persist
even
after
robust
data
harmonization using common data models (Li et al., 2021). Uncertainties and Limitations With the use
of Historical Rate Comparisons for Vaccine
Safety Monitoring Several studies have acknowledged uncertainties associated with the
use of background rates relating to temporal and geographical
variations. In one study that applied both historical comparisons
and self-controlled methods, a signal of seizure in the 2014–2015 flu
season was detected in the latter analysis but not the former. The
authors explained that one possible reason was that the historical
rates used might not reflect the expected baseline rate in the absence
of vaccination. A second explanation was a falsely elevated November 2021 | Volume 12 | Article 773875 Frontiers in Pharmacology | www.frontiersin.org 2 Historical Comparisons for Vaccine Safety Li et al. background compared to post-vaccination rates of safety events
using “real world” (electronic health records and administrative
health claims) databases from the US. Our study protocol is
available in the EU PAS Register (EUPAS40259) (European
Network
of
Centres
of
Pharmacoepidemiology
and
Pharmacovigilance, 2021), and all our analytical code is in
GitHub
(https://github.com/ohdsi-studies/Eumaeus). These
data were previously mapped to the OMOP common data
model (OHDSI, 2019). The list of included data sources, with
a brief description, is available in - Supplementary Appendix
Table S1. background rate because of the inclusion of events induced by a
previous vaccine season. Other studies have highlighted the
importance of accounting for demographic, secular and seasonal
trends to appropriately interpret historical rates (Buttery et al., 2011;
Yih et al., 2011). Nevertheless, the influence of such trends has not
been studied systematically despite observed heterogeneity in
historical incidence rates (Yih et al., 2011). It is also essential to consider the data source since there are
differences in case ascertainment. This might lead to uncertainty
in background rate estimates, especially in rare events (Mahaux
et al., 2016). In addition, there might also be differences in the use
of dictionary or codes to define an AEFI. For example, the
spontaneous reporting system generally uses the Medical
Dictionary for Regulatory Activities (MedDRA), while in
observational
databases
different
codes
are
used
(e.g.,
International Classification of Diseases (ICD), SNOMED-CT,
READ) and the granularity of available coding can impact the
sensitivity and specificity of phenotype algorithms. The use of Optum and IBM Marketscan databases was
reviewed by the New England Institution Review Board (IRB)
and was determined to be exempt from broad IRB approval, as
this research project did not involve human subjects research. Performance Metrics The estimated effect size for the association vaccine-outcome was
based on IRR by dividing the observed (post-vaccine) over
expected (historical) incidence rates. To account for systematic
error, we employ empirical calibration: we firstly compute the
distribution of systematic error using the estimates for the
negative and positive control outcomes. We then use the
distributions and their standard deviations to adjust effect-size
estimates, confidence intervals, p-values, and log likelihood ratios
(LLRs) to restore type 1 error to nominal. We used a leave-one-
out strategy for this evaluation, calibrating the estimate for a
control outcome using the systematic error distribution fitted on
all control outcomes except the one being calibrated. IRR were
computed both with and without empirical calibration (Schuemie
et al., 2016; Schuemie et al., 2018). Historical rate comparisons were –even in their simplest
form— associated with low type 2 error (0–10%), but led to
type 1 errors ranging between 30% (HPV in MDCD) and 100%
(H1N1 and seasonal flu in Optum EHR). Adjustment for age and
sex reduced type 1 error in some but not all scenarios, and had
limited impact on type 2 error (maximum 20% in all the
conducted
analyses). However,
age
and
sex
adjusted
comparisons were still prone to type 1 error, with most (12/
13) analyses still incorrectly identifying ≥40% negative controls as
potential safety signals. Anchoring the estimation of background
rates around a healthcare visit helped reduce type 1 error in some
scenarios (e.g., H1N1 in Optum EHR went from 100 to 50%), but
increased it in others (e.g., H1N1 in CCAE increased from 50% in
the unadjusted to 80% in the anchored analysis). In addition,
anchoring increased type 2 error in most of our analyses,
although none exceeded 20% in any of the analyses. Finally,
the analyses combining anchoring and age-sex adjustment led to
observable reductions in type 1 error (e.g., from 70 to 30% for
HPV in CCAE), with negligible increases in type 2 error in most
instances (e.g., from 10 to 20% for HPV in MDCD). Detailed
results for unadjusted, age-sex adjusted, and anchoring scenarios
are demonstrated in Figure 1. METHODS Data Sources and Data Access Approval
We aimed to fill a gap in the existing literature by estimating the
bias, precision and timeliness associated with the use of historical/ November 2021 | Volume 12 | Article 773875 Frontiers in Pharmacology | www.frontiersin.org 3 Li et al. Li et al. Historical Comparisons for Vaccine Safety outcomes to be included. Details of the selection process,
including the candidate outcomes, reasons for exclusion, and
the final negative control outcomes list are available in -
Supplementary Appendix Table S3. Finally, we studied the proportion of controls for which IRR
were not estimable due to lack of participants exposed to the
vaccine of interest. We also considered as not estimable (and
therefore did not report) results for negative control outcomes
with a population-based incidence rate changing >50% over time
during the study period. In
addition,
synthetic
positive
control
outcomes
were
generated to measure type 2 error (OHDSI, 2019). Given the
limited knowledge of such events and the lack of consistency in
the true causal association amongst other problems [6], we
computed synthetic positive controls with known (albeit in
silico) causal associations with the vaccines under study [5,7]. Positive outcomes were generated by modifying negative control
outcomes through injection of additional simulated occurrences
of the outcome, with effect sizes equivalent to true incidence rate
ratios (IRR) of 1.5, 2, and 4. With the 3 mentioned true IRR, 93
negative controls were used to construct at most 93 × 3 279
positive control outcomes, although no positive controls were
synthesized if for the negative control the number of outcomes
was smaller than 25. The hazard for these outcomes was
simulated
to
be
increased
for
the
period
1 day
after
vaccination until 28 days after vaccination, with a constant
hazard ratio during that time. For all the estimated metrics, we reported the results for each
database – vaccine group – method group. Bias and Precision A total of four large databases were included, most including all
four vaccines of interest: IBM MarketScan Commercial Claims
and Encounters (CCAE), IBM MarketScan Multi-state Medicaid
(MDCD), IBM MarketScan Medicare Supplemental Beneficiaries
(MDCR), and Optum© de-identified Electronic Health Record
dataset
(Optum
EHR). The
basic
socio-demographics
of
participants registered in each of these databases are reported
in Supplementary Appenix S1. All data sources had a majority of
women, from 51.1% in CCAE to 56.23% in MDCD. As expected,
data sources with older populations (e.g., IBM MDCR) had little
exposure to HPV vaccination, but high numbers of participants
exposed to seasonal influenza vaccination. All four data sources
contributed information based on healthcare encounters in
emergency rooms, outpatient as well as inpatient settings. Frontiers in Pharmacology | www.frontiersin.org Performance Metrics Bias was measured using: 1) Type 1 error, based on the
proportion of negative control outcomes identified as safety
signals according to p-value < 0.05; 2) Type 2 error, based on
how often positive control outcomes were missed (not identified)
as safety signals (p > 0.05); 3) Area Under the receiver-operator
Curve (AUC) for the discrimination of effect size estimate
between positive and negative controls; and 4) Coverage,
defined by how often the true IRR was within the 95%
confidence interval of the estimated IRR. Precision was measured using mean precision and mean
squared
error
(MSE). Geometric
mean
precision
was
computed as 1 / (standard error)^2, with higher precision
equivalent to narrower confidence intervals. MSE was obtained
from the log of estimated IRRs and the log of the true HR. Historical rates comparison had overall good discrimination
to distinguish true safety signals (i.e., positive control outcomes),
with AUCs of 80% or over in all the analyses and databases. Age-
sex adjustment and anchoring had little impact on this. Conversely, coverage was low, with many analyses failing to
accurately measure and include the true effect of our negative
and positive control outcomes (Supplementary Appendix S2). Coverage in unadjusted analyses ranged from 0 (H1N1 vaccines To understand the time it took the analysis method to identify
a safety signal (aka timeliness), the follow-up (up to 12 months)
occurring after each vaccine was divided into calendar months. For each month, the analyses were executed using the data
accumulated until the end of that month, and bias and
precision metrics were estimated. November 2021 | Volume 12 | Article 773875 4 Historical Comparisons for Vaccine Safety Li et al. FIGURE 1 | Type 1 and Type 2 error in unadjusted, age-sex adjusted, and anchored background rate analyses CCAE: IBM MarketScan Commercial Claims and
Encounters; MDCR: IBM Health MarketScan Medicare Supplemental; MDCD: IBM Health MarketScan Multi-state Medicaid; Optum EHR: Optum© de-identified
Electronic Health Record Dataset. FIGURE 1 | Type 1 and Type 2 error in unadjusted, age-sex adjusted, and anchored background rate analyses CCAE: IBM MarketScan Commercial Claims and
Encounters; MDCR: IBM Health MarketScan Medicare Supplemental; MDCD: IBM Health MarketScan Multi-state Medicaid; Optum EHR: Optum© de-identified
Electronic Health Record Dataset. The Effect of Empirical Calibration and uncalibrated estimates and p-values cannot be interpreted as
intended; while it may be encouraging that most positive effects can
be identified at a decision threshold of p < 0.05, this threshold will also
yield a substantial proportion of false positive findings. Age-sex
adjustment and anchoring background rate estimation around a
healthcare visit were useful strategies to reduce type 1 error to around
50%, while maintaining sensitivity. Empirical calibration led to
restoration of type 1 error to nominal but correction for positive
bias necessitates increasing type 2 error. In terms of timeliness,
background rate comparisons were sensitive methods for the early
identification of potential safety signals. However, most associations
were exaggerated and unstable in the first few months of vaccination
campaign. Vaccines with higher uptake, such as H1N1 or seasonal
flu, were associated with earlier identification of safety outcomes after
launch in the analyses of vaccines with rapid uptake like H1N1 or
seasonal influenza. Timeliness Most observed associations were unstable in the first few months
of study, and stabilised around the true effect size in the first
2–3 months after campaign initiation for vaccines with rapid
uptake like H1N1 or seasonal influenza. This stability was,
however, not seen until much later, and sometimes not seen at
all in the 12-months study period for vaccines with slower uptake
like HPV or varicella-zoster. This is depicted in Figure 3 using
data from CCAE as an illustrative example, and for all other
databases in Supplementary Figures S1–S3. Previous studies have shown that background incidence rates of
AESI vary between age and sex (Black et al., 2009). For example, the
incidence of Bell’s palsy in adults aged over 65 years is 4 times that in
paediatric population in the United Kingdom; whereas the risk of
optic neuritis is higher in females than males with the same age group
in Sweden. Therefore, it is crucial that age and sex are adjusted for
when using background incidence rates for comparison. Nonetheless,
Li et al. (Li et al., 2021) found considerable heterogeneity in incidence
rates of AESI within age-sex stratified subgroups. This suggests that Performance Metrics FIGURE 2 | Type 1 and type 2 error before vs after empirical calibration *CCAE: IBM MarketScan Commercial Claims and Encounters; MDCR: IBM Health
MarketScan Medicare Supplemental; MDCD: IBM Health MarketScan Multi-state Medicaid; Optum EHR: Optum© de-identified Electronic Health Record Dataset. FIGURE 2 | Type 1 and type 2 error before vs after empirical calibration *CCAE: IBM MarketScan Commercial Claims and Encounters; MDCR: IBM Health
MarketScan Medicare Supplemental; MDCD: IBM Health MarketScan Multi-state Medicaid; Optum EHR: Optum© de-identified Electronic Health Record Dataset. mean precision and MSE, varied by database and vaccine
exposure
as
reported
in
Supplementary
Appendix
S2. Adjustment for age and sex and anchoring improved precision
in most scenarios. in Optum EHR) to 0.51 (seasonal influenza vaccine in MDCR). Age-sex adjustment and anchoring had overall a positive effect on
coverage,
with
little
or
no
effect
on
discrimination
(Supplementary Appendix S2). Precision, as measured by November 2021 | Volume 12 | Article 773875 Frontiers in Pharmacology | www.frontiersin.org 5 Historical Comparisons for Vaccine Safety Li et al. FIGURE 3 | Observed effect size for negative control outcomes (true effect size 1) and positive control outcomes (true effect size 1.5, 2 and 4) [left Y axis] and
vaccine uptake [right Y axis and shaded orange area] over time in months [X axis] based on analyses of CCAE data with age-sex adjusted, and using the visit-anchored
time-at-risk definition. FIGURE 3 | Observed effect size for negative control outcomes (true effect size 1) and positive control outcomes (true effect size 1.5, 2 and 4) [left Y axis] and
vaccine uptake [right Y axis and shaded orange area] over time in months [X axis] based on analyses of CCAE data with age-sex adjusted, and using the visit-anchored
time-at-risk definition. The Effect of Empirical Calibration
Empirical calibration reduced type 1 error substantially, but
increased
type
2
error
in
all
the
tested
scenarios
(see
Figure 2). In addition to this, calibration improved coverage
without impacting AUC, and decreased precision in most
scenarios (Supplementary Appendix S3). AUTHOR’S NOTE The views expressed in this article are the personal views of the
authors and may not be understood or quoted as being made on
behalf of or reflecting the position of the European Medicines
Agency or one of its committees or working parties. Strengths and Limitations The strength of this study lies in the implementation of a
harmonised protocol across multiple databases, which allows
us to compare the findings across different healthcare systems. The use of a common data model allows the experiment to be
replicated in future databases while maintaining patient privacy
as patient-level data will not be shared outside of each institution. Use of real negative and synthetic positive control outcomes Future Research and Recommendations Future Research and Recommendations
When using background rate comparison for post-vaccine safety
surveillance, age-sex adjustment in combination with anchoring
time-at-risk around an outpatient visit resulted in somewhat
reduced type 1 error, without much impact on type 2 error. Residual bias, nonetheless, remained using this design, with very
high levels of type 1 error observed in most analyses. Calibration
is useful for reducing Type 1 error but at the expense of
decreasing precision and consequently increasing type II error. Future studies using cohort and SCCS self-controlled cased series
methods with empirical calibration will be evaluated. DATA AVAILABILITY STATEMENT The patient-level data used for these analyses cannot be provided
due to information governance restrictions. All our analytical
code is available to enable the replication of our analyses at
https://github.com/ohdsi-studies/Eumaeus. Key Results Our study found that unadjusted background rates comparison had
low type 2 error of <10% in all analyses but unacceptably high type 1
error, up to 100% in some scenarios. The method is positively biased November 2021 | Volume 12 | Article 773875 Frontiers in Pharmacology | www.frontiersin.org 6 Historical Comparisons for Vaccine Safety Li et al. Li et al. residual
patient-level
differences
in
characteristics
such
as
comorbidities and medication use remained. Background rates
comparison assumes that the background incidence in the overall
population is similar to the vaccinated population. This assumption
may not be valid because of confounding by indication, where the
vaccinated population has more chronic conditions than the
unvaccinated population. Conversely, the healthy vaccinee effect
could occur, where on average healthier patients are more likely
to adhere to annual influenza vaccination (Remschmidt et al., 2015). provides an independent estimate of residual bias in the study
design and data source. The fully specified study protocol was
published before analysis began and dissemination of the results
did not depend on estimated effects, thus avoiding publication
bias. All codes used to define the cohort, exposures, and outcomes
as well as analytical code are made open source to enhance
transparency and reproducibility. In our analysis, while using negative control outcomes can
reflect the real confounding and measurement error, the
approach of simulating positive control outcomes relied on
assumptions about systematic error. It is assumed that the
systematic error does not change when the true effect size is
greater than 1, rather than as a function of the true effect size. Furthermore, positive control synthesis assumes the positive
predictive value and sensitivity of the outcomes is the same
for background outcome events and the outcome events
simulated to be caused by the vaccine, which may not be true
in the real world (Schuemie et al., 2020). Research in Context For the Optum EHR data, we may miss the care episodes when
patients seeking care outside the respective health system, this will
cause bias towards the null. All these limitations needed to be
considered while interpreting our results. Research in Context Post-marketing surveillance is required to ensure the safety of
vaccines, so that the public do not avoid getting life-saving
vaccinations because of concerns that vaccine risks are not
monitored, and that any potential risks do not outweigh the
vaccine’s benefits. The goal of these surveillance systems is to
detect safety signals in a timely manner without raising excessive
false alarms. There is an implicit trade-off between sensitivity (type 2
error) and specificity (type 1 error). Claims extending from a false
positive result that is suggestive of an adverse event of a vaccine,
fueled by sensationalism and unbalanced reporting in the media,
could have devastating consequences on public health. A classic
example of harm is the link between the MMR vaccine and autism. Although the fraudulent report by Wakefield has been retracted and
many subsequent studies found no association, its lasting effects can
be seen in falling MMR vaccination rates below the recommend levels
from the World Health Organization (Godlee et al., 2011). Expert
consensus alleged that this was a contributing factor in measles being
declared endemic in the United Kingdom in 2008 (Jolley and
Douglas, 2014) and sporadic outbreaks in the United States in
recent years (Benecke and DeYoung, 2019). On the other hand,
missing safety signals could put patients at risk as well as dampen
public confidence in vaccination. Transparency is needed when
communicating vaccination results to the public. However, it is a
tricky balance to put both the benefits and harms of vaccination in
context. The urgency to act quickly on the basis of incomplete real-
world data could lead to confusion about vaccination safety. Negative
perceptions about vaccination can be deeply entrenched and difficult
to address. A starting point could be to include relevant background
rates to provide comparison to other scenarios. As reported in our
study, age and sex-adjusted rates are crucial to minimise false positive
safety signals. Another form of communication could be using
infographics to weigh harms versus benefits, illustrating the
differential risks in various age groups as was shown by
researchers from the University of Cambridge who contrasted the
prevention of ICU admissions due to COVID-19 against the risk of
blood clots due to the vaccine in specific age groups (Winton Centre
for Risk and Evidence Communication, 2021). REFERENCES Lieu, T. A., Kulldorff, M., Davis, R. L., Lewis, E. M., Weintraub, E., Yih, K., et al. (2007). Real-time Vaccine Safety Surveillance for the Early Detection of
Adverse Events. Med. Care 45 (10 Suppl. 2), S89–S95. doi:10.1097/
MLR.0b013e3180616c0a Barker, C. I., and Snape, M. D. (2014). Pandemic Influenza A H1N1 Vaccines and
Narcolepsy: Vaccine Safety Surveillance in Action. Lancet Infect. Dis. 14 (3),
227–238. doi:10.1016/S1473-3099(13)70238-X Mahaux, O., Bauchau, V., and Van Holle, L. (2016). Pharmacoepidemiological
Considerations
in
Observed-To-Expected
Analyses
for
Vaccines. Pharmacoepidemiol. Drug Saf. 25 (2), 215–222. doi:10.1002/pds.3918 Belongia, E. A., Irving, S. A., Shui, I. M., Kulldorff, M., Lewis, E., Yin, R., et al. (2010). Real-time Surveillance to Assess Risk of Intussusception and
Other Adverse Events after Pentavalent, Bovine-Derived Rotavirus
Vaccine. Pediatr. Infect. Dis. J. 29
(1),
1–5. doi:10.1097/
INF.0b013e3181af8605 Mesfin, Y. M., Cheng, A., Lawrie, J., and Buttery, J. (2019). Use of Routinely
Collected Electronic Healthcare Data for Postlicensure Vaccine Safety Signal
Detection: a Systematic Review. BMJ Glob. Health 4 (4), e001065. doi:10.1136/
bmjgh-2018-001065 Benecke, O., and DeYoung, S. E. (2019). Anti-Vaccine Decision-Making and
Measles
Resurgence
in
the
United
States. Glob. Pediatr. Health
6,
2333794X19862949–2333794X. doi:10.1177/2333794X19862949 OHDSI (2019). The Book of OHDSI: Observational Health Data Sciences and
Informatics. OHDSI. Available at: https://ohdsi.github.io/TheBookOfOhdsi/
TheBookOfOhdsi.pdf Black, S., Eskola, J., Siegrist, C. A., Halsey, N., MacDonald, N., Law, B., et al. (2009). Importance of Background Rates of Disease in Assessment of
Vaccine
Safety
during
Mass
Immunisation
with
Pandemic
H1N1
Influenza Vaccines. Lancet 374 (9707), 2115–2122. doi:10.1016/S0140-
6736(09)61877-8 Remschmidt, C., Wichmann, O., and Harder, T. (2015). Frequency and Impact of
Confounding by Indication and Healthy Vaccinee Bias in Observational Studies
Assessing Influenza Vaccine Effectiveness: a Systematic Review. BMC Infect. Dis. 15, 429. doi:10.1186/s12879-015-1154-y Schuemie, M. J., Cepeda, M. S., Suchard, M. A., Yang, J., Tian, Y., Schuler, A.,
et al. (2020). How Confident Are We about Observational Findings in
Healthcare: A Benchmark Study. Harv. Data Sci. Rev. 2 (1). doi:10.1162/
99608f92.147cc28e Buttery, J. P., Danchin, M. H., Lee, K. J., Carlin, J. B., McIntyre, P. B., Elliott, E. J.,
et
al. (2011). Intussusception
Following
Rotavirus
Vaccine
Administration:
post-marketing
Surveillance
in
the
National
Immunization
Program
in
Australia. Vaccine
29
(16),
3061–3066. doi:10.1016/j.vaccine.2011.01.088 Schuemie, M. J., Hripcsak, G., Ryan, P. B., Madigan, D., and Suchard, M. A. (2018). Empirical
Confidence
Interval
Calibration
for
Population-Level
Effect
Estimation Studies in Observational Healthcare Data. Proc. Natl. Acad. Sci. U S A. 115 (11), 2571–2577. ETHICS STATEMENT Ethical review and approval was not required for the study on
human participants in accordance with the local legislation and
institutional
requirements. Written
informed
consent
for November 2021 | Volume 12 | Article 773875 Frontiers in Pharmacology | www.frontiersin.org 7 Historical Comparisons for Vaccine Safety Li et al. participation was not required for this study in accordance with
the national legislation and the institutional requirements. the manuscript. All co-authors reviewed, provided feedback,
and finally gave approval for the submission of the current
version of the manuscript. FUNDING XL is a PhD student in the Pharmacoepidemiology group at the
Centre for Statistics in Medicine (CSM) (Oxford) and an MHS. She contributed to drafting the manuscript, creating tables and
figures, and reviewing relevant literature. LL has experience in
the analysis and interpretation of routine health data, and was
responsible for leading the literature review of similar studies,
and for the drafting of the first version of the manuscript. AO is
a PhD student and MD at Columbia University; she contributed
to the analysis and the drafting of the paper. FA is an
Undergraduate Student and contributed to the drafting of the
manuscript and reviewing the final version of the manuscript. MS is the guarantor of the study, and led the data analyses. DP is
also study guarantor and led the preparation and final review of UKNational Instituteof Health Research (NIHR), EuropeanMedicines
Agency, Innovative Medicines Initiative 2 (806968), US Food and Drug
Administration CBER BEST Initiative (75F40120D00039), and US
National Library of Medicine (R01 LM006910). Australian National
Health and Medical Research Council grant GNT1157506. UKNational Instituteof Health Research (NIHR), EuropeanMedicines
Agency, Innovative Medicines Initiative 2 (806968), US Food and Drug
Administration CBER BEST Initiative (75F40120D00039), and US
National Library of Medicine (R01 LM006910). Australian National
Health and Medical Research Council grant GNT1157506. SUPPLEMENTARY MATERIAL The Supplementary Material for this article can be found online at:
https://www.frontiersin.org/articles/10.3389/fphar.2021.773875/
full#supplementary-material REFERENCES doi:10.1073/pnas.1708282114 European Network of Centres of Pharmacoepidemiology and Pharmacovigilance
(2021). EUMAEUS: Evaluating Use of Methods for Adverse Event under
Surveillance (For Vaccines). 2021 [updated 15 April 2021. Available at:
http://www.encepp.eu/encepp/viewResource.htm?id40341. Schuemie, M. J., Hripcsak, G., Ryan, P. B., Madigan, D., and Suchard, M. A. (2016). Robust Empirical Calibration of P-Values Using Observational Data. Stat. Med. 35 (22), 3883–3888. doi:10.1002/sim.6977 Gee, J., Naleway, A., Shui, I., Baggs, J., Yin, R., Li, R., et al. (2011). Monitoring the
Safety of Quadrivalent Human Papillomavirus Vaccine: Findings from the
Vaccine
Safety
Datalink. Vaccine
29
(46),
8279–8284. doi:10.1016/
j.vaccine.2011.08.106 Wijnans, L., Lecomte, C., de Vries, C., Weibel, D., Sammon, C., Hviid, A., et al. (2013). The Incidence of Narcolepsy in Europe: before, during, and after the
Influenza A(H1N1)pdm09 Pandemic and Vaccination Campaigns. Vaccine 31
(8), 1246–1254. doi:10.1016/j.vaccine.2012.12.015 Godlee, F., Smith, J., and Marcovitch, H. (2011). Wakefield’s Article Linking
MMR Vaccine and Autism Was Fraudulent. BMJ 342, c7452. doi:10.1136/
bmj.c7452 Winton Centre for Risk and Evidence Communication (2021). Communicating
the Potential Benefits and Harms of the Astra-Zeneca COVID-19 Vaccine. 2021
[updated 7 April 2021. Available at: https://wintoncentre.maths.cam.ac.uk/
news/communicating-potential-benefits-and-harms-astra-zeneca-covid-19-
vaccine/. Jolley, D., and Douglas, K. M. (2014). The Effects of Anti-vaccine Conspiracy
Theories on Vaccination Intentions. PLOS ONE 9 (2), e89177. doi:10.1371/
journal.pone.0089177 Li, X., Ostropolets, A., Makadia, R., Shaoibi, A., Rao, G., Sena, A. G., et al. (2021). Characterizing the Incidence of Adverse Events of Special Interest for COVID-
19 Vaccines across Eight Countries: a Multinational Network Cohort Study. medRxiv, 2021. doi:10.1101/2021.03.25.21254315 Yih, W. K., Kulldorff, M., Fireman, B. H., Shui, I. M., Lewis, E. M., Klein, N. P., et al. (2011). Active Surveillance for Adverse Events: the Experience of the Vaccine
Safety Datalink Project. Pediatrics 127 (Suppl. 1), S54–S64. doi:10.1542/
peds.2010-1722I November 2021 | Volume 12 | Article 773875 Frontiers in Pharmacology | www.frontiersin.org 8 Historical Comparisons for Vaccine Safety Li et al. Li et al. Yih, W. K., Nordin, J. D., Kulldorff, M., Lewis, E., Lieu, T. A., Shi, P., et al. (2009). An Assessment of the Safety of Adolescent and Adult Tetanus-Diphtheria-
Acellular Pertussis (Tdap) Vaccine, Using Active Surveillance for Adverse
Events in the Vaccine Safety Datalink. Vaccine 27 (32), 4257–4262. doi:10.1016/j.vaccine.2009.05.036 Publisher’s Note: All claims expressed in this article are solely those of the authors
and do not necessarily represent those of their affiliated organizations, or those of
the publisher, the editors and the reviewers. Frontiers in Pharmacology | www.frontiersin.org REFERENCES Any product that may be evaluated in
this article, orclaim that may be made by its manufacturer, is not guaranteed or
endorsed by the publisher. Conflict of Interest: DP's research group has received grants for unrelated work
from Amgen, Chiesi-Taylor, and UCB Biopharma SRL; and his department has
received speaker/consultancy fees from Amgen, Astra-Zeneca, Astellas, Janssen,
and UCB Bioparhma SRL. Authors PR and MJS were employed by Janssen R&D. Copyright © 2021 Li, Lai, Ostropolets, Arshad, Tan, Casajust, Alshammari, Duarte-
Salles, Minty, Areia, Pratt, Ryan, Hripcsak, Suchard, Schuemie and Prieto-
Alhambra. This is an open-access article distributed under the terms of the
Creative Commons Attribution License (CC BY). The use, distribution or
reproduction in other forums is permitted, provided the original author(s) and
the copyright owner(s) are credited and that the original publication in this journal is
cited, in accordance with accepted academic practice. No use, distribution or
reproduction is permitted which does not comply with these terms. The remaining authors declare that the research was conducted in the absence of
any commercial or financial relationships that could be construed as a potential
conflict of interest. November 2021 | Volume 12 | Article 773875 Frontiers in Pharmacology | www.frontiersin.org 9
|
https://openalex.org/W1518674354
|
https://hal.science/hal-01171571/document
|
English
| null |
A Simplified Formulation for Rough Surface Cross-Polarized Backscattering Under the Second-Order Small-Slope Approximation
|
IEEE transactions on geoscience and remote sensing
| 2,015
|
cc-by
| 5,969
|
A Simplified Formulation for Rough Surface
Cross-Polarized Backscattering Under the Second-Order
Small-Slope Approximation
Charles-Antoine Guérin, Joel T. Johnson A Simplified Formulation for Rough Surface
Cross-Polarized Backscattering Under the Second-Order
Small-Slope Approximation
Charles-Antoine Guérin, Joel T. Johnson Charles-Antoine Guérin, Joel T. Johnson To cite this version: Charles-Antoine Guérin, Joel T. Johnson. A Simplified Formulation for Rough Surface Cross-Polarized
Backscattering Under the Second-Order Small-Slope Approximation . IEEE Transactions on Geo-
science and Remote Sensing, 2015, 53 (11), pp.6308-6314. 10.1109/TGRS.2015.2440443. hal-
01171571 Distributed under a Creative Commons Attribution 4.0 International License A Simplified Formulation for Rough Surface
Cross-Polarized Backscattering Under the
Second-Order Small-Slope Approximation Charles-Antoine Guérin and Joel T. Johnson, Fellow, IEEE Computation of SSA2 cross-polarized backscatter predictions,
however, remains difficult and computationally demanding be-
cause it requires integrations in both the space (over the surface
correlation function) and wavenumber (over the surface power
spectrum multiplied with an SSA2 “kernel” function) domains,
with a slowly decreasing and oscillating integrand in space. Abstract—We present simplified expressions for the cross-
polarized backscatter of a randomly rough surface predicted by
the second-order small-slope approximation (SSA2). The simplifi-
cation is based on appropriate polynomial approximations of the
SSA2 kernel function. We obtain numerically efficient expressions
for the cross-polarized backscattering amplitude of a deterministic
surface in the form of a single space integral involving only the
surface elevation and the second (mixed) derivative of the surface
elevation. The ensemble average normalized radar cross section
is then derived under a Gaussian random process assumption for
the surface. The resulting expression has the form of a Kirchhoff
integral involving the roughness correlation function and its
second- and fourth-order cross-derivatives. Further simplification
is achieved for off-nadir observations using a high-frequency ap-
proximation; the result is an analytical formula involving only the
resonant curvature and the radar-filtered mean square slope in the
out-of-plane direction. A numerical validation of the simplified ex-
pressions is provided by comparison with exact SSA2 predictions
in representative test cases. The dependence of cross-polarized
backscattering on the incidence angle as well as wind speed and
direction is then investigated for the case of a directional sea
surface model. At near nadir incidence, a clear maximum in
azimuth of the cross-polarized backscatter is observed for radar
look directions 45◦from the wind direction. One decade ago, it was shown that the copolarized computa-
tion of the SSA2 can be drastically simplified in the so-called
high-frequency approximation [9], under which a quadratic
approximation of the SSA2 kernel function makes it possible to
perform the integration in wavenumber analytically. In this pa-
per, we pursue a similar approach for cross-polarized backscat-
tering to again obtain an approximation of the full SSA2 that
involves only a single Kirchhoff-type integral in space. To
obtain more accurate predictions, differing approximations are
used depending on whether the incidence angle is close to or
away from nadir. HAL Id: hal-01171571
https://hal.science/hal-01171571v1
Submitted on 10 Jul 2016 L’archive ouverte pluridisciplinaire HAL, est
destinée au dépôt et à la diffusion de documents
scientifiques de niveau recherche, publiés ou non,
émanant des établissements d’enseignement et de
recherche français ou étrangers, des laboratoires
publics ou privés. HAL is a multi-disciplinary open access
archive for the deposit and dissemination of sci-
entific research documents, whether they are pub-
lished or not. The documents may come from
teaching and research institutions in France or
abroad, or from public or private research centers. Distributed under a Creative Commons Attribution 4.0 International License 1 I. INTRODUCTION This paper is organized as follows. General formulas and
notations for the SSA2 technique are recalled in Section II,
and the approximation of the SSA2 integral for cross-polarized
backscatter is developed in Section III. The corresponding
ensemble averaged normalized radar cross section (NRCS) is
obtained in Section IV under the assumption of a Gaussian
random process surface. High-frequency approximation of the
off-nadir NRCS is then described in Section V, and validation
tests and numerical illustrations are provided in Section VI. T T
HE use of the cross-polarized backscattering coefficient
of the ocean is of increasing recent interest, as it has been
found to be a useful proxy for wind speed, especially in high
sea states [1]–[3]. However, the interpretation and modeling of
rough surface cross-polarized backscattering is still challenging
as it involves multiple scattering as well as out-of-plane tilt-
ing effects which cannot be simultaneously accounted for by
simple analytical models [4]. Today, one of the few scattering
models capable of predicting cross-polarized backscatter is
the second-order small-slope approximation (SSA2) [5]–[8]. A Simplified Formulation for Rough Surface
Cross-Polarized Backscattering Under the
Second-Order Small-Slope Approximation In the latter case, we further show that an
additional high-frequency approximation can be used to reduce
the results to an elementary analytical formula involving the
surface spectrum at the Bragg resonant wavenumber and the
surface mean square slope. We validate the method by compar-
ison with a set of exact SSA2 computations for representative
test cases and examine the dependence of the results on wind
speed and direction. Index Terms—Cross-polarization, microwave remote sensing,
sea surface scattering. C.-A. Guárin is with the Université de Toulon at the Mediterranean Institute
of Oceanography (MIO, UM 110 UTLN/AMU/CNRS-INSU/IRD), 83957
83957 La Garde, France (e-mail: guerin@univ-tln.fr).
J. T. Johnson is with the Department of Electrical and Computer Engineering,
The Ohio State University, Columbus, OH 43210 USA. C.-A. Guárin is with the Université de Toulon at the Mediterranean Institute
of Oceanography (MIO, UM 110 UTLN/AMU/CNRS-INSU/IRD), 83957
83957 La Garde, France (e-mail: guerin@univ-tln.fr).
J. T. Johnson is with the Department of Electrical and Computer Engineering,
The Ohio State University, Columbus, OH 43210 USA. A. Off-Nadir Approximation We have used the standard conventions (e.g., [4]) for the
horizontal (QH = k −k0) and vertical (Qz = q + q0) compo-
nents of the Ewald vector Q = K −K0. The involved kernels
are, respectively, the first- and second-order Bragg kernels
B1(k, k0) and B2(k, k0, ξ), respectively, and a combination of
the latter The M12(ξ) kernel function in (II.1) multiplies η(ξ) in a
Fourier transform over ξ. An expansion of M12(ξ) that assumes
small amplitude for ξ appears appropriate for surfaces whose
roughness occurs primarily on length scales long compared to
the electromagnetic wavelength (i.e., assuming η(ξ) is larger
for small ξ amplitudes). Due to the known importance of
Bragg scattering effects for the sea surface, an approximation
of small ξ would seem undesirable in computing off-nadir sea
backscatter. However, it will be shown in what follows that this
approximation of the kernel function, when applied to the entire
range of ξ values, yields an acceptable approximation of the
complete SSA2 sea surface cross-polarized backscatter for off-
nadir angles. M(k, k0, ξ) = 1
2 (B2(k, k0, k −ξ) + B2(k, k0, k0 + ξ)
−2QzB1(k, k0)) . (II.4 M(k, k0, ξ) = 1
2 (B2(k, k0, k −ξ) + B2(k, k0, k0 + ξ) −2QzB1(k, k0)) . (II.4) We refer to [7] for the explicit expressions of these kernels
(which must be, however, corrected for a conventional factor
qq0 due to a different definition of the scattering amplitude). One of the difficulties in the numerical computation of exact
SSA2 predictions is the presence of terms of the type qξ =
K2 −ξ2 and q′
ξ =
εK2 −ξ2 in the denominator of the
kernel M which produce sharp maxima in wavenumber space. Accordingly, we assume (ξ ≪K) along with ∥k0 + ξ∥<
K, which holds at moderate incidence (k0 ≪K) and suffi-
ciently high wavenumber. Performing a Taylor expansion of the
qξ and q′
ξ variables about the origin, we obtain 1
εq(k0+ξ) + q′
(k0+ξ)
≃
1
εK + K′ ×
1+γ
1
2
k2
0
K2 + k0 · ξ
K2
(III.9) II. SSA2 SCATTERING AMPLITUDE The sea surface is described by a centered random function
z = η(x, y) in a Cartesian coordinate system with the z-axis
directed upward. The surface is illuminated from above by an
incident monochromatic plane wave (e−iωt time dependence)
with impinging wave vector K0 = (k0, −q0) and scattered
with outgoing wave vector K = (k, +q). The upper medium is
air described with vacuum wavenumber K. The lower medium
is described by a homogeneous complex relative permittivity ε
and a complex wavenumber K′ = √εK. The horizontal axis is 2 chosen in such a way that the radar look direction is along x
(k0 ∥x). The SSA2 scattering amplitude [5]–[8] is then and z[ξ, k0] denoting the vertical component of the cross prod-
uct between ξ and k0, i.e., ˆz · (ξ × k0). Due to cancellations
when the two (B2)12 terms (which have arguments of equal
amplitude but opposite direction) in (III.5) are summed, we
obtain S2=S1−i
Qz
1
(2π)2
drdξe−iQH·re−iQzηeiξ·rη(ξ)M(k, k0, ξ)
(II.1) (II.1) M12(ξ) =
2q0q′
0(ε −1)A0
εq(k0+ξ) + q′
(k0+ξ)
z[ξ, k0]
k0 + (k0ξ)
. (II where S1 is the first-order scattering amplitude (SSA1) S1 = 1
Qz
B1
1
(2π)2
dre−iQH·re−iQzη
(II.2) (II.2) (III.8) To reduce the complexity of the SSA2 integral, we seek an
accurate polynomial approximation of this kernel. For this, we
distinguish the nadir and off-nadir incidence angles. and η(ξ) is the Fourier transform of roughness and η(ξ) is the Fourier transform of roughness η(ξ) =
1
(2π)2
dre−iξ·rη(r). (II.3) (II.3) A. Off-Nadir Case Using standard techniques, we obtain Using standard techniques, we obtain z[ξ, k0](k0 · ξ) = −ξxξy. (III.15) σ12,offnad = |Gγ|2 1
π
dr e−iQH·r Using this approximation, the integration over ξ in (II.1) can
now be performed following (III.13), and the result is inserted
into (II.1) to obtain ×
∂xyη(r)∂xyη(0)e−iQz(η(r)−η(0))
(IV.23) (IV.23) (S2)12 = Gγ
1
(2π)2
dr ∂xyη(r)e−iQH·re−iQzη
(III.16) where the brackets ⟨⋆⟩denote the ensemble average. Stan-
dard manipulations on Gaussian processes (see the Appendix)
lead to with Gγ = −i(ε −1)2
ε + √ε
q0q′
0
(εq0 + q′
0) (q0 + q′
0)
1 + 3γk2
0
2K2
. (III.17)
Note that, in the limit of a perfectly conducting surface (ε →
+∞) this coefficient reduces to
∂xyη(r)∂xyη(0)e−iQz(η(r)−η(0))
=
∂xxyyρ(r) + Q2
z [∂xyρ(0)−∂xyρ(r)]2
e−Q2
z(ρ(0)−ρ(r))
(IV.24) ∂xyη(r)∂xyη(0)e−iQz(η(r)−η(0))
=
∂xxyyρ(r) + Q2
z [∂xyρ(0)−∂xyρ(r)]2
e−Q2
z(ρ(0)−ρ(r))
(IV.24) Gγ = −i(ε −1)2
ε + √ε
q0q′
0
(εq0 + q′
0) (q0 + q′
0)
1 + 3γk2
0
2K2
. (III. (III.17) Note that, in the limit of a perfectly conducting surface (ε →
+∞), this coefficient reduces to (IV.24) Note that, in the limit of a perfectly conducting surface (ε →
+∞), this coefficient reduces to for this correlator, where ρ is the autocorrelation function of
elevation G1 = −i
1 + 3k2
0
2K2
. (III.18) ρ(r) = ⟨η(r)η(0)⟩
(IV.25) (IV.25) that is the inverse Fourier transform of the power spectrum of
elevation B. Nadir Case For near-nadir angles where the condition ξ ≪k0 is no
longer valid, it is difficult to obtain a polynomial approximation
for the kernel M12. Instead, we use the approximation ρ(r) =
dξ Γ(ξ)e−ik·ξ. (IV.26) (IV.26) The cross-polarized off-nadir cross section therefore takes the
final form The cross-polarized off-nadir cross section therefore takes the
final form 1
εq(k0+ξ) + q′
(k0+ξ)
≃
1
εqξ + q′
ξ
(III.19) (III.19) σ12,offnad = |Gγ|2 1
π
dr e−iQH·re−Q2
z(ρ(0)−ρ(r))
×
∂xxyyρ(r) + Q2
z [∂xyρ(0) −∂xyρ(r)]2
. (IV.27) which is exact for k0 = 0 (nadir incidence). We then obtain which is exact for k0 = 0 (nadir incidence). We then obtain M12(ξ) ≃2q0q′
0(ε −1)A0
εqξ + q′
ξ
z[ξ, k0](k0ξ). (III.20) (IV.27) This expression is similar to the previous polynomial expres-
sion (III.11) with γ = 0 and α = 0. However, the presence
of qξ and q′
ξ in the denominator complicates the integration
over ξ. This issue can be addressed by introducing a modified
roughness function ˜η defined by the Fourier transform IV. STATISTICAL EXPRESSION Hence, z[ξ, k0] can be replaced by z[(QH/Qz), k0] = 0, so
that there is no contribution of the α term of the kernel to the
Kirchhoff integral. The preceding section provided formulations for a determin-
istic surface. We now compute the ensemble average cross-
polarized backscattering cross section assuming a Gaussian
random process surface, both for the nadir and nonnadir cases. The leading term is therefore the β term. Without loss of
generality, we can assume that k0 is directed along the x-axis,
i.e., k0 = x, so that the quadratic factor in the cross-polarization
kernel reduces to III. BACKSCATTERING CROSS-POLARIZED COEFFICIENTS We will now focus on the particular but important back-
scattering configuration (k = −k0, q = q0). For simplicity, we
adopt the notations M12(ξ) and (B2)12(ξ) to represent the
cross-polarized components of M(−k0, k0, ξ) and B2(−k0,
k0, ξ), respectively. with γ = ε3/2 + 1
ε3/2 + ε. (III.10) (III.10) It is well known that the cross-polarized components of the
first-order Bragg tensor vanish for backscattering. Hence, the
cross-polarized component (S2)12 of SSA2 is merely given by
its second-order contribution in (II.1) with This implies that the kernel M12 can be approximated by a
second-order polynomial of the variable ξ This implies that the kernel M12 can be approximated by a
second-order polynomial of the variable ξ M12(ξ) = 1
2 ((B2)12(−k0 −ξ) + (B2)12(k0 + ξ)) . (III.5) M12(ξ) ≃αz[ξ, k0] + βz[ξ, k0](k0ξ)
(III.11) (III.11) with The expression for the B2 kernel for backscattering is given by
([5] and [7], corrected for a conventional factor q2
0) The expression for the B2 kernel for backscattering is given by
([5] and [7], corrected for a conventional factor q2
0)
with α = 2q0q′
0(ε −1)
εK + K′
A0k0
1 + γk2
0
2K2
β = 2q0q′
0(ε −1)
εK + K′
A0
1 + 3γk2
0
2K2
. (III.12) (B2)12(ξ) = 2q0A0z[ξ, k0]
×
ε −1
εqξ + qξ
q′
0(k0 · ξ) −ε
qξ + q′
ξ
εqξ + q′
ξ
k0
(III.6) (III.12) The first term in (III.11) is linear in ξ. Any linear term in ξ in
the SSA2 integral (II.1) corresponds to a gradient of roughness The first term in (III.11) is linear in ξ. Any linear term in ξ in
the SSA2 integral (II.1) corresponds to a gradient of roughness with A0 =
(ε −1)q0K
(εq0 + q′
0) (q0 + q′
0) A0 =
(ε −1)q0K
(εq0 + q′
0) (q0 + q′
0)
(III.7)
∇η(r) = i
dξeiξ·rξη(ξ). (III.13) ∇η(r) = i
dξeiξ·rξη(ξ). (III.13) (III.7)
∇η(r) = i
dξeiξ·rξη(ξ). (III.13) (III.7) (III.13) 3 3 3 Any gradient of roughness involved in the Kirchhoff integral
can be integrated by parts where G0 is equal to Gγ in (III.16) with γ = 0 (e.g., in
the perfectly conducting case, G0 = −i). Both approximations
(III.19) and (III.22) are exact at nadir (k0 = 0).
(∇η)e−iQH·re−iQzη =−QH
Qz
e−iQH·re−iQzη. (III.14) B. Nadir Case The derivation of the NRCS from the scattering amplitude
(III.22) is analogous σ12 = |G0|2 1
π
dr e−iQH·r
×
∂xy˜η(r)∂xy˜η(0)e−iQz(η(r)−η(0))
. (IV.28) σ12 = |G0|2 1
π
dr e−iQH·r ˜η(ξ) = εK + K′
εqξ + q′
ξ
η(ξ). (III.21) (III.21) (III.21) (IV.28) Using the modified roughness, the cross-polarized scattering
amplitude is found analogously to the previous case as The correlator ⟨⋆⟩is now given (see the Appendix) by
∂xxyy˜˜ρ(r) + Q2
z |∂xy ˜ρ(0) −∂xy ˜ρ(r)|2
× e−Q2
z(ρ(0)−ρ(r)) The correlator ⟨⋆⟩is now given (see the Appendix) by
∂xxyy˜˜ρ(r) + Q2
z |∂xy ˜ρ(0) −∂xy ˜ρ(r)|2
× e−Q2
z(ρ(0)−ρ(r))
(IV.29) The correlator ⟨⋆⟩is now given (see the Appendix) by ⟨⟩
g
(
pp
) y
∂xxyy˜˜ρ(r) + Q2
z |∂xy ˜ρ(0) −∂xy ˜ρ(r)|2
× e−Q2
z(ρ(0)−ρ(r))
∂xxyy˜˜ρ(r) + Q2
z |∂xy ˜ρ(0) −∂xy ˜ρ(r)|2
× e−Q2
z(ρ(0)−ρ(r)) (S2)12 = G0
1
(2π)2
dr ∂xy˜η(r)e−iQH·re−iQzη
(III.22) 4 where ˜˜ρ is the autocorrelation function associated to the mod-
ified roughness ˜η and ˜ρ is the cross-covariance function of η
and ˜η where B(ξ) = ξ4Γ(ξ) is the curvature spectrum. The function
F is a Kirchhoff integral, hence a positive and rapidly de-
creasing function with its maximum at the origin which acts
as a sharp filter around the Bragg frequency QH. To push
the calculation further, we approximate the function F by a
Gaussian shape F0 about the origin ˜˜ρ(r) =
dξ
εK + K′
εqξ + q′
ξ
2
Γ(ξ)e−ik·ξ
˜ρ(r) =
dξ εK + K′
εqξ + q′
ξ
Γ(ξ)e−ik·ξ. (IV.30 Gaussian shape F0 about the origin F(r) ≃F0(r) = exp
−1
2
mssxx2 + mssyy2
(V.36) (IV.30) (V.36) Note that ˜ρ is complex. Altogether, this leads to the following
expression for the cross-polarized cross section at nadir: where the shape parameters mssx/y are optimized in order to
provide a good match of the respective Fourier transforms F(ξ)
and F0(ξ) at ξ = QH. As it is well known from the geometrical
optics approximation, the shape parameters are on the order of
the radar-filtered directional mss σ12,nad = |G0|2 1
π
dr e−iQH·re−Q2
z(ρ(0)−ρ(r))
×
∂xxyy˜˜ρ(r) + Q2
z |∂xy ˜ρ(0) −∂xy ˜ρ(r)|2
. (IV.31) mssx/y =
ξ≤K
dξxdξyξ2
x/yΓ(ξ). (V.37) (V.37) C. General Case The previous sections provide approximations for the near-
nadir and off-nadir regions. We propose a combination of these
approximations for use at general angles as On the other hand, B(ξ), ξ2
x, and ξ2 are slowly varying
functions which can thus be approximated by their value at
(ξx, ξy) = (QH, 0) in the integral, provided the filter is sharp
enough to cutoff spatial frequencies which are below the
spectral peak wavenumber. We therefore have the following
approximation: On the other hand, B(ξ), ξ2
x, and ξ2 are slowly varying
functions which can thus be approximated by their value at
(ξx, ξy) = (QH, 0) in the integral, provided the filter is sharp
enough to cutoff spatial frequencies which are below the
spectral peak wavenumber. We therefore have the following
approximation: σ12 =σ12,nad exp(−a tan2 θ)+(1−exp(−a tan2 θ))σ12,offnad
(IV.32) for a specified a which controls the angular width of the tran-
sition region between the two approximations. In the numerical
results shown, we have chosen the value a = 25 which limits
the nadir correction to about 20◦. The expressions for σ12,nad
and σ12,offnad can be evaluated numerically at the cost of a
single Kirchhoff integral, as opposed to the previous fourfold
integration required for the SSA2. This simplification is the
main result of this paper. σ12 ∼4π |Gγ|2Q−2
H B(QH)
dξxdξy F0(QH −ξ)ξ2
y
(V.38) (V.38) obtained by operating the replacements ξ2
x/ξ2 →1 and
ξ2
y/ξ2 →ξ2
y/Q2
H in the previous integral. Since QH · y = 0,
we have with a simple change of variables ξ →QH −ξ σ12 = 4π |Gγ|2Q−2
H B(QH)
dξxdξy F0(ξ)ξ2
y
(V.39) which leads in the end to Further simplification can be achieved in the high-frequency
regime (that is for large K) in the off-nadir domain using a
technique which was first introduced in [10] and termed the
“Kirchhoff filtering formula.” The squared term in (IV.24) has
a quadratic dependence ∼r2 about the origin, while |∂xxyyρ|
is maximum at zero. At high frequency where the effective
integration domain is a small interval around zero, the former
term is thus negligible, and it will be ignored in the integral
(IV.27). This is confirmed by numerical evidence. Now denote σ12 = 4π|Gγ|2cotan2(θi) Q4
HΓ(QH)mssy
(V.40) (V.40) where mssy is the radar-filtered mean square slope in the direc-
tion perpendicular to the incident wave direction. The simple
formula (IV.40) is the second main result of this paper. The
explicit appearance of the cross slope shows the importance of
the out of the incidence plane tilting in the generation of cross-
polarized power. The result represents a combination of Bragg
scatter effects [due to the presence of Γ(QH)] and “long wave
tilting” (due to the presence of mssy) and can be interpreted
as an approximation of the “two-scale” theory of sea surface
scattering in which it is assumed that long wave slopes are
small and that the sea curvature spectrum is slowly varying in
the vicinity of the Bragg wavenumber. Upon inspection of the
different approximations which have been used in deriving this
simple analytical formula, we see that it is expected to hold in
the high-frequency regime (that is for large Rayleigh parameter
Qz√ρ0) and at off-nadir incidence. F(r) = e−Q2
z(ρ0−ρ(r)). (V.33) F(r) = e−Q2
z(ρ0−ρ(r)). (V.33) Then, by the convolution theorem, we may rewrite Then, by the convolution theorem, we may rewrite σ12 =4π |Gγ|2
dξxdξy F(QH −ξ)ξ2
xξ2
yΓ(ξx, ξy)
(V.34) or equivalently σ12 =4π |Gγ|2
dξxdξy F(QH −ξ)ξ2
x
ξ2
ξ2
y
ξ2 B(ξx, ξy)
(V.35) Fig. 1. Approximated SSA2 cross-polarized NRCS according to formula
(IV.32) (G&J) and comparison with the exact SSA2 at different EM bands. The sea state is described by an isotropic Elfouhaily spectrum at 7 m/s. Fig. 2. Approximated SSA2 cross-polarized NRCS according to formulas
(IV.32) (G&J) and comparison with the exact SSA2 in X band. Superimposed
are the contributions of the nadir (IV.31) and off-nadir (IV.27) formulas. The
sea state is described by an isotropic Elfouhaily spectrum at 10 m/s. 5 5 Fig. 2. A. Isotropic Spectra Fig. 1 compares the full and simplified SSA2 cross-polarized
NRCS versus incidence angle for an isotropic Elfouhaily spec-
trum at 7-m/s wind speed. L band (λEM = 23.8 cm, ε =
75 + i61), C band (λEM = 6 cm, ε = 67 + i36), and Ku band
(λEM = 2.143 cm, ε = 42 + i39.5) frequencies are included. The agreement is found excellent over a wide range of inci-
dence angles spanning from nadir to about 60◦. VI. NUMERICAL TESTS Fig. 3. Approximation of SSA2 cross-polarized NRCS according to the ana-
lytical approximation (V.40) (SSA2-A) and comparison with the exact SSA2 in
different bands. The sea state is described by an isotropic Elfouhaily spectrum
at 7 m/s in L and Ku bands and 5 m/s in Ka band. The simplified formula (IV.32) has been implemented and
compared with an exact SSA2 computation in the case of
an Elfouhaily sea surface spectrum [11]. The technique of
implementation of the full SSA2 integral has been described
in detail elsewhere [12]. The particular method used for the
full SSA2 computation requires increasing memory storage
as the spatial integral size increases (that is wind speed) and
requires sampling on the scale of the electromagnetic wave-
length. Therefore, the comparisons were limited to moderate
wind speeds. For example, with a 64-GB-RAM computer, we
could run 7-m/s wind speed in L, C, and Ku bands. Using the
maximum available memory and a much longer computational
time, we pushed the calculation to 5-m/s wind speed in Ka band
and also ran a 10-m/s wind speed case at X band. Fig. 3. Approximation of SSA2 cross-polarized NRCS according to the ana-
lytical approximation (V.40) (SSA2-A) and comparison with the exact SSA2 in
different bands. The sea state is described by an isotropic Elfouhaily spectrum
at 7 m/s in L and Ku bands and 5 m/s in Ka band. which leads in the end to Approximated SSA2 cross-polarized NRCS according to formulas
(IV.32) (G&J) and comparison with the exact SSA2 in X band. Superimposed
are the contributions of the nadir (IV.31) and off-nadir (IV.27) formulas. The
sea state is described by an isotropic Elfouhaily spectrum at 10 m/s. Fig. 1. Approximated SSA2 cross-polarized NRCS according to formula
(IV.32) (G&J) and comparison with the exact SSA2 at different EM bands. The sea state is described by an isotropic Elfouhaily spectrum at 7 m/s. Fig. 1. Approximated SSA2 cross-polarized NRCS according to formula
(IV.32) (G&J) and comparison with the exact SSA2 at different EM bands. The sea state is described by an isotropic Elfouhaily spectrum at 7 m/s Fig. 2. Approximated SSA2 cross-polarized NRCS according to formulas
(IV.32) (G&J) and comparison with the exact SSA2 in X band. Superimposed
are the contributions of the nadir (IV31) and off-nadir (IV27) formulas The Fig. 2. Approximated SSA2 cross-polarized NRCS according to formulas
(IV.32) (G&J) and comparison with the exact SSA2 in X band. Superimposed
are the contributions of the nadir (IV.31) and off-nadir (IV.27) formulas. The
sea state is described by an isotropic Elfouhaily spectrum at 10 m/s. Fig. 1. Approximated SSA2 cross-polarized NRCS according to formula
(IV.32) (G&J) and comparison with the exact SSA2 at different EM bands. The sea state is described by an isotropic Elfouhaily spectrum at 7 m/s. A. Calculation of the Correlators A. Calculation of the Correlators To obtain (IV.29) (as well as (IV.24) which is a particular
case), we proceed in the following way. We consider the four-
point characteristic function C(α, β) = ⟨e−iQz(η−η0)+iα∂xy ˜η+iβ∂xy ˜η0⟩
(A41) (A41) where the dependence on position r is implicit and the 0
subscript refers to r = 0. We observe that the correlator (IV.29)
is given by where the dependence on position r is implicit and the 0
subscript refers to r = 0. We observe that the correlator (IV.29)
is given by −∂αβC(α, β)|α=β=0. (A42) (A42) Since the roughness processes η and ˜η are Gaussian, this
amounts to evaluating Fig. 4. Approximated SSA2 cross-polarized NRCS according to formula
(IV.32) (G & J) and comparison with the exact SSA2 X band data from [8]
(V & Z„ courtesy of V. Zavorotny and A. Voronovich). The sea state is
described by a directional Elfouhaily spectrum at 15 m/s. C(α, β) = e−1
2⟨|Qz(η−η0)−α∂xy ˜η−β∂xy ˜η0|2⟩
(A43) (A43) which involves the roughness autocorrelation function (IV.25)
as well as the modified roughness autocorrelation function
˜˜ρ(r) = ⟨˜η∗(r)˜η(0)⟩and the cross-correlation function ˜ρ(r) =
⟨˜η∗(r)η(0)⟩(the latter can be complex). Straightforward calcu-
lations then lead to the expression (IV.29). Fig. 5. Approximated SSA2 cross-polarized NRCS according to formula
(IV.32) (G & J) as a function of wind speed for two incidences (0 and 45◦). The sea state is described by a directional Elfouhaily spectrum. B. Implementation of the Kirchhoff Integrals We propose an efficient numerical implementation of the
integral (IV.27) in the case of biharmonic spectra of the form Γ(k) = Γ(k, ϕ) = Γ0(k) + Γ2(k) cos(2ϕ −2ϕw)
(A44) (A44) where ϕk is the angle of the vector k with respect to the incident
wave vector direction (which is taken to be the x axis) and
ϕw is the angle of the wind vector with respect to the same
reference. The implementation of the integral (IV.31) is similar
when using the modified spectra described in Section IV-B. Fig. 5. Approximated SSA2 cross-polarized NRCS according to formula
(IV.32) (G & J) as a function of wind speed for two incidences (0 and 45◦). The sea state is described by a directional Elfouhaily spectrum. g
p
We denote by Bn[f](u) the Bessel transform of a function f
at nth order, which is explicitly B. Anisotropic Spectra The dependence on both wind speed and direction has also
been investigated with a directional Elfouhaily spectrum (see
the Appendix). A comparison with the exact SSA2 calculations
presented in [8] is displayed in Fig. 4 for a wind speed of
15 m/s in X band. The method of performing the exact SSA2
computations in this reference allows higher wind speeds to be
considered. A good agreement is found both at nadir and larger
incidence angles with a maximum of 1-dB error at intermediate
incidence angles. Fig. 2 illustrates similar comparisons for an isotropic
Elfouhaily spectrum at 10-m/s wind speed in X band (λEM =
3 cm, ε = 60.63 + i44.97). To clarify the respective contribu-
tions of the “nadir” and “off nadir” parts, we have plotted the
results of formulas (IV.31) and (IV.27) separately and show
that an appropriate combination of both is necessary to remain
accurate over a wide range of incidence angles. Fig. 5 plots the predicted NRCS as a function of wind
speed for the upwind, crosswind, and 45◦azimuth direction
for both nadir and 45◦incidence angle. At nadir, a maximum
of the cross-polarized NRCS is found for azimuth angle 45◦
with respect to the radar polarization vector. At 45◦incidence,
the qualitative behaviors are similar to the copolarized case,
with a maximum in up/downwind direction and a minimum in
crosswind direction. Fig. 3 shows the high-frequency off-nadir approximation
(V.40) for the L, Ku, and Ka bands (λEM = 8 mm, ε = 15 +
i26) for the same spectrum at 7 m/s (L and Ku bands) and
5 m/s (Ka band). The exact SSA2 calculation is superimposed. The analytical approximation of SSA2 is found in overall very
good agreement with the latter beyond 30◦incidence and is
increasingly accurate at higher radar frequencies. 6 Fig. 4. Approximated SSA2 cross-polarized NRCS according to formula
(IV.32) (G & J) and comparison with the exact SSA2 X band data from [8]
(V & Z„ courtesy of V. Zavorotny and A. Voronovich). The sea state is
described by a directional Elfouhaily spectrum at 15 m/s. APPENDIX VII. CONCLUSION Bn[Γ](r) =
∞
0
dk jn(kr)2πk Γ(k)
(A45) (A45) We have derived a simplified formulation for rough sur-
face cross-polarized backscattering under the SSA2 analytical
model. This makes the SSA2 more tractable for numerical
applications, at the price of an approximation which has been
found reasonable for a variety of wind speeds and electro-
magnetic frequencies at nongrazing incidence angles, particu-
larly for near-nadir and moderately large incidence angles of
30–60◦. For Ka band frequencies and higher, an additional
approximation was found, which involved only the resonant
Bragg frequency of the sea surface spectrum and its cross-plane
mss, which unveils the specific contribution of the out-of-plane
tilting in the cross-polarization mechanism away from nadir. Further tests and comparisons are necessary to fully assess
these first findings, and they will be continued in the future. Although the examples illustrated focused on cross-polarized
backscatter from the sea surface, the simplified formulas devel-
oped can also be applied for other surface types and in other
applications. or or Bn[F](QH) =
∞
0
dr jn(QHr)2πr F(r)
(A46) (A46) the integration being performed with respect to the space or
wavenumber variable according to the function to which it is
applied. For the following, we recall the useful identities: (i−n)
2π
2π
0
dϕ cos(nϕ) eiu cos(ϕ−ϕ0) = cos(nϕ0)jn(u)
(i−n)
2π
2π
0
dϕ sin(nϕ) eiu cos(ϕ−ϕ0) = sin(nϕ0)jn(u). (A47) (i−n)
2π
2π
0
dϕ cos(nϕ) eiu cos(ϕ−ϕ0) = cos(nϕ0)jn(u) (i−n)
2π
2π
0
dϕ sin(nϕ) eiu cos(ϕ−ϕ0) = sin(nϕ0)jn(u). (A47) 7 7 It is then straightforward to derive the following expressions
(with x = r cos ϕr, y = r sin ϕr): [7] A. G. Voronovich and V. U. Zavorotny, “Theoretical model for scattering
of radar signals in Ku- and C-bands from a rough sea surface with break-
ing waves,” Waves Random Media, vol. 11, no. 3, pp. 247–269, 2001. [8] A. Voronovich and V. Zavorotny, “Full-polarization modeling of monos-
tatic and bistatic radar scattering from a rough sea surface,” IEEE Trans. Antennas Propag., vol. 62, no. 3, pp. 1363–1371, Mar. 2014. VII. CONCLUSION His research
interests are in the areas of microwave remote sens-
ing, propagation, and electromagnetic wave theory. Dr. Johnson is a member of commissions B and F of the International Union
of Radio Science (URSI) and a member of Tau Beta Pi, Eta Kappa Nu, and Phi
Kappa Phi. He received the 1993 Best Paper Award from the IEEE Geoscience
and Remote Sensing Society, was named an Office of Naval Research Young
Investigator, National Science Foundation Career awardee, and PECASE award
recipient in 1997, and was recognized by the U.S. National Committee of URSI
as a Booker Fellow in 2002. Joel T. Johnson (S’88–M’96–SM’03–F’08) received
the B.S. degree in electrical engineering from the
Georgia Institute of Technology, Atlanta, GA, USA,
in 1991 and the S.M. and Ph.D. degrees from the
Massachusetts Institute of Technology, Cambridge,
MA, USA, in 1993 and 1996, respectively. (A49) Dr. Johnson is a member of commissions B and F of the International Union
of Radio Science (URSI) and a member of Tau Beta Pi, Eta Kappa Nu, and Phi
Kappa Phi. He received the 1993 Best Paper Award from the IEEE Geoscience
and Remote Sensing Society, was named an Office of Naval Research Young
Investigator, National Science Foundation Career awardee, and PECASE award
recipient in 1997, and was recognized by the U.S. National Committee of URSI
as a Booker Fellow in 2002. In the case of an anisotropic spectrum, the Kirchhoff integral
is calculated with integration in polar coordinates using a
Simpson quadrature rule to accelerate the angular integration. VII. CONCLUSION ρ = B0[Γ0](r) −B2[Γ2](r) cos(2ϕr −2ϕw)
∂xyρ = 1
2B2
k2Γ0
(r) sin(2ϕr)
−1
4B0
k2Γ2
(r) sin(2ϕw)
−1
4B4
k2Γ2
(r) sin(4ϕr −2ϕw)
∂xxyyρ = 1
8B0
k4Γ0
(r) −1
8B4[k4Γ0](r) cos(4ϕr)
−1
8B2
k4Γ2
(r) cos(2ϕr −2ϕw)
+ 1
16B2
k4Γ2
(r) cos(2ϕr + 2ϕw)
+ 1
16B6
k4Γ2
(r) cos(6ϕr −2ϕw). (A4 [9] C. -A. Guérin and M. Saillard, “On the high-frequency limit of the second-
order small-slope approximation,” Waves Random Media, vol. 13, no. 2,
pp. 75–88, 2003. [10] C. A. Guérin, G. Soriano, and B. Chapron, “The weighted curvature
approximation in scattering from sea surfaces,” Waves Random Complex
Media, vol. 14, no. 3, pp. 349–363, 2010. [11] T. Elfouhaily, B. Chapron, K. Katsaros, and D. Vandemark, “A unified
directional spectrum for long and short wind-driven waves,” J. Geophys. Res., vol. 102, no. C7, pp. 15 781–15 796, Jul. 1997. [12] T. M. Elfouhaily and J. T. Johnson, “A new model for rough surface scat-
tering,” IEEE Trans. Geosci. Remote Sens., vol. 45, no. 7, pp. 2300–2308,
Jul. 2007. Charles-Antoine Guérin received the B. Eng. degree
from
Ecole
Nationale
Supérieure
de
l’Aéronautique et de l’Espace, Toulouse, France, in
1994 and the Ph.D. degree in theoretical physics
from Aix Marseille University, Marseille, France, in
1998. (A48) He is currently a Professor and a Researcher
with the Mediterranean Institute of Oceanography,
University of Toulon, La Garde, France. He is spe-
cialized in ocean remote sensing. In the case of isotropic spectrum (Γ2 = 0), an efficient cal-
culation of the Kirchhoff integral can be obtained using Bessel
transforms. Using the identities (A47), we obtain
e−iQH·r(∂xxyyρ)F = 1
8B0
B0
k4Γ0
F
(QH)
−1
8B4
B4
k4Γ0
F
(QH)
e−iQH·r(∂xyρ)2F = 1
8B0
B2
k2Γ0
2 F
(QH)
−1
8B4
B2
k2Γ0
2 F
(QH). (A49) Joel T. Johnson (S’88–M’96–SM’03–F’08) received
the B.S. degree in electrical engineering from the
Georgia Institute of Technology, Atlanta, GA, USA,
in 1991 and the S.M. and Ph.D. degrees from the
Massachusetts Institute of Technology, Cambridge,
MA, USA, in 1993 and 1996, respectively. He is currently a Professor and Department Chair
with the Department of Electrical and Computer En-
gineering and ElectroScience Laboratory, The Ohio
State University, Columbus, OH, USA. REFERENCES [1] P. A. Hwang, B. Zhang, and W. Perrie, “Depolarized radar return for
breaking wave measurement and hurricane wind retrieval,” Geophys. Res. Lett., vol. 37, no. 1, Jan. 2010, Art. ID. L01604. [2] P. W. Vachon and J. Wolfe, “C-band cross-polarization wind speed re-
trieval,” IEEE Geosci. Remote Sens. Lett., vol. 8, no. 3, pp. 456–459,
May 2011. [3] P. A. Hwang, W. Perrie, and B. Zhang, “Cross-polarization radar backscat-
tering from the ocean surface and its dependence on wind velocity,”
IEEE Trans. Geosci. Remote Sens., vol. 11, no. 12, pp. 2188–2192,
Dec. 2014. [4] T. Elfouhaily and C. A. Guérin, “A critical survey of approximate scatter-
ing wave theories from random rough surfaces,” Waves Random Complex
Media, vol. 14, no. 4, pp. 1–40, 2004. [5] A. Voronovich, “Small-slope approximation for electromagnetic wave
scattering at a rough interface of two dielectric half-spaces,” Waves
Random Complex Media, vol. 4, no. 3, pp. 337–367, 1994. [6] A. G. Voronovich, Wave Scattering From Rough Surfaces, ser. Springer
Series on Wave Phenomena. New York, NY, USA: Springer-Verlag,
1994.
|
https://openalex.org/W2907451440
|
https://europepmc.org/articles/pmc6439489?pdf=render
|
English
| null |
Pervasive admixture between eucalypt species has consequences for conservation and assisted migration
|
Evolutionary Applications
| 2,019
|
cc-by
| 14,848
|
Brenton von Takach Dukai1
| Cameron Jack2 | Justin Borevitz3,4 |
David B. Lindenmayer1 | Sam C. Banks5 1Fenner School of Environment and
Society, The Australian National University,
Canberra, Australian Capital Territory,
Australia 1Fenner School of Environment and
Society, The Australian National University,
Canberra, Australian Capital Territory,
Australia
2ANU Bioinformatics Consultancy,
John Curtin School of Medical
Research, Australian National University,
Canberra, Australian Capital Territory,
Australia
3Research School of Biology, The Australian
National University, Canberra, Australian
Capital Territory, Australia
4Centre of Excellence in Plant Energy
Biology, The Australian National University,
Canberra, Australian Capital Territory,
Australia
5Research Institute for the Environment
and Livelihoods, Charles Darwin University,
Darwin, Northwest Territories, Australia
Correspondence
Brenton von Takach Dukai, Fenner School
of Environment and Society, The Australian
National University, Canberra, ACT,
Australia. Email: brenton.takach@anu.edu.au
Funding information
Australian Research Council, Grant/Award
Number: FT130100043 Pervasive admixture between eucalypt species has
consequences for conservation and assisted migration Brenton von Takach Dukai1
| Cameron Jack2 | Justin Borevitz3,4 |
David B. Lindenmayer1 | Sam C. Banks5 Abstract Abstract
Conservation management often uses information on genetic population structure
to assess the importance of local provenancing for ecological restoration and reintro-
duction programs. For species that do not exhibit complete reproductive isolation,
the estimation of population genetic parameters may be influenced by the extent of
admixture. Therefore, to avoid perverse outcomes for conservation, genetically in-
formed management strategies must determine whether hybridization between spe-
cies is relevant, and the extent to which observed population genetic patterns are
shaped by interspecific versus intraspecific gene flow. We used genotyping by se-
quencing to identify over 2,400 informative single nucleotide polymorphisms across
18 populations of Eucalyptus regnans F. Muell., a foundation tree species of montane
forests in south‐eastern Australia. We used these data to determine the extent of
hybridization with another species, Eucalyptus obliqua L'Hér., and investigate how ad-
mixture influences genetic diversity parameters, by estimating metrics of genetic di-
versity and examining population genetic structure in datasets with and without
admixed individuals. We found hybrid individuals at all sites and two highly intro-
gressed populations. Hybrid individuals were not distributed evenly across environ-
mental gradients, with logistic regression identifying hybrids as being associated with
temperature. Removal of hybrids resulted in increases in genetic differentiation (FST),
expected heterozygosity, observed heterozygosity and the inbreeding coefficient,
and different patterns of isolation by distance. After removal of hybrids and intro-
gressed populations, mountain ash showed very little population genetic structure,
with a small effect of isolation by distance, and very low global FST(0.03). Our study
shows that, in plants, decisions around provenancing of individuals for restoration
depend on knowledge of whether hybridization is influencing population genetic
structure. For species in which most genetic variation is held within populations,
there may be little benefit in planning conservation strategies around environmental
adaptation of seed sources. The possibility for adaptive introgression may also be
relevant when species regularly hybridize. 2ANU Bioinformatics Consultancy,
John Curtin School of Medical
Research, Australian National University,
Canberra, Australian Capital Territory,
Australia 5Research Institute for the Environment
and Livelihoods, Charles Darwin University,
Darwin, Northwest Territories, Australia Evolutionary Applications. 2019;12:845–860.
| 845
wileyonlinelibrary.com/journal/eva
This is an open access article under the terms of the Creative Commons Attribution License, which permits use, distribution and reproduction in any medium,
provided the original work is properly cited.
© 2018 The Authors. Evolutionary Applications published by John Wiley & Sons Ltd eceived: 24 July 2018 | Revised: 5 December 2018 | Accepted: 21 December 2018 eceived: 24 July 2018 | Revised: 5 December 2018 | Accepted: 21 December 2018 Received: 24 July 2018 | Revised: 5 December 2018 | Accepted: 21 December 2018
DOI: 10.1111/eva.12761 DOI: 10.1111/eva.12761 1 | INTRODUCTION Using genetic approaches to inform management activities al-
lows conservation efforts to be targeted towards sites of unique
genetic composition or adaptive importance, making population
genetic studies valuable in many taxa (Ikeda et al., 2017; Maunder,
Cowan, Stranc, & Fay, 2001; McCartney‐Melstad & Shaffer, 2015;
Reynolds et al., 2015). To maximize beneficial outcomes, it is vital
that our understanding of the population genetic diversity and
structure in target species is as accurate as possible, particularly in
the context of a changing environment. For example, understanding
patterns of local adaptation across the range of a species is import-
ant for developing methods of assisted gene flow to mitigate the
impacts of climate change and other threatening processes (Kelly &
Phillips, 2016; Supple et al., 2018). Given the importance of understanding population genetic
structure for conservation, and the knowledge that hybridization
in eucalypts is a widespread and common phenomenon (Griffin,
Burgess, & Wolf, 1988), we investigated these two aspects in
Eucalyptus regnans F. Muell., (mountain ash) one of Australia's most
well‐known and economically important trees. The existence of hy-
brids between E. regnans and the frequently co‐occurring Eucalyptus
obliqua L'Hér. (messmate stringybark) has long been known (Ashton,
1956); however, the extent of hybridization across the range of the
species has never been investigated. Similarly, while the chloroplast
genetic structure of E. regnans has been studied (Nevill, Bossinger, &
Ades, 2010), the structure of the nuclear genome has not. To address
these knowledge gaps, our aims were to (a) identify the extent and
possible drivers of hybridization across the geographic distribution
of E. regnans, (b) describe how identification of admixture using ge-
nomic data may influence our understanding of population genetic
structure and (c) consider how these factors would influence current
management strategies in eucalypts. We address these aims using
genotyping by sequencing to obtain large numbers of genomewide
genetic markers on individual samples across the natural geographic Genetically informed conservation requires a detailed under-
standing of the spatial distribution of genetic diversity, particularly
as it relates to environmental adaptation. Spatial genetic structure
and population genetic differentiation are typically considered to
be driven by the influences of gene flow, genetic drift and local ad-
aptation (Orsini, Vanoverbeke, Swillen, Mergeay, & Meester, 2013). 1 | INTRODUCTION 1 have investigated patterns of nuclear genetic structure and gene
flow across large geographic regions (Hecht, Matala, Hess, & Narum,
2015; Hendricks et al., 2017; Sampson et al., 2018; Shriver et al.,
2005). Studies such as these provide critical information for the con-
servation of populations with unique genetic heritage, identification
of areas of adaptive potential for assisted migration and location of
source populations or historical refugia (Hecht et al., 2015; Supple
et al., 2018). Substantial biodiversity declines are occurring in many regions of the
world due to widespread land clearing, habitat degradation, intro-
duced species and climate change (Evans et al., 2011; Pounds et al.,
2006; Woinarski, Burbidge, & Harrison, 2015). Extensive and ongo-
ing land clearing has led to major reductions in forest cover globally
(Achard et al., 2014; Reside et al., 2017; Taubert et al., 2018), with
synergistic interactions between stressors placing some ecosys-
tems under high threat of rapid collapse or changes in ecosystem
state (Brook, Sodhi, & Bradshaw, 2008; Lindenmayer, Hobbs, Likens,
Krebs, & Banks, 2011; Lindenmayer & Sato, 2018). With such wide-
spread changes facing ecosystems, it is critical to understand how
these stressors interact with the fundamental ecological processes
operating within and between foundation species, to adequately
manage biodiversity across landscapes. Australian natural vegetation communities are dominated by
the hyperdiverse and commercially important tree genus Eucalyptus
L'Hér. With about 700 species recognized (Bayly, 2016), eucalypts
are an integral part of the Australian landscape and are foundation
species in many ecological communities. For such an important
component of Australia's vegetation, there is still much to under-
stand about gene flow, population dynamics and genetic structure
in eucalypts. Gene flow in plants is typically the result of both pol-
len and seed dispersal, with pollen typically playing a greater role in
eucalypts because it tends to disperse further than seeds (Barber,
1965; Petit et al., 2005; Potts & Wiltshire, 1997). Comparisons of the
maternally inherited chloroplast and biparentally inherited nuclear
DNA have shown that pollen‐mediated gene flow can be up to at
least 200 times greater than seed‐mediated gene flow in some spe-
cies (Bloomfield, Nevill, Potts, Vaillancourt, & Steane, 2011; Nevill,
Bradbury, Williams, Tomlinson, & Krauss, 2014), although at least
one study found that gene flow from seed dispersal is practically
equivalent to that from pollen dispersal (Jones, Shepherd, Henry, &
Delves, 2006). Correspondence Funding information
Australian Research Council, Grant/Award
Number: FT130100043 wileyonlinelibrary.com/journal/eva von TAKACH DUKAI et al. 846 K E Y W O R D S
admixture, eucalypt, Eucalyptus regnans, gene flow, hybrid, mountain ash, population genetics,
SNP K E Y W O R D S
admixture, eucalypt, Eucalyptus regnans, gene flow, hybrid, mountain ash, population genetics,
SNP admixture, eucalypt, Eucalyptus regnans, gene flow, hybrid, mountain ash, population genetics,
SNP 2.1 | Study area and species Eucalyptus regnans grows in wet forests of the south‐east Australian
states of Victoria and Tasmania. It is the tallest angiosperm in the
world, with reliable records of individuals exceeding 100 m (Beale,
2007; Hardy, 1918, 1935). It is also a serotinous obligate seeder, re-
quiring high‐intensity fires to open the understorey, create fertile
ash beds and stimulate the mass release of seeds from the forest
canopy (Ashton, 1981b; Ashton & Chinner, 1999). Without fire, trees
are typically unable to produce offspring that survive to maturity,
primarily due to predation of seeds by ants (Ashton, 1979; O'Dowd
& Gill, 1984), low availability of light (Gilbert, 1959), browsing of
seedlings by herbivores and fungal infection of seedlings (Ashton &
Macauley, 1972). FI G U R E 1 Map showing the overall distribution of Eucalyptus
regnans (green shading) and Eucalyptus obliqua (grey shading) in the
Australian states of Victoria and Tasmania, and the locations visited
for collection of tissue samples for genotyping by sequencing. Species distributions are derived from records of each species
found on the Atlas of Living Australia. The blue circles represent
sites where E. regnans samples were collected, and the red circle
represents the site where E. obliqua samples were collected. The
number of samples collected at each site is indicated by the size of
the circle FI G U R E 1 Map showing the overall distribution of Eucalyptus
regnans (green shading) and Eucalyptus obliqua (grey shading) in the
Australian states of Victoria and Tasmania, and the locations visited
for collection of tissue samples for genotyping by sequencing. Species distributions are derived from records of each species
found on the Atlas of Living Australia. The blue circles represent
sites where E. regnans samples were collected, and the red circle
represents the site where E. obliqua samples were collected. The
number of samples collected at each site is indicated by the size of
the circle FI G U R E 1 Map showing the overall distribution of Eucalyptus Eucalyptus regnans is patchily distributed through a 700 km by
500 km area, growing only where climatic conditions are suitable
(Cochrane, 1969). It reaches its highest elevations (>1,100 m ASL)
in the northernmost part of its range, on the Errinundra Plateau,
and grows near to sea level in some southern parts of its Tasmanian
distribution. 1 | INTRODUCTION However, for species that do not exist in complete reproductive iso-
lation, the estimation of population genetic parameters may be influ-
enced by the extent of hybridization and introgression with closely
related species. This could have large implications for the application
of genetic data to conservation management of species, for example,
by committing resources to putatively distinct populations, when
they may actually contain highly admixed individuals. With the advent of modern DNA genotyping techniques, stud-
ies investigating thousands of genetic markers from across the ge-
nome are becoming more common (Gaughran et al., 2018; Hand
et al., 2015; Harvey, Aleixo, Ribas, & Brumfield, 2017; Hudson,
Freeman, Myburg, Potts, & Vaillancourt, 2015), and several studies von TAKACH DUKAI et al. 847 FI G U R E 1 Map showing the overall distribution of Eucalyptus
regnans (green shading) and Eucalyptus obliqua (grey shading) in the
Australian states of Victoria and Tasmania, and the locations visited
for collection of tissue samples for genotyping by sequencing. Species distributions are derived from records of each species
found on the Atlas of Living Australia. The blue circles represent
sites where E. regnans samples were collected, and the red circle
represents the site where E. obliqua samples were collected. The
number of samples collected at each site is indicated by the size of
the circle
–37
–39
–41
–43
0 km
100 km
200 km
142
144
Longitude
Latitude
146
148
150 range of the species. We predict that (a) some individuals and popu-
lations will show greater levels of admixture with E. obliqua, (b) levels
of admixture will be driven in part by local environmental variables
and (c) the inclusion or exclusion of hybrid individuals in population
genetic analyses will lead to different strategic outcomes for man-
agement. If these predictions are true, there are implications for
future studies of population genetic structure and the planning of
restoration plantings and assisted gene flow. –37
–39
–41
–43
0 km
100 km
200 km
142
144
Longitude
Latitude
146
148
150 Latitude 2.1 | Study area and species regnans or E. obliqua using purported diagnostic morphological
characters (Brooker & Kleinig, 2006). In a number of locations throughout Victoria and Tasmania, trees
displaying intermediate characteristics between E. regnans and other
species have been recorded. These specimens have been identified
mostly as hybrid individuals between E. regnans and E. obliqua, and,
less commonly, E. regnans and E. macrorhyncha (Ashton, 1958, 1981a;
Ashton & Sandiford, 1988). At least two individuals have also been
found that appear to be tri‐hybrids—the result of a E. regnansx obli‐
qua hybrid individual mating with a E. macrorhyncha (Yorke & Ashton,
1982). As red stringybark does not occur naturally in Tasmania,
E. regnans × macrorhyncha hybrids do not occur there. 2.1 | Study area and species As the island of Tasmania has been separated from
the Australian mainland by more than 200 km for over ten thou-
sand years (Duncan, Worth, Jordan, Jones, & Vaillancourt, 2016;
Lambeck, Rouby, Purcell, Sun, & Sambridge, 2014), it is assumed that
there has been very little or no gene flow between E. regnans stands
in these two regions for at least that length of time. breast height of more than 60 cm, to avoid sampling younger trees
that were propagated after the practice of reseeding logged coupes
using seed of nonlocal provenance became common practice (Flint
& Fagg, 2007). As the combined effects of logging and wildfires have
caused a reduction in the size and frequency of old undisturbed
patches of trees (Lindenmayer, Blanchard, Blair, McBurney, & Banks,
2016), a linear transect of fixed length was sometimes impossible. At seven of the sites, we collected a second sample from the 20th
tree, to serve as a technical replicate from the field. We were also
able to incorporate an extra 42 E. regnans samples collected during
fieldwork for other studies into some analyses, taken from various
locations (Supporting Information Table S1). Twenty‐one E. obliqua
samples, taken from the Cathedral Range region in Victoria, were
also sequenced to allow us to determine the extent of hybridization
between the two species. All samples were putatively identified as
E. regnans or E. obliqua using purported diagnostic morphological
characters (Brooker & Kleinig, 2006). breast height of more than 60 cm, to avoid sampling younger trees
that were propagated after the practice of reseeding logged coupes
using seed of nonlocal provenance became common practice (Flint
& Fagg, 2007). As the combined effects of logging and wildfires have
caused a reduction in the size and frequency of old undisturbed
patches of trees (Lindenmayer, Blanchard, Blair, McBurney, & Banks,
2016), a linear transect of fixed length was sometimes impossible. At seven of the sites, we collected a second sample from the 20th
tree, to serve as a technical replicate from the field. We were also
able to incorporate an extra 42 E. regnans samples collected during
fieldwork for other studies into some analyses, taken from various
locations (Supporting Information Table S1). Twenty‐one E. obliqua
samples, taken from the Cathedral Range region in Victoria, were
also sequenced to allow us to determine the extent of hybridization
between the two species. All samples were putatively identified as
E. 2.4 | Demultiplexing and initial filters Of the 408 samples (387 E. regnans and 21 E. obliqua), sequencing re-
sulted in nearly 1.49 billion read pairs. We demultiplexed reads using
exact matches and combinatorial index mode with Axe (Murray
& Borevitz, 2018) and were unable to assign 7% of read pairs
to a sample. We then used BBDuk to remove adapters and qual-
ity‐trim (Phred score Q = 30) reads at both ends, and NextGenMap
(Sedlazeck, Rescheneder, & von Haeseler, 2013) to align reads to the
E. grandis v2.0 reference genome (Bartholome et al., 2015; Myburg
et al., 2014). We used SAMtools (Li et al., 2009) to convert the data-
set into sample‐specific Binary Alignment/Map (BAM) files and sort
reads. To create a sample‐by‐SNP matrix, we used the “ANGSD” soft-
ware package to first calculate genotype likelihoods (McKenna et al.,
2010) and used these likelihoods to call genotypes. Loci were initially
filtered based on (a) a probability of at least 99.999% that the site
was variable, (b) the site was genotyped in at least 50 individuals,
(c) the site had a minimum average sequencing depth per sample of
0.5 and (d) the site had a maximum average sequencing depth per
sample of 1,000. Genotype likelihoods were retained and exported
in BEAGLE file format for admixture analysis using “NGSadmix”
(Skotte, Korneliussen, & Albrechtsen, 2013). Called genotypes were
used for the remainder of the analyses and were derived from the
likelihoods based on a posterior genotype probability (≥0.95) and as-
suming a uniform prior. This produced a matrix containing 408 sam-
ples and 49,622 SNPs, where the mean and median read depth per
site per sample were 28.3 and 9.9, respectively. Axis 1 (2.66%) FI G U R E 2 Principal components analysis of pairwise genetic
distance between 380 putative Eucalyptus regnans and 20 putative
Eucalyptus obliqua trees. Eucalyptus regnans was sampled from
across the natural geographic range of the species. Euclidean
genetic distances were calculated using 2,192 single nucleotide
polymorphisms. Putative E. regnans samples are coloured by their
proportion of admixture with E. obliqua. The reference E. obliqua
samples are coloured black and picking two fresh, growing leaves. All samples were air‐dried in
individual brown paper bags and then stored at 4°C prior to DNA
extraction. 2.2 | Sample collection We collected 387 E. regnans tissue samples from across its geo-
graphic distribution (Figure 1). At each of 16 sites, we walked a
transect collecting tissue from trees spaced at least 20 m apart,
until we had sampled 20 trees. We targeted trees with diameters at Tissue collected was mostly cambium, obtained by cutting
through the rough and smooth bark using a machete, and slicing
off a 10 × 5 × 0.2 cm strip. A small number of samples were leaf tis-
sue, collected by climbing trees using standard arborist techniques von TAKACH DUKAI et al. 848 |
FI G U R E 2 Principal components analysis of pairwise genetic
distance between 380 putative Eucalyptus regnans and 20 putative
Eucalyptus obliqua trees. Eucalyptus regnans was sampled from
across the natural geographic range of the species. Euclidean
genetic distances were calculated using 2,192 single nucleotide
polymorphisms. Putative E. regnans samples are coloured by their
proportion of admixture with E. obliqua. The reference E. obliqua
samples are coloured black
−20
0
20
−40
−20
0
Axis 1 (2.66%)
Axis 2 (1.46%)
0.25
0.50
0.75
E. obliqua
ancestry 848 848 |
−20
0
20
−40
−20
0
Axis 1 (2 66%)
Axis 2 (1.46%)
0.25
0.50
0.75
E. obliqua
ancestry 2.3 | Sample preparation and sequencing Two separate filtering strategies were conducted on this dataset
using the statistical software package R (R Core Team, 2017). The
first of these developed a set of SNPs for investigating the extent of
admixture with E. obliqua. The second method was used to investi-
gate whether hybridization influences population genetic structure
and isolation by distance across the geographic distribution of E. reg‐
nans. Each of these filters has been discussed in the methods of the
relevant analysis. Approximately 600 mg of each tissue sample was chilled to
−65°C and homogenized using an automated tissue grinding ma-
chine (Labman Max Planck Cryogenic Grinder Dispenser, Labman
Automation & Custom Robotics), before storage at −18°C to await
DNA extraction. Samples were ordered randomly, and whole genomic DNA was
extracted in plate format by following the kit (Stratec Invisiorb DNA
Plant HTS 96) instructions. Library preparation for genotyping by
sequencing included (a) digestion using PstI restriction enzyme
(New England BioLabs, Inc.), (b) ligation using T4 DNA ligase (New
England BioLabs, Inc.), (c) a purification step (Qiagen MinElute 96‐
well PCR purification kit), (d) PCR amplification using two GBS prim-
ers (Integrated DNA Technologies), (e) postpurification quantitation
using microfluidic capillary electrophoresis (PerkinElmer LabChip
GX II), (6) pooling of 12 ng DNA per sample using an automated
robotic liquid handling machine (PerkinElmer NGS Express) and (f) a
final purification step (Sigma‐Aldrich Genelute PCR Clean‐Up Kit). and 380 samples used for the STRUCTURE analysis. Lastly, we used
the expectation–maximization algorithm in NGSadmix (Skotte et al.,
2013), using the GATK genotype likelihoods (McKenna et al., 2010)
produced by ANGSD, assuming two ancestral populations, and re-
quiring the minor allele to be present in at least eight individuals,
retaining 16,634 loci. Bar plots of admixture for all 380 samples
allowed for visual comparison of each method. As the admixture
proportions between the three methods were highly correlated (dis-
cussed in the Results section), we then averaged the admixture coef-
ficients and used these mean values to exclude or retain individuals
for the remaining analyses. Samples with a E. obliqua ancestry coef-
ficient greater than 0.1 were considered hybrids, with coefficients of
0.4–0.6 indicating intermediate levels of hybridization, and 0.1–0.4
or 0.6–1 indicating closer affinity to E. regnans or E. obliqua, respec-
tively (Field, Ayre, Whelan, & Young, 2009; Melville et al., 2017). While this method of identifying hybrid individuals is unlikely to
have completely removed admixture from the study, we considered
it sufficient to demonstrate how accounting for hybridization and
introgression in the analysis pipeline can influence the results of, and
conclusions drawn from, genetic analyses. It is unfeasible and pos-
sibly inappropriate to try and completely remove all admixture, with
gene flow between the two species possibly occurring throughout
their recent evolutionary history. minimum temperature of the coldest month (MINTEMP), as well
as topographic aspect, which was transformed into a north–south
component (NORTH) and an east–west component (EAST). A binary
response variable was also created that identified each sample as ei-
ther a hybrid individual or a “pure” E. regnans. All predictor variables
were scaled (by subtracting the mean and dividing by the standard
deviation) prior to fitting any models. To identify whether any predictor variables showed nonlinearity
on the logistic scale, we fitted a binomial generalized additive model
in R and plotted the component smooth functions for each variable
on the scale of the linear predictor. The MAXTEMP variable showed
signs of nonlinearity and so was transformed with the inverse recip-
rocal, which improved linearity substantially. A binomial generalized linear mixed‐effects model was then run
in R, using the glmer function of the “lme4” (Bates, Maechler, Bolker,
& Walker, 2015) package, using the binary response variable, the
seven unmodified environmental predictor variables and the new
MAXTEMP predictor variable. 2.7 | Influence of admixture on population structure To determine the level of influence that unrealized hybridization
can have on population structure, we first filtered SNPs using call
rate (≥0.4), minor allele frequency (MAF ≥ 0.01) and observed het-
erozygosity (≤0.5). The likelihood that each SNP does not devi-
ate from Hardy–Weinberg equilibrium (HWE) was checked using
the HWChisqStats function of the “Hardy–Weinberg” (Graffelman,
2015) package. Any SNP out of HWE in more than three popula-
tions (where n ≥ 15) was removed from further analysis. In addition,
the snpgdsLDpruning function in the “SNPRelate” (Zheng et al., 2012)
package was used to prune out SNPs using a linkage disequilibrium
threshold of 0.5 and a sliding window of 5,000 bp. Next, we calculated mean admixture proportions for every pop-
ulation and mapped this across the landscape using the “ggplot2”
and “scatterpie” (Yu, 2018) packages. The estimated tree age (AGE) for
each sample was also included as a predictor to determine whether
this was relevant, and a random effect of stand was fitted to account
for repeated sampling. Overdispersion was checked using the model
residuals and degrees of freedom, and spatial autocorrelation was
investigated in the model residuals using visual examination of a var-
iogram created with the variog function of the “geor” (Ribeiro Jr. &
Diggle, 2016) package. As principal components analysis (PCA) is often used as an initial
method of removing outlier samples (Jordan, Hoffmann, Dillon, &
Prober, 2017; Supple et al., 2018), we first checked whether PCA
would be appropriate for the identification and removal of hybrid
individuals from the dataset. We calculated pairwise Euclidean ge-
netic distances using the 2,192 SNPs and 380 samples and then per-
formed a PCA using the indpca function of the “hierfstat” (Goudet,
2005) package in R. The first two principal components for each
sample were plotted using the “ggplot2” (Wickham, 2009) package,
with samples coloured by the mean level of admixture with E. obli‐
qua, as determined above. All possible submodels of the global model were fitted and
ranked by Akaike's information criterion corrected for finite sample
sizes (AICc) using the “MuMIn” (Barton, 2016) package. Models with
AIC values ≤2 above the top‐ranked model were considered useful
for inference (Hegyi & Garamszegi, 2011). 2.5 We investigated individual admixture proportions using three
techniques and averaged the results to improve accuracy and re-
liability. Firstly, the Bayesian clustering method in STRUCTURE
v2.3.4 (Falush, Stephens, & Pritchard, 2003; Pritchard, Stephens,
& Donnelly, 2000) was used, with a 50,000 burn‐in and 200,000
Markov chain Monte Carlo (MCMC) iterations, a K value of 2, using
an admixture model and correlated allele frequencies. To obtain
SNPs used in this analysis, we filtered on call rate (≥66% of samples
genotyped) and minor allele frequency (MAF ≥ 0.01), retaining 2,192
SNPs. Samples missing more than two‐thirds of these loci were re-
moved from any further analysis, retaining 380 samples. Next, we
used an eigen‐analysis approach to investigate individual ancestries,
using the snpgdsAdmixProp function of the “SNPRelate” package
(Zheng et al., 2012; Zheng & Weir, 2016), with the same 2,192 SNPs Size fractionation, 250–450 base‐pair gel cutout, and sequenc-
ing was conducted at the Australian Cancer Research Foundation
Biomolecular Resource Facility (BRF) at the John Curtin School of
Medical Research (Australian National University) on portions of
four lanes (grouped with other E. regnans sequencing experiments)
of an Illumina HiSeq 2500 machine using a 100‐base paired‐end
read. von TAKACH DUKAI et al. 849 3.1 | Extent of admixture Despite NGSadmix using a different approach (likelihoods in-
stead of called genotypes) and much larger number of loci, there
was a very strong correlation of ancestry coefficients computed
using that method and both STRUCTURE (r = 0.94, p < 0.001) and
SNPRelate (0.95, p < 0.001), with bar plots showing near‐identi-
cal patterns between methods (Supporting Information Figure
S1). There was an even stronger correlation between the re-
sults of STRUCTURE and SNPRelate (0.98, p < 0.001). Excluding
the reference E. obliqua samples, SNPRelate, STRUCTURE and
NGSadmix identified 59, 130 and 190 hybrid individuals (>10%
admixture with E. obliqua), respectively. All hybrid samples identi-
fied in SNPRelate were also identified using STRUCTURE, but a
small proportion of samples identified as hybrids in SNPRelate
and STRUCTURE were not identified as such by NGSadmix. After
calculating the mean value of the E. obliqua admixture coefficient
for the three methods, 170 samples were identified as hybrids. Of these, 75 samples had E. obliqua admixture coefficients be-
tween 0.1 and 0.4, five had coefficients between 0.4 and 0.6, 11
had coefficients between 0.6 and 0.9, and 16 had coefficients of
greater than 0.9. Standard population genetic parameters were calculated for all
loci in both groups, using the basic.stats function in the “hierfstat”
(Goudet, 2005) package. To compare the influence of including hy-
brid samples on these parameters, the means and standard errors
of the inbreeding coefficient (FIS), genetic differentiation (FST), ex-
pected heterozygosity (HE) and observed heterozygosity (HO) were
calculated across all loci and plotted as bar charts. Prior to further analysis, all sites containing <10 individuals were
removed from both datasets, leaving the pure dataset with 15 sites
and the admixture‐inclusive dataset with 18 sites. Population ge-
netic structure was assessed in both datasets using principal coor-
dinates analysis (PCoA) of pairwise population genetic distances as
well as through isolation by distance. Pairwise population Euclidean
genetic distances were calculated by first converting genotypes to
genpop objects using the df2genind and genind2genpop functions in
the “adegenet” (Jombart, 2008) package, and then using the dist.gen‐
pop function to produce the distance matrix. Principal components
were created using the cmdscale function, with populations plotted
using the “ggplot2” package. 3.1 | Extent of admixture To investigate isolation by distance,
we calculated pairwise FST between populations using the “Nei87”
method of the genet.dist function in the “hierfstat” package, and re-
gressed these distances against pairwise population geographic dis-
tance, calculated using the earth.dist function in the “fossil” (Vavrek,
2011) package. Genetic differentiation and geographic distance
were transformed to FST/(1 − FST) and log(geographic distance), re-
spectively, to allow for linear interpretation (Rousset, 1997). Principal components analysis showed that highly admixed
samples (e.g., those from Wilsons Promontory) were able to be
easily identified due to their separation from the bulk of the
E. regnans species cluster (Figure 2). Unfortunately, PCA methods
were unable to clearly separate out individuals with low and mod-
erate levels of admixture, which account for about two‐thirds of
the admixed samples. Pairwise genetic distances involving these
individuals presumably fall within the natural variation of genetic
distance within E. regnans, meaning that distance‐based meth-
ods are more conservative in the identification and removal of
hybrids. 2.8 | Spatial structure and population genetics in
Eucalyptus regnans Spatial structuring of genotypes was investigated using two meth-
ods; firstly, a Mantel test (Mantel, 1967) comparing pairwise genetic
distance with the natural logarithm of pairwise geographic distance
was performed using the mantel.rtest function in the R package
“ade4” (Dray & Dufour, 2007). We then used the spatial structure
analysis function in GenAlEx v6.503 (Peakall & Smouse, 2006, 2012)
to determine the maximum geographic distance at which genotypes
show significant spatial autocorrelation. This was interpreted as the
maximum distance at which the lower confidence interval of the spa-
tial autocorrelation coefficient r was greater than zero on the y‐axis. The mean proportion of hybrids per sample (at sites where more
than 10 samples were analysed) was 0.24. Not a single site was com-
pletely free of hybrids, but seven sites had just one hybrid. The mean
E. obliqua ancestry coefficient of a site was 0.11 (±0.21), with sites
varying considerably in the amount of admixture (Figure 3). Samples
taken from Wilsons Promontory showed the highest degree of ad-
mixture, with a mean E. obliqua ancestry coefficient of 0.87. The
Tasmanian sites on Bruny Island and the Tasman Peninsula also had
mean E. obliqua ancestry coefficients greater than 0.1, at 0.42 and
0.15, respectively. To investigate population genetics within E. regnans, we calcu-
lated means and standard errors of the number of alleles (A), number
of effective alleles (AE), HO, HE and FIS for each of the 15 populations. We then calculated the pairwise genetic distance between all indi-
viduals in these populations using the “Dch” method of the genet. dist function. This distance matrix was then read into GenAlEx to
conduct an analysis of molecular variance (AMOVA) with 999 per-
mutations to determine the amount of genetic variation explained
within and among sites. 3 | RESULTS 3 Euclidean genetic distance matrix, showed that all biological repli-
cates were closely paired in both datasets (Supporting Information
Figure S2), indicating they were reliable and contamination was not
likely to be a factor. 2.6 | Environmental association with admixture The location of all 359 putative E. regnans samples was uploaded
into the Atlas of Living Australia's (ALA) Spatial Portal (https://spa-
tial.ala.org.au/#), and 15 environmental variables at each point were
extracted. Environmental variables used in the ALA were collated or
derived from various sources (De Vries, 2009; Williams et al., 2006,
2010; Xu & Hutchinson, 2011, 2013). Variance inflation factors were
used to remove variables that showed multicollinearity, and visual
inspection of histograms for each variable allowed removal of two
variables that showed very little variation across all individuals. This
left eight variables remaining, including mean annual rainfall (RAIN),
mean annual solar radiation (RAD), historical (pre‐European) phos-
phorus availability (PHOS), topographic wetness index (TWI), mean
maximum temperature of the hottest month (MAXTEMP), mean Filtering was done on two groups: (a) all samples including hy-
brids (but excluding the reference E. obliqua samples) and (b) pure
E. regnans individuals (i.e., those with <10% admixture with E. obli‐
qua). In the admixture‐inclusive group, 2,474 SNPs were retained. For the pure group, 2,481 SNPs were retained. Any samples with
more than 50% missing data were removed from each dataset, leav-
ing 323 samples and 228 samples in the admixture‐inclusive and
admixture‐free groups, respectively. Visual inspection of hierarchi-
cal clustering dendrograms, created using the hclust function on a von TAKACH DUKAI et al. 850 3.2 | Environmental association with admixture Model selection showed that hybrid individuals were not randomly
distributed across all environmental variables. Of the 11 top mod-
els (ΔAICc < 2), MINTEMP and MAXTEMP were identified in all 11,
suggesting that these two variables had the strongest effects on
the probability of hybrid occurrence. Sites with a high MAXTEMP | 851
FI G U R E 5 Comparison of genetic parameters, including the
inbreeding coefficient (FIS), genetic differentiation (FST), expected
heterozygosity (HE) and observed heterozygosity (HO) between two
datasets, one of which contains hybrid individuals (n = 323) and
another where hybrid individuals have been removed (n = 228). The
datasets were filtered using the same criteria, resulting in 2,474
single nucleotide polymorphisms when hybrids were included and
2,481 when hybrids were removed. Error bars represent standard
errors of the mean
0.00
0.05
0.10
FIS
FST
HE
Ho
Parameter
Value
Admixture
Absent
Present von TAKACH DUKAI et al. 851 | 851
0.00
0.05
0.10
FIS
FST
HE
Ho
Parameter
Value
Admixture
Absent
Present FI G U R E 3 Map showing the mean proportion of admixture for
18 sites (with n > 10) where Eucalyptus regnans was sampled. Pie
charts show the mean proportion of admixture between E. regnans
(yellow) and Eucalyptus obliqua (turquoise). Six geographically close
sites sampled in the central region of Victoria were not significantly
different in the amount of admixture (one‐way ANOVA,
F5,108 = 1.18, p = 0.324) and so have been pooled here for clarity
0 km
100 km
200 km
−43
−41
−39
−37
142
144
146
148
150
Longitude
Latitude 0 km
100 km
200 km
−43
−41
−39
−37
142
144
146
148
150
Longitude
Latitude Latitude Parameter FI G U R E 5 Comparison of genetic parameters, including the FI G U R E 5 Comparison of genetic parameters, including the
inbreeding coefficient (FIS), genetic differentiation (FST), expected
heterozygosity (HE) and observed heterozygosity (HO) between two
datasets, one of which contains hybrid individuals (n = 323) and
another where hybrid individuals have been removed (n = 228). The
datasets were filtered using the same criteria, resulting in 2,474
single nucleotide polymorphisms when hybrids were included and
2,481 when hybrids were removed. 3.2 | Environmental association with admixture Error bars represent standard
errors of the mean FI G U R E 5 Comparison of genetic parameters, including the
inbreeding coefficient (FIS), genetic differentiation (FST), expected
heterozygosity (HE) and observed heterozygosity (HO) between two
datasets, one of which contains hybrid individuals (n = 323) and
another where hybrid individuals have been removed (n = 228). The
datasets were filtered using the same criteria, resulting in 2,474
single nucleotide polymorphisms when hybrids were included and
2,481 when hybrids were removed. Error bars represent standard
errors of the mean FI G U R E 3 Map showing the mean proportion of admixture for
18 sites (with n > 10) where Eucalyptus regnans was sampled. Pie
charts show the mean proportion of admixture between E. regnans
(yellow) and Eucalyptus obliqua (turquoise). Six geographically close
sites sampled in the central region of Victoria were not significantly
different in the amount of admixture (one‐way ANOVA,
F5,108 = 1.18, p = 0.324) and so have been pooled here for clarity models included NORTH, RAD, PHOS, RAIN, AGE and TWI; how-
ever, none of these variables showed any clear trends with hybrid
occurrence (Supporting Information Figure S3). EAST was not identi-
fied in any top models. (hot summers) and sites with a low MINTEMP (cold winters) had
lower probabilities of hybrid occurrence, whereas sites with a high
MINTEMP and sites with a low MAXTEMP had higher probabilities
of hybrid occurrence (Figure 4). Variables that occurred in fewer top 3.3 | Influence of admixture on population structure The
bottom two plots indicate the effect
of increasing geographic distance on
genetic differentiation, with both axes
transformed to allow visualization using a
linear relationship
−0.02
0.00
0.02
−
0.05
Coordinate 2
−
0.03
0.01
0.03
0.01
0.05
Coordinate 1
R2 = 0.03
p = 0.01
0.00
0.04
0.08
0.12
6
FST/(1−FST)
R2 = 0.31
p = 0.001
2
4
2
4
6
log(geographic distance in km) 852 |
−0.02
0.00
0.02
−
0.05
Coordinate 2
−
0.03
0.01
0.03
0.01
0.05
Coordinate 1
R2 = 0.03
p = 0.01
0.00
0.04
0.08
0.12
6
FST/(1−FST)
R2 = 0.31
p = 0.001
2
4
2
4
6
log(geographic distance in km) von TAKACH DUKAI et al. Coordinate 2 FI G U R E 6 Plots of pairwise genetic
distance and isolation by distance in
Eucalyptus regnans populations across its
natural geographic distribution. Plots on
the left‐hand side of the figure include
the presence of hybrid individuals and
populations. Plots on the right‐hand
side of the figure have had all hybrid
individuals removed. The top two plots
show the first two principal coordinates
of Euclidean genetic distances, where
squares, triangles, diamonds and circles,
respectively, represent populations from
the Otways, Tasmania, Central Victoria
and Gippsland regions of Australia. The
bottom two plots indicate the effect
of increasing geographic distance on
genetic differentiation, with both axes
transformed to allow visualization using a
linear relationship FST/(1−FST) log(geographic distance in km) parameters HE, HO, FST and FIS increasing from 0.087, 0.092, 0.024
and −0.012 in the admixture‐inclusive dataset, to 0.094, 0.097, 0.03
and −0.005 in the admixture‐free dataset. A Mantel test, comparing the logarithm of geographic distance with
Nei's genetic distance for all pairs of samples, showed that there was
a relatively small but significant (r = 0.18, p = 0.001) positive effect
of isolation by distance. Removal of hybrids led to greater resolution of geographic popu-
lation structure in the PCoA of pairwise population genetic distances
and a stronger pattern of isolation by distance from the Mantel test
(Figure 6). In the admixture‐inclusive dataset, the first principal co-
ordinate was associated with the degree of admixture, with Wilsons
Promontory and Bruny Island separated out from the rest of the
sites. 3.3 | Influence of admixture on population structure The high level of admixture at Wilsons Promontory resulted
in comparatively much higher pairwise FST values between this and
other sites, leading to a reduction in fit of the isolation‐by‐distance
model when this site was included in the analysis. Within‐population genetic structure at the 15 sites was very
similar (Supporting Information Table S2), with little variation in
A, AE, HO, HE, or FIS. Most sites had higher HO than HE, and all sites
had negative FIS values, implying that outbreeding is the common
condition throughout the species distribution. Global FST was
very low (0.03), and approximately equal to the mean pairwise
FST between sites (0.028), despite some large (>600 km) pairwise
geographic distances (Figure 6). Pairwise FST values ranged from
0.003, between the geographically close Central Victorian sites
of Powelltown and Toolangi South (39 km), to 0.063, between the
geographically distant Errinundra Plateau and Gellibrand River
(489 km). 3.3 | Influence of admixture on population structure The removal of hybrid individuals from the dataset modified the
values of the genetic parameters investigated (Figure 5), with the c
0
25
50
75
100
17
19
21
23
25
Max. temp. of hottest month (°C)
Predicted probability of admixture (%)
−2
−1
0
1
2
Min. temp. of coldest month (°C) FI G U R E 4 Effect plot showing the
relationship of two predictor variables,
the mean maximum temperature of the
hottest month and the mean minimum
temperature of the coldest month,
with the probability that a Eucalyptus
regnans individual will have >10% genetic
admixture with Eucalyptus obliqua. Grey
areas indicate 95% confidence intervals
0
25
50
75
100
17
19
21
23
25
Max. temp. of hottest month (°C)
Predicted probability of admixture (%)
−2
−1
0
1
2
Min. temp. of coldest month (°C) 0
25
50
75
100
17
19
21
23
25
Max. temp. of hottest month (°C)
Predicted probability of admixture (%) −2
−1
0
1
2
Min. temp. of coldest month (°C) FI G U R E 4 Effect plot showing the
relationship of two predictor variables,
the mean maximum temperature of the
hottest month and the mean minimum
temperature of the coldest month,
with the probability that a Eucalyptus
regnans individual will have >10% genetic
admixture with Eucalyptus obliqua. Grey
areas indicate 95% confidence intervals Max. temp. of hottest month (°C) Min. temp. of coldest month (°C) 852 |
von TAKACH DUKAI et al. FI G U R E 6 Plots of pairwise genetic
distance and isolation by distance in
Eucalyptus regnans populations across its
natural geographic distribution. Plots on
the left‐hand side of the figure include
the presence of hybrid individuals and
populations. Plots on the right‐hand
side of the figure have had all hybrid
individuals removed. The top two plots
show the first two principal coordinates
of Euclidean genetic distances, where
squares, triangles, diamonds and circles,
respectively, represent populations from
the Otways, Tasmania, Central Victoria
and Gippsland regions of Australia. 4 | DISCUSSION Using a genomewide SNP dataset, we investigated the extent of
hybridization between two widespread forest trees and the way in
which admixture influences estimation of genetic parameters and in-
terpretation of population genetic structure. We found that hybrids
were not distributed evenly across geographic or environmental
space, with some populations showing more admixture than others,
and a strong association of hybrid occurrence with summer temper-
atures. Once hybrids were removed, we found very little evidence
of population genetic differentiation or local isolation of alleles, with
high levels of gene flow, slow generation times and a lack of histori-
cal geographic isolation probably responsible for these findings. Eucalypts are typically preferentially outcrossing, open‐polli-
nated, and often found in sympatry with multiple congeneric spe-
cies, which may partly explain why more than half of all species form
natural hybrid combinations (Griffin et al., 1988; Potts, Barbour,
Hingston, & Vaillancourt, 2003). In addition, many of these combi-
nations can occur at relatively high frequencies within populations
(Field et al., 2009; McKinnon et al., 2010); for example, at least
27% of Eucalyptus globulus Labill. within 450 m of Eucalyptus cor‐
data Labill. show some level of admixture (McKinnon et al., 2010). However, despite the knowledge that hybridization is not uncom-
mon, it is often not considered in population genetic studies of eu-
calypts. We also note the possibility that for some species there may
be multiple hybrid combinations with other species, which may vary
regionally (Griffin et al., 1988). In this study, we only considered ad-
mixture with a single species; however, gene flow with red stringy-
bark may regularly occur in particular sites within Victoria (Ashton &
Sandiford, 1988). This study has implications for our understanding of the pro-
cesses involved in adaptation and assisted migration. Previous work
has shown that species can not only adapt to changing environments
through the development of multiple novel genetic pathways (Steane
et al., 2017), but also through a process of adaptive introgression (De
La Torre, Wang, Jaquish, & Aitken, 2014; Suarez‐Gonzalez, Lexer, &
Cronk, 2018), whereby hybridization with congeners allows for the
capture of beneficial genetic components from parent species. For
example, hybrid individuals between Picea glauca (Moench) Voss and
Picea engelmannii Parry ex Engelm. have been shown to utilize adap-
tive introgression as a way to maximize fitness in a changing climate
(De La Torre, Wang et al., 2014). the Errinundra Plateau site as being relatively distinct from the rest
of the sites. the Errinundra Plateau site as being relatively distinct from the rest
of the sites. either. In our study, removing clear outliers using PCA or a similar
method would have retained more than half of the hybrid samples. Including highly admixed samples and populations could have led us
to make different inferences about patterns of genetic structure and
possibly evolutionary history. As many modern studies of eucalypts
(and other plant taxa) do not explicitly address the presence of hy-
bridization in their study species, it is plausible that many studies
are influenced by this issue. We propose that future studies always
be explicit in acknowledging the potential for hybridization between
their species of interest and other species, as gene flow between
some species is clearly a common, and possibly evolutionarily sig-
nificant, phenomenon (De La Torre, Roberts, & Aitken, 2014; Gerber,
Chadoeuf, Gugerli, Lascoux, & Buiteveld, 2014; Lepais et al., 2009;
Palme, Su, Palsson, & Lascoux, 2004). 4 | DISCUSSION With environmental changes out-
pacing the natural ability of many species to adapt through standing
genetic variation and mutation, adaptive introgression provides an
alternative pathway, providing faster development of adaptive traits
and rapid adaptation to novel conditions. It is in this context that
we discuss our findings and explore their implications for studies of
population genetics and genetically informed conservation. In the case of E. regnans, hybridization with E. obliqua appears
to be a more pervasive phenomenon than previously realized,
with all sampled sites containing at least one hybrid individual,
and two sites where more than half of the samples were hybrids. The Wilsons Promontory individuals contained very high levels of
E. obliqua ancestry, despite their greater morphological similarity
to E. regnans. To understand why variation in the level of admix-
ture between populations occurs, examination of the factors con-
trolling gene flow between eucalypt species is necessary. Previous
research has identified three predominant drivers of hybridization
in eucalypts, including the extent of geographic isolation, the de-
gree of overlap in flowering times, and the level of phylogenetic
divergence between species (Barbour, Potts, Vaillancourt, &
Tibbits, 2006; Butcher, McDonald, & Bell, 2009; Field et al., 2009;
McKinnon et al., 2010; Potts et al., 2003), although strong abiotic
(e.g., climatic or geological) gradients may also be explanatory in
some cases (Pryor, 1976). When stands of one species are geo-
graphically isolated from conspecifics, there is a greater oppor-
tunity for pollen from another species to successfully pollinate
flowers in the isolated stand, termed “pollen swamping” (Ellstrand
& Elam, 1993; Field et al., 2009). Additionally, environmental con-
ditions can influence the degree of overlap in flowering times, 3.4 | Spatial structure and population genetics in
Eucalyptus regnans Across the overall distribution of the species, E. regnans showed
very little population structure, with the AMOVA indicating that just
3% of the genetic variation was occurring among populations, and
97% within populations. The first two coordinates of the PCoA ex-
plained 13.2% and 11.5% of the total variation, with the Tasmanian
samples intermediate between sites in central‐eastern Victoria and
sites from the Otways region (Figure 6). The PCoA also identified After removing all hybrids and discounting sites that had <10 sam-
ples, 200 individuals from 15 sites were retained. The spatial ge-
netic structure analysis in GenAlEx identified weak but significant
autocorrelation (determined by inspection of bootstrap confidence
intervals) between genotypes at pairwise distances of up to 90 km,
with the autocorrelation coefficient r intercepting zero at 108 km. 853 von TAKACH DUKAI et al. regnans at these and
other coastal sites (e.g., Tasman Peninsula) is likely due to their
being located on the periphery of suitable climatic conditions for
this species. While the pattern of very weak genetic population structure that
we observed is indicative of a lack of local isolation of alleles, local
adaptation may still be occurring. With the majority of genetic vari-
ation present in the seed crop (or mature trees) of a stand, selection
may be acting to promote particular genotypes within a generation,
as environmental conditions filter out particular alleles. There is a
steep reduction in the stem density within E. regnans stands in the
decades after a regeneration event (von Takach Dukai, Lindenmayer,
& Banks, 2018), which suggests that selection could easily promote
genotypes that increase survival rates under local environmental
conditions (Kuparinen, Savolainen, & Schurr, 2010). Previous studies
have shown that when selection is very high, high levels of disper-
sal can maximize local adaptation, and when selection pressure is
low (but present), local adaptation is highest under moderate lev-
els of dispersal (Banks, Davies, & Cary, 2017; Forester, Jones, Joost,
Landguth, & Lasky, 2016). This is due to migration providing the ge-
netic diversity for selection to act upon, but also potentially overrid-
ing the effects of selective processes when selection is weak. 4.2 | Patterns of genetic structure The low levels of genetic population structure that we observed are
not atypical for eucalypts, with a number of studies (Broadhurst,
Mellick, Knerr, Li, & Supple, 2018; Dillon et al., 2015; Gauli, Steane,
Vaillancourt, & Potts, 2014; Sampson et al., 2018; Supple et al.,
2018) finding that geographic structure does not contribute greatly
to population differentiation. Similarly, the FST values that we ob-
served between sites are low but comparable to those found in
other eucalypts (Sampson et al., 2018; Supple et al., 2018; Yeoh,
Bell, Foley, Wallis, & Moran, 2012), and low geographic structuring
of genetic diversity is not unusual in widespread forest trees, such
as Pinus taeda L. (Eckert et al., 2010) and Quercus lobataNée (Grivet,
Sork, Westfall, & Davis, 2008; Sork et al., 2010). As gene flow in trees is often effected predominantly through
pollen dispersal (Sork, 2016), there is often a clear difference in
population structure between the nuclear genome (which is in-
herited biparentally) and genetic components that are inherited
maternally, for example, the chloroplast in angiosperms (Sampson
et al., 2018). In eucalypts, seed dispersal is highly restricted, with
individual trees distributing seeds tens of metres, and stand edges
typically only capable of moving about 1–2 m/year (Booth, 2017). While it is difficult to ascertain the upper limit of the dispersal
curve, pollen appears to regularly disperse hundreds of metres to
kilometres (Bloomfield et al., 2011; Broadhurst, 2013; Byrne, Elliott,
Yates, & Coates, 2008; Sampson et al., 2018). This explains why,
when investigating chloroplast structure for phylogeographic pur-
poses, Nevill et al. (2010) found a highly structured genetic pattern
in E. regnans, whereas our (nuclear‐based) results show very little
population structuring. Chloroplast DNA is highly conserved, with
genetic structure typically reflecting historical patterns of dispersal
and colonization. Further, only a very small proportion of chloro-
plast variation is typically contained within coding regions (Young,
Lanzatella, Sarath, & Tobias, 2011), suggesting that there is unlikely
to be substantial levels of local selection acting on the chloroplast. Australian flora. This holds true for many plant and animal species
across the Bass Strait, with species found in Victoria and Tasmania
often being identified as genetically distinct, typically to the level of
separate races or subspecies (Donsker & Gill, ; Simmons, Wapstra,
& Wapstra, 2008; van Dyck, Gynther, & Baker, 2013). Even for spe-
cies with low levels of isolation by distance, there are often clear
Victorian and Tasmanian genetic clusters (Duncan et al., 2016). Our
findings for E. regnans, with Gippsland and some Central Highlands
sites more closely affiliated with Tasmanian sites than the Otways
region, support the suggestion that, for some species, the Port
Phillip Bay and surrounding area has been a bigger obstacle to gene
flow than the Bass Strait since the last glacial maximum (Yeoh et al.,
2012). with temperature previously shown to be an important driver of
flowering phenology in eucalypts (Hudson, Kim, & Keatley, 2010;
Law, Mackowski, Schoer, & Tweedie, 2000; Rawal, Kasel, Keatley,
& Nitschke, 2015). As E. obliqua is a predominantly summer‐flow-
ering species, floral development in E. obliqua may occur earlier in
sites with warmer summers, resulting in less phenological over-
lap with the autumn‐flowering E. regnans. We therefore suggest
that the primary cause of the extensive introgression of E. obliqua
into E. regnans stands at Wilsons Promontory and Bruny Island is
the patchy distribution of E. regnans in these E. obliqua‐dominated
regions, possibly assisted by environmentally driven overlap in
flowering times. The patchy distribution of E. regnans at these and
other coastal sites (e.g., Tasman Peninsula) is likely due to their
being located on the periphery of suitable climatic conditions for
this species. with temperature previously shown to be an important driver of
flowering phenology in eucalypts (Hudson, Kim, & Keatley, 2010;
Law, Mackowski, Schoer, & Tweedie, 2000; Rawal, Kasel, Keatley,
& Nitschke, 2015). As E. obliqua is a predominantly summer‐flow-
ering species, floral development in E. obliqua may occur earlier in
sites with warmer summers, resulting in less phenological over-
lap with the autumn‐flowering E. regnans. We therefore suggest
that the primary cause of the extensive introgression of E. obliqua
into E. regnans stands at Wilsons Promontory and Bruny Island is
the patchy distribution of E. regnans in these E. obliqua‐dominated
regions, possibly assisted by environmentally driven overlap in
flowering times. The patchy distribution of E. 4.1 Had we not considered the possibility of hybrid individuals occurring
in our dataset, we would have included introgressed sites such as
Wilsons Promontory and Bruny Island in the analyses, as well as 41
hybrid samples scattered through our other sites. Many of the hy-
brid individuals showed no obvious morphological differences (e.g.,
in bud and fruit shape or rough bark height and thickness) to pure
specimens, with morphological approaches now generally regarded
as a poor way of identifying hybrids (Field et al., 2009; McKinnon,
Smith, & Potts, 2010; Rhymer & Simberloff, 1996; Schwabe, Neale,
& McGlaughlin, 2015). Despite this, the Wilsons Promontory sam-
ples, which are morphologically more similar to E. regnans but ge-
netically much closer to E. obliqua, had a different growth habit to
typical E. regnans individuals, appearing stunted in form. This was
initially thought to be the result of phenotypic plasticity rather than
genetic architecture, as the trees do not resemble typical E. obliqua von TAKACH DUKAI et al. 854 4.3 Seed used for native vegetation restoration activities has historically
been collected from small local geographic areas (Broadhurst et al.,
2008), because of the perceived risk of introducing genotypes that
are not adapted to local conditions (Hamilton, 2001). Over the past
two decades, this “local is best” approach has been criticized for a
number of reasons, including the failure to consider changing envi-
ronmental/climatic conditions and not incorporating enough genetic
variability (Broadhurst et al., 2008; Choi, 2007). By using only lo-
cally adapted genotypes, managers may be restricting the ability of
populations to survive under changing conditions. To address this
issue, the definition of what constitutes the most appropriate geo-
graphic spread for seed collection to encompass a beneficial amount
of genetic variation needs to be considered (Breed, Stead, Ottewell,
Gardner, & Lowe, 2013; Crow, Albeke, Buerkle, & Hufford, 2018;
Prober et al., 2015). As eucalypts and wattles (Acacia spp.) dominate the majority of
Australian ecological restoration schemes (Broadhurst et al., 2015),
developing knowledge for these taxa is of critical importance. Our re-
sults, and recent work by others (Bloomfield et al., 2011; Dillon et al., Broadhurst et al. (2017) identify range disjunctions as being
one of the primary predictors for genetic differentiation within the von TAKACH DUKAI et al. 855 2015; Supple et al., 2018), suggest that for many Eucalyptus species,
most of the total genetic variation is held within rather than among
populations. This has large implications for the concept of local prov-
enancing. For example, Breed et al. (2013) suggest that estimates of
historical gene flow such as FST values can be used to determine the
extent of the local (FST < 0.05), intermediate (FST > 0.05 < 0.1) or dis-
tant (FST > 0.1 < 0.2) provenances. Under this definition, the entire
range of E. regnans would be considered local provenance. Because
most of the genetic variation of the species is already contained
within local geographic areas, the source of seed used for resto-
ration activities is not likely to be as important as previously consid-
ered. 5 | CONCLUSION Our population genomic analysis of E. regnans found widespread
admixture of varying levels with a congener, suggesting regu-
lar hybridization throughout the range of the species. As many
genera of plants are known to form natural hybrids, it is critical
that admixture and its role in the adaptive process is consid-
ered appropriately in population genetic studies, as introgressed
populations may skew genetic analyses and affect management
decisions. The combination of widespread hybridization and high
levels of gene flow in E. regnans, with similar results having been
found for a number of other eucalypt species, suggests that in-
trogressive adaptation through porous genomes may be a com-
mon way for this taxon to adapt to rapid environmental change
in climate and fire regimes. Selection on hybrids expressing traits
harvested from sympatric congeners may allow for rapid adap-
tive change to new conditions. Furthermore, as the occurrence of
hybrid individuals was not distributed evenly across geographic
or climatic space, the use of climatic variables to select genotypes
for assisted migration may not be the most appropriate way to
manage eucalypts for conservation purposes and requires more
detailed consideration. There are other considerations for seed sourcing that also
need to be considered. For example, climate‐adjusted provenanc-
ing has been suggested as a way of assisting plant populations
with adaptation to climate change (Prober et al., 2015). To achieve
this, seed sourcing for restoration would have to consider the
future climate at a site and predict the best source populations
to match that climate. However, there is one important consider-
ation for this approach that has not been adequately considered
in the recent literature: the issue of hybridization and intro-
gression. The possibility that some populations contain a higher
degree of admixture with another species cannot be ignored, be-
cause interspecific gene flow is one way in which species may
cope with rapidly changing abiotic environments (De La Torre,
Wang et al., 2014). By selecting seed from populations that have
“adapted” to different climatic conditions, managers may actually
be selecting seed that is simply admixed with another species. For example, our data indicate that the occurrence of E. regnans
× obliqua hybrids is correlated with temperatures of the hottest
and coolest months, so collection of seed based on temperature
may inadvertently result in the collection of hybrid genotypes, CONFLICT OF INTEREST None declared. 4.3 To be certain of the hybrid status at a proposed seed
collection locality, genotyping of seedlings would be necessary. ACKNOWLEDGEMENTS This project was funded by an Australian Research Council Future
Fellowship (FT130100043) to SB and a Lesslie Research Scholarship
in Landscape Conservation and Ecology to BvTD. Thanks to Rebecca
Taylor and Oliver King for their assistance and company during field-
work, and Wade Blanchard for statistical advice. We also appreciate
the helpful and constructive comments provided by three anony-
mous reviewers. 4.3 While we acknowledge that reciprocal transplant experiments
show a strong effect of provenance in some species (Wang, O'Neill,
& Aitken, 2010), any negative effects of using distant seed sources
must be considered in light of considerable recent evidence show-
ing that locally adapted traits can not only be maintained despite
gene flow (Fitzpatrick, Gerberich, Kronenberger, Angeloni, & Funk,
2015), but also that the potential benefits of gene flow are large and
often outweigh the risk of negative impacts associated with anthro-
pogenic disturbances or novel selective pressures (Fisher, Garner, &
Walker, 2009; Harrisson et al., 2016). In many cases, the preserva-
tion of genetic uniqueness and taxonomic integrity is no longer con-
sidered scientifically justifiable (Ralls et al., 2018). Further, as rare
new variants are unlikely to be the cause of beneficial adaptation to
local conditions (Alberto et al., 2013; Savolainen, Lascoux, & Merilä,
2013), adaptive alleles are still likely to be present in seed collected
from distant localities, simply at different frequencies. Reductions in
the frequency of some adaptive alleles are typically of minor impor-
tance and can be naturally corrected over a small number of gener-
ations (Fitzpatrick et al., 2015; Ralls et al., 2018). Thus, we suggest
that sourcing seed from trees across a range of environments (local
or distant) will ensure adaptive potential for restoration into a chang-
ing and challenging future. or even near‐pure E. obliqua, which is likely not the desired out-
come of management actions. It should also be noted that while
our sampling avoided young trees, contemporary seed crops may
show different levels of admixture in some populations. For ex-
ample, those in landscapes where logging and fire have modified
the extent and age structure of E. regnans compared to E. obliqua,
or where changes in temperature have increased overlap in flow-
ering times. To be certain of the hybrid status at a proposed seed
collection locality, genotyping of seedlings would be necessary. or even near‐pure E. obliqua, which is likely not the desired out-
come of management actions. It should also be noted that while
our sampling avoided young trees, contemporary seed crops may
show different levels of admixture in some populations. For ex-
ample, those in landscapes where logging and fire have modified
the extent and age structure of E. regnans compared to E. obliqua,
or where changes in temperature have increased overlap in flow-
ering times. REFERENCES Achard, F., Beuchle, R., Mayaux, P., Stibig, H. J., Bodart, C., Brink, A., …
Simonetti, D. (2014). Determination of tropical deforestation rates
and related carbon losses from 1990 to 2010. Global Change Biology,
20(8), 2540–2554. https://doi.org/10.1111/gcb.12605 Booth, T. H. (2017). Going nowhere fast: A review of seed dispersal in
eucalypts. Australian Journal of Botany, 65(5), 401–410. https://doi. org/10.1071/BT17019 Alberto, F. J., Aitken, S. N., Alía, R., González‐Martínez, S. C., Hänninen,
H., Kremer, A., … Savolainen, O. (2013). Potential for evolutionary
responses to climate change – evidence from tree populations. Global Change Biology, 19(6), 1645–1661. https://doi.org/10.1111/
gcb.12181 Breed, M. F., Stead, M. G., Ottewell, K. M., Gardner, M. G., & Lowe, A. J. (2013). Which provenance and where? Seed sourcing strategies for
revegetation in a changing environment. Conservation Genetics, 14(1),
1–10. https://doi.org/10.1007/s10592-012-0425-z Ashton, D. H. (1956). Studies on the autecology of Eucalyptus regnans
F.v.M. Ph.D. Thesis. Melbourne, Australia. Broadhurst, L. M. (2013). A genetic analysis of scattered Yellow Box
trees (Eucalyptus melliodora A.Cunn. ex Schauer, Myrtaceae) and
their restored cohorts. Biological Conservation, 161, 48–57. https://
doi.org/10.1016/j.biocon.2013.02.016 Ashton, D. H. (1958). The ecology of Eucalyptus regnans F. Muell: The
species and its frost resistance. Australian Journal of Botany, 6(2),
154–176. https://doi.org/10.1071/BT9580154 Broadhurst, L. M., Lowe, A., Coates, D. J., Cunningham, S. A., McDonald,
M., Vesk, P. A., & Yates, C. (2008). Seed supply for broadscale resto-
ration: Maximizing evolutionary potential. Evolutionary Applications,
1(4), 587–597. https://doi.org/10.1111/j.1752-4571.2008.00045.x Ashton, D. H. (1979). Seed harvesting by ants in forests of Eucalyptus
regnans F. Muell. in central Victoria. Australian Journal of Ecology, 4,
265–277. https://doi.org/10.1111/j.1442-9993.1979.tb01218.x Ashton, D. H. (1981a). The ecology of the boundary between E. reg-
nans F. Muell. and E. obliqua L'Herit. in Victoria. Proceedings of the
Ecological Society of Australia, 11, 75–94. Broadhurst, L. M., Mellick, R., Knerr, N., Li, L., & Supple, M. A. (2018). Land availability may be more important than genetic diversity in the
range shift response of a widely distributed eucalypt, Eucalyptus mel‐
liodora. Forest Ecology and Management, 409(October 2017), 38–46. https://doi.org/10.1016/j.foreco.2017.10.024 Ashton, D. H. (1981b). Fire in tall open forests. In R. H. Groves, I. R. Noble, & A. M. Gill (Eds.), Fire and the Australian Biota (pp. 339–366). Canberra, ACT: Australian Academy of Science. Broadhurst, L. M., Breed, M., Lowe, A., Bragg, J., Catullo, R., Coates,
D., … Byrne, M. (2017). Genetic diversity and structure of the
Australian flora. Diversity and Distributions, 23(1), 41–52. https://doi. org/10.1111/ddi.12505 Ashton, D. REFERENCES H., & Chinner, J. H. (1999). Problems of regeneration of the
mature Eucalyptus regnans F. Muell, (The Big Ash) forest, in the ab-
sence of fire at Wallaby Creek, Victoria, Australia. Australian Forestry,
62(3), 265–280. https://doi.org/10.1080/00049158.1999.10674791 Broadhurst, L. M., Driver, M., Guja, L., North, T., Vanzella, B., Fifield, G.,
… Bush, D. (2015). Seeding the future – the issues of supply and de-
mand in restoration in Australia. Ecological Management & Restoration,
16(1), 29–32. https://doi.org/10.1111/emr.12148 Ashton, D. H., & Macauley, B. J. (1972). Winter leaf spot disease of seed-
lings of Eucalyptus regnans and its relation to forest litter. Transactions
of the British Mycological Society, 58(3), 377–386. https://doi. org/10.1016/S0007-1536(72)80087-1 Brook, B. W., Sodhi, N. S., & Bradshaw, C. J. A. (2008). Synergies among
extinction drivers under global change. Trends in Ecology & Evolution,
23(8), 453–460. https://doi.org/10.1016/j.tree.2008.03.011 Ashton, D. H., & Sandiford, E. M. (1988). Natural hybridisation between
Eucalyptus regnans F. Muell. and E. macrorhyncha F. Muell. in the
Cathedral Range, Victoria. Australian Journal of Botany, 36(1), 1–22. https://doi.org/10.1071/BT9880001 Brooker, M. I. H., & Kleinig, D. A. (2006). Field guide to eucalypts. Volume
1: south‐eastern Australia (3rd ed.). Melbourne, Australia: Bloomings
Books Pty Ltd. Banks, S. C., Davies, I. D., & Cary, G. J. (2017). When can refuges mediate
the genetic effects of fire regimes? A simulation study of the effects
of topography and weather on neutral and adaptive genetic diver-
sity in fire‐prone landscapes. Molecular Ecology, 26(19), 4935–4954. https://doi.org/10.1111/mec.14250 Butcher, P. A., McDonald, M. W., & Bell, J. C. (2009). Congruence be-
tween environmental parameters, morphology and genetic struc-
ture in Australia's most widely distributed eucalypt, Eucalyptus
camaldulensis. Tree Genetics & Genomes, 5(1), 189–210. https://doi. org/10.1007/s11295-008-0169-6 Barber, H. N. (1965). Selection in natural populations. Heredity, 20(4),
551–572. https://doi.org/10.1038/hdy.1965.68 Byrne, M., Elliott, C. P., Yates, C. J., & Coates, D. J. (2008). Maintenance
of high pollen dispersal in Eucalyptus wandoo, a dominant tree of the
fragmented agricultural region in Western Australia. Conservation
Genetics, 9(1), 97–105. https://doi.org/10.1007/s10592-007-9311-5 Barbour, R. C., Potts, B. M., Vaillancourt, R. E., & Tibbits, W. N. (2006). Gene flow between introduced and native Eucalyptus species:
Flowering asynchrony as a barrier to F‐1 hybridisation between ex-
otic E‐nitens and native Tasmanian Symphyomyrtus species. Forest
Ecology and Management, 226(1–3), 9–21. https://doi.org/10.1016/j. foreco.2006.01.017 Choi, Y. D. (2007). Restoration ecology to the future: A call for
new paradigm. Restoration Ecology, 15(2), 351–353. https://doi. org/10.1111/j.1526-100X.2007.00224.x Cochrane, G. R. (1969). Brenton von Takach Dukai
https://orcid.
org/0000-0002-7999-3521 Beale, B. (2007). If trees could speak: Stories of Australia's greatest trees. Crows Nest, Sydney: Allen & Unwin. Bloomfield, J. A., Nevill, P., Potts, B. M., Vaillancourt, R. E., & Steane,
D. A. (2011). Molecular genetic variation in a widespread forest tree
species Eucalyptus obliqua (Myrtaceae) on the island of Tasmania. Australian Journal of Botany, 59(3), 226–237. https://doi.org/10.1071/
BT10315 available at the Dryad Digital Repository: https://doi.org/10.5061/
dryad.445m9j4. available at the Dryad Digital Repository: https://doi.org/10.5061/
dryad.445m9j4. Barton, K. (2016). MuMIn: Multi‐model inference. R package version 1.15.6. Retrieved from http://cran.r-project.org/package=MuMIn Bates, D., Maechler, M., Bolker, B., & Walker, S. (2015). Fitting linear
mixed‐effects models using lme4. Journal of Statistical Software,
67(1), 1–48. https://doi.org/10.18637/jss.v067.i01 ORCID Bayly, M. J. (2016). Phylogenetic studies of eucalypts: fossils, mor-
phology and genomes. Proceedings of the Royal Society of Victoria,
128(CHAH), 12–24. DATA ARCHIVING Sequence data for all individuals are deposited to the NCBI
Sequence Read Archive (SRA). A CSV file containing location
data and called genotypes from ANGSD for all individuals are 856 von TAKACH DUKAI et al. transfer in ecological restoration. Ecosphere, 9(1), e02059. https://
doi.org/10.1002/ecs2.2059 Gaughran, S. J., Quinzin, M. C., Miller, J. M., Garrick, R. C., Edwards, D. L., Russello, M. A., … Caccone, A. (2018). Theory, practice, and con-
servation in the age of genomics: The Galápagos giant tortoise as a
case study. Evolutionary Applications, 11(7), 1084–1093. https://doi. org/10.1111/eva.12551 De La Torre, A. R., Roberts, D. R., & Aitken, S. N. (2014). Genome‐wide
admixture and ecological niche modelling reveal the maintenance
of species boundaries despite long history of interspecific gene
flow. Molecular Ecology, 23(8), 2046–2059. https://doi.org/10.1111/
mec.12710 Gauli, A., Steane, D. A., Vaillancourt, R. E., & Potts, B. M. (2014). Molecular genetic diversity and population structure in
Eucalyptus pauciflora subsp pauciflora (Myrtaceae) on the is-
land of Tasmania. Australian Journal of Botany, 62(3), 175–188. https://doi.org/10.1071/Bt14036 De La Torre, A. R., Wang, T., Jaquish, B., & Aitken, S. N. (2014). Adaptation
and exogenous selection in a Picea glauca × Picea engelmannii hybrid
zone: Implications for forest management under climate change. New
Phytologist, 201(2), 687–699. https://doi.org/10.1111/nph.12540 Gerber, S., Chadoeuf, J., Gugerli, F., Lascoux, M., & Buiteveld, J. (2014). High rates of gene flow by pollen and seed in oak populations across
Europe. PLoS ONE, 9(2), e85130. https://doi.org/10.1371/journal. pone.0085130 De Vries, R. (2009). Australia's substrate fertility – version 0.8. Canberra,
Australia: Department of the Environment, Water, Heritage and the
Arts. Dillon, S., McEvoy, R., Baldwin, D. S., Southerton, S., Campbell, C.,
Parsons, Y., & Rees, G. N. (2015). Genetic diversity of Eucalyptus
camaldulensis Dehnh. following population decline in response to
drought and altered hydrological regime. Austral Ecology, 40, 558–
572. https://doi.org/10.1111/aec.12223 Gilbert, J. M. (1959). Forest succession in the Florentine Valley, Tasmania. Papers and Proceedings of the Royal Society of Tasmania, 93, 129–151. Goudet, J. (2005). hierfstat, a package for r to compute and test hier-
archical F‐statistics. Molecular Ecology Notes, 5(1), 184–186. https://
doi.org/10.1111/j.1471-8286.2004.00828.x Donsker, D., & Gill, F. (Eds.) (2018). IOC World Bird List (v8.1). In IOC
World Bird List (v8.2). Retrieved from https://www.worldbirdnames. org/ioc-lists/crossref/. Graffelman, J. (2015). Exploring diallelic genetic markers: The
HardyWeinberg package. Journal of Statistical Software, 64(3), 1–23. https://doi.org/10.18637/jss.v064.i03 Dray, S., & Dufour, A. B. (2007). The ade4 package: Implementing the
duality diagram for ecologists. Journal of Statistical Software, 22(4),
1–20. Griffin, A. R., Burgess, I. P., & Wolf, L. (1988). Patterns of natural and
manipulated hybridization in the genus Eucalyptus L'Herit.: A review. Australian Journal of Botany, 36(1), 41–66. Duncan, C. J., Worth, J. R. P., Jordan, G. J., Jones, R. C., & Vaillancourt,
R. E. (2016). Genetic differentiation in spite of high gene flow in
the dominant rainforest tree of southeastern Australia, Nothofagus
cunninghamii. Heredity, 116(1), 99–106. https://doi.org/10.1038/
hdy.2015.77 Grivet, D., Sork, V. L., Westfall, R. D., & Davis, F. W. (2008). Conserving
the evolutionary potential of California valley oak (Quercus lobata
Nee): A multivariate genetic approach to conservation planning. Molecular Ecology, 17(1), 139–156. Hamilton, N. R. S. (2001). Is local provenance important in habitat cre-
ation? A Reply. Journal of Applied Ecology, 38(6), 1374–1376. https://
doi.org/10.1046/j.0021-8901.2001.00670.x Eckert, A. J., van Heerwaarden, J., Wegrzyn, J. L., Nelson, C. D., Ross‐
Ibarra, J., Gonzalez‐Martinez, S. C., & Neale, D. B. (2010). Patterns
of population structure and environmental associations to aridity
across the range of loblolly pine (Pinus taeda L., Pinaceae). Genetics,
185(3), 969–982. https://doi.org/10.1534/genetics.110.115543 Hand, B. K., Hether, T. D., Kovach, R. P., Muhlfeld, C. C., Amish, S. J.,
Boyer, M. C., … Luikart, G. (2015). Genomics and introgression:
Discovery and mapping of thousands of species‐diagnostic SNPs
using RAD sequencing. Current Zoology, 61(1), 146–154. https://doi. org/10.1093/czoolo/61.1.146 Ellstrand, N. C., & Elam, D. R. (1993). Population genetic consequences
of small population size: Implications for plant conservation. Annual
Review of Ecology and Systematics, 24(1), 217–242. https://doi. org/10.1146/annurev.es.24.110193.001245 Hardy, A. D. (1918). Tall Trees of Australia. Victorian Naturalist, 35, 46–55. Hardy, A. D. (1935). Victoria's Great Trees. Victorian Naturalist, 51,
231–241. Evans, M. C., Watson, J. E. M., Fuller, R. A., Venter, O., Bennett, S. C.,
Marsack, P. R., & Possingham, H. P. (2011). The spatial distribution of
threats to species in Australia. BioScience, 61(4), 281–289. https://doi. org/10.1525/bio.2011.61.4.8 Harrisson, K. A., Pavlova, A., Gonçalves da Silva, A., Rose, R., Bull, J. K.,
Lancaster, M. L., … Sunnucks, P. (2016). Scope for genetic rescue
of an endangered subspecies though re‐establishing natural gene
flow with another subspecies. Molecular Ecology, 25(6), 1242–1258. https://doi.org/10.1111/mec.13547 Falush, D., Stephens, M., & Pritchard, J. K. (2003). Inference of popula-
tion structure using multilocus genotype data: Linked loci and cor-
related allele frequencies. Genetics, 164(4), 1567–1587. Harvey, M. G., Aleixo, A., Ribas, C. C., & Brumfield, R. T. (2017). Habitat
association predicts genetic diversity and population divergence in
Amazonian birds. The American Naturalist, 190(5), 631–648. https://
doi.org/10.1086/693856 Field, D. L., Ayre, D. J., Whelan, R. J., & Young, A. G. (2009). REFERENCES Ecological valence of mountain ash (Eucalyptus reg‐
nans F.Muell.) as a key to its distribution. Victorian Naturalist, 86, 6–24. Bartholome, J., Mandrou, E., Mabiala, A., Jenkins, J., Nabihoudine, I.,
Klopp, C., … Gion, J. M. (2015). High‐resolution genetic maps of
Eucalyptus improve Eucalyptus grandis genome assembly. New
Phytologist, 206(4), 1283–1296. Crow, T. M., Albeke, S. E., Buerkle, C. A., & Hufford, K. M. (2018). Provisional methods to guide species‐specific seed von TAKACH DUKAI et al. 857 G., Bossinger, G., & Ades, P. K. (2010). Phylogeography of the
world's tallest angiosperm, Eucalyptus regnans: Evidence for multiple
isolated Quaternary refugia. Journal of Biogeography, 37(1), 179–192. https://doi.org/10.1111/j.1365-2699.2009.02193.x Lambeck, K., Rouby, H., Purcell, A., Sun, Y. Y., & Sambridge, M. (2014). Sea level and global ice volumes from the Last Glacial Maximum
to the Holocene. Proceedings of the National Academy of Sciences
of the United States of America, 111(43), 15296–15303. https://doi. org/10.1073/pnas.1411762111 Nevill, P. G., Bradbury, D., Williams, A., Tomlinson, S., & Krauss, S. L. (2014). Genetic and palaeo‐climatic evidence for widespread per-
sistence of the coastal tree species Eucalyptus gomphocephala
(Myrtaceae) during the Last Glacial Maximum. Annals of Botany,
113(1), 55–67. https://doi.org/10.1093/aob/mct253 Law, B., Mackowski, C., Schoer, L., & Tweedie, T. (2000). Flowering
phenology of myrtaceous trees and their relation to cli-
matic, environmental and disturbance variables in northern
New South Wales. Austral Ecology, 25(2), 160–178. https://doi. org/10.1046/j.1442-9993.2000.01009.x O'Dowd, D. J., & Gill, A. M. (1984). Predator satiation and site alteration
following fire: Mass reproduction of alpine ash (Eucalyptus delegaten‐
sis) in Southeastern Australia. Ecology, 65(4), 1052–1066. https://doi. org/10.2307/1938313 Orsini, L., Vanoverbeke, J., Swillen, I., Mergeay, J., & Meester, L. (2013). Drivers of population genetic differentiation in the wild: Isolation by
dispersal limitation, isolation by adaptation and isolation by coloniza-
tion. Molecular Ecology, 22(24), 5983–5999. https://doi.org/10.1111/
mec.12561 Lepais, O., Petit, R. J., Guichoux, E., Lavabre, J. E., Alberto, F., Kremer,
A., & Gerber, S. (2009). Species relative abundance and direction of
introgression in oaks. Molecular Ecology, 18(10), 2228–2242. https://
doi.org/10.1111/j.1365-294X.2009.04137.x Li, H., Handsaker, B., Wysoker, A., Fennell, T., Ruan, J., Homer, N.,
… Durbin, R. (2009). The Sequence Alignment/Map format
and SAMtools. Bioinformatics, 25(16), 2078–2079. https://doi. org/10.1093/bioinformatics/btp352 Palme, A. E., Su, Q., Palsson, S., & Lascoux, M. (2004). Extensive sharing of
chloroplast haplotypes among European birches indicates hybridiza-
tion among Betula pendula, B. pubescens and B. nana. Molecular Ecology,
13(1), 167–178. https://doi.org/10.1046/j.1365-294X.2003.02034.x Lindenmayer, D. B., Blanchard, W., Blair, D., McBurney, L., & Banks, S. C. (2016). Environmental and human drivers influencing large old tree
abundance in Australian wet forests. Forest Ecology and Management,
372, 226–235. https://doi.org/10.1016/j.foreco.2016.04.017 Peakall, R., & Smouse, P. E. (2006). GENALEX 6: Genetic analy-
sis in Excel. Population genetic software for teaching and re-
search. Molecular Ecology Notes, 6(1), 288–295. https://doi. org/10.1111/j.1471-8286.2005.01155.x Lindenmayer, D. B., Hobbs, R. J., Likens, G. E., Krebs, C. J., & Banks, S. C. (2011). McCartney‐Melstad, E., & Shaffer, H. B. (2015). Amphibian molecular
ecology and how it has informed conservation. Molecular Ecology,
24(20), 5084–5109. https://doi.org/10.1111/mec.13391 Hudson, I. L., Kim, S. W., & Keatley, M. R. (2010). Climatic influences on
the flowering phenology of four eucalypts: A GAMLSS approach. In
I. L. Hudson, & M. R. Keatley (Eds.), Phenological Research (pp. 209–
228). Dordrecht, the Netherland: Springer. McKenna, A., Hanna, M., Banks, E., Sivachenko, A., Cibulskis, K.,
Kernytsky, A., … DePristo, M. A. (2010). The Genome Analysis
Toolkit: A MapReduce framework for analyzing next‐generation
DNA sequencing data. Genome Research, 20(9), 1297–1303. https://
doi.org/10.1101/gr.107524.110 228). Dordrecht, the Netherland: Springer. Ikeda, D. H., Max, T. L., Allan, G. J., Lau, M. K., Shuster, S. M., & Whitham,
T. G. (2017). Genetically informed ecological niche models improve
climate change predictions. Global Change Biology, 23(1), 164–176. https://doi.org/10.1111/gcb.13470 Jombart, T. (2008). adegenet: A R package for the multivariate analysis
of genetic markers. Bioinformatics, 24(11), 1403–1405. https://doi. org/10.1093/bioinformatics/btn129 McKinnon, G. E., Smith, J. J., & Potts, B. M. (2010). Recurrent nuclear
DNA introgression accompanies chloroplast DNA exchange be-
tween two eucalypt species. Molecular Ecology, 19(7), 1367–1380. https://doi.org/10.1111/j.1365-294X.2010.04579.x Jones, M. E., Shepherd, M., Henry, R. J., & Delves, A. (2006). Chloroplast
DNA variation and population structure in the widespread for-
est tree, Eucalyptus grandis. Conservation Genetics, 7(5), 691–703. https://doi.org/10.1007/s10592-005-9104-7 Melville, J., Haines, M. L., Boysen, K., Hodkinson, L., Kilian, A., Smith
Date, K. L., … Parris, K. M. (2017). Identifying hybridization and ad-
mixture using SNPs: Application of the DArTseq platform in phylo-
geographic research on vertebrates. Royal Society Open Science, 4(7),
161061. https://doi.org/10.1098/rsos.161061 Jordan, R., Hoffmann, A. A., Dillon, S. K., & Prober, S. M. (2017). Evidence of genomic adaptation to climate in Eucalyptus microcarpa:
Implications for adaptive potential to projected climate change. Molecular Ecology, 26(21), 6002–6020. https://doi.org/10.1111/
mec.14341 Murray, K. D., & Borevitz, J. O. (2018). Axe: Rapid, competitive sequence
read demultiplexing using a trie. Bioinformatics, 1–2, https://doi. org/10.1093/bioinformatics/bty432 Kelly, E., & Phillips, B. L. (2016). Targeted gene flow for conservation. Conservation Biology, 30(2), 259–267. https://doi.org/10.1111/
cobi.12623 Myburg, A. A., Grattapaglia, D., Tuskan, G. A., Hellsten, U., Hayes, R. D., Grimwood, J., … Schmutz, J. (2014). The genome of Eucalyptus
grandis. Nature, 510(7505), 356–362. https://doi.org/10.1038/
Nature13308 Kuparinen, A., Savolainen, O., & Schurr, F. M. (2010). Increased mortal-
ity can promote evolutionary adaptation of forest trees to climate
change. Forest Ecology and Management, 259(5), 1003–1008. https://
doi.org/10.1016/j.foreco.2009.12.006 Nevill, P. Molecular
and morphological evidence of natural interspecific hybridization
between the uncommon Eucalyptus aggregata and the widespread E. rubida and E. viminalis. Conservation Genetics, 10(4), 881–896. https://
doi.org/10.1007/s10592-008-9649-3 Hecht, B. C., Matala, A. P., Hess, J. E., & Narum, S. R. (2015). Environmental
adaptation in Chinook salmon (Oncorhynchus tshawytscha) through-
out their North American range. Molecular Ecology, 24(22), 5573–
5595. https://doi.org/10.1111/mec.13409 Fisher, M. C., Garner, T. W. J., & Walker, S. F. (2009). Global emergence of
Batrachochytrium dendrobatidis and amphibian chytridiomycosis in
space, time, and host. Annual Review of Microbiology, 63(1), 291–310. https://doi.org/10.1146/annurev.micro.091208.073435 Hegyi, G., & Garamszegi, L. Z. (2011). Using information theory as a sub-
stitute for stepwise regression in ecology and behavior. Behavioral
Ecology and Sociobiology, 65(1), 69–76. https://doi.org/10.1007/
s00265-010-1036-7 Fitzpatrick, S. W., Gerberich, J. C., Kronenberger, J. A., Angeloni, L. M., &
Funk, W. C. (2015). Locally adapted traits maintained in the face of
high gene flow. Ecology Letters, 18(1), 37–47. https://doi.org/10.1111/
ele.12388 Hendricks, S., Epstein, B., Schönfeld, B., Wiench, C., Hamede, R., Jones,
M., … Hohenlohe, P. (2017). Conservation implications of limited
genetic diversity and population structure in Tasmanian devils
(Sarcophilus harrisii). Conservation Genetics, 18(4), 977–982. https://
doi.org/10.1007/s10592-017-0939-5 Flint, A. W., & Fagg, P. C. (2007). Mountain ash in Victoria’s state forests,
silviculture reference manual no. 1. Hawthorn, Victoria. Forester, B. R., Jones, M. R., Joost, S., Landguth, E. L., & Lasky, J. R. (2016). Detecting spatial genetic signatures of local adaptation in heteroge-
neous landscapes. Molecular Ecology, 25(1), 104–120. https://doi. org/10.1111/mec.13476 Hudson, C. J., Freeman, J. S., Myburg, A. A., Potts, B. M., & Vaillancourt,
R. E. (2015). Genomic patterns of species diversity and divergence in
Eucalyptus. New Phytologist, 206(4), 1378–1390. 858 von TAKACH DUKAI et al. Herbarium records identify sensitivity of flowering phenology of
eucalypts to climate: Implications for species response to climate
change. Austral Ecology, 40(2), 117–125. https://doi.org/10.1111/
aec.12183 Suarez‐Gonzalez, A., Lexer, C., & Cronk, Q. C. B. (2018). Adaptive in-
trogression: A plant perspective. Biology Letters, 14(3), https://doi. org/10.1098/rsbl.2017.0688 Supple, M. A., Bragg, J. G., Broadhurst, L. M., Nicotra, A. B., Byrne, M.,
Andrew, R. L., … Borevitz, J. O. (2018). Landscape genomic predic-
tion for restoration of a Eucalyptus foundation species under climate
change. eLIFE Sciences, 1–22. https://doi.org/10.1101/200352 Reside, A. E., Beher, J., Cosgrove, A. J., Evans, M. C., Seabrook, L., Silcock,
J. L., … Maron, M. (2017). Ecological consequences of land clearing
and policy reform in Queensland. Pacific Conservation Biology, 23(3),
219–230. https://doi.org/10.1071/PC17001 Taubert, F., Fischer, R., Groeneveld, J., Lehmann, S., Müller, M. S., Rödig,
E., … Huth, A. (2018). Global patterns of tropical forest fragmenta-
tion. Nature, 554, 519. https://doi.org/10.1038/nature25508 Reynolds, R. G., Puente‐Rolón, A. R., Platenberg, R., Tyler, R. K.,
Tolson, P. J., & Revell, L. J. (2015). Large divergence and low di-
versity suggest genetically informed conservation strategies for
the endangered Virgin Islands Boa (Chilabothrus monensis). Global
Ecology and Conservation, 3, 487–502. https://doi.org/10.1016/j. gecco.2015.02.003 van Dyck, S., Gynther, I., & Baker, A. (Eds.) (2013). Field Companion to
the Mammals of Australia. London, UK and Sydney, Australia: New
Holland Publishers. Vavrek, M. (2011). Fossil: Palaeoecological and palaeogeographical anal-
ysis tools. Palaeontologia Electronica, 14(1), 16. Rhymer, J. M., & Simberloff, D. (1996). Extinction by hybridization and
introgression. Annual Review of Ecology and Systematics, 27, 83–109. https://doi.org/10.1146/annurev.ecolsys.27.1.83 von Takach Dukai, B., Lindenmayer, D. B., & Banks, S. C. (2018). Environmental influences on growth and reproductive maturation of
a keystone forest tree: Implications for obligate seeder susceptibil-
ity to frequent fire. Forest Ecology and Management, 411(December
2017), 108–119. https://doi.org/10.1016/j.foreco.2018.01.014 Ribeiro, P. J. Jr, & Diggle, P. J. (2016). geoR: A package for geostatistical
analysis. R Package Version, 1(7‐5), 2. Rousset, F. (1997). Genetic differentiation and estimation of gene
flow from F‐statistics under isolation by distance. Genetics, 145(4),
1219–1228. Wang, T., O'Neill, G. A., & Aitken, S. N. (2010). Integrating environ-
mental and genetic effects to predict responses of tree popula-
tions to climate. Ecological Applications, 20(1), 153–163. https://doi. org/10.1890/08-2257.1 Sampson, J., Tapper, S., Coates, D., Hankinson, M., McArthur, S., &
Byrne, M. (2018). Individuals to ecosystems (pp. 56–91). Cambridge, UK: Cambridge
University Press. continuous patterns of human genetic variation. Human Genomics,
2(2), 81. https://doi.org/10.1186/1479-7364-2-2-81 continuous patterns of human genetic variation. Human Genomics,
2(2), 81. https://doi.org/10.1186/1479-7364-2-2-81 Simmons, M. H., Wapstra, H., & Wapstra, A. (2008). A guide to flowers and
plants of Tasmania (4th ed.). Chatswood, NSW: Reed New Holland. Pounds, A. J., Bustamante, M. R., Coloma, L. A., Consuegra, J. A., Fogden,
M. P. L., Foster, P. N., … Young, B. E. (2006). Widespread amphibian
extinctions from epidemic disease driven by global warming. Nature,
439, 161. https://doi.org/10.1038/nature04246 Skotte, L., Korneliussen, T. S., & Albrechtsen, A. (2013). Estimating in-
dividual admixture proportions from next generation sequenc-
ing data. Genetics, 195(3), 693–702. https://doi.org/10.1534/
genetics.113.154138 Pritchard, J. K., Stephens, M., & Donnelly, P. (2000). Inference of pop-
ulation structure using multilocus genotype data. Genetics, 155(2),
945–959. Sork, V. L. (2016). Gene flow and natural selection shape spatial patterns
of genes in tree populations: Implications for evolutionary processes
and applications. Evolutionary Applications, 9(1), 291–310. https://doi. org/10.1111/eva.12316 Prober, S., Byrne, M., McLean, E., Steane, D. A., Potts, B., Vaillancourt,
R., & Stock, W. (2015). Climate‐adjusted provenancing: A strategy
for climate‐resilient ecological restoration. Frontiers in Ecology and
Evolution, 3, 65. https://doi.org/10.3389/fevo.2015.00065 Sork, V. L., Davis, F. W., Westfall, R., Flint, A., Ikegami, M., Wang, H. F., & Grivet, D. (2010). Gene movement and genetic association
with regional climate gradients in California valley oak (Quercus lo‐
bata Nee) in the face of climate change. Molecular Ecology, 19(17),
3806–3823. Pryor, L. D. (1976). The biology of eucalypts (Vol. 61). London, UK: Edward
Arnold (Publishers) Ltd. R Core Team. (2017). R: A language and environment for statistical com‐
puting. Vienna, Austria: R Foundation for Statistical Computing. Retrieved from http://www.r-project.org/ Steane, D. A., Potts, B. M., McLean, E. H., Collins, L., Holland, B. R.,
Prober, S. M., … Byrne, M. (2017). Genomic scans across three eu-
calypts suggest that adaptation to aridity is a genome‐wide phe-
nomenon. Genome Biology and Evolution, 9(2), 253–265. https://doi. org/10.1093/gbe/evw290 Ralls, K., Ballou, J. D., Dudash, M. R., Eldridge, M. D. B., Fenster, C. B.,
Lacy, R. C., … Frankham, R. (2018). Call for a paradigm shift in the ge-
netic management of fragmented populations. Conservation Letters,
11(2), e12412. https://doi.org/10.1111/conl.12412 Rawal, D. S., Kasel, S., Keatley, M. R., & Nitschke, C. R. (2015). Newly discovered landscape traps produce regime shifts
in wet forests. Proceedings of the National Academy of Sciences of
the United States of America, 108(38), 15887–15891. https://doi. org/10.1073/pnas.1110245108 Peakall, R., & Smouse, P. E. (2012). GenAlEx 6.5: Genetic analysis in
Excel. Population genetic software for teaching and research‐an
update. Bioinformatics, 28(19), 2537–2539. https://doi.org/10.1093/
bioinformatics/bts460 Petit, R. J., Duminil, J., Fineschi, S., Hampe, A., Salvini, D., &
Vendramin, G. G. (2005). Invited review: Comparative orga-
nization of chloroplast, mitochondrial and nuclear diversity in
plant populations. Molecular Ecology, 14(3), 689–701. https://doi. org/10.1111/j.1365-294X.2004.02410.x Lindenmayer, D. B., & Sato, C. (2018). Hidden collapse is driven by fire
and logging in a socioecological forest ecosystem. Proceedings of the
National Academy of Sciences of the United States of America, 115(20),
5181–5186. https://doi.org/10.1073/pnas.1721738115 Mantel, N. (1967). The detection of disease clustering and a generalized
regression approach. Cancer Research, 27(2), 209–220. Potts, B. M., Barbour, R. C., Hingston, A. B., & Vaillancourt, R. E. (2003). Turner review no. 6: Genetic pollution of native eucalypt gene pools
– Identifying the risks. Australian Journal of Botany, 51(1), 1–25. https://doi.org/10.1071/BT02035 Maunder, M., Cowan, R. S., Stranc, P., & Fay, M. F. (2001). The genetic
status and conservation management of two cultivated bulb spe-
cies extinct in the wild: Tecophilaea cyanocrocus (Chile) and Tulipa
sprengeri (Turkey). Conservation Genetics, 2(3), 193–201. https://doi. org/10.1023/a:1012281827757 Potts, B. M., & Wiltshire, R. J. E. (1997). Eucalypt genetics and genecol-
ogy. In J. E. Williams, & J. C. Z. Woinarski (Eds.), Eucalypt ecology: von TAKACH DUKAI et al. 859 Persistence with episodic range expansion from
the early Pleistocene: The distribution of genetic variation in the for-
est tree Corymbia calophylla (Myrtaceae) in south‐western Australia. Biological Journal of the Linnean Society, 123(May), 545–560. https://
doi.org/10.1093/biolinnean/blx168 Wickham, H. (2009). ggplot2: Elegant graphics for data analysis. New York,
NY: Springer‐Verlag Williams, K. J., Biggs, I., McConchie, C., Briggs, P. R., Underhill, J., Ryan, P. … Thorburn, P. (2006). Identification of potential new growing areas for
macadamias. Sydeny, NSW: Horticulture Australia Limited. Savolainen, O., Lascoux, M., & Merilä, J. (2013). Ecological genomics
of local adaptation. Nature Reviews Genetics, 14, 807. https://doi. org/10.1038/nrg3522 Williams, K. J., Ferrier, S., Rosauer, D. F., Yeates, D., Manion, G., Harwood,
T., … Whalen, A. (2010). Harnessing continent‐wide biodiversity
datasets for prioritising national conservation investment. A report
prepared for the Department of Sustainability, Environment, Water,
Population and Communities (978 0 643 10342 9). Retrieved from
Canberra. Schwabe, A. L., Neale, J. R., & McGlaughlin, M. E. (2015). Examining the
genetic integrity of a rare endemic Colorado cactus (Sclerocactus
glaucus) in the face of hybridization threats from a close and wide-
spread congener (Sclerocactus parviflorus). Conservation Genetics,
16(2), 443–457. https://doi.org/10.1007/s10592-014-0671-3 Woinarski, J. C. Z., Burbidge, A. A., & Harrison, P. L. (2015). Ongoing un-
raveling of a continental fauna: Decline and extinction of Australian
mammals since European settlement. Proceedings of the National
Academy of Sciences of the United States of America, 112(15), 4531–
4540. https://doi.org/10.1073/pnas.1417301112 Sedlazeck, F. J., Rescheneder, P., & von Haeseler, A. (2013). NextGenMap:
Fast and accurate read mapping in highly polymorphic genomes. Bioinformatics,
29(21),
2790–2791. https://doi.org/10.1093/
bioinformatics/btt468 Xu, T., & Hutchinson, M. F. (2011). ANUCLIM (Version 6.1). Australian
National University, Canberra: Fenner School of Environment and
Society. Shriver, M. D., Mei, R., Parra, E. J., Sonpar, V., Halder, I., Tishkoff, S. A., …
Jones, K. W. (2005). Large‐scale SNP analysis reveals clustered and 860 von TAKACH DUKAI et al. Zheng, X., & Weir, B. S. (2016). Eigenanalysis of SNP data with an identity
by descent interpretation. Theoretical Population Biology, 107, 65–76. https://doi.org/10.1016/j.tpb.2015.09.004 Xu, T., & Hutchinson, M. F. (2013). New developments and applications
in the ANUCLIM spatial climatic and bioclimatic modelling pack-
age. Environmental Modelling and Software, 40, 267–279. https://doi. org/10.1016/j.envsoft.2012.10.003 Yeoh, S. H., Bell, J. C., Foley, W. J., Wallis, I. R., & Moran, G. F. (2012). Estimating population boundaries using regional and local‐scale
spatial genetic structure: An example in Eucalyptus globulus. Tree Genetics & Genomes, 8(4), 695–708. https://doi.org/10.1007/
s11295-011-0457-4 SUPPORTING INFORMATION Additional supporting information may be found online in the
Supporting Information section at the end of the article. Yorke, S., & Ashton, D. H. (1982). A possible tri‐hybrid eucalypt near the
Cathedral Range, Victoria. Victorian Naturalist, 99, 209–213. Young, H. A., Lanzatella, C. L., Sarath, G., & Tobias, C. M. (2011). Chloroplast genome variation in upland and lowland switch-
grass. PLoS ONE, 6(8), e23980. https://doi.org/10.1371/journal. pone.0023980 How to cite this article: von Takach Dukai B, Jack C, Borevitz J,
Lindenmayer DB, Banks SC. Pervasive admixture between
eucalypt species has consequences for conservation and
assisted migration. Evol Appl. 2019;12:845–860. https://doi. org/10.1111/eva.12761 Yu, G. (2018). Package 'scatterpie'. R Package Version, (1), 2. Zheng, X., Levine, D., Shen, J., Gogarten, S. M., Laurie, C., & Weir, B. S. (2012). A high‐performance computing toolset for relatedness and
principal component analysis of SNP data. Bioinformatics, 28(24),
3326–3328. https://doi.org/10.1093/bioinformatics/bts606
|
https://openalex.org/W2564131168
|
https://ria.ua.pt/bitstream/10773/26613/1/heusler_2.pdf
|
English
| null |
Volume dependence of magnetic properties in <mml:math xmlns:mml="http://www.w3.org/1998/Math/MathML" altimg="si0021.gif" overflow="scroll"><mml:mrow><mml:msub><mml:mrow><mml:mi>Co</mml:mi></mml:mrow><mml:mrow><mml:mn>2</mml:mn></mml:mrow></mml:msub><mml:msub><mml:mrow><mml:mi>Cr</mml:mi></mml:mrow><mml:mrow><mml:mn>1</mml:mn><mml:mo>−</mml:mo><mml:mi>x</mml:mi></mml:mrow></mml:msub><mml:msub><mml:mrow><mml:mi>Y</mml:mi></mml:mrow><mml:mrow><mml:mi>x</mml:mi></mml:mrow></mml:msub><mml:mi>Ga</mml…
|
Journal of magnetism and magnetic materials
| 2,017
|
cc-by
| 6,507
|
Abstract The magnetic properties tuning and volume dependence in the series of qua-
ternary full Heusler alloys with formula Co2Cr1−xYxGa (Y = Ti, V, Mn, Fe,
Co, Ni) were studied with a detailed first-principles exploration. We employ
the density functional KKR method with the coherent potential approximation,
estimating effective Heisenberg exchange constants via the magnetic force the-
orem together with mean-field Curie temperature (TC) and magnetic moment
for compositions in the whole concentration range. The volumetric dependency
of these magnetic properties is studied, particularly the pressure derivatives of
TC at equilibrium. Our ternary alloy calculations show good agreement with
local-density and generalized gradient approximations in the literature. The
quaternary alloys show a wide range of tunable magnetic properties, where
magnetic moments range from 0.8 to 4.9 µB, TC from 130 K to 1250 K, and
dTC/dV values range from −7 to +6.3 K ˚A−3. Keywords:
Heusler alloys, Curie temperature, pressure, volume dependence Volume dependence of magnetic properties in
Co2Cr1−xYxGa (Y =Ti-Ni) Heusler alloys: a
first-principles study J. N. Gon¸calvesa, N. M. Fortunatoa, J. S. Amarala,b, V. S. Amarala J. N. Gon¸calvesa, N. M. Fortunatoa, J. S. Amarala,b, V. S. Amarala
aCICECO - Aveiro Institute of Materials and Departamento de F´ısica, Universidade de
Aveiro, 3810-193 Aveiro, Portugal
bIFIMUP and IN-Institute of Nanoscience and Nanotechnology, Rua do Campo Alegre, 678,
4169-007 Porto, Portugal Email address: joaonsg@ua.pt (J. N. Gon¸calves) J. N. Gon¸calvesa, N. M. Fortunatoa, J. S. Amarala,b, V. S. Amarala
aCICECO - Aveiro Institute of Materials and Departamento de F´ısica, Universidade de
Aveiro, 3810-193 Aveiro, Portugal
bIFIMUP and IN-Institute of Nanoscience and Nanotechnology, Rua do Campo Alegre, 678,
4169-007 Porto, Portugal Preprint submitted to Elsevier 1. Introduction Heusler alloys are known for the great variety of compositions and properties,
since many elements can occupy the X, Y and Z sites in the L21 structure with
fractional occupations [1], allowing a great flexibility. Improved properties may
also be obtained by exploring disorder, non-stoichiometric compositions, and
variation with external parameters such as magnetic field, temperature and
pressure [2]. The exploration of several independent factors can be achieved in
a controlled way using first-principles calculations. Interest in Heusler alloys has
mainly been concerning electronic structure and transport properties, motivated
by application in spintronic devices [3], especially in half-Heusler XY Z and September 13, 2017 Preprint submitted to Elsevier full-Heusler Co2Y Z alloys. Other aspects of Heusler alloys have also motivated
first-principles studies, including exchange interactions [4, 5], to calculations of
the magnetocaloric effect [6, 7] with the aid of Monte Carlo simulations. In this context, a related aspect which benefits from computation but has
been less studied is the variation of exchange interactions, and ferromagnetic
Curie temperature (TC) with volume, or hydrostatic pressure. An initial de-
tailed theoretical study of this subject was concerned with Ni2MnSn [8]. More
recently, Mn2CoZ (Z = Al, Ga, In, Si, Ge, Sn, Sb) compounds have been an-
alyzed [9]. (Ni1−xTx)2MnSn (T = Pd, Cu) alloys have also been studied as a
function of concentration and pressure [10]. For Cu alloying, a crossover behav-
ior was found, with dTC/dP changing sign as a function of Cu concentration. ,
/
g
g
g
Experimental investigations of TC with pressure have also been specially
focused in the Mn based systems, due to coupling between structure and mag-
netism (with, e.g. the shape memory and magnetocaloric effects), including
Ni2MnZ (Z = Al, Ga, In, Sn, and Sb) [11, 12], Au2MnAl and Pd2MnZ (Z =
Sn and Sb) [13]. For all these alloys, wherein Mn has localized moments, the
positive dTC/dP (negative dTC/dV ) measured is consistent with an empirical
interaction curve as a function of the distance between Mn atoms [12]. [
]
For cobalt-based (Co2Y Z) alloys there have been experiments with Co2TiAl/Co2TiGa [14,
15], Co2ZrAl [16], and Co2VAl/Co2VGa [17]. These measurements revealed a
negative dTC/dP, which is typical of itinerant magnetism. The Co2MnZ (Z =
Ga, Si, Ge, Sn) alloys, however, display a positive dTC/dP, in conformity with
the other Mn-based systems [18]. ©2016. This manuscript version is made available under the CC-BY-NC-ND
4.0 license http://creativecommons.org/licenses/by-nc-nd/4.0/ 1. Introduction In a previous study [19] we have made an investigation of different Co2Y Z
compounds, including the variation of lattice parameters, and its effect on mag-
netic interactions and TC. In this work we focus in particular on stoichiometries
based in the Co2CrGa compound. Cr is substituted with one of six different
elements: Ti, V, Mn, Fe, Co, and Ni, and these alloys are studied with fractional
concentrations covering the whole concentration range. For this kind of system, previous GGA/LDA calculations show good agree-
ment with experimental data. For example, Ref. [20] uses GGA, Ref. [21] is a
compilation showing the success of LDA/GGA in Co2Y Z systems, where the
moment is well represented as well as TC (in this case using the random phase
approximation). Ref. [22] also shows that the LDA spin moment is consistent
with experiment. According to the discussion of Ref. [23], LDA/GGA is suffi-
cient for accurate Curie temperatures (mean field) and magnetic moments, for
several Co2Y Z and two Ni2MnZ alloys, with the only exception for the case of
Co2FeSi (which we do not study here), for which LSDA+U is found to better
describe the magnetic moment. However, with respect to magnetovolume effects, previous calculations have
shown significant differences between different exchange-correlation approxima-
tions. In bcc-Fe, dTC/dP was found to be -1.4 K/GPa with GGA and -6 K/GPa
with LDA [24]. These results are much smaller than previous first-principles re-
sults (16 and 18 K/GPa) with a different method [25], and thus the authors
claimed a correct prediction of the experimental dTC/dP ≈0. Nevertheless,
these are still significant changes, both relative and absolute. Values of the same order of magnitude have been found for Heusler alloys (for example, 6.2 and 5.6
K/GPa respectively for Ni2MnSn and Ni2MnIn [26]). Therefore, we found it
necessary to study the influence of the exchange-correlation approximation in a
series of LDA and GGA calculations. We will show that, for the compositions
containing Ti and V, dTC/dV is strongly dependent on the exchange-correlation
functional: the measured sign is obtained with the local density approximation,
but not with the generalized gradient approximation. Using the LDA, we find that those compositions close to Co2TiGa and
Co2VGa display positive (negative) dTC/dV (dTC/dP), as measured, and ex-
pected for itinerant magnetism. Moving forward in the transition metal period
to Cr, Mn, and Fe, dTC/dV (dTC/dP) becomes negative (positive), consistent
with localized moments. ©2016. This manuscript version is made available under the CC-BY-NC-ND
4.0 license http://creativecommons.org/licenses/by-nc-nd/4.0/ 1The use of the full potential by Thoene et al. [23] was found important in some cases to
obtain accurate Curie temperatures. 1. Introduction To our knowledge Co2NiGa forms in a disordered fcc or ordered Pt2FeCu-
type structure [27], while Co3Ga is not known in any form. We will also analyze
the alloying of Co and Ni in the quaternary alloys with the L21 structure, bear-
ing in mind that we are not describing known compounds in these cases. A
previous first-principles study has already analyzed the stability of Co2NiGa,
and found, as expected, that the ground state cubic structure is metastable
with respect to the tetragonal distortion [28]. However, L21-Co2NiGa may be
formed with novel synthesis methods, and the fractional alloys with Cr should
be synthesized more easily than the ternary compounds, especially for low sub-
stitution concentrations. For the compositions close to Co3Ga and Co2NiGa
our calculations show dTC/dV (dTC/dP) again positive (negative). Therefore,
substituting Ti, V, Co, or Ni for Cr gradually decreases dTC/dV and eventually
changes its sign. 2It was previously shown [36], by comparison with supercell calculations, that short-range
interactions should not affect the magnetic properties significantly for the type of alloys we
are considering. ©2016. This manuscript version is made available under the CC-BY-NC-ND
4.0 license http://creativecommons.org/licenses/by-nc-nd/4.0/ 2. Calculation Details We use the KKR Green’s function density functional theory, as implemented
in the SPR-KKR code [29, 30]. The full potential1 spin-polarized scalar-relativistic
method is used. As the exchange-correlation approximation we consider the gen-
eralized gradient approximation (GGA) with the Perdew-Burke-Ernzerhof pa-
rameterization [31] and the local density approximation (LDA) with the Vosko-
Wilk-Nusair parameterization [32] for the ternary compounds. The LDA is
used for the detailed study of the alloys. The reciprocal space is sampled with a
22 × 22 × 22 k-mesh, the calculation in the complex energy path uses 32 points,
and the angular momentum cutoffis l = 3. The magnetic order is assumed
ferromagnetic, which is consistent with the known measured cases, such as the
ternary compounds (also under pressure), and with Fe or Mn doping [33, 34]. For the quaternary, fractional alloys, we sample variation of different concentra-
tions of transition metal element substitutions in the Y site of the L21 structure,
with chemical formula Co2Cr1−xYxGa, and considering six neighbor elements (Y =Ti, V, Mn, Fe, Co, Ni). The coherent potential approximation is used to
simulate the disorder in that site [35]. 2 Steps of ∆x = 0.1 were considered. (Y =Ti, V, Mn, Fe, Co, Ni). The coherent potential approximation is used to
simulate the disorder in that site [35]. 2 Steps of ∆x = 0.1 were considered. Self-consistent total energy calculations at different volumes are used with
the Murnaghan equation of state [37] to fit the energy-volume curve and find
the corresponding pressures. (There are no internal relaxations involved due to
the symmetries of the L21 structure.) )
The exchange constants of an effective Heisenberg model are calculated with
the magnetic force theorem of Liechtenstein et al. [38]. For this calculation a
cluster of radius four times larger than the L21 cell lattice constant is considered
around each site and all the pairwise interactions in that cluster are calculated,
considering the four sublattices. TC is determined in the mean field approxima-
tion for a multisublattice system [39, 8], i.e., using the highest eigenvalue of the
matrix J containing the pairwise exchange interaction sums. 3. Results 3.1. Ternary compounds: exchange-correlation approximation and comparison
with experiments 3.1. Ternary compounds: exchange-correlation approximation and comparison
with experiments The calculated lattice parameters and magnetic moments are shown and
compared with experiments [17, 22, 33, 40, 41] in Table 1. GGA is more accurate
for the lattice parameters. The bulk moduli are also shown, and follow the usual
tendency of larger values for LDA, with smaller lattice parameters. With regards
to the magnetic moments, there is not one approximation consistently better
for all compounds: LDA is better for Co2TiGa, GGA is better for Co2FeGa and
slightly better for Co2CrGa, while for the other compounds both approximations
have similar accuracies. For all compositions except Co2CrGa, the TC obtained
with LDA is closer to the experimental [40, 42, 43, 44, 33] one than that obtained
with GGA. This does not necessarily mean that LDA produces a better estimate
of the exchange constant values than GGA, since the mean field approach used
here usually overestimates TC. Although the mean field approach to Curie temperatures usually overesti-
mates Curie temperatures, Thoene et al. [23] have obtained TC close to the ex-
perimental values for a variety of Heusler compounds. For the case of Co2CrGa,
it was even underestimated, 366 K with respect to the experimental value of
495 K [43]. However, this comes from using the experimental lattice parameters
instead of the theoretical ones. As shown in Tab. 1, the TC calculated at the
theoretical LDA lattice is close to the experimental one. We also find similar
values of the derivatives dTC/dV with both approximations at the equilibrium
values. For other compositions, dTC/dV does not always show such a good agree-
ment between functionals. This is shown here for Co2TiGa, Co2VGa, Co2MnGa, Table 1: Comparison of the calculated (GGA and LDA) equilibrium lattice parameters a0 (˚A), bulk moduli B (GPa), Curie temperature TC (K),
magnetic moment m (µB), dm/dV (10−3µB ˚A−3), and dTC/dV (K ˚A−3) at the equilibrium lattice parameters, and experimental TC and a0. GGA
LDA
exp. 3.1. Ternary compounds: exchange-correlation approximation and comparison
with experiments Compound
a0
B
TC
m
dTC/dV
dm/dV
a0
B
TC
m
dTC/dV
dm/dV
a0
TC
m
Co2TiGa
5.87
179
243
0.97
-1.2
-0.2
5.72
216
130
0.82
4.5
5.1
5.86
128
0.82
Co2VGa
5.80
194
469
1.96
-1.5
-2.8
5.66
237
345
1.96
1.7
-1.2
5.78
352
2.04
Co2CrGa
5.75
195
485
2.99
-5.3
9.9
5.61
247
470
2.96
-4.6
0.4
5.81
495
3.01
Co2MnGa
5.75
187
662
4.09
-8.2
13.8
5.60
237
670
4.01
-7.0
9.2
5.77
685
4.05
Co2FeGa
5.75
187
1317
4.99
-4.9
17.4
5.60
231
1252
4.87
-3.2
26.5
5.74
1093
5.15
pt version is made available unde
//creativecommons.org/licenses/ ©2016. This manuscript version is made available under the CC-BY-NC-ND
4.0 license http://creativecommons.org/licenses/by-nc-nd/4.0/ and Co2FeGa. For Co2MnGa and Co2FeGa, similarly to Co2CrGa, there are
quantitative differences but the results are still consistent, since the derivatives
are not far in the two cases and have the same sign. For Co2VGa, however, the
TC dependence changes sign with the different approximations, at the theoreti-
cal lattice parameters. The difference of the derivative between approximations
is not far from the other cases, but some factors contribute to the change of sign:
small values near the theoretical parameters, and the large variation of the value
with volume in LDA, plus the usual overbinding of LDA. For Co2TiGa, the vol-
ume dependence of the Curie temperature is even more distinct for the two
functionals, an almost constant behavior with negative slope with GGA, but a
steeper, positive dependence with LDA. (Note: dTC/dV values were calculated
with respect to the L21 unit cell volume differences. If we used the volume per
formula unit, the values would be four times larger.) It is also interesting to analyze the change of the magnetic moments with
volume. For Y =Mn and Fe the results are consistent for both approximations
only with small quantitative differences, as in dTC/dV , but with opposite sign. For the case of Co2TiGa we have exactly the same sign change as in dTC/dV ,
indicating the change of m as the main factor in the variation of the exchange
interactions. In contrast, for Co2VGa dm/dV has the same sign between func-
tionals, opposing the dTC/dV behavior. For Co2CrGa, dm/dV is much higher
for GGA than for LDA (where it is almost zero since m is a minimum close to
equilibrium). ©2016. This manuscript version is made available under the CC-BY-NC-ND
4.0 license http://creativecommons.org/licenses/by-nc-nd/4.0/ 3.1. Ternary compounds: exchange-correlation approximation and comparison
with experiments This also contrasts with the TC dependence, which is basically
the same, negative and not insignificant, for both functionals. The TC values and its variation with volume and approximation (LDA/GGA)
can be explained as a result of the underlying exchange interactions. The two
largest exchange interactions are shown in Table 2. For Co2FeGa and Co2MnGa,
JCo−Y is much larger than the next largest interaction, Co-Co, and both interac-
tions decrease with volume, consistent with TC. JCo−F e decreases slightly faster
for GGA, as with TC. For Co2CrGa, JCo−Cr is only slightly larger than the
Co-Co interaction. The dependence of the first interaction is slightly steeper for
LDA, consistent with the small quantitative difference between approximations. Co2VGa and Co2TiGa are different, since here the first Co-Co is much larger
than the first Co-Y interaction, which is already insignificant. For Co2VGa the
first Co-Co interaction has the same variation in both approximations, although
with GGA the variation is less steep, and it is not enough to account for the
sign change with dTC/dV (it must be accounted by considering Co-Co at larger
distances, which also have distinct volume changes between LDA and GGA). For Co2TiGa the first two Co-Co interactions show the same behavior as TC. The distinct behaviors between functionals for Y=Ti, V can then be ascribed
to the different descriptions of Co-Co interactions. Table 3 summarizes present experimental knowledge of dTC/dP in our alloys,
including Co2TiGa and Co2VGa, along with our results. The sign is correct,
but only if we consider the LDA calculations, as the GGA calculations give the
opposite sign. Even considering the LDA results, the calculated values are ∼3
and ∼5 times smaller for Co2TiGa and Co2VGa, respectively. The main reason for this disagreement is unclear at present. Alling et al. Table 2:
Two largest pairwise exchange interactions (meV) for the ternary compounds
Co2Y Ga (Y =Ti, V, Cr, Mn, Fe), calculated with LDA and GGA (in parenthesis). J1 is
JCo−Y for Y =Cr, Mn, Fe, and JCo−Co for Y =Ti, V. ,
,
,
,
Compound
J1
J2
Co−Co
Co2TiGa
2.56 (4.59)
1.26 (2.35)
Co2VGa
6.74 (8.29)
1.54 (2.17)
Co2CrGa
4.55 (5.88)
3.72 (4.35)
Co2MnGa
10.62 (11.50)
1.49 (1.92)
Co2FeGa
21.75 (22.62)
2.30 (2.56) Table 3: Comparison of the measured and calculated (LDA and GGA) dTC/dP (K/GPa) in
Co2VGa and Co2TiGa. 3.1. Ternary compounds: exchange-correlation approximation and comparison
with experiments Compound
Measured
LDA
GGA
Co2TiGa
−13 [14]; −12.7 [40]
−4.0
+0.4
Co2VGa
−7.8 [17]
−1.5
+1.2 Table 3: Comparison of the measured and calculated (LDA and GGA) dTC/dP (K/GPa) in
Co2VGa and Co2TiGa. Table 3: Comparison of the measured and calculated (LDA and GGA) dTC/dP (K/GPa) in
Co2VGa and Co2TiGa. have suggested that to improve TC estimates it should be necessary to consider
also finite-temperature effects, such as magnetic excitations and chemical dis-
order [45]. In that work, it was also shown that the use of the disordered local
moment state as the reference for the calculation of exchange interactions may
bring some change in the dependence of TC with volume. For Ni2MnSn [8],
the calculated value 3.22 was also smaller than the experimental 7.44 (K/GPa),
and it was calculated that intersublattice interchange between Mn and Ni can
strongly increase dTC/dP. A similar antisite disorder may affect the samples
discussed here. Theoretical improvements may also play a role, such as more so-
phisticated Curie temperature calculation approaches (such as the random phase
approximation) and, as we have shown here, different exchange-correlation func-
tionals. We will continue with a detailed study of the fractional alloys using LDA,
due to its better agreement with experiment for dTC/dP. ©2016. This manuscript version is made available under the CC-BY-NC-ND
4.0 license http://creativecommons.org/licenses/by-nc-nd/4.0/ 3.2. Quaternary alloys Fig. 1 shows the optimized lattice parameters for all the compositions stud-
ied. The variation of the lattice parameter with the concentration of the tran-
sition metal is approximately linear. In what follows we will consider the magnetic properties, in order of atomic
number for the six neighbor elements of Cr which we considered in our calcula-
tions. Figs. 2(a) and 2(b) show the magnetic spin moments for small volume vari-
ations (lattice parameter with ±1%) around the theoretical equilibrium values. The approximate linear variations as a function of the valence electron con-
centration are in line with a generalized Slater-Pauling rule (also observed in
many other Co-based full-Heusler compounds [46]) with the slight deviations
for the Ti and Fe rich compositions showing respectively slopes dM/dnel > 1 ©2016. This manuscript version is made available under the CC-BY-NC-ND
4.0 license http://creativecommons.org/licenses/by-nc-nd/4.0/ 0.0
0.2
0.4
0.6
0.8
1.0
5.48
5.52
5.56
5.60
5.64
5.68
5.72
a (Å)
x
Ti
Fe
V
Co
Mn
Ni
Figure 1: Theoretical lattice parameters for different compositions Co2Cr1−xYxGa (Y =Ti,
V, Mn, Fe, Co, Ni). The lines connect the endpoints. (The Mn and Fe points are superposed
in the graph, showing almost the same values.) 0.0
0.2
0.4
0.6
0.8
1.0
5.48
5.52
5.56
5.60
5.64
5.68
5.72
a (Å)
x
Ti
Fe
V
Co
Mn
Ni Figure 1: Theoretical lattice parameters for different compositions Co2Cr1−xYxGa (Y =Ti,
V, Mn, Fe, Co, Ni). The lines connect the endpoints. (The Mn and Fe points are superposed
in the graph, showing almost the same values.) 0.0
0.2
0.4
0.6
0.8
1.0
0
1
2
3
4
5
Mom. (
B
)
x
a-1%
Ti
a0%
V
a+1%
Mn
Fe
(a)
0.0
0.2
0.4
0.6
0.8
1.0
0
1
2
3
4
5
Mom. (
B
)
x
a-1%
Co
a0%
Ni
a+1%
(b)
Figure 2: Magnetic spin moments (µB) for the alloys Co2Cr1−xYxGa (Y = Ti, V, Mn, Fe,
Co Ni), as a function of the concentration x for different calculated lattice parameters around
the equilibrium values. 0.0
0.2
0.4
0.6
0.8
1.0
0
1
2
3
4
5
Mom. (
B
)
x
a-1%
Co
a0%
Ni
a+1%
(b) 0.0
0.2
0.4
0.6
0.8
1.0
0
1
2
3
4
5
Mom. (
B
)
x
a-1%
Ti
a0%
V
a+1%
Mn
Fe
(a) Mom. 3.2. Quaternary alloys (
B
) (a) (b) Figure 2: Magnetic spin moments (µB) for the alloys Co2Cr1−xYxGa (Y = Ti, V, Mn, Fe,
Co Ni), as a function of the concentration x for different calculated lattice parameters around
the equilibrium values. ©2016. This manuscript version is made available under the CC-BY-NC-ND
4.0 license http://creativecommons.org/licenses/by-nc-nd/4.0/ ©2016. This manuscript version is made available under the CC-BY-NC-ND
4.0 license http://creativecommons.org/licenses/by-nc-nd/4.0/ 0.0
0.2
0.4
0.6
0.8
1.0
0
200
400
600
800
1000
1200
1400
T
C
(K)
x
Ti
V
Mn
Fe
Co
Ni
(a)
(b)
Figure 3: a) Curie temperature (K) for the alloys Co2Cr1−xYxGa (Y =Ti, V, Mn, Fe, Co,
Ni) at the equilibrium lattice parameters, as a function of the element concentration. The
open symbols correspond to experimental values. b) corresponding derivative of the Curie
temperature with respect to volume (K ˚A−3). The lines are guides to the eye. (b) 0.0
0.2
0.4
0.6
0.8
1.0
0
200
400
600
800
1000
1200
1400
T
C
(K)
x
Ti
V
Mn
Fe
Co
Ni
(a) (a) (b) Figure 3: a) Curie temperature (K) for the alloys Co2Cr1−xYxGa (Y =Ti, V, Mn, Fe, Co,
Ni) at the equilibrium lattice parameters, as a function of the element concentration. The
open symbols correspond to experimental values. b) corresponding derivative of the Curie
temperature with respect to volume (K ˚A−3). The lines are guides to the eye. and dM/dnel < 1, resulting in moments slightly less than 1 µB and 5 µB re-
spectively. Alloying with Ti strongly reduces the spin moment. Co2TiGa has a
moment close to 0.8 µB at the theoretical lattice parameter, and we have fur-
ther calculated that with smaller lattice parameters the magnetic moment falls
quickly (not shown) and at a critical pressure of about 30 GPa it vanishes. For compounds close to the Cr and V rich compositions the magnetic mo-
ment is insensitive (changing less than 0.01 µB) to the volume changes consid-
ered (±1% in the lattice parameter). It is also almost insensitive (changing
0.02 µB) for Mn rich compositions. For Fe rich compositions it becomes slightly
sensitive (0.07 µB). For high concentrations of Ti and even at low concentrations
of the other transition metals (Co, Ni) the spin moments become significantly
(more than 0.1 µB) dependent on volume. ©2016. This manuscript version is made available under the CC-BY-NC-ND
4.0 license http://creativecommons.org/licenses/by-nc-nd/4.0/ 3.2. Quaternary alloys The spin moment with full Co sub-
stitution is close to 4 µB, consistent with the decrease of moment with electrons
after Fe (≈5µB), following the rule M = 34 −Z for itinerant metals [47]. Co2NiGa follows the same decreasing trend, but reaches a lower value than
expected, closer to 2µB than 3 µB. In all cases, at a given composition the dependence is such that higher vol-
umes always lead to higher moments. The variation of spin moments with lattice
parameters is usually negligible when compared with the changes that can be
achieved with changing composition. However, this is not so in the compositions
close to Co2TiGa, where a relatively small pressure can decrease the moments
significantly, and for compositions close to the Co3Ga and Co2NiGa cases, where
a change in the lattice parameter of 1% can produce moment changes of 0.15 µB,
comparable to a change of x = 0.3 in composition. The calculated Curie temperatures at the equilibrium lattice parameters are
shown in Fig. 3(a). dTC/dV is shown in Fig. 3(b), also for the equilibrium lattice
parameters. For the Ti case it shows negative values on the Cr rich side and positive values on the Ti rich side. At xTi ≈0.6 (where TC is close to 300 K),
dTC/dV is very small. With V substitution TC also falls, and similarly dTC/dV
also changes sign, in this case at V rich compositions (dTC/dV ≈0 at x ≈0.8). With V doping, it is possible to get TC close to room temperature (∼350 K)
at high vanadium concentrations, and with compression it is possible to further
reduce TC. With Mn doping dTC/dV becomes more negative. In spite of the
great range of Curie temperatures for Fe alloying, the respective change with
volume is the most constant in all the substitutions. With Co, TC increases
from 450 to 900 K. Just as with Ti and V, dTC/dV changes sign from negative
to positive, and at xCo ≈0.5, dTC/dV ≈0. With Ni TC changes little, while
dTC/dV changes from negative to positive with increasing Ni concentration,
reaching the highest values for Ni-rich compositions. TC varies monotonically with concentration x of the elements, with no
anomalies between the extremes, except where the minimum TC is achieved
slightly away from full V substitution, and a local TC maximum for low Ni con-
centrations. 3.2. Quaternary alloys dTC/dV usually also presents extremes at the endpoints, except
for Fe and Ni substitutions. In the case of Fe the derivative is not monotonic,
nevertheless varying little, in the interval [-4.4,-3.2] K ˚A−3. For Ni the maximum
and minimum derivatives are reached slightly away from the endpoints. For the cases of Fe and Mn doping there are previous experimental measure-
ments and calculations, showing a progressive and large increase of the Curie
temperature [33, 34], consistent with our results, although there are differences
in the values and curvature of the TC(x) dependence. For the end compounds
there are also measurements of TC [33, 40, 41, 44], and we find very good agree-
ment with our results, except for Co2FeGa, where the calculated value is higher. The suppression of magnetism with pressure in the (Ti,V) compounds is
concomitant with a high positive dTC/dV , with the largest value achieved for
Co2TiGa, which is nevertheless smaller in magnitude than the opposite sign
value −7 achieved in Co2MnGa. However, the (Ti,V) compounds present values
of TC closer to room temperature, which may be useful for devices, with appli-
cations in room temperature refrigeration or spintronics, for example. Finally,
we also predict that Co2Cr1−xNixGa alloys present high values of dTC/dV ≈6
near the rich Ni side, if these composition are synthesized in the L21 structure,
while Co rich compositions reach values close to Co2TiGa. ©2016. This manuscript version is made available under the CC-BY-NC-ND
4.0 license http://creativecommons.org/licenses/by-nc-nd/4.0/ ©2016. This manuscript version is made available under the CC-BY-NC-ND
4.0 license http://creativecommons.org/licenses/by-nc-nd/4.0/ 4. Conclusions We studied magnetic and magnetovolume coupling properties of fractional
Co2Cr1−xYxGa alloys (Y =Ti, V, Mn, Fe, Co, and Ni). The equilibrium lattice
parameters follow approximate linear variations with changing element concen-
trations, as well as the magnetic moment (in line with the generalized Slater-
Pauling rule, due to the addition/subtraction of valence electrons) except for
Y =Co and Ni. Ti, V, Co and Ni substitutions change the sign of dTC/dV with
respect to the parent Co2CrGa compound. However, at least for the Ti and V
cases the values are dependent of the exchange-correlation approximation, and the measured sign is only found with the LDA, which is related to the different
descriptions of Co-Co exchange interactions. The wide TC range of the studied compositions (100 −1300 K) and range
of magnetovolume couplings (dTC/dV from −7 to +6 V ˚A−3) establishes the
rich behavior of this compound family. Optimization with other elements or
different substitutions, disorder and volume optimization may be performed to
find alloys with optimized TC and dTC/dV values for specific applications (for
example, the magnetocaloric effect benefits from a high |dTC/dV | [48]). Further
experimental work would be valuable to compare with theory and study the
effects the various approximations on dTC/dV , which was shown to be very
sensitive to the theoretical approach employed. Acknowledgments This work was supported by the project RECI/CTM-CER/0336/2012 co-
financed by FEDER, QREN reference COMPETE: FCOMP-01-0124-FEDER-
027465 and was developed within the scope of the project CICECO-Aveiro Insti-
tute of Materials, POCI-01-0145-FEDER-007679 (FCT Ref. UID/CTM/50011/2013),
financed by national funds through the FCT/MEC and co-financed by FEDER
under the PT2020 Partnership Agreement. We acknowledge FCT grants SFRH/BPD/111270/2015
(J. S. Amaral) and SFRH/BPD/82059/2011 (J. N. Gon¸calves). [1] T. Graf, C. Felser, S. S. P. Parkin, Simple rules for the understanding of
Heusler compounds, Progress in Solid State Chemistry 39 (1) (2011) 1–50. doi:10.1016/j.progsolidstchem.2011.02.001. URL
http://www.sciencedirect.com/science/article/pii/
S0079678611000021 [2] P. Entel, M. E. Gruner, A. Hucht, A. Dannenberg, M. Siewert, H. C. Her-
per, T. Kakeshita, T. Fukuda, V. V. Sokolovskiy, V. D. Buchelnikov, Phase
Diagrams of Conventional and Inverse Functional Magnetic Heusler Al-
loys: New Theoretical and Experimental Investigations, in: T. Kakeshita,
T. Fukuda, A. Saxena, A. Planes (Eds.), Disorder and Strain-Induced
Complexity in Functional Materials, Springer Series in Materials Science,
Springer Berlin Heidelberg, 2012, pp. 19–47. [3] C. Felser, G. H. Fecher (Eds.), Spintronics, Springer Netherlands, Dor-
drecht, 2013. URL http://link.springer.com/10.1007/978-90-481-3832-6 [4] A. Jakobsson, P. Mavropoulos, E. S¸a¸sıo˘glu, S. Bl¨ugel, M. Leˇzai´c, B. Sanyal,
I. Galanakis, First-principles calculations of exchange interactions, spin
waves, and temperature dependence of magnetization in inverse-Heusler-
based spin gapless semiconductors, Physical Review B 91 (17) (2015)
174439. doi:10.1103/PhysRevB.91.174439. URL http://link.aps.org/doi/10.1103/PhysRevB.91.174439 [5] E. S¸a¸sıo˘glu, L. M. Sandratskii, P. Bruno, First-principles calculation of the
intersublattice exchange interactions and Curie temperatures of the full
Heusler alloys Ni2MnX (X=Ga,In,Sn,Sb), Physical Review B 70 (2) (2004)
024427. doi:10.1103/PhysRevB.70.024427. URL http://link.aps.org/doi/10.1103/PhysRevB.70.024427 [6] V. V. Sokolovskiy, P. Entel, V. D. Buchelnikov, M. E. Gruner, Achieving
large magnetocaloric effects in Co- and Cr-substituted Heusler alloys: Pre-
dictions from first-principles and Monte Carlo studies, Physical Review B
91 (22) (2015) 220409. doi:10.1103/PhysRevB.91.220409. URL http://link.aps.org/doi/10.1103/PhysRevB.91.220409 [7] V. D. Buchelnikov, V. V. Sokolovskiy, H. C. Herper, H. Ebert, M. E. Gruner, S. V. Taskaev, V. V. Khovaylo, A. Hucht, A. Dannenberg,
M. Ogura, H. Akai, M. Acet, P. Entel, First-principles and Monte Carlo
study of magnetostructural transition and magnetocaloric properties of
Ni2+xMn1−xGa, Physical Review B 81 (9) (2010) 094411. doi:10.1103/
PhysRevB.81.094411. URL http://link.aps.org/doi/10.1103/PhysRevB.81.094411 [8] E. S¸a¸sıo˘glu, L. M. Sandratskii, P. Bruno, Pressure dependence of the Curie
temperature in Ni2MnSn Heusler alloy: A first-principles study, Phys. Rev. B 71 (2005) 214412. doi:10.1103/PhysRevB.71.214412. URL http://link.aps.org/doi/10.1103/PhysRevB.71.214412 [9] M. Meinert, J.-M. Schmalhorst, G. ©2016. This manuscript version is made available under the CC-BY-NC-ND
4.0 license http://creativecommons.org/licenses/by-nc-nd/4.0/ Acknowledgments Reiss, Exchange interactions and Curie
temperatures of Mn2CoZ compounds, Journal of Physics: Condensed Mat-
ter 23 (11) (2011) 116005. doi:10.1088/0953-8984/23/11/116005. URL http://iopscience.iop.org/0953-8984/23/11/116005 [10] S. K. Bose, J. Kudrnovsk´y, V. Drchal, I. Turek, Pressure dependence of
Curie temperature and resistivity in complex Heusler alloys, Phys. Rev. B
84 (2011) 174422. doi:10.1103/PhysRevB.84.174422. URL http://link.aps.org/doi/10.1103/PhysRevB.84.174422 [11] T. Kaneko, H. Yoshida, S. Abe, K. Kamigaki, Pressure effect on the Curie
point of the Heusler alloys Ni2MnSn and Ni2MnSb, Journal of Applied
Physics 52 (3) (1981) 2046–2048. doi:10.1063/1.329610. URL http://scitation.aip.org/content/aip/journal/jap/52/3/10. 1063/1.329610 [12] T. Kanomata, K. Shirakawa, T. Kaneko, Effect of hydrostatic pressure
on the Curie temperature of the Heusler alloys Ni2MnZ(Z = Al, Ga, In,
Sn and Sb), Journal of Magnetism and Magnetic Materials 65 (1) (1987)
76–82. doi:10.1016/0304-8853(87)90312-X. URL
http://www.sciencedirect.com/science/article/pii/
030488538790312X [13] K. Shirakawa, T. Kanomata, T. Kaneko, Effect of hydrostatic pressure on
the Curie temperature of the Heusler alloys Au2MnAl and Pd2MnZ (Z =
Sn and Sb), Journal of Magnetism and Magnetic Materials 70 (13) (1987)
421 – 422. doi:http://dx.doi.org/10.1016/0304-8853(87)90489-6. URL
http://www.sciencedirect.com/science/article/pii/
0304885387904896 [14] T. Kanomata, K. Shirakawa, T. Kaneko, Effect of hydrostatic pressure
on the Curie temperature of the Heusler alloys Co2TiAl and Co2TiGa,
Le Journal de Physique Colloques 49 (C8) (1988) C8–143–C8–144. doi:
10.1051/jphyscol:1988858. URL http://www.edpsciences.org/10.1051/jphyscol:1988858 [15] E. DiMasi, M. C. Aronson, B. R. Coles, Pressure dependence of the Curie
temperature of Co2TiAl, Physical Review B 47 (21) (1993) 14301–14305. doi:10.1103/PhysRevB.47.14301. URL http://link.aps.org/doi/10.1103/PhysRevB.47.14301 [16] T. Kanomata, T. Sasaki, H. Nishihara, H. Yoshida, T. Kaneko, S. Hane,
T. Goto, N. Takeishi, S. Ishida, Magnetic properties of ferromagnetic
Heusler alloy Co2ZrAl, Journal of Alloys and Compounds 393 (12) (2005)
26 – 33. doi:http://dx.doi.org/10.1016/j.jallcom.2004.09.071. URL
http://www.sciencedirect.com/science/article/pii/
S0925838804013350 [17] T. Kanomata, Y. Chieda, K. Endo, H. Okada, M. Nagasako, K. Kobayashi,
R. Kainuma, R. Y. Umetsu, H. Takahashi, Y. Furutani, H. Nishihara,
K. Abe, Y. Miura, M. Shirai, Magnetic properties of the half-metallic
Heusler alloys Co2VAl and Co2VGa under pressure, Physical Review B
82 (14) (2010) 144415. doi:10.1103/PhysRevB.82.144415. URL http://link.aps.org/doi/10.1103/PhysRevB.82.144415 [18] Y. Kurtulus, R. Dronskowski, G. D. Samolyuk, V. P. Antropov, Electronic
structure and magnetic exchange coupling in ferromagnetic full Heusler
alloys, Physical Review B 71 (1) (2005) 014425. doi:10.1103/PhysRevB. 71.014425. URL http://link.aps.org/doi/10.1103/PhysRevB.71.014425 [19] J. N. Gon¸calves, J. S. Amaral, V. S. Amaral, Magnetovolume Effects in
Heusler Compounds via First-Principles Calculations, IEEE Transactions
on Magnetics 50 (11) (2014) 1301104. doi:10.1109/TMAG.2014.2326892. [20] K. Seema, R. ©2016. This manuscript version is made available under the CC-BY-NC-ND
4.0 license http://creativecommons.org/licenses/by-nc-nd/4.0/ ©2016. This manuscript version is made available under the CC-BY-NC-ND
4.0 license http://creativecommons.org/licenses/by-nc-nd/4.0/ Acknowledgments Kumar, Electronic structure and magnetic properties of qua-
ternary Heusler alloy Co2CrGa1−xGex (x=0-1), Journal of Magnetism and
Magnetic Materials 377 (2015) 70–76. doi:10.1016/j.jmmm.2014.10.076. URL
http://www.sciencedirect.com/science/article/pii/
S0304885314009883 [21] J. K¨ubler, G. H. Fecher, C. Felser, Understanding the trend in the Curie
temperatures of Co2-based Heusler compounds:
Ab initio calculations,
Phys. Rev. B 76 (2007) 024414. doi:10.1103/PhysRevB.76.024414. URL http://link.aps.org/doi/10.1103/PhysRevB.76.024414 [22] M. Zhang, E. Br¨uck, F. R. d. Boer, Z. Li, G. Wu, The magnetic and
transport properties of the Co2FeGa Heusler alloy, Journal of Physics D:
Applied Physics 37 (15) (2004) 2049. doi:10.1088/0022-3727/37/15/
001. URL http://stacks.iop.org/0022-3727/37/i=15/a=001 [23] J. Thoene, S. Chadov, G. Fecher, C. Felser, J. K¨ubler, Exchange energies,
Curie temperatures and magnons in Heusler compounds, Journal of Physics
D: Applied Physics 42 (8) (2009) 084013. doi:10.1088/0022-3727/42/8/
084013. URL http://iopscience.iop.org/0022-3727/42/8/084013 [24] F. K¨ormann, A. Dick, T. Hickel, J. Neugebauer, Pressure dependence of
the Curie temperature in bcc iron studied by ab initio simulations, Phys. Rev. B 79 (2009) 184406. doi:10.1103/PhysRevB.79.184406. URL http://link.aps.org/doi/10.1103/PhysRevB.79.184406 [25] S. Mor´an, C. Ederer, M. F¨ahnle, Ab initio electron theory for magnetism
in Fe: Pressure dependence of spin-wave energies, exchange parameters,
and Curie temperature, Phys. Rev. B 67 (2003) 012407. doi:10.1103/
PhysRevB.67.012407. URL http://link.aps.org/doi/10.1103/PhysRevB.67.012407 [26] I. G. Austin, P. K. Mishra, The effect of pressure on the Curie temperature
and resistivity of some rare-earth metals and Heusler alloys, Philosophical
Magazine 15 (135) (1967) 529–538. doi:10.1080/14786436708220900. URL http://dx.doi.org/10.1080/14786436708220900 [27] X. Dai, G. Liu, Y. Li, J. Qu, J. Li, J. Chen, G. Wu, Structure and
magnetic properties of highly ordered Co2NiGa alloys, Journal of Applied
Physics 101 (9) (2007) 09N503. doi:10.1063/1.2709417. URL
http://scitation.aip.org/content/aip/journal/jap/101/9/
10.1063/1.2709417 [28] R. Arr´oyave, A. Junkaew, A. Chivukula, S. Bajaj, C. Y. Yao, A. Garay,
Investigation of the structural stability of Co2NiGa shape memory al-
loys via ab initio methods, Acta Materialia 58 (16) (2010) 5220–5231. doi:10.1016/j.actamat.2010.05.033. URL
http://www.sciencedirect.com/science/article/pii/
S1359645410003186 [29] H. Ebert, et al., The Munich SPR-KKR package, version 6.3. URL http://ebert.cup.uni-muenchen.de/SPRKKR [30] H. Ebert, D. K¨odderitzsch, J. Min´ar, Calculating condensed matter prop-
erties using the KKR-Green’s function method-recent developments and
applications, Reports on Progress in Physics 74 (9) (2011) 096501. doi:
10.1088/0034-4885/74/9/096501. URL http://iopscience.iop.org/0034-4885/74/9/096501 [31] J. P. Perdew, K. Burke, M. Ernzerhof, Generalized Gradient Approxima-
tion Made Simple, Phys. Rev. Lett. 77 (1996) 3865–3868. doi:10.1103/
PhysRevLett.77.3865. URL http://link.aps.org/doi/10.1103/PhysRevLett.77.3865 [32] S. H. Vosko, L. Wilk, M. ©2016. This manuscript version is made available under the CC-BY-NC-ND
4.0 license http://creativecommons.org/licenses/by-nc-nd/4.0/ Acknowledgments Nusair, Accurate spin-dependent electron liquid
correlation energies for local spin density calculations: a critical analy-
sis, Canadian Journal of Physics 58 (8) (1980) 1200–1211. doi:10.1139/
p80-159. URL http://www.nrcresearchpress.com/doi/abs/10.1139/p80-159 [33] R. Y. Umetsu, K. Kobayashi, A. Fujita, K. Oikawa, R. Kainuma,
K. Ishida, N. Endo, K. Fukamichi, A. Sakuma, Half-metallic properties of
Co2Cr1−xFexGa Heusler alloys, Physical Review B 72 (21) (2005) 214412. doi:10.1103/PhysRevB.72.214412. URL http://link.aps.org/doi/10.1103/PhysRevB.72.214412 [34] R. Y. Umetsu, Magnetic properties and phase stability of Half-metal-type
ferromagnetic Co-based Heusler alloys of Co2(Cr1−xMnx)Ga, Journal of
The Society of Japanese Women Scientists 15 (1) (2015) 1–8. doi:10. 5939/sjws.15001. [35] P. Soven, Coherent-Potential Model of Substitutional Disordered Alloys,
Physical Review 156 (3) (1967) 809–813. doi:10.1103/PhysRev.156.809. URL http://link.aps.org/doi/10.1103/PhysRev.156.809 [36] K. ¨Ozdoˇgan,
E. S¸a¸sıoglu,
I. Galanakis,
Engineering the electronic,
magnetic, and gap-related properties of the quinternary half-metallic
Heusler alloys,
Journal of Applied Physics 103 (2) (2008) 023503. doi:10.1063/1.2831224. URL
http://scitation.aip.org/content/aip/journal/jap/103/2/
10.1063/1.2831224 [37] F. D. Murnaghan, The Compressibility of Media under Extreme Pressures,
Proceedings of the National Academy of Sciences of the United States of
America 30 (9) (1944) 244–247. URL http://www.ncbi.nlm.nih.gov/pmc/articles/PMC1078704/ [38] A. I. Liechtenstein, M. I. Katsnelson, V. P. Antropov, V. A. Gubanov, Lo-
cal spin density functional approach to the theory of exchange interactions
in ferromagnetic metals and alloys, Journal of Magnetism and Magnetic
Materials 67 (1) (1987) 65–74. doi:10.1016/0304-8853(87)90721-9. URL
http://www.sciencedirect.com/science/article/pii/
0304885387907219 [39] P. W. Anderson, Theory of magnetic exchange interactions: Exchange in
insulators and semiconductors, Solid State Physics 14 (1963) 99–214. [40] T. Sasaki,
T. Kanomata,
T. Narita,
H. Nishihara,
R. Note,
H. Yoshida,
T. Kaneko,
Magnetic
properties
of
Co2TiGa
com-
pound,
Journal of Alloys and Compounds 317318 (2001) 406–410. doi:10.1016/S0925-8388(00)01417-1. URL
http://www.sciencedirect.com/science/article/pii/
S0925838800014171 [41] P. J. Webster, Magnetic and chemical order in Heusler alloys containing
cobalt and manganese, Journal of Physics and Chemistry of Solids 32 (6)
(1971) 1221–1231. doi:10.1016/S0022-3697(71)80180-4. URL
http://www.sciencedirect.com/science/article/pii/
S0022369771801804 [42] K. H. J. Buschow, P. G. van Engen, Magnetic and magneto-optical
properties of heusler alloys based on aluminium and gallium, Jour-
nal
of
Magnetism
and
Magnetic
Materials
25
(1)
(1981)
90–96. doi:10.1016/0304-8853(81)90151-7. URL
http://www.sciencedirect.com/science/article/pii/
0304885381901517 [43] R. Y. Umetsu, K. Kobayashi, R. Kainuma, A. Fujita, K. Fukamichi,
K. Ishida, A. Sakuma, Magnetic properties and band structures of
half-metal-type Co2CrGa Heusler alloy, Applied Physics Letters 85 (11)
(2004) 2011–2013. doi:10.1063/1.1790029. URL
http://scitation.aip.org/content/aip/journal/apl/85/11/
10.1063/1.1790029 [44] R. Y. Umetsu,
K. Kobayashi,
A. Fujita,
R. Kainuma,
K. Ishida,
K. Fukamichi, A. ©2016. This manuscript version is made available under the CC-BY-NC-ND
4.0 license http://creativecommons.org/licenses/by-nc-nd/4.0/ ©2016. This manuscript version is made available under the CC-BY-NC-ND
4.0 license http://creativecommons.org/licenses/by-nc-nd/4.0/ B 66 (17) (2002) 174429. doi:10.1103/PhysRevB.66.174429.
URL http://link.aps.org/doi/10.1103/PhysRevB.66.174429 Acknowledgments Sakuma, Magnetic properties, phase stability, electronic
structure, and half-metallicity of L21-type Co2V1−xMnxGa Heusler alloys,
Physical Review B 77 (10) (2008) 104422. doi:10.1103/PhysRevB.77. 104422. URL http://link.aps.org/doi/10.1103/PhysRevB.77.104422 [45] B. Alling, A. V. Ruban, I. A. Abrikosov, Effect of thermal expan-
sion, electronic excitations, and disorder on the Curie temperature of
Ni1−xCuxMnSb alloys, Physical Review B 79 (13) (2009) 134417. doi:
10.1103/PhysRevB.79.134417. URL http://link.aps.org/doi/10.1103/PhysRevB.79.134417 [46] I. Galanakis, P. H. Dederichs, N. Papanikolaou, Slater-Pauling behavior
and origin of the half-metallicity of the full-Heusler alloys, Physical Review ©2016. This manuscript version is made available under the CC-BY-NC-ND
4.0 license http://creativecommons.org/licenses/by-nc-nd/4.0/ B 66 (17) (2002) 174429. doi:10.1103/PhysRevB.66.174429. URL http://link.aps.org/doi/10.1103/PhysRevB.66.174429 B 66 (17) (2002) 174429. doi:10.1103/PhysRevB.66.174429. URL http://link.aps.org/doi/10.1103/PhysRevB.66.174429 [47] G. H. Fecher, H. C. Kandpal, S. Wurmehl, C. Felser, G. Sch¨onhense, Slater-
Pauling rule and Curie temperature of Co2-based Heusler compounds,
Journal of Applied Physics 99 (8) (2006) 08J106. doi:10.1063/1.2167629. URL http://scitation.aip.org/content/aip/journal/jap/99/8/10. 1063/1.2167629 [48] J. S. Amaral, J. N. Gon¸calves, V. S. Amaral, Thermodynamics of the 2-D
Ising Model From a Random Path Sampling Method, IEEE Transactions
on Magnetics 50 (11) (2014) 1002204. doi:10.1109/TMAG.2014.2326713. ©2016. This manuscript version is made available under the CC-BY-NC-ND
4.0 license http://creativecommons.org/licenses/by-nc-nd/4.0/
|
https://openalex.org/W2132577746
|
https://hal-polytechnique.archives-ouvertes.fr/hal-00488197/file/1472-6807-7-79.pdf
|
English
| null |
Neutral evolution of protein-protein interactions: a computational study using simple models
|
BMC structural biology
| 2,007
|
cc-by
| 14,311
|
To cite this version: Josselin Noirel, Thomas Simonson. Neutral evolution of protein-protein interactions: a computational
study using simple models.. BMC Structural Biology, 2007, 7, pp.79-93. 10.1186/1472-6807-7-79. hal-00488197 HAL Id: hal-00488197
https://polytechnique.hal.science/hal-00488197v1
Submitted on 13 Feb 2013 L’archive ouverte pluridisciplinaire HAL, est
destinée au dépôt et à la diffusion de documents
scientifiques de niveau recherche, publiés ou non,
émanant des établissements d’enseignement et de
recherche français ou étrangers, des laboratoires
publics ou privés. HAL is a multi-disciplinary open access
archive for the deposit and dissemination of sci-
entific research documents, whether they are pub-
lished or not. The documents may come from
teaching and research institutions in France or
abroad, or from public or private research centers. BioMed Central Open Acc
Research article
Neutral evolution of Protein-protein interactions: a computational
study using simple models
Josselin Noirel and Thomas Simonson* Open Access Address: Laboratoire de Biochimie, École polytechnique, route de Saclay, 91128 Palaiseau Cedex, France
Email: Josselin Noirel - j.noirel@sheffield.ac.uk; Thomas Simonson* - thomas.simonson@polytechnique.fr
* Corresponding author Received: 23 April 2007
Accepted: 19 November 2007 © 2007 Noirel and Simonson; licensee BioMed Central Ltd. This is an Open Access article distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/2.0),
which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited. ;
This is an Open Access article distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/2.0),
which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited. Abstract Background: Protein-protein interactions are central to cellular organization, and must have
appeared at an early stage of evolution. To understand better their role, we consider a simple
model of protein evolution and determine the effect of an explicit selection for Protein-protein
interactions. Results: In the model, viable sequences all have the same fitness, following the neutral evolution
theory. A very simple, two-dimensional lattice representation of the protein structures is used, and
the model only considers two kinds of amino acids: hydrophobic and polar. With these
approximations, exact calculations are performed. The results do not depend too strongly on these
assumptions, since a model using a 3D, off-lattice representation of the proteins gives results in
qualitative agreement with the 2D one. With both models, the evolutionary dynamics lead to a
steady state population that is enriched in sequences that dimerize with a high affinity, well beyond
the minimal level needed to survive. Correspondingly, sequences close to the viability threshold
are less abundant in the steady state, being subject to a larger proportion of lethal mutations. The
set of viable sequences has a "funnel" shape, consistent with earlier studies: sequences that are
highly populated in the steady state are "close" to each other (with proximity being measured by
the number of amino acids that differ). Conclusion: This bias in the the steady state sequences should lead to an increased resistance of
the population to environmental change and an increased ability to evolve. tral if it leads to an organism that is viable, but does not
increase or decrease the fitness [1,2]. A mutation in non-
coding DNA will often be neutral. A mutation in a protein
coding sequence can also be neutral if it does not signifi-
cantly affect the structure, stability, or biochemical func-
tion of the protein. http://www.biomedcentral.com/1472-6807/7/79 BMC Structural Biology 2007, 7:79 molecular sequences and structures be explored together. This suggests that simple models should be used, so that
a precise mapping can be defined between genotype, phe-
notype, and fitness [3]. An important example is the use
of lattice models to represent protein structures. In these
models, the polypeptide is treated as a chain of beads,
with one bead per amino acid. The allowed conforma-
tions are defined by a simple two- or three-dimensional
lattice, with the beads occupying nodes of the lattice. Lat-
tice models first revealed, for example, that for a typical
small protein, only a few sequences fold rapidly into a
well-defined, stable structure. These sequences tend to
adopt structures that have a high stability and are espe-
cially "robust" with respect to point mutations: many
amino acids can be mutated without disrupting the folded
structure. The corresponding folded conformation is said
to be highly "designable", since it is shared by many dif-
ferent sequences (corresponding to all the allowed point
mutations) [4-6]. molecular sequences and structures be explored together. This suggests that simple models should be used, so that
a precise mapping can be defined between genotype, phe-
notype, and fitness [3]. An important example is the use
of lattice models to represent protein structures. In these
models, the polypeptide is treated as a chain of beads,
with one bead per amino acid. The allowed conforma-
tions are defined by a simple two- or three-dimensional
lattice, with the beads occupying nodes of the lattice. Lat-
tice models first revealed, for example, that for a typical
small protein, only a few sequences fold rapidly into a
well-defined, stable structure. These sequences tend to
adopt structures that have a high stability and are espe-
cially "robust" with respect to point mutations: many
amino acids can be mutated without disrupting the folded
structure. The corresponding folded conformation is said
to be highly "designable", since it is shared by many dif-
ferent sequences (corresponding to all the allowed point
mutations) [4-6]. because they act as a basin of attraction for the population
dynamics: in the steady state, population accumulates in
these basins. In contrast, sequences that are not very
robust to mutations lie mostly outside these regions,
forming the outer "edge" of the neutral network. Sequence funneling was recently observed experimentally
by directed evolution [14]. http://www.biomedcentral.com/1472-6807/7/79 Because protein functionality is very complex, evolution-
ary models usually assume that protein structure can be
used as a proxy for function: proteins that adopt the cor-
rect structure are assumed viable [10,15]. More recently,
explicit models of functionality have been introduced,
involving the ability of the protein to bind a small ligand
[16-19]. While these models have been very useful, it is increas-
ingly clear that most proteins must interact with other
proteins to function, and co-evolve with them [20,21]. The set of Protein-protein interactions has been studied
exhaustively for several organisms, and some of its topo-
logical properties established [22,23]. Its complexity is
thought to correlate with the overall complexity of an
organism. In recent years, both on-lattice and off-lattice models of
protein structure have been employed in evolutionary
models [7-10]. The simplest models focus on a particular
protein, and allow evolution only through random point
mutations. Such models usually define a set of viable
sequences, immersed in a "sea" of non-viable sequences. An individual that undergoes a mutation to a non-viable
sequence dies. The set of viable sequences can be thought
of as a cluster within the larger space of all possible
sequences (for the protein of interest). The set of viable
sequences is often viewed as a graph, or network, where
the sequences are the nodes, and a point mutation
between two viable sequences defines an edge connecting
the sequences. In an evolving population, the population
dynamics can be viewed as a set of individuals randomly
diffusing over the graph [1,11,12]. Here, we extend previous evolutionary models to take
into account explicitly the essential role of Protein-protein
interactions. We model the neutral evolution of two pro-
teins, coupled by a selection criterion that requires the for-
mation of a specific Protein-protein interaction (with a
specific, predefined interaction mode). We require only a
transient interaction, present around 10–20% of the time. This is meant to mimic the behavior of proteins involved
in information transfer and signalling, rather than pro-
teins involved in long-lived, multi-protein complexes. We
only consider neutral evolution through point mutations. This mechanism, though simple, is nevertheless impor-
tant for the evolution of individual protein domains. More complex events like recombination, essential for the
creation or rearrangement of entire domains in higher
organisms, are neglected here. Background g
Modern genomics and molecular biology have trans-
formed our understanding of molecular evolution. The
diversity of modern proteins is illustrated by the millions
of known gene sequences and thousands of known pro-
tein structures. It has become clear that proteins are
remarkably robust with respect to mutations, retaining
structure and function in many cases. This has helped
renew interest in theories of evolution that explore the
role of "neutral" mutations. A mutation is selectively neu- In recent years, computer models have proved useful to
explore some of the principles of molecular evolution. To
model evolution at the molecular level requires that Page 1 of 15
(page number not for citation purposes) Page 1 of 15
(page number not for citation purposes) http://www.biomedcentral.com/1472-6807/7/79 http://www.biomedcentral.com/1472-6807/7/79 BMC Structural Biology 2007, 7:79 proteins is considered. The two proteins must not only
fold, but specifically associate to perform a vital function. Thus, both the stability of the individual monomers and
that of the dimer are subjected to negative selection. In
contrast to most earlier protein-ligand studies, the two
proteins are both allowed to evolve. Using the 2D lattice
description of the structures, the chemical equations for
dimerization can be solved exactly for any particular
sequence pair. Pairs that have a sufficient dimerization
ability are viable. The viable sequences can thus be enu-
merated and the evolutionary dynamics characterized. Viable sequences, which ensure folding of the two part-
ners along with a sufficient degree of dimerization, are all
assumed to be equally fit. We refer to this as a neutral evo-
lution model, in the spirit of several models studied by
Kimura [1,12]. Under conditions of moderate selection,
where only weak dimerization is required, we find that
neutral evolution increases the functional effectiveness of
the proteins considered: the steady state population is
enriched in sequences coding for proteins that readily
dimerize. Using a more realistic, 3D, off-lattice descrip-
tion, a similar effect is observed. This result is analogous
to the result described above for individual proteins: the
(monomeric) steady state was enriched in mutationally
robust sequences. In both the monomeric and dimeric
cases, sequences in the core of the neutral network are
overpopulated, while sequences at the edge are rare. Depletion of sequences at the edge leads to a reduced
mutational load [1]. In practice, it has the same effect as a
positive adaptation: an enhanced functional ability. The
enhancement emerges from a neutral model that requires
only a minimal ability to function, through the funneled
shape of the network of viable sequences. complexes are formed by associating two square mono-
mers side by side. A pair of such proteins can adopt well
over one billion possible sequence pairs (40,0002). For
these pairs, the selection stringency is characterized by the
fractional population δ required for the functional dimer
AB. A value of δ = 0.1, for example, means that at chemical
equilibrium, the dimer must be present at least 10% of the
time. For reasons of computational cost, the analysis is
limited to 16 2D dimers. They all involve monomeric
neutral networks of about 10,000 sequences. http://www.biomedcentral.com/1472-6807/7/79 The largest
monomeric neutral networks (40,000–68,000 sequences)
are too large to allow complete dimerization studies. Fig. 1 shows the effect of the selection criterion on a typi-
cal dimer. For δ = 0, all the pairs of sequences formed from
the viable monomeric sequences of A and B are viable. As
δ increases, sequences that dimerize poorly are increas-
ingly eliminated, and the number of viable sequence pairs
decreases rapidly. This decrease is accompanied by a frag-
mentation of the dimer's neutral network into smaller,
disconnected pieces, as shown in Fig. 2. Interestingly,
there is always one very large connected component,
along with a number of much smaller components. The
existence of a single large component implies that many
sequences can be explored even though only point muta-
tions are allowed. The sequences eliminated by selection are those with too
few hydrophobic residues at the functional interface. This
follows from our energy function (Eq. 3), where hydro-
phobic-hydrophobic interactions are the most favorable. Indeed, Fig. 1 shows that the average sequence, weighted
by the steady state population, has an interface that is
increasingly hydrophobic (darker) as δ increases (e.g. Fig. 1A,B,C). The neutral network for the pair is increasingly
depleted. This is seen by the decreasing number of red
dots going from left to right in Fig. 1D,E,F. Despite this
depletion, the viable sequences of A and B remain very
diverse: the red dots are not grouped in one part of their
respective neutral networks, but are widely distributed
throughout the network. Sequence diversity and the pressure to dimerize Sequence diversity and the pressure to dimerize Two structural models of a protein were considered in this
work: a 2D and a 3D model. With either model, accepta-
ble sequence pairs are those that not only fold, but also
form a functional dimer with a sufficient cellular concen-
tration (see Methods). In this section, we consider how
the selective pressure to dimerize affects the sequence
diversity. Another, more quantitative measure of sequence diversity
is given by the network diameters. The diameter of a neu-
tral network is defined as the largest number of point
mutations separating any two viable sequences [23]. In
Fig. 3A, the neutral network diameters in the absence (D)
and presence (D') of selection for dimerization are shown
as a histogram. We consider each 2D protein in turn, with
its neutral network of sequences (1081 networks in all). The dimerization condition (when applied) requires that
these protein dimerize specifically with another, particu-
lar protein (not shown), chosen arbitrarily. The dimer
concentration threshold for viability was set to δ = 0.2. Although the networks shrink when the dimerization con- http://www.biomedcentral.com/1472-6807/7/79 Protein structure is repre-
sented through two simple, very different models: a two-
dimensional lattice model and a three-dimensional off-
lattice model. The structural models are thus highly sim-
plified and are sometimes referred to as "toy" models. Nevertheless, these and similar models have been shown
in the past to provide useful insights. The main qualitative
results below appear to be robust with respect to model
details. In particular, we have done detailed studies of
"monomeric" evolution with several different amino acid
alphabets and interaction models that will be published
elsewhere. In a neutral evolutionary model, all the viable sequences
have the same fitness. Therefore, the graph just defined is
referred to as a "neutral network". We noted above that
some viable sequences are especially "robust" with respect
to point mutations. In fact, with a neutral evolutionary
model limited to point mutations, the steady state has a
remarkable property: sequences with a high tolerance of
mutations are overrepresented within the population,
compared to a random selection of viable sequences. There is a corresponding depletion in sequences that have
a low tolerance of mutations, since they undergo a larger
proportion of lethal mutations. Furthermore, within the
set of viable sequences, the mutationally robust sequences
are "close" to other robust sequences: a small number of
mutations is needed to transform one into the other. Con-
sequently, the most robust sequences form one or more
clusters within the neutral network [7,13]. These clusters
are referred to as funnels (or "superfunnels" [7,13]), Only limited studies of Protein-protein pairs have been
reported [24]. Here, a functional coupling between two Page 2 of 15
(page number not for citation purposes) Page 2 of 15
(page number not for citation purposes) http://www.biomedcentral.com/1472-6807/7/79 Page 3 of 15
(page number not for citation purposes) 2D on-lattice proteins Monomer evolution has been extensively studied with the
2D lattice model [10,25]. Using this model, 12,386,286
out of 33,554,432 sequences fold into one of the 1081
possible conformations (unrelated by symmetry) [8]. The
fraction of sequences able to fold (about 37%) is unreal-
istically high, compared to real proteins. This is due to the
simple HP model and the limited space of allowed con-
formations. Ten conformations are especially robust
towards mutations, or "designable", with neutral net-
works of 40,000–68,000 sequences. We consider that the Page 3 of 15
(page number not for citation purposes) Page 3 of 15
(page number not for citation purposes) BMC Structural Biology 2007, 7:79 http://www.biomedcentral.com/1472-6807/7/79 Example of a 2D dimer
Figure 1
Example of a 2D dimer. A) Low selective pressure for dimerization: δ = 0.04 (i.e., only sequences that lead to a protein frac-
tion of at least 4% engaged in the functional dimer are viable). Amino acids are colored according to the mean sequence in the
steady state (hydrophobic: dark; polar: light). B) The same dimer under a moderate selective pressure: δ = 0.1 This leads to a
more hydrophobic interface. C) The same dimer with δ = 0.2. D) The neutral network for one of the protein partners when δ =
0.04. Black dots represent viable monomer sequences; red dots represent sequences that survive under the dimerization condi-
tion. Connections between sequences are omitted for clarity. The radial position of each sequence reflects its distance from an
arbitrary center (the most populated sequence when δ = 0). E) Idem, δ = 0.1. F) Idem, δ = 0.2. A
B
C
D
delta = 4%
E
delta = 10%
F
delta = 20% delta = 10% delta = 20% F E D Example o
Figure 1 p
g
Example of a 2D dimer. A) Low selective pressure for dimerization: δ = 0.04 (i.e., only sequences that lead to a protein frac-
tion of at least 4% engaged in the functional dimer are viable). Amino acids are colored according to the mean sequence in the
steady state (hydrophobic: dark; polar: light). B) The same dimer under a moderate selective pressure: δ = 0.1 This leads to a
more hydrophobic interface. C) The same dimer with δ = 0.2. D) The neutral network for one of the protein partners when δ =
0.04. 2D on-lattice proteins Black dots represent viable monomer sequences; red dots represent sequences that survive under the dimerization condi-
tion. Connections between sequences are omitted for clarity. The radial position of each sequence reflects its distance from an
arbitrary center (the most populated sequence when δ = 0). E) Idem, δ = 0.1. F) Idem, δ = 0.2. dition is applied (many sequences are no longer viable),
the diameters shrink very little: D' is typically only 1–2
units (amino acids) smaller than D. Similarly, the "dis-
tance" between two protein folds can be defined as the
number of mutations needed to convert one fold into the
other. Fig. 3B shows that for the 2D model, the distances
between folds increase only slightly (by 1–2 amino acids)
under the dimerization condition. In fact, the sequence
diversity is such that for moderate values of δ, and for typ-
ical pairs of 2D proteins, almost every sequence in the
neutral network of A has at least one B sequence with
which it can form a viable dimer. Size of neutral network components
Figure 2
Size of neutral network components. For a representa-
tive 2D dimer, the size of the four largest components as a
function of the selective pressure δ. As δ increases, there are
fewer viable sequence pairs, but there is always a single con-
nected component that is much larger than the other, small
components. Dashed vertical lines are visual aids to show how
the small components progressively break off from the largest
one. Dimeric constraint delta (%)
5
10
15
20
25
30
102
104
106
Size of connected components Dimeric constraint delta (%)
5
10
15
20
25
30
102
104
106
Size of connected components Size of connected components Size of connected components 3D off-lattice proteins The second structural model is the three-dimensional, off-
lattice model [9,26]. For Grb2 and Vav, it gives 31,469
and 29,667 different HP profiles (according to the classi-
fication given in the Methods section). These profiles lead
to almost 109 possible pairs of HP profiles. The selection
for dimerization is determined by the Z-score of the
native, functional structure, compared to the Z-score of
the decoys. Any pair of sequences whose interaction is
weaker than that of the natural sequences is discarded. For
the functional structure to be populated at least 10% of
the time in the cell, there cannot be more than 9 alternate Page 4 of 15
(page number not for citation purposes) Diversity o
Figure 3 The same reduction is seen in the 2D case with
a δ of about 0.25 (Fig. 2; Tab. 1). The rather small, viable,
3D fraction is related to the larger size of the 3D dimer
interface. The Grb2-Vav complex involves 14–16 amino
acids on each partner. A reduction factor of
for the
number of sequence profiles can be obtained by fixing the
profile (H or P) of just 11 positions in the dimer (since 211
≈ 2000), or 5–6 positions on each monomer. These posi-
tions are chosen according to their proximity to the inter-
face. Fixing 11 positions appears reasonable with respect
to the size and diversity of typical Protein-protein inter-
faces [27]. 1
2000 10. This may be an indication of insufficient sequence
sampling during the Monte Carlo simulation of the Vav
monomer. Longer (and expensive) simulations are
needed to test this further. However, the cost of the
present calculations is already close to the limit of what is
feasible (weeks of CPU time to construct the dimeric neu-
tral network using ~10 recent processors). Diversity o
Figure 3 Diversity of viable genotypes in the neutral networks
Figure 3
Diversity of viable genotypes in the neutral networks. A) The neutral network diameters in the absence (D) and presence
(D') of selection for dimerization, shown as a 2D histogram. We consider each 2D protein structure in turn, with its neutral net-
work of sequences. The dimerization condition (when applied) requires that this protein dimerize specifically with another, partic-
ular protein (not shown), chosen arbitrarily. The dimer concentration threshold for viability was set to δ = 0.2. The diameter
represents the largest "distance" between any two sequences in the neutral network (the number of amino acid mutations that
separate them) [23]. The darkest squares are above the diagonal, indicating that among the 1081 structures, most have neutral
networks that shrink when the dimerization condition is applied. However the diameters shrink very little: D' is typically only 1–2
units (amino acids) smaller than D. B) The distance between neutral networks in the absence (D) and presence (D') of selection
for dimerization, shown as a 2D histogram. The dimerization selection criterion is the same as above. The darkest squares are
above the diagonal, indicating for most pairs of structures, under the dimerization selection criterion, the corresponding neutral
networks shrink so that their mutual distance increases (D' > D). However, the increase is small, only 1–2 units. structures of lower energy. Therefore, the rank k we require
for the native Z-score was varied from 1 to 10. With k = 1,
only 537 sequence pairs were viable. With k = 3–10, there
were between 67,291 and 470,334 viable pairs of HP pro-
files. The latter value represents just one 2000th of all pos-
sible pairs. The same reduction is seen in the 2D case with
a δ of about 0.25 (Fig. 2; Tab. 1). The rather small, viable,
3D fraction is related to the larger size of the 3D dimer
interface. The Grb2-Vav complex involves 14–16 amino
acids on each partner. A reduction factor of
for the
1
2000 structures of lower energy. Therefore, the rank k we require
for the native Z-score was varied from 1 to 10. With k = 1,
only 537 sequence pairs were viable. With k = 3–10, there
were between 67,291 and 470,334 viable pairs of HP pro-
files. The latter value represents just one 2000th of all pos-
sible pairs. Page 5 of 15
(page number not for citation purposes) Size of neu
Figure 2 p
g
Size of neutral network components. For a representa-
tive 2D dimer, the size of the four largest components as a
function of the selective pressure δ. As δ increases, there are
fewer viable sequence pairs, but there is always a single con-
nected component that is much larger than the other, small
components. Dashed vertical lines are visual aids to show how
the small components progressively break off from the largest
one. Page 4 of 15
(page number not for citation purposes) Page 4 of 15
(page number not for citation purposes) http://www.biomedcentral.com/1472-6807/7/79 BMC Structural Biology 2007, 7:79 Diversity of viable genotypes in the neutral networks
Figure 3
Diversity of viable genotypes in the neutral networks. A) The neutral network diameters in the absence (D) and presence
(D') of selection for dimerization, shown as a 2D histogram. We consider each 2D protein structure in turn, with its neutral net-
work of sequences. The dimerization condition (when applied) requires that this protein dimerize specifically with another, partic-
ular protein (not shown), chosen arbitrarily. The dimer concentration threshold for viability was set to δ = 0.2. The diameter
represents the largest "distance" between any two sequences in the neutral network (the number of amino acid mutations that
separate them) [23]. The darkest squares are above the diagonal, indicating that among the 1081 structures, most have neutral
networks that shrink when the dimerization condition is applied. However the diameters shrink very little: D' is typically only 1–2
units (amino acids) smaller than D. B) The distance between neutral networks in the absence (D) and presence (D') of selection
for dimerization, shown as a 2D histogram. The dimerization selection criterion is the same as above. The darkest squares are
above the diagonal, indicating for most pairs of structures, under the dimerization selection criterion, the corresponding neutral
networks shrink so that their mutual distance increases (D' > D). However, the increase is small, only 1–2 units. A
B
D
D′
5
10
15
20
5
10
15
20
D
D′
5
10
15
20
5
10
15
20 A
D
D′
5
10
15
20
5
10
15
20 B
D
D′
5
10
15
20
5
10
15
20 B 5 5 D 5 Independence between mutational robustness and
dimerization ability The selective pressure to expose hydrophobic residues
might be expected to correlate with a lower mutational
robustness of the two dimerizing proteins. Indeed, the
dimerizing sequences are more constrained by negative
selection, so that they might have fewer mutations that
lead to viable sequences. To quantify this idea, we define
more precisely the mutational robustness of a particular
viable sequence as the number n of its single mutants that
are also viable. With respect to the neutral network and its
graph structure (see Methods), n represents the number of
neighboring nodes the node is connected to and can be
identified with the "mutational robustness" of the
sequence pair. The robustness n depends on the protein
(A or B), on the particular sequence, and on the level of Similar to the 2D case, most (70%) of the Grb2 sequences
have at least one Vav sequence with which they are able to
form a viable dimer. The Vav sequences are less diverse:
only 8% of the monomeric sequences survive when k = Page 5 of 15
(page number not for citation purposes) http://www.biomedcentral.com/1472-6807/7/79 BMC Structural Biology 2007, 7:79 Table 1: Mutational robustness as a function of the functional constraint
δ
nrandom
nss
Survival #
Survival %
0.00
9.18
8.96
18.14
11.48
11.84
23.32
96,108,582
100
0.01
8.47
8.36
16.84
10.95
11.39
22.34
68,290,887
71
0.02
8.03
7.96
16.00
10.61
11.00
21.61
47,695,260
50
0.03
7.67
7.63
15.31
10.30
10.71
21.01
34,273,488
36
0.04
7.43
7.42
14.85
10.10
10.57
20.66
25,785,086
27
0.06
7.06
7.01
14.07
9.69
9.99
19.68
14,825,259
15
0.08
6.71
6.68
13.39
9.38
9.69
19.07
8,983,407
9.3
0.10
6.45
6.42
12.87
9.08
9.40
18.48
5,522,233
5.8
0.15
5.86
5.78
11.64
8.32
8.53
16.86
1,626,428
1.7
0.20
5.18
5.02
10.20
7.48
7.52
15.01
353,538
0.37
0.25
4.43
4.16
8.60
6.21
6.20
12.41
47,192
0.05
Mutational robustness n as a function of the functional constraint (measured by δ). Data are shown for a representative 2D dimer. The two
partners each have about ten thousand viable sequences. The mean mutational robustness for randomly chosen sequences is nrandom. The
robustness averaged over the steady state is nss. These values correspond to the overall mutational robustness of the AB dimer. The separate
contributions of each partner, A and B, are also shown. The number and percentage of viable sequence pairs are shown. Independence between mutational robustness and
dimerization ability 〈〉
n A
random
〈〉
n B
random
〈〉
n ss
A
〈〉
n ss
B Table 1: Mutational robustness as a function of the functional constraint Mutational robustness n as a function of the functional constraint (measured by δ). Data are shown for a representative 2D dimer. The two
partners each have about ten thousand viable sequences. The mean mutational robustness for randomly chosen sequences is nrandom. The
robustness averaged over the steady state is nss. These values correspond to the overall mutational robustness of the AB dimer. The separate
contributions of each partner, A and B, are also shown. The number and percentage of viable sequence pairs are shown. stringency of the dimerization condition. In the limit
where δ = 0 (no dimerization required), n becomes the
mutational robustness of the protein (A or B) considered
as a monomer. As δ increases, the monomeric networks
are increasingly depleted (Fig. 1), and typical values of n
may be expected to decrease for both A and B. Another
useful quantity to characterize a particular sequence is the
folding temperature Tf of the protein. Tf is a measure of
protein stability (see Methods), and might also be
expected to decrease as the stringency of selection for
dimerization increases, since exposing hydrophobic resi-
dues tends to lower stability with our energy function. viewing them as two independent monomers. The muta-
tional robustness n and the folding temperature Tf are
defined for each sequence of A or B as described above. Next, we take an A sequence and a B sequence; we identify
their functional interface (see Methods), and we compute
the concentration [AB]func of the functional dimer in the
cell at chemical equilibrium (Eqs. 4). No negative selec-
tion is applied for dimerization; i.e., δ = 0. Considering all
pairs of A, B sequences, we find that the ability to dimerize
is actually not correlated with either n (Fig. 4A) or the
folding temperature Tf (Fig. 4B). Sequences with very
diverse values of n and Tf have the same ability to dimer-
ize, as measured by [AB]func. We consider first the neutral networks of A and B sepa-
rately, in the absence of any dimerization requirement, Absence of correlation of n and Tf with dimerization ability
Figure 4
Absence of correlation of n and Tf with dimerization ability. Independence between mutational robustness and
dimerization ability A) For a representative 2D protein dimer, we show the
mutational robustness n of each viable sequence pair, versus the cellular concentration [AB]func of the functional dimer at chemical
equilibrium for that pair. No selective pressure for dimerization is applied. The black line represents the average over the distribu-
tion of n values for each value of [AB]func. n and [AB]func are seen to be uncorrelated. B) For the same 2D dimer, we show the fold-
ing temperature of each viable sequence pair as a function of [AB]func. C) For the Grb2-Vav 3D dimer, we show the Z value (which
measures the dimerization ability) of each viable sequence pair as a function of n. The two are seen to be uncorrelated, even
though a selective pressure for dimerization is present in this case (see text). A
B
C
Dimer concentration [AB]
0.5
1
1.5
2
2.5
3
10
20
30
40
Mutational robustness n
Dimer concentration [AB]
0.5
1
1.5
2
2.5
3
0.1
0.2
0.3
0.4
0.5
0.6
Folding temperature Tf
Mutational robustness n
10
20
30
40
–4.0
–3.5
–3.0
–2.5
–2.0
–1.5
Z-score of the functional dimer B
Dimer concentration [AB]
0.5
1
1.5
2
2.5
3
0.1
0.2
0.3
0.4
0.5
0.6
Folding temperature Tf A
Dimer concentration [AB]
0.5
1
1.5
2
2.5
3
10
20
30
40
Mutational robustness n C
Mutational robustness n
10
20
30
40
–4.0
–3.5
–3.0
–2.5
–2.0
–1.5
Z-score of the functional dimer http://www.biomedcentral.com/1472-6807/7/79 In
other words, the sequences that form high affinity com-
plexes are overpopulated. Thus, neutral evolution leads
not only to increased mutational robustness, but to
increased concentrations of the functional species present
in the average cell. This effect and its extent were some-
what harder to anticipate, despite the analogy to the mon-
omeric result (enrichment in n). Like the mutational
robustness n in the monomeric case, [AB]func is directly
selected for in the dimer case. However, the mode of selec-
tion is quite different for the two quantities: [AB]func is
subjected to a threshold, while selection for n acts in a
more continuous manner. In addition, we saw that the
concentration of the functional dimer is not correlated
with, or closely-related to either n or Tf. The enrichment in
n arises because highly-connected sequences are grouped
in the middle of the neutral network. In effect, the enrich-
ment arises because n varies in a smooth, continuous
manner over the network, so that robust sequences are
close to other robust sequences. But we saw above that the
sequences satisfying the dimerization threshold are
widely distributed throughout the underlying monomeric
network (Fig. 1). Therefore, it was not obvious ahead of
time that dimerization ability would vary sufficiently
smoothly and continuously. For the three-dimensional Grb2-Vav dimer, the correla-
tion between mutational robustness and functionality is
very weak, even in the presence of a selective pressure for
dimerization. In Fig. 4, the mutational robustness n is
plotted against the dimerization energy Z-score, denoted
Z. The data correspond to a dimerization selection thresh-
old of k = 10, for a total of 470,334 viable dimer sequence
profiles. The correlation coefficient between n and Z is
low, less than 5%. This independence between the Z and n (or Tf) is likely to
hold qualitatively for real proteins. For a given dimer
interface AB, we expect that the interface sequences will be
rather contrained by natural selection [27,28], whereas a
wider range of amino acid types may be found on the
remaining parts of the surfaces and in the proteins' core,
leading to a wide range of protein stabilities. However, a
systematic analysis of both sequence conservation and
protein stability in families of dimerizing proteins would
be needed to make this statement quantitative. Thermo-
dynamic data are scarce, and such an analysis is beyond
the scope of this study. In Fig. http://www.biomedcentral.com/1472-6807/7/79 BMC Structural Biology 2007, 7:79 The data in Fig. 4A,B correspond to one representative
dimer, made of a particular pair of protein structures. For
the 16 dimer structures we analyzed, the correlation coef-
ficients range from -0.026 to 0.042 for n (respectively, -
0.035 to 0.100 for Tf). Inspecting the sequences in more
detail, we find that, in fact, dimerization can be enhanced
without increasing the number of exposed hydrophobic
residues. Instead, hydrophobic residues can be moved to
the interface region from another part of the protein sur-
face. Within the simple 2D model, this operation has very
little effect on the protein stability and folding tempera-
ture, which explains that Tf and [AB]func are uncorrelated. Since Tf and n are known to be strongly correlated [13], n
must also be uncorrelated with [AB]func. state [7,8,13]. Sequences in the core of the neutral net-
work are overpopulated, while those at the edge of the
network, with fewer graph connections, are underpopu-
lated. This steady state enrichment is preserved under the
dimerization constraint, as shown in Fig. 5A. The extent of
enrichment is similar to the pure monomeric case; see
Table 1 for illustrative, numerical values for a particular
complex. A similar enrichment is observed for the 3D pro-
teins (Fig. 5B). The agreement between the 2D and 3D
models provides encouraging evidence that this behavior
does not depend on model details. The dimer folding
temperature is also enriched in the steady state (Fig. 5C). This is consistent with the known correlation between n
and Tf. We consider next the neutral networks of A and B in the
presence of a selective pressure for dimerization. For the
same 16 representative dimer structures, we consider a
series of selection thresholds δ. By imposing a particular δ,
we effectively discard all the points in Fig. 4 to the left of
δ and all the corresponding sequence pairs. Surviving
pairs close to the dimerization threshold tend to lose
some of their neighboring sequences, so that their muta-
tional robustness n decreases. The net effect is a rather
strong correlation between [AB]func and n. For δ =
10–30%, the correlation coefficient is about 50–65%. At
a low selective threshold of δ = 2%, the correlation is
about 30%. In a similar way, the cellular concentration of the func-
tional dimer is enriched in the steady state (Fig. 5D). Absence o
Figure 4 Absence of correlation of n and Tf with dimerization ability
Figure 4
Absence of correlation of n and Tf with dimerization ability. A) For a representative 2D protein dimer, we show the
mutational robustness n of each viable sequence pair, versus the cellular concentration [AB]func of the functional dimer at chemical
equilibrium for that pair. No selective pressure for dimerization is applied. The black line represents the average over the distribu-
tion of n values for each value of [AB]func. n and [AB]func are seen to be uncorrelated. B) For the same 2D dimer, we show the fold-
ing temperature of each viable sequence pair as a function of [AB]func. C) For the Grb2-Vav 3D dimer, we show the Z value (which
measures the dimerization ability) of each viable sequence pair as a function of n. The two are seen to be uncorrelated, even
though a selective pressure for dimerization is present in this case (see text). Page 6 of 15
(page number not for citation purposes) http://www.biomedcentral.com/1472-6807/7/79 http://www.biomedcentral.com/1472-6807/7/79 6, we define an enrichment factor for dimerization
ability, Φ([AB]) = [AB]ss/[AB]random, where [AB]ss is the
cellular concentration of the functional dimer averaged
over the steady state sequences, and [AB]random is the con-
centration averaged over all the viable sequences, regard-
less of their population. Typical values of Φ([AB]) are greater than 1, correspond-
ing to enrichment. The enrichment in functional species is
strongest when the selection criterion is only moderately
stringent: Φ([AB]) ≈ 1.2–1.3 when δ ≈ 0.01–0.10. As δ
increases, selection becomes more stringent and the set of
viable sequences is increasingly depleted (Fig. 5). Enrich- The steady state is enriched in functional sequences The steady state is enriched in functional sequences Previous studies of single protein evolution have revealed
an enrichment in mutational robustness in the steady Page 7 of 15
(page number not for citation purposes) BMC Structural Biology 2007, 7:79 http://www.biomedcentral.com/1472-6807/7/79 The population dynamics enhance robustness and functionality
Figure 5
The population dynamics enhance robustness and functionality. A) The distribution of mutational robustness n for a
given 2D protein dimer. Solid line: the steady state population. Dashed line: a population drawn randomly from the neutral net-
work. B) Idem for the 3D, off-lattice Grb2-Vav 3D dimer. C) The folding temperature distribution for the same 2D dimer; solid
line: steady state population; dashed line: random population. D) The distribution of the equilibrium concentration [AB]func of the
functional dimer, for the same 2D dimer; solid line: steady state population; dashed line: random population. A
B
Mutational robustness n
5
10
15
20
25
30
35
0.02
0.04
0.06
0.08
Frequency of mutational robustness
Mutational robustness n
5
10
15
20
0.05
0.1
0.15
0.2
Frequency of mutational robustness
C
D
Folding temperature Tf
0.1
0.2
0.3
0.4
0.5
0.6
0.02
0.04
0.06
0.08
Frequency of folding temperature
Dimer concentration [AB]
0.5
1
1.5
2
2.5
3
0.05
0.1
0.15
0.2
0.25
Frequency of dimer concentration A
B
Mutational robustness n
5
10
15
20
25
30
35
0.02
0.04
0.06
0.08
Frequency of mutational robustness
Mutational robustness n
5
10
15
20
0.05
0.1
0.15
0.2
Frequency of mutational robustness A
B
Mutational robustness n
5
10
15
20
25
30
35
0.02
0.04
0.06
0.08
Frequency of mutational robustness Mutational robustness n
5
10
15
20
05
.1
5
.2
equency of mutational robustness B Frequency of mutational robustness B The population dynamics enhance robustness and functionality
Figure 5
C
D
Folding temperature Tf
0.1
0.2
0.3
0.4
0.5
0.6
0.02
0.04
0.06
0.08
Frequency of folding temperature
Dimer concentration [AB]
0.5
1
1.5
2
2.5
3
0.05
0.1
0.15
0.2
0.25
Frequency of dimer concentration C
D
Folding temperature Tf
0.1
0.2
0.3
0.4
0.5
0.6
0.02
0.04
0.06
0.08
Frequency of folding temperature C Frequency of folding temperature D
Dimer concentration [AB]
0.5
1
1.5
2
2.5
3
0.05
0.1
0.15
0.2
0.25
Frequency of dimer concentration The popul
Figure 5 The population dynamics enhance robustness and functionality
Figure 5
The population dynamics enhance robustness and functionality. The steady state is enriched in functional sequences A) The distribution of mutational robustness n for a
given 2D protein dimer. Solid line: the steady state population. Dashed line: a population drawn randomly from the neutral net-
work. B) Idem for the 3D, off-lattice Grb2-Vav 3D dimer. C) The folding temperature distribution for the same 2D dimer; solid
line: steady state population; dashed line: random population. D) The distribution of the equilibrium concentration [AB]func of the
functional dimer, for the same 2D dimer; solid line: steady state population; dashed line: random population. ment then decreases. A model with larger proteins and/or
a more complex amino acid alphabet would probably
allow a greater enrichment, extending to higher selection
stringencies. The requirement of a 10% population for the
functional dimer (δ = 0.10) appears reasonable if the
dimer's function is to form transiently and transmit a sig-
nal. dimer concentrations cannot be readily computed. Never-
theless, the steady state populations of the viable
sequences are available, so that we can compare the typi-
cal Z-scores, Z, in the steady state population and a ran-
dom population. This is done in Table 2. An enrichment
factor Φ(Z) is defined in the same way as Φ([AB]), above. We considered selection thresholds k between 2 and 10. We recall that a value of k = 4 implies that the functional
dimer is among the four lowest-energy structures, out of a
total of 784 structures. As k increases, Φ(Z) first increases
from 1.07 to 1.18 (k = 4 or 5), then decreases to 1.04 (k =
9 or 10). As in the 2D case, the enrichment is maximal for
an intermediate selection stringency. The maximum
enrichment factor is roughly comparable in the 2D and
3D cases, even though the measures of dimerization
([AB]func and Z) are obviously different. Again, the quali-
tative 2D-3D agreement is encouraging. For one particular dimer, we constrained protein B to have
a single, fixed sequence, so that it no longer evolves and
plays the role of a simple ligand. In this case, the steady
state enrichment in the concentration of the functional
dimer is reduced by half (Φ([AB]) ≈ 1.15 instead of 1.3;
data not shown). Page 8 of 15
(page number not for citation purposes) The popul
Figure 6 p p
y
y
g
The population dynamics enhance robustness and functionality. A) Φ(n) = nss/nrandom measures the enrichment in
mutational robustness due to the steady state dynamics. Each colored point corresponds to a 2D protein pair subjected to a
dimerization condition. Each color corresponds to a particular level of the selection stringency, indicated by the value of the con-
centration threshold δ (in %, legend on right). Data are shown for a selection of 15 dimers. B) Φ([AB]) = [AB]ss/[AB]random meas-
ures the enrichment in dimerization ability due to the steady state dynamics: [AB]ss is the cellular concentration of the functional
dimer averaged over the steady state sequences; [AB]random is the value averaged over the viable sequences, regardless of their
population. A value of Φ([AB]) greater than 1 indicates an enrichment of the steady state population in sequences that readily
dimerize. Colors indicate the selection stringency. Data are shown for a selection of 10 dimers. Functional sequences form a funnel ment are grouped in the center of the neutral network,
forming a basin, or funnel in sequence space. The enrichment in functional species is strongest for
sequence pairs near the "prototype" pair, defined as the
most populated pair in the steady state [8]. This is illus-
trated in Fig. 7A for the 2D proteins. The mean concentra-
tion [AB]func of the functional dimer is plotted for each
viable sequence pair, as a function of its distance from the
prototype pair (for a representative dimer and a few values
of the selection threshold, δ). The concentration [AB]func
varies widely, but the mean value drops off rapidly and
smoothly as one moves away from the prototype
sequence. Similar behavior is seen for other 2D dimers. Thus, the sequences responsible for the functional enrich- With the 3D model, dimerizing ability is measured by the
Z-score. We saw above that the corresponding enrichment
factor, Φ(Z), was slightly smaller than Φ([AB]) in the 2D
case. Nevertheless, a basin of high affinity complexes is
also seen with the 3D model, although the funnel shape
is somewhat less pronounced (Fig. 7B). The funnel shape
flattens out at a distance of about 6–7 from the prototype
sequence. A small number of sequences forming high
affinity complexes are actually found outside the basin, at
distances of 13–14 from the prototype. The steady state is enriched in functional sequences For the 3D model, it is harder to characterize the enrich-
ment (if any) in steady state dimerization ability, because
there are too many (784) possible structures and the Page 8 of 15
(page number not for citation purposes) Page 8 of 15
(page number not for citation purposes) http://www.biomedcentral.com/1472-6807/7/79 BMC Structural Biology 2007, 7:79 The population dynamics enhance robustness and functionality
Figure 6
The population dynamics enhance robustness and functionality. A) Φ(n) = nss/nrandom measures the enrichment in
mutational robustness due to the steady state dynamics. Each colored point corresponds to a 2D protein pair subjected to a
dimerization condition. Each color corresponds to a particular level of the selection stringency, indicated by the value of the con-
centration threshold δ (in %, legend on right). Data are shown for a selection of 15 dimers. B) Φ([AB]) = [AB]ss/[AB]random meas-
ures the enrichment in dimerization ability due to the steady state dynamics: [AB]ss is the cellular concentration of the functional
dimer averaged over the steady state sequences; [AB]random is the value averaged over the viable sequences, regardless of their
population. A value of Φ([AB]) greater than 1 indicates an enrichment of the steady state population in sequences that readily
dimerize. Colors indicate the selection stringency. Data are shown for a selection of 10 dimers. A
B
Number of viable sequence pairs
102
104
106
108
1.0
1.1
1.2
1.3
1.4
1.5
Enrichment of mutational robustness Phi(n)
0
1
2
3
4
6
8
10
12.5
15
20
25
26
27
28
29
30
delta (%)
Number of viable sequence pairs
102
104
106
108
1
1.1
1.2
1.2
1.3
Enrichment of functional dimer Phi([AB])
0
1
2
3
4
6
8
10
12.5
15
20
25
26
27
28
29
30
delta (%) A
Number of viable sequence pairs
102
104
106
108
1.0
1.1
1.2
1.3
1.4
1.5
Enrichment of mutational robustness Phi(n)
0
1
2
3
4
6
8
10
12.5
15
20
25
26
27
28
29
30
delta (%) B
Number of viable sequence pairs
102
104
106
108
1
1.1
1.2
1.2
1.3
Enrichment of functional dimer Phi([AB])
0
1
2
3
4
6
8
10
12.5
15
20
25
26
27
28
29
30
delta (%) A Enrichment of mutational robustness Phi(n) Discussion B) Similar representation for the 3D Grb2-Vav
dimer: the dimerization Z score as a function of the distance from the prototype sequence. A
B
Distance from prototype sequence pair
[AB]
5
10
15
2
3
ffi
Distance from prototype sequence pair
2
4
6
8
10
12
14
–3.4
–3.2
–3.0
–2.8
–2.6
–2.4
Z-score of the functional dimer B
Distance from prototype sequence pair
2
4
6
8
10
12
14
–3.4
–3.2
–3.0
–2.8
–2.6
–2.4
Z-score of the functional dimer A
Distance from prototype sequence pair
[AB]
5
10
15
2
3
ffi A B Distance from prototype sequence pair
5
10
15 Emergence of a functional funnel in sequence space
Figure 7
Emergence of a "functional funnel" in sequence space. A) 2D dimer: The mean concentration [AB]func of the functional
dimer as a function of the distance of each viable sequence pair from the prototype pair (the most populated pair in the steady
state). Data are shown for a representative dimer and a selection threshold of δ = 0.08 (dotted horizontal line). There are no
sequence pairs below the dotted line, because such sequences are not viable, by definition. Black curve: the mean value for each
distance. Black vertical bars indicate the standard deviation at each distance. Dashed horizontal line: average over a random set of
sequences. Solid horizontal line: overall steady state average. The concentration [AB]func varies widely, but the mean value drops
off rapidly and smoothly as one moves away from the prototype sequence pair. B) Similar representation for the 3D Grb2-Vav
dimer: the dimerization Z score as a function of the distance from the prototype sequence. functions as a simple ligand, this property would proba-
bly not hold. A single, continuous network makes it easier
to explore sequence space, since single mutations can be
used more extensively, rather than large hops involving
several mutations at a time. the restriction to maximally compact structures [31,32],
and the use of a highly-simplified energy function [33]. Despite their simplicity, these models have some impor-
tant protein-like features, such as hydrophobic-polar seg-
regation. The simple, pairwise energy function allowed
England and collaborators to devise a determinant of pro-
tein designability that is applicable to real proteins [5]. Discussion Table 2: Stability enrichment for the Grb2-Vav 3D dimer Table 2: Stability enrichment for the Grb2-Vav 3D dimer
k
viable sequences
nss
Φ(Z)
1
537
14.13
1.05
2
30801
4.65
1.07
3
67291
3.50
1.08
4
109954
3.07
1.18
5
157903
3.07
1.18
6
211133
3.01
1.06
7
269022
3.00
1.07
8
331852
2.96
1.07
9
398776
2.99
1.04
10
470334
3.00
1.04
Steady state enrichment factor Φ(Z) of the dimerization ability
(measured by the Z-score) for the Grb2-Vav 3D dimer as a function of
the functional constraint (measured by k). The number of viable
sequence profile pairs and the mean mutational robustness nss are
also shown. Protein-protein interactions are central to cellular organi-
zation, and must have appeared at a very early stage of
evolution. To understand better their effects, we consid-
ered here two simple, "toy" models of protein structure
and evolution, and determined the effect of explicitly
selecting for Protein-protein interactions. By employing a
2D, lattice representation of protein structure and binary,
hydrophobic/polar sequences, exact calculations could be
performed. The 3D, off-lattice model gives a similar qual-
itative picture. For example, the 3D model also predicts
that mutationally robust sequences are overrepresented in
the steady state, in agreement with the well-known result
of lattice models [7,13,29]. Lattice models like the present one have been a subject of
debate, because of their use of limited alphabets [25,30], Page 9 of 15
(page number not for citation purposes) Page 9 of 15
(page number not for citation purposes) http://www.biomedcentral.com/1472-6807/7/79 BMC Structural Biology 2007, 7:79 Emergence of a "functional funnel" in sequence space
Figure 7
Emergence of a "functional funnel" in sequence space. A) 2D dimer: The mean concentration [AB]func of the functional
dimer as a function of the distance of each viable sequence pair from the prototype pair (the most populated pair in the steady
state). Data are shown for a representative dimer and a selection threshold of δ = 0.08 (dotted horizontal line). There are no
sequence pairs below the dotted line, because such sequences are not viable, by definition. Black curve: the mean value for each
distance. Black vertical bars indicate the standard deviation at each distance. Dashed horizontal line: average over a random set of
sequences. Solid horizontal line: overall steady state average. The concentration [AB]func varies widely, but the mean value drops
off rapidly and smoothly as one moves away from the prototype sequence pair. Discussion Nonetheless, artifacts certainly arise from these models,
and that is why we also used a significantly different, 3D,
off-lattice structural model, to corroborate the conclu-
sions drawn from the lattice model. The present treatment corresponds to neutral evolution,
in the sense that it treats all viable sequences as equally fit. The model has a at fitness plateau–the neutral network,
surrounded by a sea of non-viable sequence pairs. Evolu-
tion takes the form of a random diffusion throughout the
neutral network. This neutral picture should be in rough
qualitative agreement with real proteins. Neutral muta-
tions are very common in proteins, as shown by the
sequence diversity associated with modern protein folds. The neutral model predicts that the probability for a pro-
tein to retain its native fold decreases exponentially with
the number of mutations, at least for the first few muta-
tions; this prediction agrees with recent experimental
observations [34]. The proportions of tolerated mutations
computed here for the individual "proteins" are also com-
parable to those of several real protein folds [35]. Here, a functional coupling between pairs of genes was
added to two previous evolutionary models: the two pro-
teins of interest must associate in order to function. The
steady state enrichment in mutational robustness is pre-
served under this additional constraint. Sequence diver-
sity remains very large when dimerization is required,
even though only a fraction of sequences survive under
these more selective conditions. The sequence diversity is
reflected, for example, by the wide range of protein
robustnesses and folding temperatures that can lead to the
same dimerization ability. It is somewhat unexpected that
as the pressure to dimerize is increased and more and
more sequence pairs are eliminated, the viable sequences
continue to be largely grouped in a single, continuous net-
work (Fig. 2), instead of splitting into many small, discon-
nected networks. If one of the proteins were constrained
to have a fixed ligand, so that it no longer evolves but Our model selects for a minimal level of functional abil-
ity, determined by the chosen dimer concentration
threshold, δ, or the Z-score rank, k. The steady state
dynamics then lead to a population that is enriched in
sequences that form high affinity dimers, well beyond the
minimal ability needed to survive. Page 10 of 15
(page number not for citation purposes) Conclusion
F
h From the present models, the sequences that are popu-
lated in the steady state are enhanced in their functional
ability. This should allow an increased resistance to envi-
ronmental change, or adaptability [14]. Indeed, a strong
dimer is more likely to be preserved under a change in the
surrounding temperature or pH, for example. They should
also provide an increased ability to evolve and comply
with newly imposed functional requirements. Indeed,
after a gene duplication event, a protein A that starts out
with the ability to bind strongly to its partner B will be
better able to explore mutations that allow it to co-evolve
with B, or to dimerize with other, existing, homologues of
B (using the ancestral binding mode, at least at the begin-
ning). This effect, which arises from a very simple, mini-
mal model of protein evolution, should lead to an
enhanced ability to create homologous interacting pairs
of proteins, and could have played a role in the early
emergence of Protein-protein interaction networks. The timescale of the present model is set by the muta-
tional probability per unit time, α in Eq. 2. The enrich-
ment in [AB]func is then obtained in the evolutionary
steady state. In real systems, the timescale to reach the
steady state depends on the population size N and the
mutation rate, μ. Previous simulations have identified two
regimes, characterized by the product N μ [36]. When N μ
> 100 (large population and/or mutation rate), the popu-
lation is expected to rapidly sample the steady state, so
that the enrichment phenomena predicted above should
be visible. For lower populations and/or mutation rates,
N μ << 100, the population is expected to behave like a
random sample drawn from the neutral network, so that
no enrichment should be observed. If μ is a mutation rate
per individual, then N μ represents the mutation rate
within the entire population. The product N μ is small for
eukaryotic populations and large for RNA viruses. For
eubacteria with a generation time of minutes, a neutral
mutation should appear in a typical protein about N μ =
101–103 times a day within a large colony [37]. The given
range corresponds to different colony sizes (105–107 indi-
viduals); it can be expanded if one considers longer gener-
ation times or artificially accelerated mutation rates (e.g.,
in the presence of chemical mutagens). http://www.biomedcentral.com/1472-6807/7/79 BMC Structural Biology 2007, 7:79 functionality of the typical sequence pairs has been
enhanced by the evolutionary dynamics. The enhance-
ment occurs through negative selection, and can be
viewed as a reduction of the mutational load [1]. In prac-
tice, it has a similar effect to the adaptation that would
occur in response to positive selection; namely, enhanced
functionality. The functional enrichment arises because of
the plateau form of the neutral network and because of
the funneled organization of the sequences within the
network. The enrichment is analogous to the enrichment
in mutational tolerance seen previously for single protein
models. Nevertheless, the extent of the enrichment and its
qualitative response as a function of the stringency of the
selection needed to be investigated. Indeed, the mode of
selection for mutational robustness and that for dimeriza-
tion are mathematically quite different. Dimer sequences
are widely dispersed throughout the monomeric net-
works, whereas the steady state enrichment in a given
property (mutational tolerance or dimerization ability) is
related to its continuity over the set of viable sequences. Little or no correlation is seen between the "monomeric"
properties, n and Tf, and the dimerization ability. In addi-
tion, while the functional enrichment first increases with
δ, it then decreases for larger values of δ. Back-of-the-enve-
lope predictions for dimers are difficult because of the
complex chemical equilibria involved (see Methods). dictions could also be tested by analyzing experimental
protein sequences. The weak correlation between dimeriz-
ing ability and protein stability is in accord with our
knowledge of Protein-protein interfaces. Typical Protein-
protein interfaces have a few amino acids forming a cen-
tral hydrophobic patch; small, polar, mutational
"hotspots" are also frequent [27]. The evolutionary con-
traints on these local surface patches should have a lim-
ited effect on other surface and core regions, so that a large
range of protein stabilities can be achieved despite the
constraints. Conversely, it would be interesting to com-
pare the dimerization abilities of very stable proteins,
such as those of thermophilic organisms. Conclusion
F
h This range for N μ
should encompass the two regimes just discussed [36]. Thus, the role of the steady state dynamics in elevating the
average functionality could be experimentally tested by
comparing two such bacterial colonies. Some of our pre- Discussion In other words, the Page 10 of 15
(page number not for citation purposes) Page 10 of 15
(page number not for citation purposes) http://www.biomedcentral.com/1472-6807/7/79 Page 11 of 15
(page number not for citation purposes) Structural models: 2D lattice model and 3D off-lattice
model 2 describes the
flow of population within the neutral network. The first
term in parentheses on the right represents sequences
flowing into a given node i; the second term represents
sequences owing out of i, taking into account the fraction
of viable and lethal mutations. An important property of
Eq. 2 is that there is a single stable steady state. The steady
state probability vector, p = pss, is an eigenvector of C, asso-
ciated with the largest eigenvalue, ν = νss. Remarkably, the
steady state can be shown to be not only stable with
respect to small fluctuations, but globally stable. A
detailed proof will be published elsewhere; see [3] for a
detailed treatment of related mathematical models. ν
ν
=
=
∑
i i
i
S
p
1 It is easy to show that ν is the mean number of neighbors
of the sequences in the network:
, where νi is
the number of neighbors of sequence i. Eq. 2 describes the
flow of population within the neutral network. The first
term in parentheses on the right represents sequences
flowing into a given node i; the second term represents
sequences owing out of i, taking into account the fraction
of viable and lethal mutations. An important property of
Eq. 2 is that there is a single stable steady state. The steady
state probability vector, p = pss, is an eigenvector of C, asso-
ciated with the largest eigenvalue, ν = νss. Remarkably, the
steady state can be shown to be not only stable with
respect to small fluctuations, but globally stable. A
detailed proof will be published elsewhere; see [3] for a
detailed treatment of related mathematical models. ν
ν
=
=
∑
i i
i
S
p
1 The second physical model is a three-dimensional, off-lat-
tice model [9,26]. Two proteins are considered: the 69-res-
idue SH3 domain of Vav and the 57-residue SH3 domain
of Grb2. These two form a Protein-protein complex (PDB
accession number 1gcq). For each one, the experimental,
3D structure is considered, along with over 1200 "decoy"
structures, whose backbone geometries are taken from
completely different proteins [39,40]. The sidechains are
built assuming the most common rotamer for each amino
acid type [41]. For each protein, 100 additional decoys,
with more native-like structures, were produced by molec-
ular dynamics in vacuo at 310 K. Amino acids interact
through Eq. The evolutionary model and its properties y
p
p
Following [3,8] and others, we first assume that all genes
evolve independently, and we focus arbitrarily on one of
them. In a second step, below, we will consider co-evolu-
tion of two interacting proteins. For now, the single gene
of interest is assumed viable if the corresponding protein
folds into its correct conformation. The S sequences that
adopt this conformation are all assumed to be equally fit. We assume evolution can only occur through point muta-
tions; i.e., substitutions of a single amino acid. Frameshift
and nonsense mutations are assumed to be lethal. The
complete set of viable sequences defines a graph, contain-
ing S nodes. Each node represents a viable sequence; links
between nodes represent point mutations. The graph may
not be fully connected; i.e., it may be impossible to con-
nect two viable sequences by a series of point mutations. If the entire population starts out with the same, "native"
protein sequence, then future evolution will only explore
the corresponding, connected subgraph. Therefore, we
can assume without loss of generality that there is only Page 11 of 15
(page number not for citation purposes) Page 11 of 15
(page number not for citation purposes) BMC Structural Biology 2007, 7:79 http://www.biomedcentral.com/1472-6807/7/79 Structural models: 2D lattice model and 3D off-lattice
model one connected graph, which is referred to as a "neutral
network". Two physical models of a protein are considered. The first
treats the "protein" as a chain of L = 25 beads, or amino
acids, which can be either polar (P) or hydrophobic (H). Acceptable conformations occupy a two-dimensional, 5 ×
5, square lattice. Thus, only maximally compact confor-
mations are allowed. The energy is The following, discrete-time, evolutionary model is ana-
lyzed [3,8]. For simplicity, we describe it in detail for the
present, single gene case. The case of interacting proteins
is considered further on. The model behavior is simplest
in the case of a very large (essentially infinite) population,
and we limit ourselves to this case. The effect of a finite
population size is considered in the Discussion. Between
times t and t + 1, an individual with sequence i has a prob-
ability α of undergoing a point mutation and a probabil-
ity β of dying. A new individual has a probability γpi of
appearing spontaneously by birth; with probability 1 - (α
+ β + γ), the individual continues unchanged. After each
generation, populations are rescaled to maintain a con-
stant total. The probability to find a given individual with
sequence i at time t is denoted pi; the change between t
and t + 1 is denoted δpi. We consider here the limit of a
large population, in which case these probabilities follow
the equation: E
eij
ij
i j
=
<∑
Δ ,
(3) (3) where Δij = 1 if the beads i, j are neighbors on the lattice
and zero otherwise, and the interaction coefficients
depend on the type (P or H) of each bead. The values eHH
= -2.3, eHP = -1, and ePP = 0 are used, following [8,38], to
favor compact conformations with a hydrophobic core. A
particular sequence is considered to fold if its lowest
energy conformation is unique (i.e., non-degenerate). It is
viable if it folds into a particular, preselected conforma-
tion. The protein chain is considered to have a direction
(even though the energy function does not); e.g., the
sequences HPP and PPH are different. Sequence explora-
tion is done by exhaustive enumeration. where Δij = 1 if the beads i, j are neighbors on the lattice
and zero otherwise, and the interaction coefficients
depend on the type (P or H) of each bead. Structural models: 2D lattice model and 3D off-lattice
model The values eHH
= -2.3, eHP = -1, and ePP = 0 are used, following [8,38], to
favor compact conformations with a hydrophobic core. A
particular sequence is considered to fold if its lowest
energy conformation is unique (i.e., non-degenerate). It is
viable if it folds into a particular, preselected conforma-
tion. The protein chain is considered to have a direction
(even though the energy function does not); e.g., the
sequences HPP and PPH are different. Sequence explora-
tion is done by exhaustive enumeration. δ
α
β
γ
α
p
p
M
p
i
i
k
k i
= −
+
−
+ ∑
(
)
. ~
(1) (1) Here k ~ i means that k and i are neighbors in the neutral
network and M = 25 denotes the chain length of the pro-
tein. The S × S adjacency matrix C [23] is defined by: Cij =
1 if i ~ j and zero otherwise; the vector of sequence prob-
abilities is p = (p1, p2, ... , pS). Eq. 1 can be rearranged into
the following vector form: Here k ~ i means that k and i are neighbors in the neutral
network and M = 25 denotes the chain length of the pro-
tein. The S × S adjacency matrix C [23] is defined by: Cij =
1 if i ~ j and zero otherwise; the vector of sequence prob-
abilities is p = (p1, p2, ... , pS). Eq. 1 can be rearranged into
the following vector form: With the 2D model, we can calculate exactly the folding
temperature Tf of each structure and sequence. By defini-
tion, Tf is the temperature at which the native conforma-
tion is populated 50% of the time. It is straightforward to
compute it numerically from Boltzmann's law and the
energy spectrum of the 1081 possible conformations. Because we consider only the maximally compact confor-
mations during the computation, the value of Tf is overes-
timated. For a protein dimer, we define Tf as the minimum
of the folding temperatures of the two separate partners. δ
α
ν
p
M C
I p
=
−
(
) . (2) (2) It is easy to show that ν is the mean number of neighbors
of the sequences in the network:
, where νi is
the number of neighbors of sequence i. Eq. Page 12 of 15
(page number not for citation purposes) http://www.biomedcentral.com/1472-6807/7/79 BMC Structural Biology 2007, 7:79 concentration of proteins A and B. In what follows, we
always explore a wide range of protein concentrations,
and so interactions with other proteins are not considered
further. There are then 38 possible chemical species,
whose equilibrium concentrations are obtained by solv-
ing the system: concentration of proteins A and B. In what follows, we
always explore a wide range of protein concentrations,
and so interactions with other proteins are not considered
further. There are then 38 possible chemical species,
whose equilibrium concentrations are obtained by solv-
ing the system: The energy parameters are eHH = -8.5, eHP = 9.0, ePP = -3.5,
These parameters were optimized to discriminate experi-
mental protein structures from large sets of decoys
[39,40]. With this model, we first enumerate sequences
that are viable as monomers; i.e., they fold into the
desired, native structure. For this step, the monomer
sequence space is explored by a Monte Carlo method [9]. A "move" consists in a random point mutation, which is
accepted if the desired, functional structure has a suffi-
ciently low energy, compared to the non-functional,
decoy structures. Specifically, the functional fold must
have an energy gap (energy difference from the lowest
decoy) and a Z-score (energy difference from the average
decoy, in standard deviation units) as large as those of the
starting sequence. The starting sequence is slightly differ-
ent from the native sequence. It is obtained by minimiz-
ing the latter through several thousand Monte Carlo
moves. A trajectory of one hundred million mutations is
then performed. For each accepted mutation, we also
explore systematically its nearest "neighbors" (all its sin-
gle mutations), thus generating a large, representative sub-
set of the relevant neutral network in monomeric
sequence space [9]. http://www.biomedcentral.com/1472-6807/7/79 [
]
[ ] ,
[
]
[ ] ,
[
]
[ ][ ],
AA
a A
I
BB
a B
J
AB
c
A B
K
I
I
J
J
K
K
=
≤
≤
=
≤
≤
=
≤
≤
2
2
1
10
1
10
1
166
(4) [
]
[ ] ,
[
]
[ ] ,
[
]
[ ][ ],
AA
a A
I
BB
a B
J
AB
c
A B
K
I
I
J
J
K
K
=
≤
≤
=
≤
≤
=
≤
≤
2
2
1
10
1
10
1
166
(4) (4) with fixed total concentrations [A]tot and [B]tot. Here, aI, bJ,
cK are equilibrium constants; for example aI = exp(-ΔEI/
kT), where k is Boltzmann's constant, T the temperature,
and ΔEI is the association free energy of the dimer AAI. The
aI, bJ, cK depend on the sequences of A, B through the asso-
ciation free energies. The chemical equations 4 can be
reduced to a fourth-order polynomial, by grouping all the
AAI (respectively, BBJ or ABK) dimers into a single species
(resp.,
,
) and solving for them. The relative con-
centrations of the various AAI subspecies, for example, are
then obtained immediately from [
]. The whole system
can thus be solved numerically very efficiently. An A, B
pair with particular sequences is then considered viable if
the functional dimer has an equilibrium concentration
greater than a chosen threshold δ. The functional dimer is
the one that minimizes the dimerization energy, averaged
over all the A, B sequences that fold (into their designated
native conformations).
Once the neutral network has been constructed (either for
a 2D or a 3D protein), the steady state distribution of
sequences is computed by an iterative, shifted power
method [42], which yields the eigenvector of C corre-
sponding to the largest eigenvalue. Interacting genes: the 2D on-lattice case In a second step, an evolutionary scenario with interacting
genes is explored. We describe first the 2D lattice case. It is
assumed that a vital function can only be performed when
two proteins A, B not only fold, but specifically dimerize. The sequences that are viable as monomers are first
obtained by the procedure described above. In addition,
the two proteins, because of their square-lattice structure,
can form ten homodimers AA, BB and 16 heterodimers
AB, just one of which is functional. Inter-protein interac-
tions are described by Eq. 3. In addition, dimerization is
opposed by a constant entropic penalty, ε. We consider
here only sequences that are known to form viable mon-
omers (see above), so that their unfolded conformations
are unstable and can be neglected. We only consider asso-
ciation between the two proteins A and B; association of
A with a third protein C, corresponding to a different 2D
structure, is largely neglected. We can always view the
pool of other proteins C, D, ... as the source of a compet-
ing, background interaction. Our model can incorporate
these interactions only in an average way, by replacing the
protein pool by a single competing protein, whose
sequence is an average over all viable monomer sequences
of all structures C, D, ... In that case, the competing pro-
teins have the same effect as a modification of the total Structural models: 2D lattice model and 3D off-lattice
model 3, with Δij = 1 if they have two nonhydrogen
atoms within 4.5Å of each other, and zero otherwise. The
amino acids are divided into two classes: H = {LVIMCAST-
PGFWY} and P = {EDNQKRH}. The first, "hydrophobic"
class includes amino acids usually considered hydropho-
bic or neutral; the second class includes amino acids con-
sidered polar. Page 12 of 15
(page number not for citation purposes) Page 12 of 15
(page number not for citation purposes) http://www.biomedcentral.com/1472-6807/7/79 Page 13 of 15
(page number not for citation purposes) http://www.biomedcentral.com/1472-6807/7/79 http://www.biomedcentral.com/1472-6807/7/79 BMC Structural Biology 2007, 7:79 2. Ohta T: Near-neutrality in evolution of genes and gene regu-
lation. Proc Natl Acad Sci USA 2002, 99(25):16134-16137. (rms deviation of more than 3.5Å with respect to the start-
ing monomer conformation) are discarded. Overall, we
produced a total of 1695 decoys, of which 912 were dis-
carded because they lead to a lower interaction energy
than the native structure. This is due to the simplicity of
the energy function. We are left with 783 decoys (com-
pared to 35 non-native structures in the 2D dimer case). 3. Eigen M, McCaskill J, Schuster P: The molecular quasipecies. In
Advances in Chemical Physics Volume 75. New York: Wiley;
1989:149-263. 4. Sali A, Shakhnovich E, Karplus M: How does a protein fold? Nature
1994, 369(6477):248-251. 5. England JL, Shakhnovich BE, Shakhnovich EI: Natural selection of
more designable folds: a mechanism for thermophilic adap-
tation. Proc Natl Acad Sci USA 2003, 100(15):8727-8731. 6. Wolynes P: Energy landscapes and solved protein folding
problems. Phil Trans Roy Soc London A 2005, 363(1827):453-464. To determine the viable dimer sequences, we start from
the sequences that are viable as monomers, generated by
the Monte Carlo method described above. Because the
number of accepted monomer sequences is very high, we
only keep one sequence per hydrophobicity profile
(which is computed using the classification shown in the
Methods section), picked arbitrarily from the available
sequences. There were a total of 31,469 and 29,667 pro-
files for Grb2 and Vav, respectively. We then consider the
ability to dimerize, by comparing the energy of the native
dimer structure to the energy of all the decoy structures. For a given pair of Vav and Grb2 sequences and a given
dimer structure, the energy is obtained by threading the
sequence onto the dimer structure. Sidechains are posi-
tioned in their most common rotamer, as described above
for the monomer case. For a given sequence pair, a Z-score
is calculated for each dimeric structure. The Z-score is
defined as the energy of the structure relative to the aver-
age energy, measured in standard deviation units. A pair
of sequences is considered to form a viable dimer if the
native structure has a sufficiently low Z-score. Specifically,
its Z-score should be among the top k values, where k is an
integer between 3 and 10. http://www.biomedcentral.com/1472-6807/7/79 Choosing k = 3, for example,
means that for a sequence pair to be viable, at most two
decoys should have a lower Z-score than the native dimer
structure. By varying k, we can explore different stringen-
cies for the selection criterion. This is analogous to varying
δ in the 2D lattice case. With k = 10, there are 470,334 via-
ble pairs of HP profiles. To test the viability of the
sequences pairs required several weeks of CPU time using
ten computer processors. Once the viable sequences are
known, the steady state is computed as for the monomer
case. 7. Bornberg-Bauer E, Chan HS: Modeling evolutionary landscapes:
Mutational stability, topology, and superfunnels in sequence
space. Proc Natl Acad Sci USA 1999, 96(19):10689-10694. 8
X
Y L
M R l
f
i
d
bi
i
i
h p
(
)
8. Xia Y, Levitt M: Roles of mutation and recombination in the
evolution of protein thermodynamics. Proc Natl Acad Sci USA
2002, 99(16):10382-10387. (
)
9. Bastolla U, Porto M, Roman E, Vendruscolo M: Lack of self-averag-
ing in neutral evolution of proteins. Phys Rev Lett 2001,
89(20):208101/1-208101/4. 10. Xia Y, Levitt M: Simulating protein evolution in sequence
space and structure space. Curr Opin Struct Biol 2004, 14:202-207. 11. Kauffman SA: The origins of order. Self-organization and selection in evo-
lution New York: Oxford University Press; 1993. p
p
p
11. Kauffman SA: The origins of order. Self-organization and selection in evo-
lution New York: Oxford University Press; 1993. 12. Kimura M: Evolutionary rate at the molecular level. Nature
1968, 217(5129):624-626. (
)
13. Taverna DM, Goldstein RA: Why are proteins so robust to site
mutations? J Mol Biol 2002, 315(3):479-484. J
( )
14. Bloom JD, Labthavikul ST, Otey CR, Arnold FH: Protein stability
promotes evolvability. Proc Natl Acad Sci USA 2006,
103(15):5869-5874. (
)
15. Chan HS, Bornberg-Bauer E: Perspectives on protein evolution
from simple exact models. Applied Bioinformatics 2002,
1(3):121-144. ( )
16. Bloom JD, Wilke CO, Arnold FH, Adami C: Stability and the
evolvability of function in a model protein. Biophys J 2004,
86:2758-2764. 17. Williams PD, Pollock DD, Goldstein RA: Evolution of functional-
ity in lattice proteins. J Molec Graph Model 2001, 19:150-156. 18. Blackburne BP, Hirst JD: Evolution of functional model proteins. J Chem Phys 2001, 115(4):. J
y
( )
19. Blackburne BP, Hirst JD: Population dynamics simulations of
functional model proteins. http://www.biomedcentral.com/1472-6807/7/79 J Chem Phys 2005, 123:154907. p
J
y
20. Gavin AC, Superti-Furga G: Protein complexes and proteome
organization from yeast to man. Curr Opin Chem Biol 2003,
7:21-27. 21. Pazos F, Valencia A: Similarity of phylogenetic trees as indica-
tor of Protein-protein interaction. Protein Engineering Design and
Selection 2001, 14(9):609-614. ( )
22. Mika S, Rost B: Protein-protein interactions more conserved
within species than across species. PLOS Comp Biol 2006,
2(7):698-709. ( )
23. Albert R, Barabasi AL: Statistical mechanics of complex net-
works. Rev Mod Phys 2002, 74:47-97. y
24. Tiana G, Provasi D, Broglia RA: Role of bulk and of interface con-
tacts in the behaviour of lattice model dimeric proteins. Phys
Rev Lett 2003, 67(5):. Authors' contributions ( )
25. Buchler NEG, Goldstein RA: Effect of alphabet size and foldabil-
ity requirements on protein structure designability. Proteins
1999, 34:113-124. TS conceived of the study and its design, and participated
in the numerical implementation. JN participated in the
design and did the bulk of the programming, computa-
tions, and analysis. JN and TS drafted, wrote, and
approved the manuscript. 26. Ueda Y, Taketomi H, Go N: Studies on protein folding, unfold-
ing, and fluctuations by computer simulation. 2. 3-dimen-
sional lattice model of lysozyme. Biopolymers 1978,
17(6):1531-1548. 27. Wodak SJ, Janin J: Structural basis of macromolecular recogni-
tion. Adv Prot Chem 2002, 61:9-73. 28. Teichmann SA: The constraints Protein-protein interactions
place on sequence divergence. J Mol Biol 2002, 324(3):399-407. 1.
Kimura M: The neutral theory of molecular evolution Cambridge Univer-
sity Press; 1983. Acknowledgements g
The authors are grateful to the anonymous reviewers for their remarks. Support was provided by the French research program IMPBio. p
q
g
J
( )
29. Xia Y, Levitt M: Funnel-like organization in sequence space
determines the distributions of protein stability and folding
rate preferred by evolution. Proteins 2004, 55:107-114. 30. Li H, Tang C, Wingreen NS: Designability of protein structures:
A lattice-model study using the Miyazawa-Jernigan matrix. Proteins 2002, 49(3):403-412. Page 14 of 15
(page number not for citation purposes) Interacting genes: the 3D off-lattice case g g
We now turn to the 3D, off-lattice case. We consider the
Grb2-Vav complex [43] which plays a role in tissue spe-
cific signaling in the hematopoietic lineage [44]. In con-
trast to the lattice case above, there are far too many
possible dimer structures for an exact enumeration to be
done. Instead, we consider a limited set of dimeric decoy
structures. These were generated by a docking procedure
described in detail elsewhere [40]. Briefly, we start from
the two separate proteins, with their native sequences,
positioned randomly with respect to each other. They are
then docked together with a molecular mechanics energy
function [45], using restrained energy minimization. In
an initial phase, the restraint consists in a harmonic spring
that pulls their centers of mass together. In a second
phase, the restraint corresponds to an electrostatic con-
trast introduced artificially between the two proteins:
charges on one are slightly increased; charges on the other
are slightly decreased. The pair is energy-minimized,
allowing for limited intra-protein deformations. Struc-
tures that involve too large a deformation of either partner Page 13 of 15
(page number not for citation purposes) Page 13 of 15
(page number not for citation purposes) http://www.biomedcentral.com/1472-6807/7/79 BMC Structural Biology 2007, 7:79 http://www.biomedcentral.com/1472-6807/7/79 http://www.biomedcentral.com/1472-6807/7/79 http://www.biomedcentral.com/1472-6807/7/79 31. Wroe R, Bornberg-Bauer E, Chan HS: Comparing Folding Codes
in Simple Heteropolymer Models of Protein Evolutionary
Landscape: Robustness of the Superfunnel Paradigm. Biophys
J 2005, 88:118-131. J
32. Bornberg-Bauer E: How are model protein structures distrib-
uted in sequence space? Biophys J 1997, 73(5):2393-2403. q
p
p y J
( )
33. Vendruscolo M, Domany E: Pairwise contact potentials are
unsuitable for protein folding. J Chem Phys 1998,
109:11101-11108. 34. Bloom J, Silberg J, Wilke C, Drummond D, Adami C, Arnold F: Ther-
modynamic prediction of protein neutrality. Proc Natl Acad Sci
USA 2005, 102(3):606-611. ( )
35. Guo H, Choe J, Loeb L: Protein tolerance to random amino
acid change. Proc Natl Acad Sci USA 2004, 101(25):9205-9210. 36. van Nimwegen E, Crutchfield JP, Huygen M: Neutral evolution of
mutational robustness. Proc Natl Acad Sci USA 1999,
96(17):9716-9720. (
)
37. Elena SF, Cooper V, Lenski R: Punctuated evolution caused by
selection of rare beneficial mutations. Science 1996,
272(5269):1802-1804. (
)
38. Li H, Helling R, Tang C, Wingreen NS: Why do proteins look like
proteins? Science 1996, 273(5275):666-669. p
(
)
39. Bastolla U, Farwer J, Knapp E, Vendruscolo M: How to guarantee
optimal stability for the most representative structures in
the Protein Data Bank. Proteins 2001, 44(2):79-96. ( )
40. Launay G, Mendez R, Wodak S, Simonson T: Recognizing protein-
protein interfaces with reduced amino acid alphabets. BMC
Bioinformatics 2007 in press. p
41. Tuffery P, Etchebest C, Hazout S, Lavery R: A new approach to the
rapid determination of protein side chain conformations. J
Biomol Struct Dyn 1991, 8(6):1267-1289. y
( )
42. Press W, Flannery B, Teukolsky S, Vetterling W: Numerical Recipes
Cambridge University Press, Cambridge; 1986. y
( )
42. Press W, Flannery B, Teukolsky S, Vetterling W 42. Press W, Flannery B, Teukolsky S, Vetterling W: Nu
Cambridge University Press, Cambridge; 1986. Cambridge University Press, Cambridge; 1986. 43. Nishida M, Nagata K, Hachimori Y, Horiuchi M, Ogura K, Mandiyan
V, Schlessinger J, Inagaki F: Novel recognition mode between
Vav and Grb2 SH3 domains. EMBO Journal 2001, 20:2995-3007. J
44. Ye Z, Baltimore D: Binding of Vav to Grb2 Through Dimeriza-
tion of Src Homology 3 Domains. Proc Natl Acad Sci USA 1994,
91(26):12629-12633. 45. Brooks B, Bruccoleri R, Olafson B, States D, Swaminathan S, Karplus
M: Charmm: a program for macromolecular energy, minimi-
zation, and molecular dynamics calculations. References 1. Kimura M: The neutral theory of molecular evolution Cambridge Univer-
sity Press; 1983. Page 14 of 15
(page number not for citation purposes) Page 14 of 15
(page number not for citation purposes) BMC Structural Biology 2007, 7:79 BMC Structural Biology 2007, 7:79 J Comp Chem
1983, 4:187-217. Publish with BioMed Central and every
scientist can read your work free of charge
"BioMed Central will be the most significant development for
disseminating the results of biomedical research in our lifetime."
Sir Paul Nurse, Cancer Research UK
Your research papers will be:
available free of charge to the entire biomedical community
peer reviewed and published immediately upon acceptance
cited in PubMed and archived on PubMed Central
yours — you keep the copyright
Submit your manuscript here:
http://www.biomedcentral.com/info/publishing_adv.asp
BioMedcentral Publish with BioMed Central and every
scientist can read your work free of charge
"BioMed Central will be the most significant development for
disseminating the results of biomedical research in our lifetime."
Sir Paul Nurse, Cancer Research UK
Your research papers will be:
available free of charge to the entire biomedical community
peer reviewed and published immediately upon acceptance
cited in PubMed and archived on PubMed Central
yours — you keep the copyright
Submit your manuscript here:
http://www.biomedcentral.com/info/publishing_adv.asp
BioMedcentral Publish with BioMed Central and every
scientist can read your work free of charge Page 15 of 15
(page number not for citation purposes)
|
https://openalex.org/W2253961736
|
https://journals.plos.org/plosone/article/file?id=10.1371/journal.pone.0148642&type=printable
|
English
| null |
Evaluating Cell Processes, Quality, and Biomarkers in Pluripotent Stem Cells Using Video Bioinformatics
|
PloS one
| 2,016
|
cc-by
| 13,423
|
RESEARCH ARTICLE OPEN ACCESS OPEN ACCESS
Citation: Zahedi A, On V, Lin SC, Bays BC, Omaiye
E, Bhanu B, et al. (2016) Evaluating Cell Processes,
Quality, and Biomarkers in Pluripotent Stem Cells
Using Video Bioinformatics. PLoS ONE 11(2):
e0148642. doi:10.1371/journal.pone.0148642
Editor: Johnson Rajasingh, University of Kansas
Medical Center, UNITED STATES
Received: July 15, 2015
Accepted: January 20, 2016
Published: February 5, 2016
Copyright: © 2016 Zahedi et al. This is an open
access article distributed under the terms of the
Creative Commons Attribution License, which permits
unrestricted use, distribution, and reproduction in any
medium, provided the original author and source are
credited. Abstract Citation: Zahedi A, On V, Lin SC, Bays BC, Omaiye
E, Bhanu B, et al. (2016) Evaluating Cell Processes,
Quality, and Biomarkers in Pluripotent Stem Cells
Using Video Bioinformatics. PLoS ONE 11(2):
e0148642. doi:10.1371/journal.pone.0148642 There is a foundational need for quality control tools in stem cell laboratories engaged in
basic research, regenerative therapies, and toxicological studies. These tools require auto-
mated methods for evaluating cell processes and quality during in vitro passaging, expan-
sion, maintenance, and differentiation. In this paper, an unbiased, automated high-content
profiling toolkit, StemCellQC, is presented that non-invasively extracts information on cell
quality and cellular processes from time-lapse phase-contrast videos. Twenty four (24) mor-
phological and dynamic features were analyzed in healthy, unhealthy, and dying human
embryonic stem cell (hESC) colonies to identify those features that were affected in each
group. Multiple features differed in the healthy versus unhealthy/dying groups, and these fea-
tures were linked to growth, motility, and death. Biomarkers were discovered that predicted
cell processes before they were detectable by manual observation. StemCellQC distin-
guished healthy and unhealthy/dying hESC colonies with 96% accuracy by non-invasively
measuring and tracking dynamic and morphological features over 48 hours. Changes in cel-
lular processes can be monitored by StemCellQC and predictions can be made about the
quality of pluripotent stem cell colonies. This toolkit reduced the time and resources required
to track multiple pluripotent stem cell colonies and eliminated handling errors and false clas-
sifications due to human bias. StemCellQC provided both user-specified and classifier-deter-
mined analysis in cases where the affected features are not intuitive or anticipated. Video
analysis algorithms allowed assessment of biological phenomena using automatic detection
analysis, which can aid facilities where maintaining stem cell quality and/or monitoring
changes in cellular processes are essential. In the future StemCellQC can be expanded to
include other features, cell types, treatments, and differentiating cells. Editor: Johnson Rajasingh, University of Kansas
Medical Center, UNITED STATES Editor: Johnson Rajasingh, University of Kansas
Medical Center, UNITED STATES
Received: July 15, 2015
Accepted: January 20, 2016
Published: February 5, 2016 Editor: Johnson Rajasingh, University of Kansas
Medical Center, UNITED STATES
Received: July 15, 2015
Accepted: January 20, 2016
Published: February 5, 2016 Copyright: © 2016 Zahedi et al. This is an open
access article distributed under the terms of the
Creative Commons Attribution License, which permits
unrestricted use, distribution, and reproduction in any
medium, provided the original author and source are
credited. Evaluating Cell Processes, Quality, and
Biomarkers in Pluripotent Stem Cells Using
Video Bioinformatics Atena Zahedi1,4,5☯‡, Vincent On2,3☯‡, Sabrina C. Lin4,5, Brett C. Bays6, Esther Omaiye4,5,
Bir Bhanu1,2,3, Prue Talbot4,5* Atena Zahedi1,4,5☯‡, Vincent On2,3☯‡, Sabrina C. Lin4,5, Brett C. Bays6, Esther Omaiye4,5,
Bir Bhanu1,2,3, Prue Talbot4,5* 1 Bioengineering Department, University of California, Riverside, California, United States of America,
2 Electrical and Computer Engineering Department, University of California, Riverside, California, United
States of America, 3 Center for Research in Intelligent Systems, University of California, Riverside,
California, United States of America, 4 Department of Cell Biology and Neuroscience, University of
California, Riverside, California, United States of America, 5 UCR Stem Cell Center and Core, University of
California, Riverside, California, United States of America, 6 Psychology Department, University of
California, Riverside, California, United States of America ☯These authors contributed equally to this work. ‡ These authors are co-first authors on this work. * talbot@ucr.edu OPEN ACCESS
Citation: Zahedi A, On V, Lin SC, Bays BC, Omaiye
E, Bhanu B, et al. (2016) Evaluating Cell Processes,
Quality, and Biomarkers in Pluripotent Stem Cells
Using Video Bioinformatics. PLoS ONE 11(2):
e0148642. doi:10.1371/journal.pone.0148642
Editor: Johnson Rajasingh, University of Kansas
Medical Center, UNITED STATES
Received: July 15, 2015
Accepted: January 20, 2016
Published: February 5, 2016
Copyright: © 2016 Zahedi et al. This is an open
access article distributed under the terms of the
Creative Commons Attribution License, which permits
unrestricted use, distribution, and reproduction in any
medium, provided the original author and source are
credited. Introduction Competing Interests: The authors have declared
that no competing interests exist. Human pluripotent stem cells (hPSC) have enormous potential for enhancing our understand-
ing of human prenatal development, modeling diseases-in-a-dish, treating patients with degen-
erative diseases, and evaluating the effects of drugs and environmental chemicals on cells that
model human embryos and fetuses [1–3]. In each of these applications, there is a foundational
unmet need for technology to non-invasively monitor the quality of hPSC during passaging,
expansion, growth, experimentation, and differentiation [4, 5]. Ideally such tools should be
rapid, non-invasive, resource saving, and non-biased. Video bioinformatics, which involves
mining data from video images using algorithms that speed analysis and eliminate human bias,
offers a solution to this problem and can be used to produce high quality software for stem cell
applications [6–13]. Prior applications of video bioinformatics tools have successfully identified pluripotent
stem cell colonies based on colony morphology [14], thereby speeding induced pluripotent
stem cell (iPSC) derivation and reducing cost. Another study applied image processing soft-
ware to fluorescent videos to identify iPSC after reprogramming [15], and a video bioinformat-
ics method was developed to identify in vitro fertilized human embryos that will progress to
blastocysts by 2 days after fertilization [16]. A recent report used phase-contrast video segmen-
tation to generate lineage trees of neural stem cells using cell location, shape, movement, and
size [17]. We previously developed in-house video segmentation tools to analyze single hESC
and small colonies [8–11]. In a toxicological application of video bioinformatics using
CL-Quant software [18], cigarette smoke treatment altered hESC colony growth (area) and
health [7]. While the above studies looked at single endpoints, multiple features related to cel-
lular processes and health can be extracted from video data thereby enhancing the depth of
analysis and providing data on the kinetics of each endpoint. However, no software currently
exists for automatic detection of pluripotent stem cell processes and quality in culture. It is highly desirable to be able to multiplex multiple endpoints from a single experiment. The purpose of this study was to develop a high-content profiling software platform, Stem-
CellQC, to automatically identify cell processes affected by culture/treatment and to classify
the health of individual hESC colonies based on features extracted from phase contrast micro-
scope video data. Introduction The method automatically segments the input colonies (non-labeled phase
contrast images), extracts relevant novel features for each colony, utilizes the changes in fea-
tures over time to identify cell processes that are affected by treatment, and statistically classi-
fies healthy and unhealthy/dying colonies. StemCellQC’s feature analysis and classification
system provide an effective method to evaluate pluripotent stem cell colony processes and qual-
ity before use in experiments or clinical applications. Because hESC model the epiblast cells of
embryos [19], which if harmed can lead to embryonic death or development of congenital
defect(s), StemCellQC has the potential to be used as a novel technology to identify toxicants
or drugs that could affect cellular processes in young embryos. Overall Design of the System The overall design of the system will be discussed in Results (S1 Fig). Each component is pre-
sented in the following Methods section. Evaluating Stem Cells with Video Bioinformatics Abstract Data Availability Statement: All relevant data are
within the paper and its Supporting Information files. Funding: This work was supported by a NSF IGERT
grant in Video Bioinformatics (DGE 093667) to BB,
the California Institute for Regenerative Medicine
(#NE-A0005A-1E) to PT, and grants from the
Tobacco-Related Disease Research Program of CA
(# 22RT-0127 and #20PT-0184) to PT. The funders
had no role in study design, data collection and
analysis, decision to publish, or preparation of the
manuscript. 1 / 22 PLOS ONE | DOI:10.1371/journal.pone.0148642
February 5, 2016 Development and Use of StemCellQC Software StemCellQC was written and developed with MATLAB 2015a programming environment. The MATLAB source code, a stand-alone executable version of this algorithm, and supplied
test data are available online at http://vislab.ucr.edu/SOFTWARE/software.php. Scqc_multi.m
is the main program of the code and requires the following MATLAB toolboxes: Statistics and
Machine Learning, Bioinformatics, System Identification, Image Processing, and Model-Based
Calibration. The standalone alone executable requires the installation of the 64-bit version of
MATLAB Runtime R2015a (8.5) available at http://www.mathworks.com/products/compiler/
mcr/. Categorization of hESC as Healthy, Unhealthy or Dying Before the StemCellQC software is run, a training dataset was collected. Categorization of colo-
nies as healthy, unhealthy, or dying by the end of 48 hours was provided by experts in culturing
hESC and was based on biological observations listed in a decision tree (S2 Fig). Categorization
was used to validate the program’s predictions of colony health. Culture and Collection of hESC Videos H9 hESC, purchased from WiCell Stem Cell Institute (Madison, Wisconsin), were maintained
on Matrigel coated 6-well plates in mTeSR complete medium (Stem Cell Technologies, 2 / 22 PLOS ONE | DOI:10.1371/journal.pone.0148642
February 5, 2016 Evaluating Stem Cells with Video Bioinformatics Vancouver, Canada) in a 37°C incubator with 90% humidity and 5% CO2 [7,20]. When cul-
tures reached 80–85% confluency, they were detached using Accutase (eBiosciences, San
Diego, CA) for 1 minute and used in experiments. Vancouver, Canada) in a 37°C incubator with 90% humidity and 5% CO2 [7,20]. When cul-
tures reached 80–85% confluency, they were detached using Accutase (eBiosciences, San
Diego, CA) for 1 minute and used in experiments. For live cell imaging in the BioStation CT (Nikon Instruments, Melville NY), hESC colonies
were plated at 25–30% confluency and allowed to attach for 24 hours. To create groups of colo-
nies that were healthy, unhealthy and dying, hESC were treated with Marlboro Red cigarette
smoke solutions as described previously [7,21]. Sidestream smoke (SS) was used at a dose of
0.1 puff equivalent (PE), where 1 PE is the amount of smoke that dissolves in 1 ml of medium
in 60 sec. The 0.1PE concentration of SS smoke has an estimated concentration of nicotine of
0.2 μg/ml [22,23], which is within the estimated tissue range of passive smokers [21]. All imaging was done using a 10x phase contrast objective in the BioStation CT using auto-
matic Z-focus; cells were not stained, labeled, or genetically modified thereby permitting non-
invasive analysis of cells. StemCellQC was tested on different magnifications (4x and 20x), and
performs well. The dataset for feature analysis was made up of 34 videos of individual hESC
colonies. 23 colonies were treated for 48 hours with sidestream cigarette smoke, while 11 con-
trol colonies were incubated in culture medium only. PLOS ONE | DOI:10.1371/journal.pone.0148642
February 5, 2016 Feature Extraction Features based on appearance, morphology, and dynamics were extracted from segmented col-
onies. Dynamic features were obtained from morphological features by computing the rate of
change of a feature over time. For example, to segment the protrusions extending from hESC,
the main body of the colony (obtained using a morphological open algorithmic operation) was
subtracted from the total colony segmentation. Also, the bright-to-total area ratio is the num-
ber of bright pixels divided by the area. Bright pixels were found by first computing a histogram
of the intensities in the segmented colony to acquire the mean (μ) and standard deviation (σ). Then, a range of intensities (lower bound threshold = μ + 3 σ, upper bound threshold = μ + 6
σ) that best described dead cells in colonies were taken as the bright pixels. A complete list of
features and their definitions is provided in S9 Fig. Validation of Segmentation The accuracy of segmentation was determined by manually segmenting hESC colonies using
ImageJ and comparing the area and perimeter values to those obtained from the automatic seg-
mentation (S3 Fig). Evaluating Stem Cells with Video Bioinformatics smoother than the edge-based method. The main purpose of using solidity was to detect dead
cells that were extruded from the stem cell colonies. Edge-based segmentation provides tight
edge boundaries, which does not include the dead cells that are in the process of being expelled
from the colony; whereas, Otsu’s “larger” segmentation includes the dead cells. The concave
regions of the segmentation that are produced by the dead cells affect the sensitivity of solidity. Therefore, solidity changes (primarily due to dead cells) are more pronounced when using
Otsu’s method. This larger boundary results in a larger convex hull which increases the sensi-
tivity of the solidity feature. This allowed for better distinction between the peaks and valleys in
the solidity plot. For the region-based method, initially frames were smoothed to remove a small amount of
noise using a 3x3 Gaussian filter. Next, Ostu’s thresholding-based method was used to separate
the pixels into the background and foreground by finding the optimal threshold for segmenting
an image [25,26]. Connected components were then found in the binary image. A morphologi-
cal open operation (used to open gaps between loosely connected objects) was performed on
the binary image to disconnect loosely connected pixels in the foreground. This is carried out
by first eroding an object of interest (a connected component) and then dilating the output
with a structuring element. To disconnect the objects, a circle with a radius of 12 pixels worked
the best, and it was kept fixed for all the experiments. For both erosion and dilation, every pixel
in the object was individually probed by the structuring element. The end result was a set of fil-
tered connected components. Any holes in this region were filled, and the final ROIs were used
to extract features. Video Segmentation An edge-based method and a region-based method were used to segment colonies in video
images (frames). The edge-based method convolves the Sobel edge operator with the image to
produce a gradient magnitude image and a gradient direction image. All gradient magnitudes
below a calculated threshold were ignored and the rest were used as edges. This calculated
threshold was automatically computed by the “edge” function in the Image Processing
Toolbox from MATLAB. The edges in the image were dilated using two line structuring ele-
ments (vertical and horizontal) of three pixel lengths to merge connected regions of the colony. Connected components were then filled and smoothed with image erosion using a diamond
structuring element of one pixel radius to produce the segmentations. Segmented objects that
are smaller than a user specified threshold, 3000 pixels in our case, were removed [24]. The Otsu’s region-based method for segmentation [25,26], which was used to compute the
solidity feature, is the ratio of the colony area divided by the area of the convex hull. The con-
vex hull can be visualized as the shape enclosed by a rubber band stretched around a region-of-
interest (ROI) [27]. Otsu’s method produces a slightly larger segmented boundary which is 3 / 22 PLOS ONE | DOI:10.1371/journal.pone.0148642
February 5, 2016 Identification of Key Features Both biologically-based feature selection and statistical-based feature selection were used to
identify those features that provided information on affected cellular processes and to distin-
guish healthy, unhealthy, and dying colonies. In addition, all the features were exhaustively
applied to the classifier. For biologically-based feature selection, plot observations over time can be interpreted by
the user. A subset of features are shown as graphical plots in Figs 1–4. From the feature plots,
the user can observe the non-overlapping standard error of the means (SEMs) to get a good
indication of groups that are significantly different from each other. For a more rigorous 4 / 22 PLOS ONE | DOI:10.1371/journal.pone.0148642
February 5, 2016 Evaluating Stem Cells with Video Bioinformatics Fig 1. Features related to hESC colony growth. The outline of segmentation for a healthy (A), unhealthy (B), and a dying colony (C) at the last recorded
frame. (D) Area normalized to the first time point for colonies that were healthy, unhealthy, and dying. Colonies first became significantly different by 2-way
ANOVA at 37.6 hours for healthy vs unhealthy (green arrow), at 33.5 hours for unhealthy versus dying (blue arrow), and at 26.2 hours for healthy versus
dying colonies (red arrow). (E) Change in area over time showing second contraction of dying colonies at 30–32 hours. (F) Perimeter over time normalized to
the first time point for colonies that were healthy, unhealthy, and dying. Colonies first became significantly different by 2-way ANOVA at 46.6 hours for healthy
versus unhealthy groups (green arrow), at 35.4 hours for unhealthy versus dying groups (blue arrow), and at 28.9 hours for healthy versus dying groups (red
arrow). (G) Minor axis normalized to the first time point for colonies that were healthy, unhealthy, and dying. Colonies first became significantly different by
2-way ANOVA at 44.3 hour for healthy versus unhealthy groups (green arrow), at 36.5 hour for unhealthy versus dying groups (blue arrow), and at 28.9 hour
for healthy versus dying group (red arrow). Number of colonies per group = 16 healthy, 12 unhealthy, 6 dying. Data are plotted as means ± SEM for each
group. Arrows indicate first values that differed significantly from the control by 2-way ANOVA (* = P < 0.05). d i 10 1371/j
l
0148642 001 Fig 1. Features related to hESC colony growth. Identification of Key Features The outline of segmentation for a healthy (A), unhealthy (B), and a dying colony (C) at the last recorded
frame. (D) Area normalized to the first time point for colonies that were healthy, unhealthy, and dying. Colonies first became significantly different by 2-way
ANOVA at 37.6 hours for healthy vs unhealthy (green arrow), at 33.5 hours for unhealthy versus dying (blue arrow), and at 26.2 hours for healthy versus
dying colonies (red arrow). (E) Change in area over time showing second contraction of dying colonies at 30–32 hours. (F) Perimeter over time normalized to
the first time point for colonies that were healthy, unhealthy, and dying. Colonies first became significantly different by 2-way ANOVA at 46.6 hours for healthy
versus unhealthy groups (green arrow), at 35.4 hours for unhealthy versus dying groups (blue arrow), and at 28.9 hours for healthy versus dying groups (red
arrow). (G) Minor axis normalized to the first time point for colonies that were healthy, unhealthy, and dying. Colonies first became significantly different by
2-way ANOVA at 44.3 hour for healthy versus unhealthy groups (green arrow), at 36.5 hour for unhealthy versus dying groups (blue arrow), and at 28.9 hour
for healthy versus dying group (red arrow). Number of colonies per group = 16 healthy, 12 unhealthy, 6 dying. Data are plotted as means ± SEM for each
group. Arrows indicate first values that differed significantly from the control by 2-way ANOVA (* = P < 0.05). doi:10.1371/journal.pone.0148642.g001 PLOS ONE | DOI:10.1371/journal.pone.0148642
February 5, 2016 5 / 22 Evaluating Stem Cells with Video Bioinformatics Fig 2. Surface protrusions on colonies can be used to study cell morphology and growth. (A) Segmentation of protrusions (red outline) for a healthy
colony (A), unhealthy colony (B), and dying colony (C) at the last recorded time frame. (D) Number of protrusions over time normalized to the initial time point
for healthy, unhealthy, and dying colonies. Colonies first became significantly different by 2-way ANOVA at 43 hours for healthy versus unhealthy groups
(green arrow), at 33.3 hours for unhealthy versus dying groups (blue arrow), and at 27.1 hours for healthy versus dying group (red arrow). (E) Protruding-to-
total area ratio for healthy, unhealthy, and dying colonies. Colonies first became significantly different by 2-way ANOVA at 6.8 hours for unhealthy versus
dying groups (blue arrow), and at 4.8 hours for healthy versus dying groups (red arrow). Identification of Key Features Number of colonies per group = 16 healthy, 12 unhealthy, 6 dying. Data are plotted as means ± SEM for each group. Arrows indicate first values that differed significant from the control by 2-way ANOVA (* = P < 0.05). doi:10.1371/journal.pone.0148642.g002 Fig 2. Surface protrusions on colonies can be used to study cell morphology and growth. (A) Segmentation of protrusions (red outline) for a healthy
colony (A), unhealthy colony (B), and dying colony (C) at the last recorded time frame. (D) Number of protrusions over time normalized to the initial time point
for healthy, unhealthy, and dying colonies. Colonies first became significantly different by 2-way ANOVA at 43 hours for healthy versus unhealthy groups
(green arrow), at 33.3 hours for unhealthy versus dying groups (blue arrow), and at 27.1 hours for healthy versus dying group (red arrow). (E) Protruding-to-
total area ratio for healthy, unhealthy, and dying colonies. Colonies first became significantly different by 2-way ANOVA at 6.8 hours for unhealthy versus
dying groups (blue arrow), and at 4.8 hours for healthy versus dying groups (red arrow). Number of colonies per group = 16 healthy, 12 unhealthy, 6 dying. Data are plotted as means ± SEM for each group. Arrows indicate first values that differed significant from the control by 2-way ANOVA (* = P < 0.05). doi:10.1371/journal.pone.0148642.g002 statistical analysis, two-way ANOVAs with the Bonferroni post-test were performed to identify
those features that were significantly different in the plotted data. For the solidity feature, a
one-tailed independent samples t-test was used to determine if significant differences existed
between means of healthy/unhealthy versus /dying colonies at 12 hours. Statistical-based methods are useful in cases where the graphs for features may not reveal
obvious effects, and they are good starting points to identify combinations or subsets of useful
features. Filter methods which select variables regardless of the classification model are prefera-
ble for StemCellQC because of the use of multiple classifiers. 11 feature selection algorithms
(10 methods from the Feature Selection @ Arizona State University toolbox [28]) and 6 / 22 PLOS ONE | DOI:10.1371/journal.pone.0148642
February 5, 2016 Evaluating Stem Cells with Video Bioinformatics Fig 3. Features related to hESC colony motility. Extracted contour of a healthy colony at 16 hours (yellow line) and 24 (green line) hrs. The distance
between the centroids is indicated by the white line. (B) Change in centroid over time for healthy, unhealthy, and dying colonies. Identification of Key Features Colonies first became
significantly different by 2-way ANOVA at 7.7 hours for healthy versus dying groups (red arrow), and at 20.6 hours for unhealthy versus dying groups (blue
arrow). (C) The total displacement for healthy, unhealthy, and dying colonies. (D) The total distance traveled for healthy, unhealthy, and dying colonies. Colonies first became significantly different by 2-way ANOVA at 34.7 hours for healthy versus unhealthy groups (green arrow). (E) The mean squared
displacement (MSD) for healthy, unhealthy, and dying colonies. (F-G) Localization of actin microfilaments in a healthy (F) and an unhealthy (G) colony which
had fewer microfilaments than the untreated control. Number of colonies per group = 16 healthy, 12 unhealthy, 6 dying. Data are plotted as means ± SEM for
each group. Arrows indicate first values that differed significant from the control by 2-way ANOVA (* = P < 0.05). doi:10.1371/journal.pone.0148642.g003 Fig 3. Features related to hESC colony motility. Extracted contour of a healthy colony at 16 hours (yellow line) and 24 (green line) hrs. The distance
between the centroids is indicated by the white line. (B) Change in centroid over time for healthy, unhealthy, and dying colonies. Colonies first became
significantly different by 2-way ANOVA at 7.7 hours for healthy versus dying groups (red arrow), and at 20.6 hours for unhealthy versus dying groups (blue
arrow). (C) The total displacement for healthy, unhealthy, and dying colonies. (D) The total distance traveled for healthy, unhealthy, and dying colonies. Colonies first became significantly different by 2-way ANOVA at 34.7 hours for healthy versus unhealthy groups (green arrow). (E) The mean squared
displacement (MSD) for healthy, unhealthy, and dying colonies. (F-G) Localization of actin microfilaments in a healthy (F) and an unhealthy (G) colony which
had fewer microfilaments than the untreated control. Number of colonies per group = 16 healthy, 12 unhealthy, 6 dying. Data are plotted as means ± SEM for
each group. Arrows indicate first values that differed significant from the control by 2-way ANOVA (* = P < 0.05). doi:10.1371/journal.pone.0148642.g003 doi:10.1371/journal.pone.0148642.g003 doi:10.1371/journal.pone.0148642.g003 PLOS ONE | DOI:10.1371/journal.pone.0148642
February 5, 2016 7 / 22 Evaluating Stem Cells with Video Bioinformatics Fig 4. Features related to cell death. (A-D) Frames representing the beginning of the video (A), the highes
and lowest solidity values respectively (B and C), and the time of death of dying colonies (D). E) Solidity
values over time for healthy/unhealthy (blue) versus dying colonies (red). Correlation of Key Features to Cellular Processes Key features were interpreted to identify cellular processes such as growth, motility, and apo-
ptosis, which differed in the healthy, unhealthy and dying groups. Also, two-feature plots were
constructed to detect correlations between features and identify temporal patterns over time
(Fig 5, S1 and S2 Videos). Identification of Key Features Colonies first became significantly
different by 2-way ANOVA at 31.5 hour healthy versus unhealthy groups (green arrow), and at 24 hours for
healthy versus dying groups (red arrow). (K-L) A healthy (K) and an unhealthy (L) colony incubated with
Magic Red to identify activated caspases 3&7. Number of colonies per group = 16 healthy, 12 unhealthy, 6
dying. Data are plotted as means ± SEM for each group. Arrows indicate first values that differed significant
from the control by 2-way ANOVA (* = P < 0.05). doi:10.1371/journal.pone.0148642.g004 quadratic programming feature selection [29] were run on our dataset. These methods include
Correlation-based Feature Selection (CFS) [30], Chi Square (Chi2) [31], Fast Correlation-
based Filter (FCBF) [32], Fisher 9 [33], Gini Index 16 [34], InfoGain 6 [35], Sparse Multinomial
Logistic Regression (SBMLR 3) [36], t-test [37], Kruskal Wallis [38], and Minimal-Redun-
dancy-Maximal-Relevance [39]. The inputs for the feature selection algorithms are the average
slope of each feature. The slope (incremental difference) is computed for each pair of adjacent
frames for individual features. Next, the mean of these slopes is calculated for all 24 features. These features individually or in groups can be used to train the classifier. Classification as Healthy or Unhealthy/Dying To automatically classify the dataset, all features measurements were first normalized with
maximum-minimum normalization. 410 frames (collected over a 48 hour period) multiplied
by 24 features results in a 9840 feature space. Therefore, in order to reduce the number of
dimensions, the input value to the classifier was the mean slope of the data. Classes used by the
classifier were healthy and unhealthy (the unhealthy group included dying colonies which were
not classified separately since this group contained only 6 colonies). To test the ability of the features to distinguish healthy versus unhealthy/dying colonies, sev-
eral biologically selected features and additional features selected by 11 statistical methods were
used to train three classifiers: (1) support vector machines (SVM), (2) K-nearest neighbor
(KNN), and (3) naïve Bayes [33]. SVM uses the training data to create a boundary in multi-
dimensional space, which can be used to classify future data samples. KNN takes a test sample
and compares it to the K-nearest training samples in a multi-dimensional space. The KNN
algorithm was used with k = 3 (the 3 closest neighbors to the sample). A majority vote is taken
by these neighbors and used as a label for the test sample. Naïve Bayes is a probabilistic classi-
fier based on Bayes’ theorem that uses strong assumptions that features are independent from
one another. An exhaustive test using all possible combinations of features was performed to
determine the best classification results using combination of features. A summary of the classification results using single features, combinations of features, and
statistically determined features are shown in Tables 1–4 in the Results section. Identification of Key Features Colonies that eventually died are
distinguished by a large peak in solidity between 8–24 hours. A one-tailed independent sample t-test at 12
hours revealed that the two groups were significantly different (P = 0.0285). (F-H) White regions on top of
hESC colonies (outlined in red) represent dead cells, shown at the end of recording for a healthy colony (F),
unhealthy colony (G), and dying colony (H). (I) Bright-to-total area ratio over time for healthy, unhealthy, and
dying colonies. Colonies first became significantly different by 2-way ANOVA at 4 hours for healthy versus
unhealthy groups (green arrow), and at 11.5 hour for the healthy versus dying groups (red arrow). (J) Fig 4. Features related to cell death. (A-D) Frames representing the beginning of the video (A), the highest
and lowest solidity values respectively (B and C), and the time of death of dying colonies (D). E) Solidity
values over time for healthy/unhealthy (blue) versus dying colonies (red). Colonies that eventually died are
distinguished by a large peak in solidity between 8–24 hours. A one-tailed independent sample t-test at 12
hours revealed that the two groups were significantly different (P = 0.0285). (F-H) White regions on top of
hESC colonies (outlined in red) represent dead cells, shown at the end of recording for a healthy colony (F),
unhealthy colony (G), and dying colony (H). (I) Bright-to-total area ratio over time for healthy, unhealthy, and
dying colonies. Colonies first became significantly different by 2-way ANOVA at 4 hours for healthy versus
unhealthy groups (green arrow), and at 11.5 hour for the healthy versus dying groups (red arrow). (J) 8 / 22 PLOS ONE | DOI:10.1371/journal.pone.0148642
February 5, 2016 Evaluating Stem Cells with Video Bioinformatics Minimum intensity values for healthy, unhealthy, and dying colonies. Colonies first became significantly
different by 2-way ANOVA at 31.5 hour healthy versus unhealthy groups (green arrow), and at 24 hours for
healthy versus dying groups (red arrow). (K-L) A healthy (K) and an unhealthy (L) colony incubated with
Magic Red to identify activated caspases 3&7. Number of colonies per group = 16 healthy, 12 unhealthy, 6
dying. Data are plotted as means ± SEM for each group. Arrows indicate first values that differed significant
from the control by 2-way ANOVA (* = P < 0.05). Minimum intensity values for healthy, unhealthy, and dying colonies. Fig 5. Two-Feature Plot Analysis. (A) Average intensity compared to perimeter running plot shown at approximately at 16 hours of incubation for all
individual healthy (green), unhealthy (blue), and dying (red) hESC colonies. (B) Mean-squared displacement compared to area running plot shown at
approximately 16 hours for all individual healthy (green), unhealthy (blue), and dying (red) hESC colonies. (C) User derived equation (perimeter divided by
number of protrusions) plotted for healthy, unhealthy, and dying colonies. doi:10.1371/journal.pone.0148642.g005 test data once, while the other nine parts were used as training data. The partitions were ran-
domized, and this process was repeated with 5 random permutations of the data. A percentage
of correctly classified samples was calculated for each permutation by comparison to the man-
ual labels. The classification results were then used to find the mean and standard deviation. Classification Validation The classification experiments were run with 10-fold cross validation where the dataset was
partitioned into 10 parts. The 34 videos gave six partitions containing three videos each and
four partitions containing four videos each. A 10-fold partition of the training data was used,
which allowed for computation of a standard deviation of the results. One part was used as the 9 / 22 PLOS ONE | DOI:10.1371/journal.pone.0148642
February 5, 2016 Evaluating Stem Cells with Video Bioinformatics Fig 5. Two-Feature Plot Analysis. (A) Average intensity compared to perimeter running plot shown at approximately at 16 hours of incubation for all
individual healthy (green), unhealthy (blue), and dying (red) hESC colonies. (B) Mean-squared displacement compared to area running plot shown at
approximately 16 hours for all individual healthy (green), unhealthy (blue), and dying (red) hESC colonies. (C) User derived equation (perimeter divided by
number of protrusions) plotted for healthy, unhealthy, and dying colonies. doi:10.1371/journal.pone.0148642.g005 Molecular Validation Apoptotic activity was detected using the Magic Red Caspases 3&7 Detection FLICA Kit
(Immunochemistry Technologies, LLC, Bloomington, MN) as described previously [7]. 10 / 22 PLOS ONE | DOI:10.1371/journal.pone.0148642
February 5, 2016 PLOS ONE | DOI:10.1371/journal.pone.0148642
February 5, 2016 PLOS ONE | DOI:10.1371/journal.pone.0148642
February 5, 2016 Perimeter, Total Distance Travelled.
***SMBLR selected the following features: Number of Protrusions, Change in Perimeter, Minor Axis Length, Bright Area Ratio, Total Distance Travelled.
****FCBF selected the following features: Area, Perimeter, Minimum Intensity, Bright Area Ratio, Change in Perimeter, total Distance Travelled. Evaluating Stem Cells with Video Bioinformatics Table 1. Classification Results Using 48 Hours of Video. 48 Hours
*Classification Techniques
Single Features
SVM
K-NN, k = 3
Naïve Bayes
1) Area
94.12 ± 0.00
94.12 ± 0.00
94.00 ± 0.91
2) Number of Protrusions
90.71 ± 1.35
96.06 ± 1.32
91.24 ± 0.65
3) Total Distance Travelled
84.24 ± 1.20
74.06 ± 1.48
84.88 ± 1.01
Combination of Features
1) Area, Orientation, Num. of Protrusions
94.12 ± 0.00
94.71 ± 1.15
94.12 ± 0.00
2) Num. of Protrusions, Min. Intensity
97.06 ± 0.00
97.06 ± 0.00
96.47 ± 1.15
3) Major Axis, Minor Axis, Change in Centroid
93.53 ± 1.57
92.94 ± 1.62
90.00 ± 1.27
Feature Selection Methods
**CFS
91.76 ± 1.32
96.47 ± 1.32
91.76 ± 1.32
***Chi Square
91.76 ± 0.00
91.76 ± 0.00
95.29 ± 1.61
****QPFS
91.76 ± 1.32
94.12 ± 3.60
91.76 ± 2.46
*Classification of colonies as healthy or unhealthy using three different classification techniques: SVM, KNN, and Naive Bayes. **CFS selected the following features: Area, Number of Protrusions, and Change in Area. ***ChiSquare selected the following features: Area, Number of Protrusions, and Major Axis Length
****Quadratic Programming Feature Selection selected the following features: Total Distance Travelled, Major Axis Length, Minimum Radius. Table 1. Classification Results Using 48 Hours of Video. Table 1. Classification Results Using 48 Hours of Video *Classification of colonies as healthy or unhealthy using three different classification techniques: SVM, KNN, and Naive Bayes. **CFS selected the following features: Area Number of Protrusions and Change in Area *Classification of colonies as healthy or unhealthy using three different classification techniques: SVM, KNN, and Naive Bayes. **CFS selected the following features: Area, Number of Protrusions, and Change in Area. ***ChiSquare selected the following features: Area, Number of Protrusions, and Major Axis Length
****Quadratic Programming Feature Selection selected the following features: Total Distance Travelled Major Axis Length Minimum Radius doi:10.1371/journal.pone.0148642.t001 Fluorescent staining of F-actin was performed using a phalloidin-Alexa Fluor 488 conjugate
(Invitrogen, Carlsbad, CA) diluted 1:200 in 1% goat serum in phosphate buffered saline. hESC
colonies in chamber slides were fixed using 4% paraformaldehyde for 10 minutes, incubated in
blocking solution (3% goat serum in PBS) at room temperature for 1 hour, washed 5 times, and
incubated in phalloidin-Alexa Fluor 488 for 1 hour at room temperature. Samples were Fluorescent staining of F-actin was performed using a phalloidin-Alexa Fluor 488 conjugate
(Invitrogen, Carlsbad, CA) diluted 1:200 in 1% goat serum in phosphate buffered saline. hESC
colonies in chamber slides were fixed using 4% paraformaldehyde for 10 minutes, incubated in
blocking solution (3% goat serum in PBS) at room temperature for 1 hour, washed 5 times, and
incubated in phalloidin-Alexa Fluor 488 for 1 hour at room temperature. Samples were Table 2. Classification Results Using 36 Hours of Video. 36 Hours
*Classification Techniques
Single Features
SVM
K-NN, k = 3
Naïve Bayes
1) Area
85.18 ± 1.25
88.00 ± 1.49
85.18 ± 1.49
2) Number of Protrusions
81.94 ± 1.10
73.65 ± 1.72
81.00 ± 1.34
3) Bright Area Ratio
81.76 ± 3.22
85.29 ± 0.00
76.47 ± 2.08
Combination of Features
1) Area, Min. Radius, Num. of Protrusions, Change in Area, Change in Perimeter
96.47 ± 1.32
94.71 ± 1.32
94.71 ± 1.32
2) Area, Min. Radius, Change in Area, Change in Perimeter
95.88 ± 1.61
92.94 ± 1.61
95.88 ± 1.61
3) Area, Avg. Radius, Change in Area, Change in Perimeter
95.29 ± 1.61
92.94 ± 1.61
95.88 ± 2.63
Feature Selection Methods
**CFS
91.18 ± 2.08
88.24 ± 2.08
90.00 ± 1.61
***SMBLR
85.88 ± 1.32
91.76 ± 1.32
90.59 ± 1.32
****FCBF
86.47 ± 1.61
87.65 ± 1.32
92.94 ± 1.61
*Classification of colonies as healthy or unhealthy using three different classification techniques: SVM, KNN, and Naive Bayes. **CFS selected the following features: Area, Perimeter, Minor Axis Length, Minimum Intensity, Bright Area Ratio, Number of Protrusions, Change in
Perimeter, Total Distance Travelled. ***SMBLR selected the following features: Number of Protrusions, Change in Perimeter, Minor Axis Length, Bright Area Ratio, Total Distance Travelled. ****FCBF selected the following features: Area, Perimeter, Minimum Intensity, Bright Area Ratio, Change in Perimeter, total Distance Travelled. Table 2. Classification Results Using 36 Hours of Video. olonies as healthy or unhealthy using three different classification techniques: SVM, KNN, and Naive Bayes. *Classification of colonies as healthy or unhealthy using three different classification techniques: SVM, KNN, and Naive Bayes. **CFS selected the following features: Area, Perimeter, Minor Axis Length, Minimum Intensity, Bright Area Ratio, Number of Protrusions, Change in
Perimeter, Total Distance Travelled. ***SMBLR selected the following features: Number of Protrusions, Change in Perimeter, Minor Axis Length, Bright Area Ratio, Total Distance Travelled. **CFS selected the following features: Area, Perimeter, Minor Axis Length, Minimum Intensity, Bright Area Ratio, Number of Protrusions, Change in doi:10.1371/journal.pone.0148642.t002 ****FCBF selected the following features: Area, Perimeter, Minimum Intensity, Bright Area Ratio, Change in Perimeter, total Distance Travelled. *Classification of colonies as healthy or unhealthy using three different classification techniques: SVM, KNN, and Naive Bayes. **CFS selected the following features: Area, Perimeter, Minor Axis Length, Minimum Intensity, Bright Area Ratio, Number of Protrusions, Change in
Perimeter, Total Distance Travelled. ***SMBLR selected the following features: Number of Protrusions, Change in Perimeter, Minor Axis Length, Bright Area Ratio, Total Distance Travelled. ****FCBF selected the following features: Area, Perimeter, Minimum Intensity, Bright Area Ratio, Change in Perimeter, total Distance Travelled. Perimeter, Total Distance Travelled. ***SMBLR selected the following features: Number of Protrusions, Change in Perimeter, Minor Axis Length, Bright Area Ratio, Total Distance Travelled. ****FCBF selected the following features: Area, Perimeter, Minimum Intensity, Bright Area Ratio, Change in Perimeter, total Distance Travelled. PLOS ONE | DOI:10.1371/journal.pone.0148642
February 5, 2016 PLOS ONE | DOI:10.1371/journal.pone.0148642
February 5, 2016 11 / 22 Single Features *Classification of colonies as healthy or unhealthy using three different classification techniques: SVM, KNN, and Naive Bayes. **CFS selected the following features: Area, Minor Axis Length, Bright Area Ratio, Change in Perimeter, Total Distance Travelled. ***SMBLR selected the following features: Area, Orientation, Number of Protrusions, Total Distance Travelled. ****FCBF selected the following features: Area, Minor Axis Length, Bright Area Ratio, Change in Perimeter, Total Distance Travelled. *Classification of colonies as healthy or unhealthy using three different classification techniques: SVM, KNN, and Naive Bayes. **CFS selected the following features: Area, Minor Axis Length, Bright Area Ratio, Change in Perimeter, Total Distance Travelled. ***SMBLR selected the following features: Area, Orientation, Number of Protrusions, Total Distance Travelled. ****FCBF selected the following features: Area, Minor Axis Length, Bright Area Ratio, Change in Perimeter, Total Distance Travelled. doi:10.1371/journal.pone.0148642.t003 mounted using Vectashield with DAPI (Vector Laboratories, Burlingame, CA) and imaged
with a Nikon Eclipse Ti fluorescent microscope (Nikon, Melville, NY). mounted using Vectashield with DAPI (Vector Laboratories, Burlingame, CA) and imaged
with a Nikon Eclipse Ti fluorescent microscope (Nikon, Melville, NY). Feature Analysis Features were analyzed graphically to identify those that differed in the healthy, unhealthy, and
dying groups. Sets of affected features were then grouped according to the biological processes
urs of Video Table 4. Classification Results Using 12 Hours of Video. 12 Hours
*Classification Techniques
Single Features
SVM
K-NN, k = 3
Naïve Bayes
1) Bright Area Ratio
67.06 ± 1.32
54.12 ± 1.61
64.70 ± 2.94
2) Total Distance Travelled
71.76 ± 2.63
65.29 ± 2.46
61.18 ± 3.83
3) Change in Centroid
64.71 ± 0.00
57.06 ± 3.35
67.06 ± 2.46
Combination of Features
1) Max. Radius, Ratio of Protrusion Area, Change in Area, Change in Centroid
74.12 ± 2.46
75.88 ± 1.31
79.41 ± 3.60
2) Eccentricity, Min. Radius, Ratio of Protrusion Area, Change in Perimeter, Change in Centroid
71.18 ± 3.83
78.24 ± 3.35
78.24 ± 1.61
3) Min. Radius, Max Intensity, Change in Centroid
71.18 ± 2.46
80.59 ± 2.63
73.53 ± 0.00
Feature Selection Methods
*CFS
71.76 ± 2.63
65.29 ± 2.46
61.18 ± 3.83
**SBMLR
67.65 ± 2.94
77.06 ± 2.46
64.12 ± 3.22
***FCBF
71.76 ± 2.63
65.29 ± 2.46
61.18 ± 3.83
*Classification of colonies as healthy or unhealthy using three different classification techniques: SVM, KNN, and Naive Bayes. **CFS selected the following features: Total Distance Travelled. ***SMBLR selected the following features: Change in Area, Total Distance Travelled. ****FCBF selected the following features: Total Distance Travelled. doi:10 1371/journal pone 0148642 t004 Table 4. Classification Results Using 12 Hours of Video. 12 Hours
*Classification Techniques
Single Features
SVM
K-NN, k = 3
Naïve Bayes
1) Bright Area Ratio
67.06 ± 1.32
54.12 ± 1.61
64.70 ± 2.94
2) Total Distance Travelled
71.76 ± 2.63
65.29 ± 2.46
61.18 ± 3.83
3) Change in Centroid
64.71 ± 0.00
57.06 ± 3.35
67.06 ± 2.46
Combination of Features
1) Max. Radius, Ratio of Protrusion Area, Change in Area, Change in Centroid
74.12 ± 2.46
75.88 ± 1.31
79.41 ± 3.60
2) Eccentricity, Min. Radius, Ratio of Protrusion Area, Change in Perimeter, Change in Centroid
71.18 ± 3.83
78.24 ± 3.35
78.24 ± 1.61
3) Min. Evaluating Stem Cells with Video Bioinformatics Table 3. Classification Results Using 24 Hours of Video. 24 Hours
*Classification Techniques
Single Features
SVM
K-NN, k = 3
Naïve Bayes
1) Area
83.35 ± 1.49
74.94 ± 2.00
82.47 ± 1.37
2) Number of Protrusions
75.29 ± 1.85
67.41 ± 1.95
71.35 ± 1.51
3) Perimeter
75.29 ± 4.36
73.53 ± 2.08
75.29 ± 1.61
Combination of Features
1) Area, Orientation, Num. of Protrusions, Change in Perimeter, Change in Centroid
83.53 ± 1.61
91.17 ± 0.00
88.82 ± 1.31
2) Area, Avg. Radius, Change in Perimeter, Change in Centroid
88.24 ± 2.08
83.53 ± 1.61
90.59 ± 1.32
3) Orientation, Bright Area Ratio, Change in Perimeter, Change in Centroid
87.65 ± 1.32
88.24 ± 2.94
84.12 ± 1.61
Feature Selection Methods
*CFS
84.71 ± 1.32
72.35 ± 4.46
78.24 ± 1.61
**SBMLR
78.24 ± 5.73
69.42 ± 3.35
83.53 ± 3.35
***FCBF
84.71 ± 1.32
72.35 ± 4.46
78.24 ± 1.61
*Classification of colonies as healthy or unhealthy using three different classification techniques: SVM, KNN, and Naive Bayes. **CFS selected the following features: Area, Minor Axis Length, Bright Area Ratio, Change in Perimeter, Total Distance Travelled. ***SMBLR selected the following features: Area, Orientation, Number of Protrusions, Total Distance Travelled. ****FCBF selected the following features: Area, Minor Axis Length, Bright Area Ratio, Change in Perimeter, Total Distance Travelled. Table 3. Classification Results Using 24 Hours of Video *Classification Techniques Features Related to Colony Growth as Biomarkers of hESC Health Extracted features related to colony growth (area, perimeter, minor axis, protrusions) were
evaluated in healthy, unhealthy, and dying colonies. Area (the total number of pixels inside a
segmented colony; S5 Fig) differed in healthy, unhealthy and dying colonies (Fig 1A–1E). When area was normalized to the initial time point to account for variability in the starting
size, all groups displayed an initial contraction which decreased area and lasted about 8 hours
(Fig 1D). Contraction was likely caused by changes in temperature/CO2 during transfer to the
BioStation. After contraction, healthy colonies displayed a steady increase in area until the end
of recording, while unhealthy colonies grew at a slower and variable rate. Both groups followed
a similar trend up to 16 hours, after which healthy and unhealthy growth rates deviated and
become distinguishable by about 24 hours. Dying colonies could be distinguished from the
healthy and unhealthy groups by about 10 hours when growth rates for the dying group clearly
diverged. At 30 hours, dying colonies underwent a second contraction leading to a sharp
decrease in area (Fig 1D and 1E). These colonies were interpreted to be dead based on this pro-
nounced decrease in size and shedding of dead cells. g
Perimeter, defined as the number of pixels constituting the colony periphery (red outline in
Fig 1A–1C and 1F; S5 Fig), provided additional information about colony growth. There was
divergence in the perimeters of healthy and unhealthy colonies at approximately 25 hours (Fig
1F), after which the rate of change in perimeter for unhealthy colonies slowed until about 33
hours when it underwent a growth spurt that lasted 3 hours. The dying colonies diverged from
the other two groups at 8 hours and had an abrupt decrease in size at 30 hours, as was seen
with area. Minor axis (smaller axis of an ellipse fitted to a colony) (S6 Fig) was affected in unhealthy/
dying colonies (Fig 1G). After 23 hours, healthy colonies showed a steep increase in minor axis,
suggesting that once healthy colonies reach a critical size, they have a less-elongated
morphology. Protrusions are dynamic cell processes that extend off colonies and take a variety of shapes
(Fig 2A–2C; S6 Fig). They allow colonies to attach, spread, and migrate, [40,41]. The number
of protrusions increased on healthy and unhealthy colonies and decreased on dying colonies
during incubation (Fig 2D). Feature Analysis Radius, Max Intensity, Change in Centroid
71.18 ± 2.46
80.59 ± 2.63
73.53 ± 0.00
Feature Selection Methods
*CFS
71.76 ± 2.63
65.29 ± 2.46
61.18 ± 3.83
**SBMLR
67.65 ± 2.94
77.06 ± 2.46
64.12 ± 3.22
***FCBF
71.76 ± 2.63
65.29 ± 2.46
61.18 ± 3.83
*Classification of colonies as healthy or unhealthy using three different classification techniques: SVM, KNN, and Naive Bayes. **CFS selected the following features: Total Distance Travelled. ***SMBLR selected the following features: Change in Area, Total Distance Travelled. ****FCBF
l
t d th
f ll
i
f
t
T t l Di t
T
ll d Table 4. Classification Results Using 12 Hours of Video. *Classification of colonies as healthy or unhealthy using three different classification techniques: SVM, KNN, and Naive Bayes. **CFS selected the following features: Total Distance Travelled. ***SMBLR selected the following features: Change in Area, Total Distance Travelled. ****FCBF selected the following features: Total Distance Travelled. *Classification of colonies as healthy or unhealthy using three different classification techniques: SVM, KNN, and Naive Bayes. **CFS selected the following features: Total Distance Travelled. ***SMBLR selected the following features: Change in Area, Total Distance Travelled. ****FCBF selected the following features: Total Distance Travelled. PLOS ONE | DOI:10.1371/journal.pone.0148642
February 5, 2016 12 / 22 PLOS ONE | DOI:10.1371/journal.pone.0148642
February 5, 2016 Evaluating Stem Cells with Video Bioinformatics they represented (morphology, growth, motility, death) (Figs 1–4, S4 Fig). The classifiers were
run with 48, 36, 24, and 12 hours of video to show their effectiveness at different time points. For all durations, all 24 features were run singularly through the classifiers and the ones with
the highest accuracy are shown in Tables 1–4 (Single Features). Additionally, exhaustive
searches for combinations of up to 5 features were run to identify the most accurate results
(Table 1 Combination of Features). Lastly, the best results from the 11 existing feature selection
algorithms are also shown in Tables 1–4 (Feature Selection Methods). Features Related to Colony Growth as Biomarkers of hESC Health The protruding-to-total area ratio, which is defined as the ratio of
protrusion area divided by total colony area, had an inverse relationship with colony growth. Protrusion area decreased gradually in healthy and unhealthy groups, but increased slightly in
the dying group (Fig 2E). Colony Motility Although F-actin was partially depolymerized by smoke
treatment, there was sufficient functional F-actin in the treated colonies to allow colony move-
ment. A decrease in F-actin may be linked to a decrease in the number of focal adhesions, which
may facilitate motility in the unhealthy group [44,45]. Other studies have reported the inverse
correlation between cell motility and polymerization state of the actin cytoskeleton [46]. Colony Motility β < 1 indicates sub-diffusive movement, defined as a tendency for the colonies
not to diffuse due to trapping (inability to move). For Brownian motion, or a random walk, β is
approximately 1. The MSD feature is robust because it uses the squared value of displacement, The mean squared displacement (MSD) feature measures Brownian motion [42] and can be
used to study cellular migration [43]. MSD is defined by the equatio: MSD(t) = ([x(t+t0) −x
(t0)]2+[y(t+t0) −y(t0)]2), where MSD (t) can be approximated as ~ tβ(t). The logarithmic deriv-
ative exponent β can be used to determine the particular mode of motility, with β > 1 indicat-
ing super-diffusive movement, a form of diffusion where the colonies occasionally undergo
very long steps. β < 1 indicates sub-diffusive movement, defined as a tendency for the colonies
not to diffuse due to trapping (inability to move). For Brownian motion, or a random walk, β is
approximately 1. The MSD feature is robust because it uses the squared value of displacement,
making it less sensitive to small fluctuations. The MSD plot shows a similar trajectory for all
three groups up until about 11 hours (Fig 3E), after which the healthy colonies display Brown-
ian motion (β = 1.04). For dying colonies, sub-diffusive motility (β = 0.21) was observed from
23–50 hours, consist with their death after 30 hours. The unhealthy group demonstrated sub-
diffusive motility from 22–35 hours and 35–50 hours (β values = 0.76 and 0.52, respectively). not to diffuse due to trapping (inability to move). For Brownian motion, or a random walk, β is
approximately 1. The MSD feature is robust because it uses the squared value of displacement,
making it less sensitive to small fluctuations. The MSD plot shows a similar trajectory for all
three groups up until about 11 hours (Fig 3E), after which the healthy colonies display Brown-
ian motion (β = 1.04). For dying colonies, sub-diffusive motility (β = 0.21) was observed from
23–50 hours, consist with their death after 30 hours. The unhealthy group demonstrated sub-
diffusive motility from 22–35 hours and 35–50 hours (β values = 0.76 and 0.52, respectively). To investigate the molecular basis of the aforementioned effects on motility, F-actin was
labeled with phalloidin-Alexa 488. Healthy colonies (Fig 3F) had a more robust actin cytoskele-
ton than unhealthy colonies (Fig 3G). Colony Motility The change in centroid feature allowed tracking of stem cell colony movement. This feature is
determined by finding the centroid of each colony and calculating the distance between two
successive frames (S7 Fig). Outlines of a hESC colony at two times and the change in centroids
are shown in Fig 3A. Change in centroid oscillations were smaller in the healthy and unhealthy
groups than in the dying group (Fig 3B). The unhealthy and healthy groups were similar in the PLOS ONE | DOI:10.1371/journal.pone.0148642
February 5, 2016 13 / 22 Evaluating Stem Cells with Video Bioinformatics magnitude of their oscillations, but overall motility was less in the healthy colonies, probably
because the larger sized healthy colonies required more energy and coordination for directed
movement. After a certain area was reached, the center of the healthy colonies moved very little
as the colony continued to expand. It is also possible that smoke stimulated motility in
unhealthy hESC to facilitate escape from exposure. The dying colonies displayed erratic motil-
ity and showed a significant decrease in movement after 20 hours as they were approaching
death. Movements detected after death (30 hours) are due to slight segmentation differences
between frames. Total displacement detected how far a colony moved from its original starting point (Fig
3C; S7 Fig), while total distance traveled is the sum of the entire trajectory of movement (Fig
3D; S7 Fig). These features revealed information on the pattern of travel. Dying colonies trav-
eled more up to 30 hours (when they died) than the other two groups, but their displacement
was low indicating that they moved erratically near their original starting point. Unhealthy col-
onies moved further from their point of origin and travelled a longer total distance than healthy
colonies. Both the healthy and unhealthy colonies displayed remarkably little variance in total
distance travelled (Fig 3D). The mean squared displacement (MSD) feature measures Brownian motion [42] and can be
used to study cellular migration [43]. MSD is defined by the equatio: MSD(t) = ([x(t+t0) −x
(t0)]2+[y(t+t0) −y(t0)]2), where MSD (t) can be approximated as ~ tβ(t). The logarithmic deriv-
ative exponent β can be used to determine the particular mode of motility, with β > 1 indicat-
ing super-diffusive movement, a form of diffusion where the colonies occasionally undergo
very long steps. Solidity as Predictor of Apoptosis As colonies became rounder or more convex, their solidity increased and approached 1. Fig
4A–4D show hESC colonies at different times with outlines of their segmentations (red lines)
and convex hulls (white lines). Solidity, which measured convexity (Fig 4E; S8 Fig), identified
colonies that were destined to die by 48 hours. Solidity for the healthy/unhealthy groups com-
bined changed little during 48 hours (Fig 4E). These two groups were combined since solidity
was a predictor of colony death, not health. In contrast, dying colonies had a significant spike
in solidity at about 12 hours due to contraction and rounding of the colonies (B label in Fig
4E). This was followed by a drop that reached a minimum at 30 hours (C label in Fig 4E), when
death occurred and extrusion of dead cells caused the convex hull to be less circular. Graphs of
solidity can be used to identify at 12 hours, colonies that will die by 48 hours. 14 / 22 PLOS ONE | DOI:10.1371/journal.pone.0148642
February 5, 2016 Evaluating Stem Cells with Video Bioinformatics 3D Visualization of Features and Custom Features Enhance Data
Mining To mine additional biological information such as correlation of features, StemCellQC can plot
features against each other and play the plot as a video over time (S1 and S2 Videos). In Fig 5A,
perimeter and average intensity, when plotted against each other, showed an inverse relation-
ship (indicative of dead cells). In Fig 5B, area and the mean-squared displacement features
were plotted against each other to highlight individual colonies with elevated MSD values
(mainly colonies from the unhealthy group). This type of analysis can also reveal outlier colo-
nies within a group. In addition, StemCellQC is able to plot mathematical equations using the
original 24 features. In Fig 5C, a user derived equation, ratio of perimeter to the number of pro-
trusions, is plotted. This plot displays an estimate for the average length of a protrusion for
each class and shows that the protrusions on dying colonies are about twice as long as those on
healthy/unhealthy colonies (Fig 5C). Colony Brightness Identifies Dying Cells As cells within a colony die, they are extruded to the top of the colony where their brightness
increases. The white areas in Fig 4F are dead cells on a healthy colony at the end of incubation. Significantly more dead cells were present on the unhealthy (Fig 4G) and dying colonies (Fig
4H). To quantify dead cells on top of colonies, a bright-to-total area ratio feature was used. This feature measured the number of bright pixels in the colony as a ratio to the total area and
is an indicator of cell death. All groups exhibited an increase in bright-to-total area ratio during
the first 6 hours when the colonies contracted (Fig 4I), after which the bright-to-total area
ratios of healthy and unhealthy colonies decreased and the ratio for the dead colonies increased
up to 16 hours and stayed elevated. To compare the progression of colony brightness over time, a minimum intensity feature
(lowest pixel intensity in the colony) was monitored (Fig 4J). Throughout incubation, healthy
colonies displayed a lower minimum intensity than the unhealthy and dying colonies. These
data support the idea that the unhealthy and dying colonies failed to spread as well on Matrigel
as healthy colonies. To confirm cell death, colonies were labeled with Magic Red which detects activated cas-
pases 3&7, biomarkers for apoptosis. As shown by the red staining in Fig 4K and 4L, unhealthy
colonies exhibited more caspase 3&7 activity than the healthy colonies. Discussion StemCellQC is an innovative, cost effective, non-invasive software tool that utilizes bioinfor-
matics to automatically monitor dynamic cell processes, cell morphology, and cell health dur-
ing passaging, culture, expansion, maintenance, or experimental treatment of pluripotent stem
cells. StemCellQC eliminates the need for labeling with dyes or fluorescent probes and elimi-
nates tedious manual classification, which significantly decreases analysis time and classifica-
tion errors due to observer bias. Graphical plots of features provide quantifiable, real-time data on living hESC and are excel-
lent analytical tools for comparing features across treatments and cell types. The plots can help
users visualize trends or features that are not easily detectable by manual inspection. Cell pro-
cess analysis is especially valuable in toxicological or drug studies as it provides insight into the
mode of action of the treatment. For example, smoke treatment inhibited growth (area, perim-
eter, minor axis and protrusions), increased motility (change in centroid, total displacement,
total distance traveled and MSD), and increased apoptosis (solidity and intensity features). Chemical treatments other than cigarette smoke may affect other features, and in such cases,
other cell processes could be revealed by feature analysis. Multiplexing cellular process infor-
mation (colony growth rate, motility, and apoptosis) increases the power of analysis, and in
toxicological studies, this greatly increases the probability of detecting an effect if one exists. Graphical plots of features provide quantifiable, real-time data on living hESC and are excel-
lent analytical tools for comparing features across treatments and cell types. The plots can help
users visualize trends or features that are not easily detectable by manual inspection. Cell pro-
cess analysis is especially valuable in toxicological or drug studies as it provides insight into the
mode of action of the treatment. For example, smoke treatment inhibited growth (area, perim-
eter, minor axis and protrusions), increased motility (change in centroid, total displacement,
total distance traveled and MSD), and increased apoptosis (solidity and intensity features). Ch
i
l t
t
t
th
th
i
tt
k
ff
t th
f
t
d i
h StemCellQC can plot user-derived equations of features (for example: perimeter/number of
protrusions) for customized types of analysis. The software can also create videos of various
features plotted against each other. These plots enable correlation between features and help
determine how biological processes are related over time. Classification Results The input values used by the classifiers were the mean slopes of each feature. The nine individ-
ual features found by user-interpreted feature selection were tested separately giving each fea-
ture a classification rate (Table 1). Area was the best individual feature at predicting health
with a 94% accuracy when using any classifier. By combining features that are not related to
the same process, accuracy increased. When the number of protrusions and minimum inten-
sity were combined, the system’s ability to distinguish hESC colony health improved to 97%
accuracy when using any classifier. Results for feature selection algorithms (CFS, ChiSquare
and QPFS) were also shown. All three had at least 91% accuracy and CFS was 96.47% accurate
with KNN. The classifiers were also run with the first 36, 24, and 24 hours, which are shown in Tables
2–4. For 36 hours (Table 2), area was the best feature with 88% accuracy, and combination of
features improved results to 96.47%. For 24 hours (Table 3), area was still the strongest feature
with 83.35% accuracy and a combination improved results to 91.17% accuracy. For 12 hours 15 / 22 PLOS ONE | DOI:10.1371/journal.pone.0148642
February 5, 2016 Evaluating Stem Cells with Video Bioinformatics (Table 4), however, total distance travelled is the strongest feature with 71.76% accuracy. It
should be noted that for 12 hours most individual features performed at about 50–60% accu-
racy which is slightly better than chance, however when we combine features, we are able to
improved classification to 80.59% accuracy. Judging colony health by eye after a mere 12 hours
of time is biased and difficult, making an 80.59% classification rate very useful. These tables
show that while a certain combination of features work best using the full 48 hours of time,
another combination may produce a more accurate classification if less time is used. With
shorter video duration, there is an increase in accuracy using a combination of features;
whereas, with longer durations, a single strong feature (such as area) can be sufficient to get
accurate results. PLOS ONE | DOI:10.1371/journal.pone.0148642
February 5, 2016 Discussion For example, an inverse relationship
was found between colony size and colony brightness (dead cells on a colonies’ surface). Feature analysis, when combined with a classifier, enabled identification of healthy,
unhealthy, and dying colonies. Area, which classified with 94% accuracy, was the strongest fea-
ture for predicting colony health. Changes in area are not always detected by human observa-
tion, especially when colonies do not die but experience stunted growth. In clinics or research
laboratories, a decrease in growth rate may signal a problem with the culture or cell quality,
and this would be rapidly detected in cultures monitored using StemCellQC. While smoke
treatment slowed colony growth, factors that increase growth rate may be equally important
and detectable by StemCellQC. For example, when chromosomal translocations occur in
hESC, growth can be accelerated [47], and this would not be desirable in clinical or research
labs. Combinations of features successfully increased the accuracy of classification of
unhealthy/dying colonies to 97%. Depending on the rigor needed, change in area by itself will
usually be sufficient to distinguish healthy from unhealthy/dying colonies. By comparing changes in features over 48 hours, biomarkers that predict biological out-
comes were found at early time points (Fig 6). For example, growth rate separated dying from PLOS ONE | DOI:10.1371/journal.pone.0148642
February 5, 2016 16 / 22 Evaluating Stem Cells with Video Bioinformatics healthy/unhealthy colonies by 16 hours and further separated healthy from unhealthy colonies
Fig 6. hESC Health Timeline and Biomarkers. (A) Changes in several feature values and biological events during 48 hours of incubation for healthy,
unhealthy and control groups. This type of plot can be used to compare events in different groups. (B) Biomarkers that can be used to identify healthy,
unhealthy, and dying colonies and their earliest detection times. doi:10.1371/journal.pone.0148642.g006 Fig 6. hESC Health Timeline and Biomarkers. (A) Changes in several feature values and biological events during 48 hours of incubation for healthy,
unhealthy and control groups. This type of plot can be used to compare events in different groups. (B) Biomarkers that can be used to identify healthy,
unhealthy, and dying colonies and their earliest detection times. doi:10.1371/journal.pone.0148642.g006 healthy/unhealthy colonies by 16 hours and further separated healthy from unhealthy colonies
by 26 hours (Fig 1D; Fig 6A and 6B). Similar distinctions can be made from graphs for other
growth features (perimeter, minor axis, and number of protrusions) (Figs 1F and 1G and 2D
and 2E). Conclusions StemCellQC is a versatile toolkit for analyzing cell processes, evaluating cell quality, and dis-
covering biomarkers. It is designed for use with pluripotent stem cell colonies in culture, and is
adaptable to other cell types. It can be used retrospectively or on-the-fly to solve numerous
problems. There are at least four applications for StemCellQC. First, core facilities that culture
pluripotent cells for distribution to research labs could monitor cell quality using non-invasive
morphological tools to guarantee that distributed cells meet an acceptable uniform standard
from day-to-day. This is especially important when the results of a research study may ulti-
mately affect a patient’s health. Second, StemCellQC can serve as a quality control tool in future
clinics that deliver therapies based on pluripotent stem cells. Such clinics will need to maintain
and differentiate cells that meet future FDA criteria for transplantation to patients. A record of
cell quality produced by StemCellQC would be an important part of a patient’s medical record,
and could be mined after cell transfer to patients to better understand those features that work
best for patient treatment. Third, hPSC can differentiate into specific cell types that can be used
for studying genetic disorders, such as Huntington’s disease [48]. StemCellQC can monitor the
behavior of cells/colonies in disease-in-a-dish models to determine how cells respond to drug
treatments [49]. Fourth, StemCellQC could be used in laboratories that perform drug testing or
that monitor chemical toxicity. Multiplexing data enhances the discovery of toxicants and bio-
markers. hESC provide an excellent model for prenatal development, a process that cannot be
studied experimentally in humans [20] and which is generally sensitive to environmental
chemicals [50]. We are currently using StemCellQC with other pluripotent cell types and experimental con-
ditions and found that it performed very well. We have found clear cut effects on processes
such as growth, motility, death and morphology using StemCellQC with cells grown in optimal
and suboptimal media, indicating StemCellQC will be useful for recognizing culture conditions
that are not satisfactory. As more treatments are used, we anticipate that other processes or
effects may be observed. In the future, StemCellQC software can be enhanced by adapting it to
single cells and including features that correlate to cell processes such as stress, differentiation,
and pluripotency. More classifiers can be added, and additional biomarkers will likely be dis-
covered with new applications of the software. Discussion Change in centroid was the strongest motility biomarker which cleanly separated
healthy from dying colonies as early as 8 hours (Fig 3D). Solidity successfully separated dying
colonies from healthy/unhealthy by 12 hours when used with Otsu’s segmentation (Fig 4E),
and bright-to-total area ratio separated all three groups from each other by 14 hours (Fig 4I). The biomarkers for dying colonies are powerful tools for monitoring apoptosis in living cul-
tures without use of labels or probes, which themselves often produce unwanted effects. PLOS ONE | DOI:10.1371/journal.pone.0148642
February 5, 2016 PLOS ONE | DOI:10.1371/journal.pone.0148642
February 5, 2016 17 / 22 Evaluating Stem Cells with Video Bioinformatics S6 Video. Solidity feature video of a representative dying colony with the convex hull
shown in white and the colony segmentation outlined in red.
(MPG) S6 Video. Solidity feature video of a representative dying colony with the convex hull
shown in white and the colony segmentation outlined in red. (MPG) Supporting Information S1 Fig. (A) Diagram showing workflow used to develop StemCellQC™. (B) Diagram showing
feature selection methods for classification. (TIF) S2 Fig. Decision tree showing method for classifying hESC colonies into healthy, unhealthy
or dying groups. Red arrows show decisions resulting in classification of a colony as unhealthy
or dying, green arrows show decisions resulting in classification as healthy, and black arrows
indicate points where the classification process was continued. (TIF) S3 Fig. Ground truth verification of colony segmentation using ImageJ to manually seg-
ment 6 representative healthy, 6 unhealthy, and 6 dying colonies. (A, B) Normalized area
and perimeter values for healthy colonies extracted by StemCellQC compared to ground truth
using ImageJ. 2-way ANOVA revealed no significant differences. (C, D) Normalized area and
perimeter values for unhealthy colonies extracted by StemCellQC compared to ground truth
using ImageJ. 2-way ANOVA revealed no significant differences. (E, F) Normalized area and
perimeter values for dying colonies extracted by StemCellQC compared to ground truth using PLOS ONE | DOI:10.1371/journal.pone.0148642
February 5, 2016 18 / 22 Evaluating Stem Cells with Video Bioinformatics ImageJ. 2-way ANOVA revealed no significant differences, except for a portion of the normal-
ized area of dying colonies. This corresponds with slight over-segmentation of software due to
detection of cellular debris ejected from dying colonies after their death at 30hours ( =
P < 0.05). (TIF)
S4 Fig. Relationship between features and cell processes. (TIF)
S5 Fig. Visual descriptors of extracted features related to area. (TIF)
S6 Fig. Visual descriptors of extracted features related to morphology and area. (TIF)
S7 Fig. Visual descriptors of extracted features related to motility. (TIF)
S8 Fig. Visual descriptors of extracted features related to apoptosis. (TIF)
S9 Fig. List of Extracted Features and Definitions. (TIF)
S1 Video. Average intensity versus perimeter running plot shown for all individual healthy
(green), unhealthy (blue), and dying (red) hESC colonies. (MPG)
S2 Video. Mean-squared displacement versus area running plot shown for all individual
healthy (green), unhealthy (blue), and dying (red) hESC colonies. (MPG)
S3 Video. Phase contrast video of a representative healthy colony with the segmentation
outlined in white. (MPG)
S4 Video. Protrusions feature video of a representative healthy colony with the protrusions
outlined in red. (MPG)
S5 Video. Bright-to-total area ratio feature video with the bright dead cells of a representa-
tive unhealthy colony highlighted in white. (MPG)
S6 Video. S6 Fig. Visual descriptors of extracted features related to morphology and area.
(TIF) S6 Fig. Visual descriptors of extracted features related to morphology and area. (TIF) S7 Fig. Visual descriptors of extracted features related to motility. (TIF) S8 Fig. Visual descriptors of extracted features related to apoptosis. (TIF) S1 Video. Average intensity versus perimeter running plot shown for all individual healthy
(green), unhealthy (blue), and dying (red) hESC colonies. (MPG) S1 Video. Average intensity versus perimeter running plot shown for all individual healthy
(green), unhealthy (blue), and dying (red) hESC colonies. (MPG) S2 Video. Mean-squared displacement versus area running plot shown for all individual
healthy (green), unhealthy (blue), and dying (red) hESC colonies. (MPG) Supporting Information Solidity feature video of a representative dying colony with the convex hull
shown in white and the colony segmentation outlined in red. (MPG)
Acknowledgments
The authors thank Dr. Albert Cruz (IGERT Fellow) and Dr. Seyyed Farhad Razavi who helped
with parts of the software implementation and Jackie Whitehead for her suggestions on the
manuscript. This work is supported by a NSF IGERT grant in Video Bioinformatics (DGE
0903667), the California Institute for Regenerative Medicine (# NE-A0005A-1E), and grants
from the Tobacco-Related Disease Research Program of CA (# 22RT-0127 and #20PT-0184). ImageJ. 2-way ANOVA revealed no significant differences, except for a portion of the normal-
ized area of dying colonies. This corresponds with slight over-segmentation of software due to
detection of cellular debris ejected from dying colonies after their death at 30hours ( =
P < 0.05). (TIF) S5 Fig. Visual descriptors of extracted features related to area. (TIF) S4 Video. Protrusions feature video of a representative healthy colony with the protrusions
outlined in red.
(MPG) S5 Video. Bright-to-total area ratio feature video with the bright dead cells of a representa-
tive unhealthy colony highlighted in white. (MPG) S5 Video. Bright-to-total area ratio feature video with the bright dead cells of a representa-
tive unhealthy colony highlighted in white. (MPG) S3 Video. Phase contrast video of a representative healthy colony with the segmentation
outlined in white.
(MPG) S3 Video. Phase contrast video of a representative healthy colony with the segmentation
outlined in white. (MPG) S4 Video. Protrusions feature video of a representative healthy colony with the protrusions
outlined in red. (MPG) S4 Video. Protrusions feature video of a representative healthy colony with the protrusions
outlined in red. (MPG) References 1. Tabar V, Studer L. Pluripotent stem cells in regenerative medicine: challenges and recent progress. Nat Rev Genet. 2014; 15(2): 82–92. doi: 10.1038/nrg3563 PMID: 24434846 2. Talbot P, Lin S. Mouse and human embryonic stem cells: can they improve human health by preventing
disease? Curr Top Med Chem. 2011; 11(13): 1638–1652. doi: 10.2174/156802611796117621 PMID:
21446909 3. Zhu Z, Huangfu D. Human pluripotent stem cells: an emerging model in developmental biology. Devel-
opment. 2013; 140(4): 705–717. doi: 10.1242/dev.086165 PMID: 23362344 4. Chen KG, Mallon BS, McKay RDG, Robey PG. Human pluripotent stem cell culture: considerations for
maintenance, expansion, and therapeutics. Cell Stem Cell. 2014; 14(1): 13–26. doi: 10.1016/j.stem. 2013.12.005 PMID: 24388173 5. Sheehy SP, Pasqualini F, Grosberg A, Park SJ, Aratyn-Schaus Y, Parker KK. Quality metrics for stem
cell-derived cardiac myocytes. Stem Cell Reports. 2014; 2(3): 282–294. doi: 10.1016/j.stemcr.2014.01. 015 PMID: 24672752 6. Lin S, Fonteno S, Satish S, Bhanu B, Talbot P. Video bioinformatics analysis of human embryonic stem
cell colony growth. J Vis Exp. 2010; 39: pii: 1933. doi: 10.3791/1933 PMID: 20495527 7. Lin S, Fonteno S, Weng J-H, Talbot P. Comparison of the toxicity of smoke from conventional and harm
reduction cigarettes using human embryonic stem cells. Toxicol Sci. 2010; 118(1): 202–212. doi: 10. 1093/toxsci/kfq241 PMID: 20702591 8. Guan BX, Bhanu B, Talbot P, Lin S. Automated Human Embryonic Stem Cell Detection. IEEE Second
Int Conf Healthc Informatics, Imaging Syst Biol. 2012; 75–82. doi: 10.1109/HISB.2012.25 9. Guan BX, Bhanu B, Thakoor N, Talbot P, Lin S. Human Embryonic Stem Cell Detection by Spatial Infor-
mation and Mixture of Gaussians. IEEE First Int Conf Healthc Informatics, Imaging Syst Biol. 2011;
307–314. doi: 10.1109/HISB.2011.30 10. Guan B, Bhanu B, Thakoor N, Talbot P, Lin S. Automatic Cell Region Detection by K-Means with
Weighted Entropy. IEEE Int Symp on Biomed Imaging. 2013; 418–421. doi: 10.1109/ISBI.2013. 6556501 11. Guan B, Bhanu B, Talbot P, Lin S. Bio-Driven Cell Region Detection in Human Embryonic Stem Cell
Assay. IEEE/ACM Trans Comput Biol Bioinforma. 2014; 1–1. doi: 10.1109/TCBB.2014.2306836 12. Talbot P, Nieden NI, Lin S, Martinez I, Guan B, Bhanu B. Use of Video Bioinformatics Tools in Stem
Cell Toxicology. In: Sahu SC, Casciano DA, editors. Handbook of Nanotoxicology, Nanomedicine and
Stem Cell Use in Toxicology. Chichester, UK; 2014. 13. The UC Riverside NSF Integrated Graduate Education Research and Training (IGERT) Program on
Video Bioinformatics (n.d.). Available: http://www.cris.ucr.edu/IGERT/index.php. 14. Acknowledgments The authors thank Dr. Albert Cruz (IGERT Fellow) and Dr. Seyyed Farhad Razavi who helped
with parts of the software implementation and Jackie Whitehead for her suggestions on the
manuscript. This work is supported by a NSF IGERT grant in Video Bioinformatics (DGE
0903667), the California Institute for Regenerative Medicine (# NE-A0005A-1E), and grants
from the Tobacco-Related Disease Research Program of CA (# 22RT-0127 and #20PT-0184). PLOS ONE | DOI:10.1371/journal.pone.0148642
February 5, 2016 19 / 22 Evaluating Stem Cells with Video Bioinformatics Author Contributions Conceived and designed the experiments: AZ VO SL BB PT. Performed the experiments: AZ
VO SL. Analyzed the data: AZ VO SL BCB EO BB PT. Contributed reagents/materials/analysis
tools: AZ VO SL BCB. Wrote the paper: AZ VO SL BB PT. PLOS ONE | DOI:10.1371/journal.pone.0148642
February 5, 2016 References De Berg M, van Krefeld M, Overmars M, Schwarzkopf O. Computational Geometry: Algorithms and
Applications. 3rd ed. Berlin: Springer Press; 2000. 28. Zhao Z, Morstatter F, Sharma S, Alelyani S, Anand A, Liu H. Advancing Feature Selection Research. ASU Featur Sel Repos Arizona State Univ; 2010. Available: http://featureselection.asu.edu/
featureselection_techreport.pdf. 29. Vinh NX, Chan J, Romano S, Bailey J. Effective Global Approaches for Mutual Information Based Fea-
ture Selection. Proc 20th ACM SIGKDD Int Conf Knowl Discov Data Min. 2014; 512–521. doi: 10.1145/
2623330.2623611 30. Hall M, Smith LA. Feature Selection for Machine Learning : Comparing a Correlation-based Filter
Approach to the Wrapper CFS : Correlation-based Feature. Int FLAIRS Conf. 1999; 5. 31. Liu H, Setiono R. Chi2: Feature Selection and Discretizatino of Numeric Attributes. Proc 7th IEEE Int
Conf Tools with Artif Intell. 1995; 388–391. 32. Bermingham ML, Pong-Wong R, Spiliopoulou A, Hayward C, Rudan I, Campbell H, et al. Application of
high-dimensional feature selection: evaluation for genomic prediction in man. Sci Rep 2015; 5: 10312. doi: 10.1038/srep10312 PMID: 25988841 33. Duda RO, Hart PE, Stork DG. Pattern Classification. 2nd ed. New York: Wiley-Interscience Press;
2000. 34. Gini C. Variabilite e mutabilita. Rome: Libr Eredi Virgilio Veschi; 1912. 35. Cover TM, Thomas JA. Elements of information theory. New York: Wiley-Interscience Press; 1991. 36. Cawley G, Talbot N, Girolami M. Sparse multinomial logistic regression via bayesian l1 regularisation. Adv Neural Inf Process Syst. 2007; 19: 209–216. 37. Montgomery DC, Runger GC, Hubele NF. Engineering Statistics. New York: John Wiley Press; 2001. 38. Wei L. Asymptotic Conservativeness and Efficiency of Kruskal-Wallis Test for K Dependent Samples. J
Am Stat Assoc. 1981; 76(376): 1006–1009. doi: 10.2307/2287603 39. Peng HC, Long FH, Ding C. Feature selection based on mutual information: Criteria of max-depen-
dency, max-relevance, and min-redundancy. IEEE Trans Pattern Anal Mach Intell. 2005; 27(8): 1226–
1238. doi: 10.1109/TPAMI.2005.159 PMID: 16119262 40. Wang X, Song W, Kawazoe N, Chen G. Influence of cell protrusion and spreading on adipogenic differ-
entiation of mesenchymal stem cells on micropatterned surfaces. Soft Matter. 2013; 9: 4160–4166. doi:
10.1039/c3sm27818b 41. McBeath R, Pirone DM, Nelson CM, Bhadriraju K, Chen CS. Cell shape, cytoskeletal tension, and
RhoA regulate stem cell lineage commitment. Dev Cell. 2004; 6(4): 483–495. PMID: 15068789 42. Uhlenbeck GE, Ornstein LS. On the theory of the Brownian motion. Phys Rev. 1930; 36: 823–841. doi:
10.1103/PhysRev.36.823 43. Dieterich P, Klages R, Preuss R, Schwab A. References Tokunaga K, Saitoh N, Goldberg IG, Sakamoto C, Yasuda Y, Yoshida Y, et al. Computational image
analysis of colony and nuclear morphology to evaluate human induced pluripotent stem cells. Sci Rep. 2014; 4:6996. doi: 10.1038/srep06996 PMID: 25385348 15. Megyola CM, Gao Y, Teixeira AM, Cheng J, Heydari K, Cheng EC, et al. Dynamic migration and cell-
cell interactions of early reprogramming revealed by high-resolution time-lapse imaging. Stem Cells. 2013; 31(5): 895–905. doi: 10.1002/stem.1323 PMID: 23335078 16. Wong CC, Loewke KE, Bossert NL, Behr B, De Jonge CJ, Baer TM, et al. Non-invasive imaging of
human embryos before embryonic genome activation predicts development to the blastocyst stage. Nat Biotechnol. 2010; 28(10): 1115–1121. doi: 10.1038/nbt.1686 PMID: 20890283 17. Winter M, Wait E, Roysam B, Goderie SK, Ali RA, Kokovay E, et al. Vertebrate neural stem cell seg-
mentation, tracking and lineaging with validation and editing. Nat Protoc. 2011; 6(12): 1942–1952. doi:
10.1038/nprot.2011.422 PMID: 22094730 18. Alworth SV, Watanabe H, Lee JSJ. Teachable, high-content analytics for live-cell, phase contrast mov-
ies. J Biomol Screen. 2010; 15(8): 968–977. doi: 10.1177/1087057110373546 PMID: 20639505 19. Nichols J, Smith A. Pluripotency in the Embryo and in Culture. Cold Spring Harb Perspect Biol. 2012; 4
(8): a008128–a008128. doi: 10.1101/cshperspect.a008128 PMID: 22855723 20 / 22 PLOS ONE | DOI:10.1371/journal.pone.0148642
February 5, 2016 Evaluating Stem Cells with Video Bioinformatics 20. Lin S, Talbot P. Methods for culturing mouse and human embryonic stem cells. Methods Mol Biol. 2011; 690: 31–56. doi: 10.1007/978-1-60761-962-8_2 PMID: 21042983 21. Lin S, Tran V, Talbot P. Comparison of toxicity of smoke from traditional and harm-reduction cigarettes
using mouse embryonic stem cells as a novel model for preimplantation development. Hum Reprod. 2009; 24(2): 386–397. doi: 10.1093/humrep/den419 PMID: 19043081 22. Knoll M, Shaoulian R, Magers T, Talbot P. Ciliary beat frequency of hamster oviducts is decreased in
vitro by exposure to solutions of mainstream and sidestream cigarette smoke. Biol Reprod 1995; 53(1):
29–37. doi: 10.1095/biolreprod53.1.29 PMID: 7669854 23. Knoll M, Talbot P. Cigarette smoke inhibits oocyte cumulus complex pick-up by the oviduct in vitro inde-
pendent of ciliary beat frequency. Reprod Toxicol 1998; 12(1): 57–68. doi: 10.1016/S0890-6238(97)
00100-7 PMID: 9431573 24. Detecting a Cell Using Image Segmentation (n.d.). Available: http://www.mathworks.com/help/images/
examples/detecting-a-cell-using-image-segmentation.html. 25. Otsu N. A Threshold Selection Method from Gray-Level Histograms. IEEE Trans on Syst, Man, and
Cybernetics. 1979; 9(1): 62–66. 26. Shapiro L, Stockman G. Computer Vision. 1st ed. New Jersey: Prentice Hall Press; 2001. 27. PLOS ONE | DOI:10.1371/journal.pone.0148642
February 5, 2016 PLOS ONE | DOI:10.1371/journal.pone.0148642
February 5, 2016 References Anomalous dynamics of cell migration. Proc Natl Acad Sci
U S A. 2008; 105(2): 459–463. doi: 10.1073/pnas.0707603105 PMID: 18182493 44. Huang C, Rajfur Z, Borchers C, Schaller M, Jacobson K. JNK phosphorylates paxillin and regulates cell
migration. Nature. 2003; 424(6945): 219–223. PMID: 12853963 45. Fogh BS, Multhaupt HA, Couchman JR. Protein kinase C, focal adhesions and the regulation of cell
migration. J Histochem Cytochem. 2014; 62(3): 172–184. doi: 10.1369/0022155413517701 PMID:
24309511 21 / 22 PLOS ONE | DOI:10.1371/journal.pone.0148642
February 5, 2016 Evaluating Stem Cells with Video Bioinformatics 46. Ke H, Zhang JY, Akiyama SK, French JE. BCL2 interaction with actin in vitro may inhibit cell motility by
enhancing actin polymerization. Cell Adh Migr. 2011; 5(1): 6–10. doi: 10.4161/cam.5.1.13175 PMID:
20716950 47. Barbaric I, Biga V, Gokhale PJ, Jones M, Stavish D, Glen A, et al. Time-Lapse Analysis of Human
Embryonic Stem Cells Reveals Multiple Bottlenecks Restricting Colony Formation and Their Relief
upon Culture Adaptation. Stem Cell Reports. 2014; 3(1): 142–155. doi: 10.1016/j.stemcr.2014.05.006
PMID: 25068128 48. Bradley CK, Scott HA, Chami O, Peura TT, Dumevska B, Schmidt U, et al. Derivation of Huntington’s
disease-affected human embryonic stem cell lines. Stem Cells Dev. 2011; 20(3): 495–502. doi: 10. 1089/scd.2010.0120 PMID: 20649476 49. Miller JD, Ganat YM, Kishinevsky S, Bowman RL, Liu B, Tu EY, et al. Human iPSC-based modeling of
late-onset disease via progerin-induced aging. Cell Stem Cell. 2013; 13(6): 691–705. doi: 10.1016/j. stem.2013.11.006 PMID: 24315443 50. Grandjean P, Bellinger D, Bergman A, Cordier S, Davey-Smith G, Eskenazi B, et al. The faroes state-
ment: human health effects of developmental exposure to chemicals in our environment. Basic Clin
Pharmacol Toxicol. 2008; 102(2): 73–75. doi: 10.1111/j.1742-7843.2007.00114.x PMID: 18226057 22 / 22
|
https://openalex.org/W2593171762
|
https://scindeks-clanci.ceon.rs/data/pdf/0040-2176/2015/0040-21761505804A.pdf
|
Croatian
| null |
Geothermal energy and its application opportunities in Serbia
|
Tehnika
| 2,015
|
cc-by
| 3,267
|
Adresa autora: Nenad Andrić, Institut „Mol”, Stara
Pazova, Nikole Tesle 15
Rad primljen: 22.01.2015.
Rad prihvaćen: 28.07.2015. GEOTERMALNA ENERGIJA I MOGUĆNOSTI NJENE PRIMENE U SRBIJI GEOTERMALNA ENERGIJA I MOGUĆNOSTI NJENE PRIMENE U SRBIJI N. ANDRIĆ Geotermalna energija i mogućnosti njene primene u Srbiji NENAD M. ANDRIĆ, Institut „Mol”, Stara Pazova Stručni rad
UDC: 620.91:550.36(497.11)
DOI: 10.5937/tehnika1505804A Geotermalna energija predstavlja prirodnu toplotu Zemlje akumuliranu u fluidima i stenskim masama
u Zemljinoj kori. Prirodno raspadanje radioaktivnih elemenata (urana, torijuma i kalijuma) koji se
nalaze u svim stenama, proizvodi ogromnu termičku energiju. Najjednostavnija upotreba geotermalne energije za grejanje je pomoću, tzv, otvorenog sistema. U
takvom sistemu geotermalna voda se koristi direktno za grejanje, posredno pomoću toplotne pumpe. Geotermalna energija može se iskoristiti i za proizvodnju električne energije. Tu se koristi vrela voda i
para iz Zemlje za pokretanje generatora turbine. Srbija ima značajne mogućnosti za korišćenje geotermalne energije, s obzirom da po potencijalu spada
u bogatije zemlje. Razvojem geotermalnih resursa do moglo bi da se zameni najmanje 500.000 tona
uvoznih tečnih goriva godišnje. [3] Ukupna količina akumulirane toplote u geotermalnim izvorima do dubine od 3 km oko 2 puta je veća od
ekvivalentne toplote koja bi se mogla dobiti sagorevanjem svih vrsta uglja iz svih njihovih nalazišta u
Srbiji! Ukupna izdašnost svih geotermalnih izvora u Srbiji je 4000 l/s. Najčešća vodoizdašnost bušotina
u Vojvodini je 10-20 l/s, a izlazna temperatura je od 40 do 60 °C. Potencijal geotermalnih izvora u
Mačvi mogli bi da koriste gradovi: Bogatić, Šabac, Sremska Mitrovica i Loznica, sa ukupno 150.000
stanovnika. Najbogatiji hidrogeotermalni resursi su Mačva, Vranjska i Jošanička Banja. Pomoću
toplotnih pumpi moguće je eksploatisati geotermalne vode na skoro celoj teritoriji Srbije! Iako je veliki
proizvođač, Srbija uvozi hranu, tj, sezonsko i vansezonsko voće i povrće. Izgradnjom staklenika koji bi
se grejali geotermalno, Srbija bi mogla da postane zemlja izvoznik. Ključne reči: geotermalna energija, toplota, Republika Srbija 2. GREJANJE POMOĆU GEOTERMALNE
ENERGIJE I PROIZVODNJA ELEKTRIČNE Razlikujemo 4 grupe geotermalnih energetskih iz-
vora:
hidrogeotermalna energija izvora vrele vode,
hidrogeotermalna energija izvora vrele vode, Uglavnom se koriste tri osnovna tipa geotermalnih
elektrana:
hidrogeotermalna energija izvora vodene pare,
hidrogeotermalna energija vrele vode u velikim
dubinama,
Princip suve pare (Dry steam) – koristi se vrela
para iznad 235 °C. Para se koristi za direktno po-
kretanje turbina generatora. Ovo je najjednosta-
vniji i najstariji način i još uvek se koristi jer pre-
dstavlja daleko najjeftiniji princip generisanja ele-
ktrične energije iz geotermalnih izvora. Trenutno
najveća elektrana koja koristi “Dry steam“ princip
nalazi se u severnoj Kaliforniji i zove se The Ge-
ysers, a proizvodi električnu energiju još od 1960. godine. Količina proizvedene električne energije
iz tog postrojenja još uvek je dovoljna za snabde-
vanje grada veličine San Francisco-a.
Princip suve pare (Dry steam) – koristi se vrela
para iznad 235 °C. Para se koristi za direktno po-
kretanje turbina generatora. Ovo je najjednosta-
vniji i najstariji način i još uvek se koristi jer pre-
dstavlja daleko najjeftiniji princip generisanja ele-
ktrične energije iz geotermalnih izvora. Trenutno
najveća elektrana koja koristi “Dry steam“ princip
nalazi se u severnoj Kaliforniji i zove se The Ge-
ysers, a proizvodi električnu energiju još od 1960. godine. Količina proizvedene električne energije
iz tog postrojenja još uvek je dovoljna za snabde-
vanje grada veličine San Francisco-a.
petrotermička energija, tj, energija vrelih i suvih
stena. Prema temperaturi geotermalnog medija, geoter-
malni izvori mogu biti: niskotemperaturni (sa gornjom
granicom temperature između 90 i 150 °C), srednje-
temperaturni (temperatura od najmanje 90 do najviše
150 °C), visokotemperaturni (s donjom granicom tem-
perature između 150 i 225 °C). Najjednostavnija upotreba geotermalne energije za
grejanje je pomoću, tzv, otvorenog sistema. U takvom
sistemu geotermalna voda se koristi direktno za gre-
janje, ili posredno pomoću toplotne pumpe. Voda mora
biti zadovoljavajućeg kvaliteta i mora postojati mo-
gućnost vraćanja vode u bušotinu.
Flash princip (Flash steam) – koristi se vrela voda
iz geotermalnih izvora koja je pod velikim pri-
tiskom i na temperaturama iznad 182 °C.
Flash princip (Flash steam) – koristi se vrela voda
iz geotermalnih izvora koja je pod velikim pri-
tiskom i na temperaturama iznad 182 °C.
Binarni princip (Binary cycle) – kao radna ma-
terija koriste se fluidi uobičajeni u rashladnoj te-
hnici (supstance koje isparavaju na nižim tempe-
raturama i na pritiscima iznad 1 bar). 1. UVOD
zahvaljujući sistemu grejanja pomoću toplotne
pumpe, osećaj ugodnosti u bilo kom delu grejanog
prostora omogućava da se temperatura prostora
održava na dva do tri stepena nižoj temperaturi u
odnosu na druge klasične sisteme grejanja, Proizvodnja električne energije putem geoter-
malne energije započela je još 1913. godine.
snižavanje temperature životnog ili radnog ambi-
jenta u proseku za 1°C donosi energetsku uštedu
od 5 do 6%, Geotermalni izvori postoje u preko 90 država, ko-
riste se u preko 70 država, a električna energija se
proizvodi u 25 država. Ukupno korišćenje geotermalne
energije 2004. godine bilo je oko 55 TWh električne
energije. Na Islandu se 89% domaćinstava greje po-
moću geotermalne energije. [1]
jednokratnom investicijom može da se reši i
grejanje i hlađenje prostora. [4] Geotermalna energija može se iskoristiti i za proi-
zvodnju električne energije. Tu se koristi vrela voda i
para iz Zemlje za pokretanje turbine generatora. Nema
spaljivanja fosilnih goriva niti emisije štetnih gasova u
atmosferu, ispušta se samo vodena para. Dodatna pre-
dnost je u tome što se takve elektrane mogu izgraditi u
najrazličitijim okruženjima, od farmi i pustinja pa sve
do šumskih područja. 1. UVOD obnovljivi izvor, iako je u suštini neobnovljiv. Tople i
suve stene, odnosno magma, nalaze se u nepropusnim
slojevima na velikim dubinama i imaju visoku tempe-
raturu, između 700 i 1200 °C. Izvori suve vodene pare
na svetu su retki, ali se smatraju najjednostavnijima i
najisplativijima za korišćenje, jer se prirodna suva vo-
dena para može koristiti za pogon parne turbine. Ležišta vode i gasova pod visokim pritiskom nalaze se
na velikim dubinama (od 3000 do 6000 m). Voda je pri
tome umerene temperature (između 90 i 200 °C) i
sadrži otopljeni metan. Zahvaljujući vrlo visokim
pritiscima bilo bi moguće koristiti mehaničku, toplotnu
pa i hemijsku energiju. Izvori tople ili vrele vode
(gejziri) najčešći su i najprepoznatljiviji način dolaska
zagrejane vode iz dubine na površinu Zemlje. Potiču
od vrele vode ili pare koja se nalazi zarobljena u
razlomljenom i poroznom stenju, na manjim ili sred-
njim dubinama (od 100 do 4500 m). [4] Geotermalna energija predstavlja prirodnu ener-
giju akumuliranu u fluidima i stenskim masama u Ze-
mljinoj kori. Zemljina kora sastoji se od stena, vode i
magme, a geotermalna energija akumulirana je u nji-
ma. [3] Prirodno raspadanje radioaktivnih elemenata (ura-
na, torijuma i kalijuma) koji se nalaze u svim stenama,
proizvodi ogromnu termičku energiju. Geotermalna
energija može se smatrati fosilnom nuklearnom ener-
gijom, ona je, tzv, unutrašnja toplotna energija. U zemljinoj kori postoje nalazišta termalnih voda i
vodene pare, toplih i vrelih stena i nalazišta magme. Spuštanjem kroz Zemljinu koru temperatura raste, otp-
rilike, 17 do 30 °C po kilometru dubine. Geotermalni
potencijali Zemlje su ogromni i zato se ona tretira kao Područja koja imaju najveći broj geotermalnih
izvora su istovremeno i ona koja su geološki vrlo ak-
tivna, tj, koja imaju aktivne vulkane ili u kojima često TEHNIKA – MAŠINSTVO 64 (2015) 5 804 GEOTERMALNA ENERGIJA I MOGUĆNOSTI NJENE PRIMENE U SRBIJI N. ANDRIĆ dolazi do potresa. To su područja oko Tihog okeana
(tzv. Pacifički vatreni krug: zapadni delovi SAD i
Kanade, Srednja Amerika, zapadne obale Južne Ame-
rike, Novi Zeland, Indonezija, Filipini, Japan i istočni
Sibir), srednje atlanski greben (Island), planinski lanci
kao što su Alpi i Himalaji, istočna Afrika, srednja Azi-
ja i neka ostrva u Tihom okeanu.
toplotna pumpa može da se koristi i za grejanje i
za hlađenje. Tokom zimske sezone, toplotna pum-
pa se koristi za grejanje, a leti za hlađenje. 2. GREJANJE POMOĆU GEOTERMALNE
ENERGIJE I PROIZVODNJA ELEKTRIČNE Da bi se podzemne vode smatrale toplotnim iz-
vorom potrebno je da njihove temperature budu više
od 10 °C. Ako se koristi geotermalna pumpa, korišćenje geo-
termalne energije pruža sledeće prednosti: Princip koji će se koristiti kod izgradnje nove ele-
ktrane zavisi od vrste geotermalnog izvora, tem-
perature, dubine i kvaliteta vode i pare. U svim
slučajevima kondenzovana para i ostaci geotermalne
tečnosti vraćaju se nazad u bušotinu i time se povećava
izdašnost
geotermalnog izvora,
štiti okolina od
zagađenja solima i sprečava sleganje tla. [2]
ako se voda koristi direktno (tj, kada je tempera-
tura dovoljno visoka, npr, preko 50° C) troškovi za
grejanje su zanemarljivi. U slučaju da se koristi
toplotna pumpa, sezonski koeficijent grejanja mo-
že da bude 3-4, pa se preko 70% potrebne toplote
za grejanje može uštedeti. Tada treba imati u vidu
da toplotna pumpa troši električnu energiju za
pogon. 3. GEOTERMALNA ENERGIJA U SRBIJI Srbija ima značajne mogućnosti za korišćenje geo-
termalne energije, s obzirom da po potencijalu spada u
skoro potpuno ekološki način korišćenja energije,
skoro potpuno ekološki način korišćenja energije, TEHNIKA – MAŠINSTVO 64 (2015) 5 805 GEOTERMALNA ENERGIJA I MOGUĆNOSTI NJENE PRIMENE U SRBIJI N. ANDRIĆ bogatije zemlje. Nažalost, koristimo samo 86 MW
geotermalne energije. Srbija bi mogla da 55% ukupno
proizvedene energije dobija iz obnovljivih izvora. [3]
Od celokupne energije, 4% bi se moglo dobiti iz
geotermalnih izvora. Razvojem geotermalnih resursa
moglo bi da se zameni najmanje 500.000 tona uvoznih
tečnih goriva godišnje. Stalni deficit fosilnih goriva,
loša ekološka situacija i porast troškova zaštite životne
sredine mogli bi uticati na intenzivnije korišćenje ovog
resursa. Korišćenje geotermalne energije imalo bi
najveći značaj za grejanje u gradu i selu. [4] snage 5 MW koriste otpadne termalne vode tempe-
rature 25°C (sistem je najveći u Evropi) za grejanje
hotela i rehabilitacionog centra. Isti princip primenjen
je i u Prolom Banji, samo što je temperatura vode oko
30°C. [1] 30°C. [1]
Tabela 1. Korišćenje hidrogeotermalne energije u Srbiji
po vrsti primene [1]
Vrsta primene
Instalisana
toplotna snaga
(MW)
Proizvedena
toplota
(TJ/godina)
Zagravanje prostorija
18.5
575
Balneologija i
rekreacija
36.0
1150
Sušenje žitarica
0.7
22
Staklenici
8.4
256
Ribarstvo i stočarstvo
6.4
211
Industrijski procesi
3.9
121
Zagrevanje sa
toplotnim pumpama
12.0
80
Ukupno
86.0
2415
Tabela 2. Temperature podzemnih voda u nekim mestima
Mesto
Temperature podzemnih
voda (°C)
Bogatić
80
Debrc
58
Lukovska banja
80
Inđija
62
Kupinovo
54
Mladenovac, Banja Selters
54 Tabela 1. Korišćenje hidrogeotermalne energije u Srbiji
po vrsti primene [1] Tabela 1. Korišćenje hidrogeotermalne energije u Srbiji
po vrsti primene [1]
Vrsta primene
Instalisana
toplotna snaga
(MW)
Proizvedena
toplota
(TJ/godina)
Zagravanje prostorija
18.5
575
Balneologija i
rekreacija
36.0
1150
Sušenje žitarica
0.7
22
Staklenici
8.4
256
Ribarstvo i stočarstvo
6.4
211
Industrijski procesi
3.9
121
Zagrevanje sa
toplotnim pumpama
12.0
80
Ukupno
86.0
2415 Gustina geotermalnog toka je glavni parametar na
osnovu koga se procenjuje geotermalni potencijal ne-
kog područja. Ona predstavlja količinu geotermalne
toplote koja svakog sekunda kroz površinu od 1 m2
dolazi iz unutrašnjosti Zemlje do njene površine. Na
najvećem delu teritorije Srbije gustina je veća od njene
prosečne vrednosti za kontinentalni deo Evrope, koji
iznosi 60 mW/m2. Panonski basen, južna i centralne Srbija imaju vre-
dnost gustine preko 100 mW/m2. 3. GEOTERMALNA ENERGIJA U SRBIJI Van Panonskog
basena nalazi se 160 geotermalnih izvora sa tempera-
turom većom od 15°C i 60 nalazišta sa temperaturom
većom od 15°C do dubine od 3 km. Neki od tih geo-
termalnih izvora su: Vranjska Banja (96-110°C), Joša-
nička Banja (78°C), Sijarinska Banja (72 °C). Vranjska
Banja je najtoplija banja u Evropi. [1] Tabela 2. Temperature podzemnih voda u nekim mestima Tabela 2. Temperature podzemnih voda u nekim mestima
Mesto
Temperature podzemnih
voda (°C)
Bogatić
80
Debrc
58
Lukovska banja
80
Inđija
62
Kupinovo
54
Mladenovac, Banja Selters
54 Ukupna količina akumulirane toplote u geoter-
malnim izvorima do dubine od 3 km oko 2 puta je veća
od ekvivalentne toplotne energije koja bi se mogla
dobiti sagorevanjem svih vrsta uglja iz svih njihovih
nalazišta u Srbiji! [1] Rezerve termalne vode su na različitim dubinama:
od 207 m u Bogatiću do 600-700 m u Sremu. Pote-
ncijal geotermalnih izvora u Mačvi mogli bi da koriste
gradovi: Bogatić, Šabac, Sremska Mitrovica i Loznica,
sa ukupno 150.000 stanovnika (tablica 2). Najbogatiji
hidrogeotermalni resursi su Mačva, Vranjska i Joša-
nička Banja. Pomoću toplotnih pumpi moguće je eks-
ploatisati geotermalne vode na skoro celoj teritoriji
Srbije! [1] Ukupna izdašnost svih geotermalnih izvora u Sr-
biji je 4000 l/s. Izdašnost 62 geotermalne bušotine u
Vojvodini je oko 550 l/s, a toplotna snaga oko 50 MW. Ostatak Srbije poseduje 48 bušotina snage oko 108
MW. Najčešća vodoizdašnost bušotina u Vojvodini je
10-20 l/s, a izlazna temperatura je od 40 do 60 °C. Najdublja bušotina (2520 m) je u Banatu, u mestu
Vrbica, temperatura vode je 82 °C na dubini od oko
1800 m. Najplića se nalazi u Novom Sadu, na 305 m,
temperatura vode je 25 °C. Vojvodina koristi geoter-
malnu energiju još od 1981. godine. U Srbiji se eks-
ploatiše oko 25% postojećih bušotina, ukupne snage
oko 19 MW. Snaga 75% nekorišćenih bušotina proce-
njuje se na oko 54 MW. [2] GEOTERMALNA ENERGIJA I MOGUĆNOSTI NJENE PRIMENE U SRBIJI N. ANDRIĆ konkurentnosti ovog energetskog izvora u odnosu na
fosilna goriva i nuklearnu energiju predstavlja
prepreku za razvoj geotermalne tehnologije. Struja iz
obnovljivih izvora energije može se prodavati samo uz
saglasnost Vlade Srbije. U svetu je dokazano da je
geotermalna energija konkurentski energetski izvor.
Ne koriste se na mestu gde se nalaze termalni
fluidi.
Vrlo visoki inicijalni troškovi (opitne bušotine). Prema podacima američkog ministarstva energije
povrat inicijalnih troškova instalacije može biti od
2 do čak 20 godina! [3]
Vrlo visoki inicijalni troškovi (opitne bušotine). Prema podacima američkog ministarstva energije
povrat inicijalnih troškova instalacije može biti od
2 do čak 20 godina! [3] Iako je veliki proizvođač, Srbija uvozi hranu, tj,
sezonsko i vansezonsko voće i povrće. Izgradnjom sta-
klenika koji bi se grejali geotermalnom energijom, Sr-
bija bi mogla da postane zemlja izvoznik. Geotermalno grejanje nije u potpunosti 100% čisti
oblik grejanja zbog toplotnih pumpi koje za pogon
koriste električnu energiju, a ta električna energija se
proizvodi iz ekološki neprihvatljivog uglja, pa se ipak
u atmosferu oslobađa određena količina gasova
staklene bašte. [1] Za jednu pijezometarsku bušotinu, čiji je cilj
istraživanje geotermalnih rezervi, potrebno je izdvojiti
oko 300 hiljada evra. Oko 200 hiljada evra je potrebno
za projektnu dokumentaciju, opremanje i zasnivanje
proizvodnje. 6. ZAKLJUČAK Ukoliko neko može priuštiti ovakav vid grejanja
(zbog velikih inicijalnih troškova), dugoročno će se
investicija svakako isplatiti. Ovo se posebno odnosi na
nisku cenu grejanja i izvrsnu efikasnost sistema greja-
nja koji iskorištava geotermalnu energiju. Takvo ulaganje, na najmanje 5 ha, može da donese
dobit već posle 2 do 3 godine, u zavisnosti od
proizvodnje i plasmana. Bušotine sa temperaturama
iznad 40 °C i izlivom oko 60 m3/h isplative su pod
uslovom da se koristi ceo potencijal bušotine na preko
6000 ha. Srbija se po pitanju obnovljivih izvora energije
konačno pokrenula sa mrtve tačke, ali to je još uvek
nedovoljno za bilo kakav iole ozbiljniji poduhvat. Pre
svega, treba maksimalno skratiti i ubrzati birokratiju i
edukovati stanovništvo o novim izvorima energije. Nesvatljivo je da cela Mačva leži na geotermalnim
izvorima, a da to još uvek nije iskorišćeno na pravi
način. Takođe, oblast oko Vranja i Vranjske Banje mo-
gla bi se bez problema snabdevati toplotnom ener-
gijom iz geotermalnih izvora. Država bi morala ozbi-
ljnije da eksploatiše geotermalne resurse, umesto da
ulaže u fosilna goriva koja imaju ograničene resurse. Takođe, dok je niska cena struje, neće biti isplativo
koristiti geotermalne izvore za generisanje električne
energije. Priča o obnovljivim izvorima energije u Srbiji
se pokrenula i ide polako, nadajmo se da će se u skorije
vreme ubrzati. “Ceo potencijal” podrazumeva hlađenje vode do
oko 15 °C, za šta je potrebna toplotna pumpa. Takvo
korišćenje geotermalne energije može ostvariti samo
potrošač koji će (pored grejne sezone) energiju koristiti
i u proleće i jesen, kao i noću tokom grejne sezone. Ovakva analiza znatno smanjuje broj realnih potro-
šača. [3] Cena bušotine zavisi od njene dubine i kreće se od
220.000 do 500.000 eura za dubine bušotina od 600 do
1100 m. Za postrojenje snage 1 do 100 MW cena
energije, dobijene iz geotermalnih izvora, kreće se od
3.0 do 5.5 ceur/kWh. Cene konkurentnih energija su:
prirodni gas 7.0 ceur/kWh,
prirodni gas 7.0 ceur/kWh,
prirodni gas 7.0 ceur/kWh,
prirodni gas 7.0 ceur/kWh,
električna energija oko 6.0 ceur/kWh,
toplotna energija iz toplana 7.0 ceur/kWh. [2] 4. ISPLATIVOST KORIŠĆENJA
GEOTERMALNIH RESURSA Korišćenje geotermalnih resursa u našoj zemlji je
veoma malo u odnosu na raspoloživi potencijal, iako
Srbija ima pojedine lokalitete među najboljima u
Evropi! Razvoj geotermalne tehnologije u Srbiji počeo
je u isto vreme kada i u zemljama Evrope u kojima je
ova tehnologija danas na najvišem stepenu razvoja. Geotermalna energija u Srbiji uglavnom se koristi
za zagrevanje prostorija i u banjskim lečilištima (bal-
neologija). Pored toga, koristi se i za grejanje stakle-
nika u proizvodnji ranog povrća, grejanje farmi za
uzgoj svinja, proizvodne procese u fabrikama kože i
tekstila itd. (tablica 1). Poznata banjska lečilišta su
Soko Banja, Lukovska Banja, Niška Banja, Prolom
Banja i dr. (tablica 2). U Niškoj Banji toplotne pumpe Razvoj bilo kog novog energetskog resursa nemo-
guće je početi bez odgovarajućih zakonskih i finan-
sijskih mera. Potrebno je obezbediti povoljne kredite,
subvencije i fondove. Nedovoljna informisanost o TEHNIKA – MAŠINSTVO 64 (2015) 5 806 GEOTERMALNA ENERGIJA I MOGUĆNOSTI NJENE PRIMENE U SRBIJI GEOTERMALNA ENERGIJA I MOGUĆNOSTI NJENE PRIMENE U SRBIJI GEOTERMALNA ENERGIJA I MOGUĆNOSTI NJENE PRIMENE U SRBIJI N. ANDRIĆ 5. PREDNOSTI I MANE GEOTERMALNE
ENERGIJE [1] Liber Perpetuum, OSCE Mission to Serbia and Mo-
ntenegro, Econimic and Environmental department,
Belgrade, 2004. Od svih obnovljivih izvora energije geotermalna
energija je možda najbolje rešenje, jer vetar nije uvek
pouzdan, Sunce ne može da osvetljava solarne ploče
noću, a i hidroenergija može da podbaci ako nema
dovoljno padavina. Dakle, nezavisna je od godišnjih
doba i doba dana. Ovo je ujedno i najveća prednost
ovog obnovljivog izvora energije. [1] [2] Mogućnosti korišćenja energetskog potencijala geo-
termalnih voda u Vojvodini, Fakultet tehničkih na-
uka, Institut za energetiku, procesnu tehniku i zaštitu
okoline, Novi Sad, 2005. [3] www.well.org.rs [3] www.well.org.rs Glavni nedostaci ovog obnovljivog resursa su
sledeći: [4] www.izvorienergije.co TEHNIKA – MAŠINSTVO 64 (2015) 5 807 GEOTHERMAL ENERGY AND ITS APPLICATION OPPORTUNITIES IN SERBIA Geothermal energy is accumulated heat in the fluid and rock masses in the Earth 's crust. The natural
decay of radioactive elements (uranium, thorium and potassium) in rocks produces heat energy. The simplest use of geothermal energy for heating is by heat pump. Geothermal energy can be used for
production of electricity. It uses hot water and steam from the earth to run the generator. Serbia has significant potential for geothermal energy. The total amount of accumulated heat in
geothermal resources in a depth of 3 km is two times higher than the equivalent thermal energy that
could be obtained by burning all types of coal from all their sites in Serbia! The total abundance of
geothermal resources in Serbia is 4000 l/s. Abundance of wells in Vojvodina is 10-20 l/s, and the
temperature is from 40 to 60 °C. Exploitation of thermal waters in Mačva could cause heating of
following cities: Bogatić, Šabac, Sremska Mitrovica and Loznica, with a total population of 150.000
people. The richest hydrogeothermal resources are in Mačva, Vranje and Jošanička Banja. Using heat
pumps, geothermal water can be exploited on the entire territory of Serbia! Although large producer , Serbia is importing food, ie., fruits and vegetables. With the construction of
of greenhouses, which will be heated with geothermal energy, Serbia can become an exporting country. Key words: geothermal energy heat Republic of Serbia TEHNIKA – MAŠINSTVO 64 (2015) 5 808
|
https://openalex.org/W4391767026
|
https://etj.uotechnology.edu.iq/article_182396_a8a2bb2b5bdc86bc1028fec1404547de.pdf
|
English
| null |
Performance of potato starch admixture on fresh and hardened behaviours of concrete at varied mix design ratios
|
Maǧallaẗ al-handasaẗ wa-al-tiknūlūǧiyā
| 2,024
|
cc-by
| 6,174
|
Performance of potato starch admixture on fresh and hardened
behaviours of concrete at varied mix design ratios Chidobere D. Nwa-David Civil Engineering Dept., Michael Okpara University of Agriculture, Umudike, Abia State, Nigeria. *Corresponding author Email: nwadavid.chidobere@mouau.edu.ng http://doi.org/10.30684/etj.2024.145272.1654
Received 16 December 2023; Received in revised form 26 January 2024; Accepted 28 January 2024; Available online 12 February 2024
2412-0758/University of Technology-Iraq, Baghdad, Iraq
This is an open access article under the CC BY 4.0 license http://creativecommons.org/licenses/by/4.0 H I G H L I G H T S Mix design ratios were varied with different
percentage fractions of potato starch dosage. The effect of potato starch admixture on the
fresh and hardened concrete properties was
investigated. g
The optimum percentage of potato starch
content for all mixes was stipulated. The optimum percentage of potato starch
content for all mixes was stipulated. H I G H L I G H T S Potato starch was adopted as a natural
admixture in concrete. The urgency to attenuate techno-economic and ecological threats constituted by
chemical admixtures has necessitated the adoption of feasible natural admixtures. The chemical interaction of admixtures with concrete ingredients affects
concrete's plastic and hardened structure. The impact of concrete properties on
construction works is of great concern to engineers as it plays a vital role towards
viable infrastructural development. This study focused on developing an eco-
friendly, sustainable structural concrete using potato starch (PS) as an admixture. PS has been included to in concrete in discrete percentages (0%, 0.4%, 0.8%,
1.2%, 1.6%, 2.0%) of cement weight employing mix ratios of 1: 1.5: 2, 1:2:4,
1 3 6
d
i
f 0 6 Thi
d
l
h
ff
f PS The urgency to attenuate techno-economic and ecological threats constituted by
chemical admixtures has necessitated the adoption of feasible natural admixtures. The chemical interaction of admixtures with concrete ingredients affects
concrete's plastic and hardened structure. The impact of concrete properties on
construction works is of great concern to engineers as it plays a vital role towards
viable infrastructural development. This study focused on developing an eco-
friendly, sustainable structural concrete using potato starch (PS) as an admixture. PS has been included to in concrete in discrete percentages (0%, 0.4%, 0.8%,
1.2%, 1.6%, 2.0%) of cement weight employing mix ratios of 1: 1.5: 2, 1:2:4,
1:3:6 and water-cement ratio of 0.6. This study evaluates the effect of PS-
admixture on concrete properties in the plastic state (setting time and
workability) and at the hardened state (compressive strength) at 7, 21, 28, 56, and
90 days of curing. Potato starch (PS) admixture showed higher compressive
strengths in concrete mixes with lower ratios (1:2:4 and 1:3:6). The strength test
results for the 1:1.5:2, 1:2:4 and 1:3:6 mix ratios show a maximum strength of
32.77 N/mm2, 35.12 N/mm2 and 38.97 N/mm2 respectively for 28 curing days, at
1.2% PS addition, after which there was a strength-reduction. Rise A rise in the
percentage inclusion of PS boosts the setting- time of cement paste and
compressive strength. An optimal- percentage of 1.2% PS by cement- weight is
advised for reliable concrete strength performance. Utilization of PS in concrete
is recommended and should be embraced, considering its low- cost, availability,
environmental amicability, and ability to enhance structural durability. Handling editor: Mahmoud S. Al-Khafaji Potato starch A R T I C L E I N F O Handling editor: Mahmoud S. Al-Khafaji
Keywords:
Concrete
Mix ratio
Admixtures
Potato starch
Compressive strength Engineering and Technology Journal 42 (05) (2024) 540-547 Engineering and Technology Journal 42 (05) (2024) 540-547 1. Introduction y
p
p
Potatoes (Solanum tuberosum) are annual underground tubers that grow on the roots of the potato plant. They are grown in
torrid nations due to their roots, which contain high starch content. The starch extracted from potatoes is regarded as potato
starch (PS). This extraction involves crushing the potatoes and the starch grains produced from the destroyed cells. Potato
starch is a thickener in many recipes because of its water-absorption tendency. PS could serve as a good fried food coating and
a suitable gluten-free flour alternative in baking recipes. This study examines the effect of PS as a substitute for chemical
admixtures in concrete production. Several studies have been done on the use of chemical admixtures as well as their supplementary materials. Adetayo and
Jubril [20] ascertained the essence of Ripe Plantain peel ash (RPPA) and Unripe Plantain peel ash (UPPA) on concrete. At 0%,
1.5%, and 2.5% by weight, RPPA and UPPA were added as admixture and blended with other constituents. Grade M15 was
addressed with a 1:2:4 mix and water-cement ratio of 0.7. Their study showed that the workability of concrete rises with the
addition of RPPA and UPPA. The compressive-strength values of RPPA at 28 and 90 days are higher than that of UPPA at
1.5% and 2.5% proportions. Andrew et al. [6] examined the impact of Na2CO3-admixture on the properties of concrete at
assorted mix ratios. Na2CO3 was added at a 0.5% interval from 0 to 2.0% by weight of OPC. The strength tests were done after
7, 28, 56, and 90 days of curing. Their results proved that Na2CO3 has an accelerating influence on setting time and enhanced
its early strength gain but had opposite long-term effects on strength with increasing dosage. The authors recommended an
optimal percentage of 1% Na2CO3 for accelerating effects in concrete. Oni et al. [9] evaluated the relevance of cassava starch
(CS)-admixture in concrete. The authors carried out their study by adding CS by weight of cement at 0.4% to 2.0%. CS
improved concrete properties at an optimum of 0.8% CS. They confirmed CS as a retarding-admixture in hot-weather-
concreting because CS increased the initial and final setting time of concrete at each rise in percentage addition. g
g
p
g
Sybis and Konował [7] analyzed the impact of the inclusion of restructured starches on rheology and the preferred
physicochemical characteristics of cement pastes and mortars. 1. Introduction Concrete strength assessment is crucial to eco-friendly infrastructural buildout, and compressive strength is the principal
index as it determines the quality of construction works [1-5]. This study is imperative to circumvent the tragedies of structural
failure emanating from poor concrete properties. The current study has formulated an experimental program to investigate the
novelty of adopting a natural biopolymer (potato starch) as an admixture and evaluating its effects on the workability, setting
time, and compressive strength of concrete at varied curing periods. Admixtures are additional substances incorporated into the concrete mix during or immediately before mixing to modify
the quality of concrete. They are commonly embraced to advance refined concrete works that can perform satisfactorily under
different loading conditions in the advent of challenging environmental factors. As they interact chemically with the
constituents of concrete, they impel the characteristics of concrete. Their impact varies with the cement type, mix ratio, and
quantity. Their applications in engineering works are on the increase due to their edge such as ameliorated workability, better
frost and sulfate-retardant, fireproof, cracking-control, acceleration or retardation in setting-time, high concrete strength,
increased flow for the equal water-cement ratio, coloring and improved quality [6-8]. Admixtures are categorized as either
mineral or chemical. Aqueous materials added to cement to transform concrete traits are regarded as chemical admixtures. These include accelerating, water-reducing, retarding admixtures and those combining two or more features [9-11]. Mineral
admixtures are pozzolanic inorganic auxiliary-cementitious elements that can boost essential attributes of concrete, such as 540 Engineering and Technology Journal 42 (05) (2024) 540-547 Chidobere D. Nwa-David durability, strength, and permeability. They include rice husk ash, sawdust ash, ground-granulated-blast furnace slag, silica
fume, and cassava peel ash [12-15]. durability, strength, and permeability. They include rice husk ash, sawdust ash, ground-granulated-blast furnace slag, silica
fume, and cassava peel ash [12-15]. The exigency of minimizing dependence on chemical admixtures and the issue of their availability and cost in developing
areas has led researchers to investigate and deploy naturally occurring admixtures that are abundantly accessible, cost-
effective, eco-friendly, and can be produced locally. Such alternatives include cassava starch, maize starch, corn starch, black
liquor, Gum Arabic Karroo, and broiler hen egg [16-19]. More studies are required on these local alternatives, which validates
the need for this study as potato starch can also be adopted. 1. Introduction The authors ascertained the feasibility of adopting customized
starches as plasticizers. Their study showed a reasonable deduction in the yield points and plastic viscosities of the cement
slurries and a rise in the flow diameters of the cement mortars and concretes due to all the tested starch hydrolysates. PS that
were not starch hydrolysates revealed the thickening of the cement mix and increased yield point. Adam et al. [21] showed that
concrete mixes containing cassava flour were stronger than the control mix, with a 3% cassava flour blend generating the best
results. The authors examined the compressive, split-tensile, and flexural strengths of concrete made with cassava flour up to
5% by weight of cement, using the water-cement ratio of 0.35 and a carboxylate-based superplasticizing admixture. Akindahunsi and Uzoegbo [22] adopted cassava and maize starch as admixtures in concrete. They considered the effect of
these starches on the sorptivity, oxygen-permeability, and strength of concrete using 0.5% to 2.0% of the starches. Incorporating admixtures reduced the cement paste's setting time but improved the concrete's strength and durability. The distinction of this study lies in the material adopted as admixture, water-cement ratio, percentage admixture addition,
mix ratios, and concrete properties addressed. This study focused on adopting a fully green natural plasticizer that participates
in ecosystem protection and sustainability. Considering its thickness-refinement and deceleration ability, investigating PS as an
admixture is worth addressing. This study intends to evaluate the effect of PS-admixture on the characteristics of concrete at
discrete mix ratios. Chidobere D. Nwa-David Table 3 shows that the grading of the fine aggregate lies between the acceptable limits as stated in BS 882_ [26]. Table 4
shows that the distribution of the coarse aggregate is between the upper and lower limits, respectively, as stipulated in BS 882_
[26]. Sand and granite have coefficients of uniformity and curvature values of 2.67 and 1.83, 0.96 and 1.05, respectively,
obtained from Tables 3 and 4. This implies that the aggregates are well-graded and suitable for workable concrete. The nature
of their gradation minimizes the tendency for concrete to bleed, segregate, and experience plastic shrinkage cracking. 2.1 Materials The materials employed in this study include Grade 42.5N, BUA brand of Ordinary Portland Cement (OPC) which
conformed to the specifications of BS EN 17075_ [23], fine aggregate (river sand), coarse aggregate with a maximum size of
14 mm, clean and drinkable water free from impurities by BS EN 197-1_ [24], and potato starch (PS) with dosage ranging
from 0.4% to 2.0% by cement weight. The potatoes were procured from a local market at Ndioru, Ikwuano, Abia State of
Nigeria. PS was dried directly under the sun before use. The starch constituents were activated with hot water at a temperature
of 80 oc for 30 minutes to ensure a complete breakdown of the intermolecular bonds of the starch molecules. The required
percentage proportion of PS by weight of cement was separately prepared and allowed to cool to room temperature before
mixing with the concrete. The chemical composition of the starch is presented in Table 1. g
p
p
Table 2 captured the physical attributes of cement, aggregates, and admixture. Tables 3 and 4 show the particle distribution
of the aggregates. The specific surface area of the cement (4150 cm2/g) was less than that of potato starch (4932 cm2/g), which
indicates that the admixture is finer than that of cement. There will be a retardation tendency during the induction phase of the
hydration of cement as the granules of PS are adsorbed onto the cement. The reactivity of tricalcium aluminate is inhibited in
this process, while the formation of ettringite in concrete is reduced. This reduction enhances the production of concrete with
high resistance to sulphate attack. The mean size of PS grains is higher than that of Oni et al. [25] (15.27µm) and Akindahunsi
and Uzoegbo [22] (14.29 µm). 541 Engineering and Technology Journal 42 (05) (2024) 540-547 Chidobere D. Nwa-David Chidobere D. Nwa-David Nwa-David 2.2.1 Experimental program The experimental program includes tests on the aggregates, cement, plastic, and hardened concrete with and without the
admixture. The workability of the fresh concrete was measured using a slump test in accordance with BS EN 12350-2_ [28]. Cured concrete cube specimens were subjected to compressive strength tests at 7, 21, 28, 56, and 90 days curing ages in
accordance with BS EN 12390-3_ [29]. 2.2.2 Design mix Concrete mixes were produced using (150×150×150) mm standard molds in ratios 1:1.5:2, 1:2:4, and 1:3:6 with the water-
binder ratio at 0.6 as the pre-established constant optimum water-binder ratio. In the absence of the admixture, control
specimens were first prepared. Subsequently, test specimens were made by adding PS to different mixes of 0.4% to 2%,
employing a 0.4% increase in dosage by cement weight. The admixture was first dissolved in water when the aqueous solution
was heated to 180 0C. Then, the dissolved liquids were included in the concrete mix. Concrete cube specimens were prepared
from the various mixes and cured in water for 7, 21, 28, 56, and 90 days. These curing durations were selected to examine the
rate of strength development, resistance to freezing and thawing, volume stability, and durability of the concrete in
construction works. The mix proportions for the experimental study were captured in Table 5. Table 5: Mix proportion of batching materials Table 5: Mix proportion of batching materials
Mix ratio
Mixture
Admixture(%)
Cement
(Kg/m3)
Sand
(Kg/m3)
Granite
(Kg/m3)
Water-
cement ratio
1:1.5:2
X0
0
654.60
982.00
1309.30
0.6
X1
0.4
654.72
982.08
1309.45
X2
0.8
655.53
983.30
1311.06
X3
1.2
656.49
984.74
1312.98
X4
1.6
657.51
986.40
1315.03
X5
2.0
657.56
986.34
1315.12
1:2:4
Y0
0
494.28
988.56
1977.12
0.6
Y1
0.4
495.75
991.50
1983.00
Y2
0.8
496.23
992.46
1984.92
Y3
1.2
496.84
993.68
1987.36
Y4
1.6
497.26
994.52
1989.04
Y5
2.0
498.21
996.42
1992.84
1:3:6
Z0
0
308.42
925.26
1850.52
0.6
Z1
0.4
321.02
963.06
1926.12
Z2
0.8
317.15
951.45
1902.90
Z3
1.2
328.42
985.26
1970.52
Z4
1.6
330.02
990.06
1980.12
Z5
2.0
324.19
972.56
1945.14 Chidobere D. Nwa-David Table 1: Chemical properties of potato starch
Symbol
Element
Concentration (ppm)
Si
Silicon
50.3
Al
Aluminum
16.2
Mg
Magnesium
80.6
P
Phosphor
42.5
Na
Sodium
550.7 Table 1: Chemical properties of potato starch
Symbol
Element
Concentration (ppm)
Si
Silicon
50.3
Al
Aluminum
16.2
Mg
Magnesium
80.6
P
Phosphor
42.5
Na
Sodium
550.7 Table 1: Chemical properties of potato starch Table 2: Physical properties of concrete constituents Table 2: Physical properties of concrete constituents
Material
Property
Value
Specification ASTM C 33_ [27]
Fine Aggregate
Moisture Content
3.54%
0 - 4%
Water Absorption
2.65%
< 4%
Bulk density
1533.7
1200 – 1800 kg/m3
Specific Gravity (SSD)
2.54
2.4 -2.9
Course Aggregate
Moisture Content
3.49%
0 - 4%
Water Absorption
3.01%
< 4%
Bulk density
1421.2
1200 – 1800 kg/m3
Specific Gravity (SSD)
2.54
2.4 -2.9
Cement
Initial Setting time
132
> 60 minutes
Final Setting time
243
< 375 minutes
Specific gravity
3.12
3.10 – 3.15
Fineness
8.4%
-
Specific surface area
4150 cm2/g
-
Potato Starch (PS)
Mean Size
15.32 µm
-
Median Size
14.21 µm
-
Specific surface area
49322/g
-
Table 3: Grain size distribution of sand
Sieve size (mm)
Mass of sand passing (g)
Mass of sand Retained (g)
% passing
4.75
760
0
100
2.36
645.32
114.68
84.91
1.18
467.56
177.76
61.52
0.850
324.18
143.38
42.65
0.6
204.46
119.72
26.90
0.425
97.18
107.28
12.78
0.3
60.51
36.67
7.96
0.212
34.65
25.86
4.56
0.15
12.45
22.20
1.64
0.075
5.64
6.81
0.74
Pan
0
5.64
0
Total
760
Table 4: Grain size distribution of gravel
Sieve size (mm)
Mass of granite passin
(g)
Mass of granite retained (g)
% passing
31.5
3560
0
100
22.4
3212.37
347.63
90.24
19
2856.34
356.03
80.24
16
1468.23
1388.11
41.25
12.5
820.72
647.51
23.06
9.5
469.56
351.16
13.20
6.3
78.41
391.15
2.21
4.75
22.35
56.06
0.63
Pan
0
0
0
Total
3560 Table 2: Physical properties of concrete constituents
Material
Property
Value
Specification ASTM C 33_ [27]
Fine Aggregate
Moisture Content
3.54%
0 - 4%
Water Absorption
2.65%
< 4%
Bulk density
1533.7
1200 – 1800 kg/m3
Specific Gravity (SSD)
2.54
2.4 -2.9
Course Aggregate
Moisture Content
3.49%
0 - 4%
Water Absorption
3.01%
< 4%
Bulk density
1421.2
1200 – 1800 kg/m3
Specific Gravity (SSD)
2.54
2.4 -2.9
Cement
Initial Setting time
132
> 60 minutes
Final Setting time
243
< 375 minutes
Specific gravity
3.12
3.10 – 3.15
Fineness
8.4%
-
Specific surface area
4150 cm2/g
-
Potato Starch (PS)
Mean Size
15.32 µm
-
Median Size
14.21 µm
-
Specific surface area
49322/g
- Table 3: Grain size distribution of sand
Sieve size (mm)
Mass of sand passing (g)
Mass of sand Retained (g)
% passing
4.75
760
0
100
2.36
645.32
114.68
84.91
1.18
467.56
177.76
61.52
0.850
324.18
143.38
42.65
0.6
204.46
119.72
26.90
0.425
97.18
107.28
12.78
0.3
60.51
36.67
7.96
0.212
34.65
25.86
4.56
0.15
12.45
22.20
1.64
0.075
5.64
6.81
0.74
Pan
0
5.64
0
Total
760
Table 4: Grain size distribution of gravel
Sieve size (mm)
Mass of granite passin
(g)
Mass of granite retained (g)
% passing
31.5
3560
0
100
22.4
3212.37
347.63
90.24
19
2856.34
356.03
80.24
16
1468.23
1388.11
41.25
12.5
820.72
647.51
23.06
9.5
469.56
351.16
13.20
6.3
78.41
391.15
2.21
4.75
22.35
56.06
0.63
Pan
0
0
0
Total
3560 542 Engineering and Technology Journal 42 (05) (2024) 540-547 Chidobere D. 2.2 Methodology 2.2.1 Experimental program 3.2 Results of workability test The workability outcome of different PS-concrete mixes measured using slump are shown in Figure. 2. There was a
reduction in slump values from 93 mm to 19 mm, 97 mm to 16 mm, and 98 mm to 20 mm for mix ratios 1:1.5:2, 1:2:4, and
1:3:6, respectively. The thickening quality of PS contributed to the decrease in slump values. The mix flowability decreased
due to the rise in the internal friction of the cement paste within the concrete. The results had correspondence and disparity
with the findings of [22] and [6] respectively. PS 0.0, PS 0.4 and 0.8, PS 1.2 and 1.6, and PS 2.0 have high, medium, low, and
very low workability, respectively, and can be adopted for piling works, normal reinforced concrete works, mass concrete
footings, and rigid pavements, respectively. Concrete at 1.6% and 2.0% inclusion of PS depicted low and very low slumps,
implying that supplemental effort is needed to consummate excellent compaction when juxtaposed with the other mixes. A
porous concrete is often a product of low workability. The interrelation of pore spaces within a concrete mix can be reduced
with PS of 1.6% and 2.0% and the permeability of concrete. Mix ratios of 1:2:4 and 1:3:6 are most suitable when used in the
construction industry as concrete block-binder, pavement works, and fire-resistant wallboard, while mix ratios of 1:1.5:2 can
be employed for paint filler, clay, and limestone binder. Figure 2: Variation of slump values with mixed ratios Figure 2: Variation of slump values with mixed ratios 3.3 Results of compressive strength test 3.1 Results of setting time test The need to effectively handle concrete during construction validates the necessity of ascertaining the initial and final
setting times when cement starts losing its plasticity and when all of its plasticity is gone. In Figure 1, the setting time tests
show a progressive increase in the setting time with the addition of potato starch (PS) admixture. Initial and final setting times
for the control mix were 132 and 243 mins and gradually increased to 229 and 415 mins, respectively, with the addition of PS
up to 2.0% by weight of cement. The reports of [9, 22] showed similar trend. The rise in the initial and final setting times for
each percentage inclusion of CS over the control are 36 minutes, 53 minutes, 60 minutes, 86 minutes, 97 minutes and 53
minutes, 73 minutes, 90 minutes, 130 minutes, 172 minutes, respectively. The presence of tricalcium aluminates (C3A) in the
cement paste and the adsorption of starch polymers to them contributed to the retardation. The delayed hydration process is
also traceable to the influence of the hydroxyl group of starch, which promotes the special complexation of calcium ions. Hydration delay is a great advantage for concrete works done in hot weather conditions. It also gives site workers more time to
place and finish the concrete. However, too much retardation can cause the concrete to be too weak and soft. The fineness
nature of the cement and potato starch increased the surface area. It quickened the hydration process, contributing to the
concrete's increased setting time and early strength development. 543 Engineering and Technology Journal 42 (05) (2024) 540-547 Chidobere D. Nwa-David Figure 1: Variation of Setting Time with % PS Content Figure 1: Variation of Setting Time with % PS Content 3.2 Results of workability test 3.3 Results of compressive strength test Figures 3 to 5 showed that the compressive strength expanded above the control at all percentage additions of PS during
the entire curing days. This enlargement is traceable to the viscosity-remodeling ability of PS, which includes minimized
tendency of bleeding and segregation in concrete, revamped degree of hydration, internal curing, and the ability to prevent the
formation of unwanted hydration products such as ettringite in concrete. The presence of tricalcium silicates (C3S) determines
the early strength, while dicalcium silicates (C2S) account for the long-term strength phenomenon. The PS admixture had higher compressive strengths for lower mix proportions (1:2:4 and 1:3:6), as shown in Figures 3-5. For all the mixes, the maximum compressive strength was found at 1.2% addition of PS at all curing days. At 28 days of
curing, the recorded strengths were at their peak at 32.77 N/mm2, 35.12 N/mm2, and 38.97 N/mm2 at 1.2% PS inclusion for 544 Chidobere D. Nwa-David Engineering and Technology Journal 42 (05) (2024) 540-547 1:1.5:2, 1:2:4, and 1:3:6 mix ratios respectively. Concretes produced with a mix ratio of 1:3:6 had the highest compressive
strength. Concrete made with a mix ratio of 1:1.5:2 had the least strength. Beyond 1.2%, there was a reduction in strength due
to a decrease in the viscosity-enhancing property of the starch. Considering the biodegradable nature of the admixture, its long-
term effect on concrete might have an adverse effect beyond a 1.2% addition. This outcome shows non-uniformity and
coincidence with the results of [6] and [9, 22, 30, 31] because of the chemical and natural composition of the admixture used. Concrete containing chemical admixtures behaved differently from those of natural biopolymers. [ ]
[
]
p
Concrete containing chemical admixtures behaved differently from those of natural biopolymers. As the curing ages increase, the strength values are added in all cases. The prolonged retardation of PS furnishes the lower
strength of the concrete at early periods of curing. Improved compaction, the dispersion of cement particles, the production of
denser gel due to delayed setting, and enhanced workability aided the elevated strength on later curing days. As the curing ages increase, the strength values are added in all cases. The prolonged retardation of PS furnishes the lower
strength of the concrete at early periods of curing. 3.3 Results of compressive strength test It was also observed that concrete made with mix 1:2:4, had 7.26% and 19.2%, 6.0% and
9.9%, 6.7% and 12.7%, 5.4% and 12.0%, 7.0% and 9.1% reductions in strength for 1.6% and 2.0% PS inclusion when collated
with 1.2% PS inclusion by cement weight at curing durations of 7, 21, 28, 56 and 90 days respectively. There were 49.1%, 41.54%, 40.0%, 28.9%, and 32.4% increments in strengths for a concrete mix of 1:2:4 at 1.2% PS
addition by cement weight in comparison with the control mixes at 0% for 7, 21, 28, 56, and 90 days curing periods
respectively, as captured in figure 4. It was also observed that concrete made with mix 1:2:4, had 7.26% and 19.2%, 6.0% and
9.9%, 6.7% and 12.7%, 5.4% and 12.0%, 7.0% and 9.1% reductions in strength for 1.6% and 2.0% PS inclusion when collated
with 1.2% PS inclusion by cement weight at curing durations of 7, 21, 28, 56 and 90 days respectively. In Figure 5, the concretes made with mix 1:3:6 had increments of 52.2%, 51.0%, 37.0%, 24.7%, and 24.2% in their
compressive strength at 1.2% PS addition by cement weight when analogized with the control mixes at 0% for 7, 21, 28, 56,
and 90 days curing periods respectively. It was also observed that concrete made with mix 1:3:6 had 4.0% and 7.5%, 6.5% and
9.7%, 10.0% and 11.2%, 7.0% and 11.8%, 7.2% and 10.8% reductions in strength for 1.6% and 2.0% PS inclusion when
contrasted with 1.2% PS inclusion by cement weight at curing durations of 7, 21, 28, 56 and 90 days respectively. In Figure 5, the concretes made with mix 1:3:6 had increments of 52.2%, 51.0%, 37.0%, 24.7%, and 24.2% in their
compressive strength at 1.2% PS addition by cement weight when analogized with the control mixes at 0% for 7, 21, 28, 56,
and 90 days curing periods respectively. It was also observed that concrete made with mix 1:3:6 had 4.0% and 7.5%, 6.5% and
9.7%, 10.0% and 11.2%, 7.0% and 11.8%, 7.2% and 10.8% reductions in strength for 1.6% and 2.0% PS inclusion when
contrasted with 1.2% PS inclusion by cement weight at curing durations of 7, 21, 28, 56 and 90 days respectively. 3.3 Results of compressive strength test Improved compaction, the dispersion of cement particles, the production of
denser gel due to delayed setting, and enhanced workability aided the elevated strength on later curing days. From Figure 3, it was observed that the concretes had 51.5%, 33.9%, 35.9%, 19.5%, and 18.4% increments in the
compressive strengths for a concrete mix of 1:1.5:2 at 1.2% PS addition by cement weight in comparison with the control
mixes at 0% for 7, 21, 28, 56, and 90 days curing periods respectively. There were 13.6%, and 22.2%, 7.9% and 15.0%, 7.9%
and 13.2%, 6.4% and 12.3%, 8.2% and 14.5% reductions in strength for 1.6% and 2.0% PS inclusion when juxtaposed with
1.2% PS inclusion by cement weight at curing durations of 7, 21, 28, 56 and 90 days respectively for concrete mix 1:1.5:2. From Figure 3, it was observed that the concretes had 51.5%, 33.9%, 35.9%, 19.5%, and 18.4% increments in the
compressive strengths for a concrete mix of 1:1.5:2 at 1.2% PS addition by cement weight in comparison with the control
mixes at 0% for 7, 21, 28, 56, and 90 days curing periods respectively. There were 13.6%, and 22.2%, 7.9% and 15.0%, 7.9%
and 13.2%, 6.4% and 12.3%, 8.2% and 14.5% reductions in strength for 1.6% and 2.0% PS inclusion when juxtaposed with
1.2% PS inclusion by cement weight at curing durations of 7, 21, 28, 56 and 90 days respectively for concrete mix 1:1.5:2. y
g
g
y
p
y
There were 49.1%, 41.54%, 40.0%, 28.9%, and 32.4% increments in strengths for a concrete mix of 1:2:4 at 1.2% PS
addition by cement weight in comparison with the control mixes at 0% for 7, 21, 28, 56, and 90 days curing periods
respectively, as captured in figure 4. It was also observed that concrete made with mix 1:2:4, had 7.26% and 19.2%, 6.0% and
9.9%, 6.7% and 12.7%, 5.4% and 12.0%, 7.0% and 9.1% reductions in strength for 1.6% and 2.0% PS inclusion when collated
with 1.2% PS inclusion by cement weight at curing durations of 7, 21, 28, 56 and 90 days respectively. y
g
g
y
p
y
There were 49.1%, 41.54%, 40.0%, 28.9%, and 32.4% increments in strengths for a concrete mix of 1:2:4 at 1.2% PS
addition by cement weight in comparison with the control mixes at 0% for 7, 21, 28, 56, and 90 days curing periods
respectively, as captured in figure 4. 3.3 Results of compressive strength test In Figure 5, the concretes made with mix 1:3:6 had increments of 52.2%, 51.0%, 37.0%, 24.7%, and 24.2% in their
compressive strength at 1.2% PS addition by cement weight when analogized with the control mixes at 0% for 7, 21, 28, 56,
and 90 days curing periods respectively. It was also observed that concrete made with mix 1:3:6 had 4.0% and 7.5%, 6.5% and
9.7%, 10.0% and 11.2%, 7.0% and 11.8%, 7.2% and 10.8% reductions in strength for 1.6% and 2.0% PS inclusion when
contrasted with 1.2% PS inclusion by cement weight at curing durations of 7, 21, 28, 56 and 90 days respectively. Figure 3: Compressive strength of PS Concrete at a mix ratio of 1:1.5:2 at varying curing days
Figure 4: Compressive strength of PS Concrete at a mix ratio of 1:2:4 at varying curing days Figure 3: Compressive strength of PS Concrete at a mix ratio of 1:1.5:2 at varying curing days 3: Compressive strength of PS Concrete at a mix ratio of 1:1.5:2 at varying curing days Figure 3: Compressive strength of PS Concrete at a mix ratio of 1:1.5:2 at varying curing days
Figure 4: Compressive strength of PS Concrete at a mix ratio of 1:2:4 at varying curing days Figure 4: Compressive strength of PS Concrete at a mix ratio of 1:2:4 at varying curing days 545 Engineering and Technology Journal 42 (05) (2024) 540-547 Chidobere D. Nwa-David Figure 5: Compressive strength of PS Concrete at a mix ratio of 1:3:6 at varying curing days Figure 5: Compressive strength of PS Concrete at a mix ratio of 1:3:6 at varying curing days 4. Conclusion From the upshot of this experimental study, potato starch (PS) can be regarded as a viscosity-permutating admixture in
concrete, along with its retarding characteristics. This property is a great advantage to concrete production under hot weather
scenarios. The rise in PS percentage addition minimized the workability of concrete but increased the setting time of cement
paste and the compressive strength. This trade-off enhances the production of self-compacting concrete. The observed increase
in strength with the addition of PS is significant in construction works where early strength is needed, such as concrete repairs
and pavement construction. Adding not more than 1.2% PS by cement weight is recommended to achieve optimum results in
better strength performance. With respect to the retarding property of PS, potential adjustments could be made during
construction depending on the peculiarity of the project. Potato starch-based concrete could benefit flooring, walkways,
soundproofing, and water-resistant backings. However, it should be applied cautiously when employed for foundations and
tunnels. In summary, adopting PS as an admixture will fix the issues of cost, sustainability, and availability associated with
dependence on chemical admixtures in concreting. The interest in the production of green concrete is also revamped through
the outcome of the current study. Conflicts of interest The authors declare that there is no conflict of interest. Data availability statement The data that support the findings of this study are available on request from the corresponding author. Conflicts of interest Funding h received no specific grant from any funding agency in the public, commercial, or not-for-profit sectors. This research received no specific grant from any funding agency in the public, commercial, or not-for-profit sectors. Data availability statement This research received no specific grant from any funding agency in the public, commercial, or not-f References http://dx.doi.org/10.2174/1874836801913010331 [10] ASTM C494, ASTM C494: Standard specification for chemical admixtures for concrete, (2005). [11] Y. Wang, F. He, J. Wang, and Q. Hu, Comparison of effects of sodium bicarbonate and sodium carbonate on the
hydration and properties of Portland cement paste, Mater., 12 (2019). http://doi.org/10.3390/ma12071033 [12] J. J. Brooks, M. A. MegatJohari, M. Mazloom, Effect of admixtures on the setting times of high-strength concrete,
Cement & Concrete Composites, 22 (2000) 293-301. https://doi.org/10.1016/S0958-9465(00)00025-1 [13] C. D. Nwa-David, and O. M. Ibearugbulem, Model for Prediction and Optimization of Compressive Strengths of Cement
Composites using Nanostructured Cassava Peel Ash as Partial Replacement of the Binder,
Niger. Res. J. Eng. Environ. Sci., 8 (2023) 82-91. http://doi.org/10.5281/zenodo.8094024 [14] A. M. Neville, and J. J. Brooks, Concrete Technology (2nd ed.). Harlow, UK: Longman Scientific & Technical., (1990). [15] A. Elahi, P. A. M. Basheer, S. V. Nanukuttan, Q. U. Z. Khan, Mechanical and durability properties of high performance
concretes
containing
supplementary
cementitious
materials,
Constr. Build. Mater.,
24
(2010)
292–299. https://doi.org/10.1016/j.conbuildmat.2009.08.045 [16] A. E. Abalaka, Comparative effects of cassava starch and simple sugar in cement mortar and concrete, ATBU J. Environ. Technol., 4 (2011) 13-22. [17] S.A. El-Mekkawi, I.M. Ismail, M.M. El-Attar, A.A. Fahmy, and S.S. Mohammed, Utilization of black liquor as concrete
admixture and set retarder aid, J. Adv. Res., 2 (2011) 163-169. http://dx.doi.org/10.1016/j.jare.2011.01.005 [18] S.M. Abd, Q.Y. Hamood, A.S. Khamees, and Z.H. Ali, Effect of using corn starch as concrete admixture, Int. J. Eng. Sci. Technol., 5 (2016) 35-44. [19] R. Mbugua, R. Salim, and J. Ndambuki, Effect of Gum Arabic Karroo as a water-reducing admixture in cement mortar,
Case Stud. Constr. Mater., 5 (2016) 100-111. http://dx.doi.org/10.1016/j.cscm.2016.09.002 [20] O. A. Adetayo, and O. Jubril, Effect of Ripe and Unripe Plantain Peel Ash on Concrete Workability and Compressive
Strength, FUOYE Eng. Technol. J., 4 (2019) 145-148. http://dx.doi.org/10.46792/fuoyejet.v4i1.318 [21] M. G. O. Adam, D. O Koteng, J. N. Thuo, M. Matallah, Analysing the Effect of Cassava Flour as a Mixture on the
Physical, Mechanical and Durability Properties of High-Strength Concrete, Civil Eng. J., 8 (2022) 3866-3879. http://dx.doi.org/10.26991/CEJ-2022-08-12-015 [22] A. A. Akindahunsi, H.C. Uzoegbo, Strength and durability properties of concrete with starch admixture, Int. J. Concr. Struct. Mater., 9 (2015) 323-335. http://dx.doi.org/10.1007/s40069-015-0103-x [23] BS EN 17075. Water quality, General requirements and performance test procedure for water monitoring equipment. European Standard Test. 2018. [24] BS EN 197-1. Cement composition, specifications and conformity criteria for common cement. References [1] C. D. Nwa-David, D. O. Onwuka, and F. C. Njoku, Prediction of Compressive Strength of Concrete containing Nanosized
Cassava Peel Ash as partial Replacement of Cement using Artificial Neural Network, Fuoye J. Eng. Technol.,
(FUOYEJET) 8 (2023) 253-259. http://doi.org/10.46792/fuoyejet.v8i2.1029 [1] C. D. Nwa-David, D. O. Onwuka, and F. C. Njoku, Prediction of Compressive Strength of Concrete containing Nanosized
Cassava Peel Ash as partial Replacement of Cement using Artificial Neural Network, Fuoye J. Eng. Technol.,
(FUOYEJET) 8 (2023) 253-259. http://doi.org/10.46792/fuoyejet.v8i2.1029 [2] C. D. Nwa-David, Investigation of Fresh and Hardened Behaviour of Nanostructured Concrete. Arid Zone J. Eng. Technol. Environ., 19 (2023) 423-436. [2] C. D. Nwa-David, Investigation of Fresh and Hardened Behaviour of Nanostructured Concrete. Arid Zone J. Eng. Technol. Environ., 19 (2023) 423-436. [3] C. D. Nwa-David, D. O. Onwuka, F. C. Njoku, O. M. Ibearugbulem, Prediction of Fresh and Hardened Properties of
Concrete Containing Nanostructured Cassava Peel Ash Using Ibearugbulem's Approach, Eng. Technol. J., 41 (2023) 1-14. http://doi.org/10.30684/etj.2023.138099.1374 [4] C. D. Nwa-David, Effect of Elevated Temperatures on Mechanical Properties of Concrete Blended with Nanosized
Sawdust Ash and Rice Husk Ash, UNIZIK J. Eng. Appl. Sci., 2 (2023) 404-415. [4] C. D. Nwa-David, Effect of Elevated Temperatures on Mechanical Properties of Concrete Blended with Nanosized
Sawdust Ash and Rice Husk Ash, UNIZIK J. Eng. Appl. Sci., 2 (2023) 404-415. [5] C. D Nwa-David, The Behaviour of Concrete Made with Nanosized Periwinkle Shell Ash as Partial Replacement of
Cement Under Varying Curing Conditions with Emphasis on its Compressive Strength, J. Mater. Eng. Comput. Struct., 2
(2023) 41-50. https://doi.org/10.5281/zenodo.8301486 [6] O. T. Andrew, O. O. Omotayo, C. Arum, and C. M. Ikumapayi, Effects of Sodium Carbonate Admixture and Mix Design Ratios
on the Compressive Strength of Concrete, Niger. J. Technol., 42 (2023) 185 – 190. http://dx.doi.org/10.4314/njt.v42i2.4 546 Engineering and Technology Journal 42 (05) (2024) 540-547 Chidobere D. Nwa-David [7] M. Sybis, and E. Konował, Influence of Modified Starch Admixtures on Selected Physicochemical Properties of Cement
Composites, Mater., 15 (2022) 1-12. https://doi.org/10.3390/ma15217604 [8] S. U. Khan, M. F. Nuruddin, T. Ayub and N. Shafiq, Effects of different mineral admixtures on the properties of fresh
concrete, Sci. World J., 201 (2014). https://doi.org/10.1155/2014/986567 [9] D. O. Oni, J. Mwero and C. Kabubo, Experimental Investigation of the Physical and Mechanical Properties of Cassava
Starch
Modified
Concrete,
Open
Constr. Build. Technol. J.,
13
(2019)
331-343. References European Standard
specifications. 2011. [25] D. Oni, J. Mwero and C. Kabubo, The Effect of Cassava Starch on the Durability Characteristics of Concrete, Open J. Civ. Eng., 14 (2020) 289-301. http://dx.doi.org/10.2174/1874149502014010289 [26] BS 882, Specification for aggregates from natural sources for concrete, British Standards Institution, London, UK, 1992. 7] ASTM C 33, Standard specification for concrete aggregates, ASTM International., West Conshohocken tandard specification for concrete aggregates, ASTM International., West Conshohocken, USA, 2003. 8] BS EN 12350-2, Testing fresh concrete — Part 2: Slump test.” British Standard, UK, 2016. [29] BS EN 12390-3, Testing hardened concrete. Compressive strength of test specimens.British Standard, UK, 2019. [30] F.O. Okafor, The Performance of Cassava Flour as a Water-Reducing Admixture for Concrete, Niger. J. Technol., 29
(2010) 106-112. [31] S. K. Joseph, and A. S. Xavier, Effect of Starch Admixtures on Fresh and Hardened Properties of Concrete, Int. J. Sci. Eng. Res., 4 (2016) 27-30 . 547
|
https://openalex.org/W2980633172
|
https://europepmc.org/articles/pmc6899522?pdf=render
|
English
| null |
Polymeric nanobiotics as a novel treatment for mycobacterial infections
|
Journal of controlled release
| 2,019
|
cc-by
| 8,145
|
Polymeric nanobiotics as a novel treatment for mycobacterial infections Iris L. Batalhaa,b, Audrey Bernutc,d, Mark Schieblera,b, Myriam M. Ouberaia, Charlotte Passemarb,
Catherine Klapholzb, Sonja Kinnaa, Sarah Michela, Kasim Sadere, Pablo Castro-Hartmanne,
Stephen A. Renshawc,d, Mark E. Wellanda,*, R. Andres Flotob,f,** a Nanoscience Centre, Department of Engineering, University of Cambridge, 11 J.J. Thomson Avenue, Cambridge, CB3 0FF, United Kingdom
b Molecular Immunity Unit, Department of Medicine, University of Cambridge, Francis Crick Avenue, Cambridge Biomedical Campus, Cambridge, CB2 0QH, United
Kingdom a Nanoscience Centre, Department of Engineering, University of Cambridge, 11 J.J. Thomson Avenue, Cambridge, CB3 0FF, United Kingdom
b Molecular Immunity Unit, Department of Medicine, University of Cambridge, Francis Crick Avenue, Cambridge Biomedical Campus, Cambridg
Kingdom c Dept. of Infection, Immunity & Cardiovascular Disease, Bateson Centre, University of Sheffield, Firth Court, Western Bank, Sheffield, S10 2TN
d Medical School, University of Sheffield, Sheffield, S10 2RX, United Kingdomi unity & Cardiovascular Disease, Bateson Centre, University of Sheffield, Firth Court, Western Bank, Sheffield, S10 2TN, United Kingdom
sity of Sheffield, Sheffield, S10 2RX, United Kingdomi p
f
f
,
y
,
,
y
f
ffi
,
,
,
ffi
,
,
g
d Medical School, University of Sheffield, Sheffield, S10 2RX, United Kingdom
e Cambridge CryoEM Pharmaceutical Consortium, Thermo Fisher Scientific, Nanoscience Centre, Department of Engineering, University of Cambridge, 11 J.J. Thomson
Avenue, Cambridge, CB3 0FF, United Kingdom
f C
b id
C
f
L
I f
i
R
l P
h H
i
l C
b id
CB23 3RE U i d Ki
d ,
y
f
ffi
,
ffi
,
,
g
e Cambridge CryoEM Pharmaceutical Consortium, Thermo Fisher Scientific, Nanoscience Centre, Department of Engineering, University of Cambridge, 11 J.J. Thomson
Avenue, Cambridge, CB3 0FF, United Kingdom f Cambridge Centre for Lung Infection, Royal Papworth Hospital, Cambridge, CB23 3RE, United Kingdom https://doi.org/10.1016/j.jconrel.2019.10.009
Received 18 June 2019; Received in revised form 10 September 2019; Accepted 3 October 2019
Corresponding author.
⁎⁎ Corresponding author at: Molecular Immunity Unit, Department of Medicine, University of Cambridge, Francis Crick Avenue, Cambridge Biomedical Campus,
Cambridge, CB2 0QH, United Kingdom.
E-mail addresses: mew10@cam.ac.uk (M.E. Welland), arf27@cam.ac.uk (R.A. Floto).
Available online 21 October 2019
0168-3659/ ©
2019 The Authors. Published by Elsevier B.V. This is an open access article under the CC BY license
(http://creativecom
m
ons.org/licenses/BY/4.0/). A R T I C L E I N F O Keywords:
Mycobacterium tuberculosis
Polymer-drug conjugate
Antibiotic
Nanoparticles
Zebrafish
Isoniazid
Clofazimine Mycobacterium tuberculosis (Mtb) remains a major challenge to global health, made worse by the spread of multi-
drug resistance. Currently, the efficacy and safety of treatment is limited by difficulties in achieving and sus-
taining adequate tissue antibiotic concentrations while limiting systemic drug exposure to tolerable levels. Here
we show that nanoparticles generated from a polymer-antibiotic conjugate (‘nanobiotics’) deliver sustained
release of active drug upon hydrolysis in acidic environments, found within Mtb-infected macrophages and
granulomas, and can, by encapsulation of a second antibiotic, provide a mechanism of synchronous drug de-
livery. Nanobiotics are avidly taken up by infected macrophages, enhance killing of intracellular Mtb, and are
efficiently delivered to granulomas and extracellular mycobacterial cords in vivo in an infected zebrafish model. We demonstrate that isoniazid (INH)-derived nanobiotics, alone or with additional encapsulation of clofazimine
(CFZ), enhance killing of mycobacteria in vitro and in infected zebrafish, supporting the use of nanobiotics for
Mtb therapy and indicating that nanoparticles generated from polymer-small molecule conjugates might provide
a more general solution to delivering co-ordinated combination chemotherapy. Due to the scarcity of new drugs against Mtb and thus limited
therapeutic options for drug-resistant Mtb, increased efforts have been
put on the development of improved formulations and delivery systems
for existing antibiotic regimens [3]. Journal of Controlled Release 314 (2019) 116–124 Journal of Controlled Release 314 (2019) 116–124 https://doi.org/10.1016/j.jconrel.2019.10.009
Received 18 June 2019; Received in revised form 10 September 2019; Accepted 3 October 2019
⁎ Corresponding author.
⁎⁎ Corresponding author at: Molecular Immunity Unit, Department of Medicine, University of Cambridge, Francis Crick Avenue, Cambridge Biomedical Campus,
Cambridge, CB2 0QH, United Kingdom.
E-mail addresses: mew10@cam.ac.uk (M.E. Welland), arf27@cam.ac.uk (R.A. Floto). https://doi.org/10.1016/j.jconrel.2019.10.009
Received 18 June 2019; Received in revised form 10 September 2019; Accepted 3 October 2019
Available online 21 October 2019
0168-3659/ ©
2019 The Authors. Published by Elsevier B.V. This is an open access article under the CC BY license
(http://creativecom
m
ons.org/licenses/BY/4.0/). 1. Introduction Over 1.6 million deaths annually are caused by Mtb infection [1]. Existing antibiotic regimens for Mtb infection require long durations of
therapy with multiple drugs and are associated with significant side
effects (due to systemic exposure), contributing to poor adherence and
treatment failure [2].fi In the last two decades, the application of polymer-drug conjugation
to drug delivery has increased noticeably, offering advantages in-
cluding enhanced drug solubilization, reduced immunogenicity, con-
trolled delivery, increased efficacy, and improved pharmacokinetics. However, most polymer-small molecule drug conjugates have to date
used non-biodegradable polymer carriers, such as polyethylene glycol
(PEG), that constrains polymer size below the molecular cut-off of
∼40 kDa required for renal elimination [4]. Alternatively, hydrolysable
hydrophobic polyesters, such as polycaprolactone (PCL) and poly(lac-
tide-co-glycolide) (PLGA), widely employed in FDA-approved devices, One of the major difficulties in treating tuberculosis is that Mtb can
survive both intracellularly within macrophages and extracellularly
within granulomas; environments where conventional drug delivery is
compromised. Bacteria are therefore exposed to sub-lethal concentra-
tions of antibiotics, permitting firstly the development of phenotypic
drug tolerance and eventually the acquisition of drug resistance mu-
tations [2]. Fig. 1. Synthesis of an α-keto polyester by (trans)esterification reaction cata-
lysed by CALB and conjugation to isoniazid (INH). Journal of Controlled Release 314 (2019) 116–124 Journal of Controlled Release 314 (2019) 116–124 I.L. Batalha, et al. present limited functionality for drug conjugation [5] and are used to
physically entrap drugs within nanoparticulate carriers [6]. However,
many anti-tuberculosis drugs are highly water soluble, making them
easily leached out from the nanocarriers during fabrication and more
prone to burst release in systemic circulation [7–10]. Polyketals, which
in contrast with polyesters yield pH neutral hydrolysis products, have
also recently been explored as new class of acid responsive and bio-
degradable polymers suitable for drug conjugation [11].i Isoniazid (INH) is a potent antibiotic universally used as a first-line
treatment of tuberculosis, either as part of combination therapy to treat
the active disease, or often used as monotherapy in cases of latent tu-
berculosis infection. Despite its high activity against Mtb, INH is rapidly
egested and highly toxic, prompting the development of delivery sys-
tems aiming for targeted and controlled release of INH [12,13]. A few
polymers have been explored for INH conjugation, including natural
polymers, such as gelatin [14] and chitosan [13], and synthetic poly-
mers, such as PLGA [15]. 2.1. Chemicals All reagents were analytical grade. 1,8-Octanediol, calcium pan-
tothenate, citric acid, chloroform, clofazimine (CFZ), coumarin 6 (Cou-
6), dimethyl 2-oxoglutarate, dimethyl sulfoxide (DMSO) anhydrous,
diphenyl ether, Dulbecco’s Modified Eagle’s Medium - high glucose
(DMEM),
hexane,
L-leucine,
poly(vinyl
alcohol)
(PVA;
Mw
31,000–50,000; 98–99% hydrolysed), phosphate buffered saline ta-
blets, phenol red, phorbol 12-myristate 13-acetate (PMA), silicon oil,
sodium phosphate dibasic, Tween 80 and tricaine were purchased from
Sigma-Aldrich, Merck (UK). Dichloromethane, formaldehyde, hygro-
mycin B, methanol, Middlebrook 7H9 with OADC, Remel Middlebrook
7H10 Agar (Dehydrated) and trifluoroacetic acid (TFA) were acquired
from Thermo Fisher Scientific (UK). Dimethyl sulphoxide-[D₆] (99.8%
D), fetal calf serum (FCS; Sera Plus, EU approved regions, special pro-
cessed FBS, 0.2 μm sterile filtered) and macrophage colony stimulating
factor (MCSF) were purchased from VWR (UK), PanBiotech (Germany)
and Peprotech (UK), respectively. Another important aspect for combination therapy is the ability to
co-deliver multiple drugs to the target sites. Manca and co-workers
prepared microparticles of gelatin-INH conjugates with encapsulated
rifampicin by spray drying technique. INH-derivatized gelatin was
prepared by heterogenous reaction of amidation, yielding an amide
bond between the terminal acyl chloride group of gelatine and the
hydrazide group of INH. They have shown good nebulization efficiency,
cell internalization, and low cytotoxicity, but they have not reported
the therapeutic efficacy of the conjugates [14]. fi
As a response to these challenges, we have developed a polymeric
nanoparticulate drug delivery system, using simple, fast and scalable
processes, where antibiotics are covalently incorporated into a polymer
chain, through a hydrolysable bond, creating ‘nanobiotics’. Multiple copies of antibiotics can be incorporated into the polymer
chain, which becomes active upon pH-triggered hydrolysis to achieve
targeted release of a high drug payload. As a proof-of-concept, we in-
corporated isoniazid (INH) by reacting its hydrazide group with the
ketone group of an α-keto polyester (Fig. 1). However, this strategy also
allows for the incorporation of other polar antibiotics, such as the first
line drug ethambutol or AZD5847 - a next generation oxazolidinone
currently in Phase II clinical trials [16], which could be used as the
polyol monomer instead of 1,8-octanediol. On the other hand, hydro-
phobic antibiotics, with poor water solubility and poor caseum pene-
tration [17], can be easily encapsulated in these systems, providing a
mechanism of synchronous nanoscale delivery of hydrophilic and hy-
drophobic payloads, while preventing undesirable drug-drug interac-
tions. 1. Introduction However, these systems involve further
chemical modifications of the polymers in order to introduce functional
groups amenable for drug conjugation. Berezin and Skorik prepared
chitosan-INH conjugates using two different synthetic routes, either by
modifying chitosan with acrylic acid or epichlorohydrin, before INH
conjugation. Modified chitosan polymers presented lower biodegrad-
ability, and either similar (for N-(2-carboxyethyl)chitosan INH con-
jugates) or higher (for N-(3-chloro-2-hydroxypropyl)chitosan INH
conjugates) minimum inhibitory concentrations compared to free drug,
possibly due to incomplete cleavage of INH from the polymer [13]. In a
different study, Huang and co-workers used an INH conjugated star
PLGA to fabricate a composite scaffold with β-tricalcium phosphate to
treat bone tuberculosis. The process involved esterification of the PLGA
and 4-carboxybenzaldehyde prior to drug conjugation. They have
produced a 4-arm PLGA-INH conjugate instead of a linear polymer in
order to achieve suitable drug loading capacity [15]. Fig. 1. Synthesis of an α-keto polyester by (trans)esterification reaction cata-
lysed by CALB and conjugation to isoniazid (INH). 1,8-Octanediol was melted in a round bottom flask at 75 °C using a 2.3. Blank and INH polymer synthesis 2.2. Antibodies The following antibodies were used for flow cytometric studies:
Brilliant Violet 785™anti-human CD3 Antibody (Biolegend®, clone
OKT3,
317330),
CD14
Monoclonal
Antibody
APC-eFluor
780
(eBioscience™, clone 61D3, 47-0149-42), PE/Cy7 anti-human CD15
(SSEA-1)
antibody
(Biolegend®,
clone
W6D3,
323029),
CD19
Monoclonal Antibody PE (eBioscience™, clone HIB19, 12-0199-42). The
following isotype controls were used: Mouse IgG1 kappa Isotype
Control, APC-eFluor 780 (eBioscience™, clone P3.6.2.8.1, 47-4714-82),
Brilliant Violet 785™Mouse IgG2a, κ Isotype Ctrl Antibody (Biolegend®,
MOPC-173, 400273), PE/Cy7 Mouse IgG1, κ Isotype Ctrl Antibody
(Biolegend®, MOPC-21, 400125), Mouse IgG1 kappa Isotype Control,
PE (eBioscience™, clone P3.6.2.8.1, 12-4714-81). All antibodies were
used at 1/250 dilution. Nanobiotics could be an invaluable tool for delivering drugs in a
spatiotemporal-controlled manner, increasing the bioavailability of
drugs in the target tissue, while simultaneously protecting drugs against
degradation and minimizing their toxic effects in patients. 2.3. Blank and INH polymer synthesis 2.3. Blank and INH polymer synthesis 1,8-Octanediol was melted in a round bottom flask at 75 °C using a 1,8-Octanediol was melted in a round bottom flask at 75 °C using a 117 Journal of Controlled Release 314 (2019) 116–124 I.L. Batalha, et al. silicon oil bath. Dimethyl 2-oxoglutarate (1:1 M ratio to 1,8-octanediol)
and Lipase acrylic resin from Candida antarctica (CALB beads; ≥5000
U/g) (10% (w/w) relative to monomers) were added to the flask and
the reaction was left for 1 h at 75 °C under vacuum with agitation. Temperature was increased to 90 °C, diphenyl ether (3× volume of
monomer) was added and reaction was incubated for another 5 h. After
the reaction mixture cooled to room temperature, chloroform (4×
volume of monomer) was added to the flask, and the solution was fil-
tered to remove the CALB beads. The crude solution was then pre-
cipitated into a 20-fold excess of hexane to remove unreacted monomer. Precipitation was repeated twice, and the obtained copolymer was
dried under vacuum overnight. A number average molecular weight
(Mn) of 5265 ± 487 g/mol and a dispersity (Ð) of 2.247 ± 0.395
(average of three different polymer batches) were determined for the
synthesized polymer (blank polymer) by gel permeation chromato-
graphy (Agilent 1260 Infinity II GPC/SEC system). Briefly, a sample of
blank polymer was dissolved in chloroform, 0.22 μm-filtered and in-
jected (50 μL) into a PLgel MiniMIX-B column (Agilent). Chloroform
was used as eluent at a flow rate of 0.3 mL/min in a 30 min run at 25 °C. 2.8. Cryo-electron microscopy (Cryo-EM) All samples were vitrified with a Thermo Fisher Vitrobot MkIV by
plunge freezing in liquid ethane. The Vitrobot blot force was calibrated
to give a "wedge" of thick ice on roughly 1/3 of the grid, with a gradient
of ice thicknesses on the other 2/3 of the grid, corresponding to a set-
ting of "-6" on this system. Other Vitrobot conditions are: temperature
4 °C, RH 100%, blot time 2.5 s, and volume of sample applied 2.5 μL. Quantifoil R1.2/1.3 300 mesh grids were used and made hydrophilic by
glow discharge in a weak vacuum in a Pelco Easiglo glow discharge unit
at 0.39 mbar for 60 s at 25 mA. Images were acquired on a Thermo
Fisher Krios G2 with the single particle data acquisition package EPU
(1.10) on a Falcon 3 direct detector at magnifications of 37,000×
(2.26 Å/pixel), 47,000× (1.77 Å /pixel), and 6,500× (24.7 Å /pixel). Tomography tilt series +/- 60 deg were acquired with Tomography 4
software on a Falcon 3 detector in counting mode at a nominal mag-
nification of 37,000× corresponding to a total accumulated dose of
∼100 e−/Å2. Tilt series were aligned by cross-correlation with a
stretching factor for tilt and reconstructed by 10 iterations of a
Simultaneous Iterative Reconstruction Technique (SIRT) algorithm in
Thermo Fisher Inspect 3D 4.3. Visualization and rendering were per-
formed in Thermo Fisher Amira 6.5. 2.9. In vitro release of INH at different pH Nano INH were resuspended in 3 different buffers: PBS pH 7.4, ci-
trate-phosphate pH 6 and citrate-phosphate pH 5. The resuspended
nanobiotics were aliquoted (100 μL volume; 1.5 mM initial INH con-
centration) in triplicates and incubated at 37 °C under mild agitation. At
pre-defined time intervals, nanobiotic suspensions were centrifuged at
14,000 rpm for 15 min. Supernatant (80 μL) was collected and analysed
for drug content by HPLC as described above. Fresh buffer (80 μL) was
added and the nanobiotics were resuspended and incubated for another
time interval. 2.7. Nanobiotics characterization by Dynamic Light Scattering (DLS) and
Electrophoretic Light Scattering (ELS) 2.7. Nanobiotics characterization by Dynamic Light Scattering (DLS) and
Electrophoretic Light Scattering (ELS) The hydrodynamic size and zeta potential of the nanobiotics were
measured by DLS and ELS, respectively, with a Zetasizer Nano ZS
system (Malvern Panalytical) fitted with a 4 mW He-Ne laser operating
at 633 nm. Measurements were performed at 25 °C and 173° angle at a
final nanobiotic concentration of 0.05 mg/mL in dd water. Data were
analysed using Zetasizer Software 7.13 (Malvern Panalytical). Blank polymer was dissolved in anhydrous DMSO to a final con-
centration of 200 mg/mL, INH (2 M equivalents excess of keto groups in
the polymer) was added and the mixture was kept at 37 °C with orbital
agitation for 72 h. After that time, the bright yellow INH-functionalized
polymeric solution was added dropwise to methanol (1:10 (v/v)) and
subsequently poured into distilled deionised (dd) water (1:2.5 (v/v)) to
remove any unreacted INH. The sample was then centrifuged for 1.5 h
at 8000 rpm, supernatant was discarded and the polymeric pellet (INH
Polymer) was dried overnight under vacuum. 2.5. Polymer characterization by 1H-NMR Polymers were dissolved in deuterated DMSO at ∼5 mg/mL con-
centration. A Bruker Avance III HD 500 MHz equipped with 1H/13C
dual cryoprobe was used to conduct
1HNMR measurements. A
10,000 Hz sweep width was observed, acquired using a digital resolu-
tion of 64 K points over 3.28 s. A 30° pulse angle was used; based on a
10.5 μs, 14 W pulse at 500.053 MHz being the nominal 90° pulse. 32
scans were accumulated; with an interpulse delay (D1) of 1 s. Data were
analysed using Mnova NMR software (Mestrelab Research). 2.6. Formulation of nanobiotics The polymer (either blank polymer or INH polymer) was dissolved
in 2 mL of dichloromethane to a final concentration of 10 mg/mL. The
polymer solution was added dropwise to 10× volume of an aqueous
solution of 1% (w/v) PVA and homogenised for 10 min at 30,000 rpm
(VWR Homogenizer VDI12). The emulsion was then probe sonicated for
3 min (35% Amplitude; Pulse: 3 s ON, 6 s OFF), and stirred overnight at
room temperature to evaporate dichloromethane. Finally, the sample
was centrifuged for 30 min at 8000 rpm and pellet was washed with
and resuspended in dd water. For nanobiotics containing CFZ or Cou-6,
the compounds were first solubilized in the INH Polymer solution to a
final concentration of 5 mg/mL and 0.1 mg/mL of each compound,
respectively, and procedure was followed as described above. INH
loading was determined by high-performance liquid chromatography
(Agilent 1260 Infinity II LC system). Briefly, a sample of nanobiotics
was diluted in 1% TFA (v/v) (1:5 or 1:10) and incubated for 48 h at 2.4. Polymer characterization by Fourier-transform infrared spectroscopy
(FTIR) The chemical fingerprints of INH, blank polymer and INH polymer
were determined by FTIR (PerkinElmer Spotlight 400 Frontier FT-IR
equipped with Universal ATR) with a scan range of 650-4000 cm−1. Data analysis was performed in PerkinElmer Spectrum 10.5.3. 2.2. Antibodies The column was calibrated using polystyrene standards (Agilent).i 37 °C, the sample was then centrifuged for 10 min at 14,000 rpm and
injected (20 μL) into a Zorbax 300SB C18 column (Agilent). Samples
were run for 10 min at 25 °C and flow rate of 0.9 mL/min using an
isocratic gradient (0.1% TFA). Absorbance was followed at 260 nm and
solutions of known concentrations of free INH were used for calibra-
tion. Data analysis was performed using OpenLAB CDS ChemStation for
LC 1.15.26 (Agilent). CFZ loading was determined by solubilising the
nanobiotics in DMSO and measuring the absorbance at 450 nm in a
microplate reader. Solutions of known concentrations of free CFZ were
used for calibration. 2.7. Nanobiotics characterization by Dynamic Light Scattering (DLS) and
Electrophoretic Light Scattering (ELS) 3.1. Synthesis and characterization of isoniazid-based polymer Both blank and INH-based polymers were characterized by Fourier-
transformed infrared spectroscopy (FTIR) (Fig. 2a) and 1H nuclear
magnetic resonance (NMR; Fig. 2b). The characteristic FTIR peaks from
the α-keto polyester blank polymer appear at 2933 and 2856 cm−1 due
to C–H stretching vibrations from CH2 and CH3 functional groups, at
1271 and 1179 cm−1 due to C–O stretching from ester groups, and at
683 cm−1 due to CeC]O bending. The INH drug has a characteristic
peak at 3303 cm−1, related to NeH stretching of the hydrazide group,
which shifts to 3255 cm−1 and reduces in intensity upon formation of
the hydrazone bond during conjugation to the polymer [22]. In addi-
tion, INH-polymer conjugation also generates peaks at 1556 cm−1
(from H-N-N bending), at 1136 cm−1 (from NeN stretching of the hy-
drazide group), and at 841 cm−1 (from ring CeCeH bending vibra-
tions). The peaks in the region of 3100-2900 cm−1 characteristic from
C–H stretching vibrations of heteroaromatic compounds, present in the 2.13. Zebrafish husbandry and ethic statements Experimental procedures were performed using the nacre line zeb-
rafish. Transgenic Tg(mpeg:mCherryCAAX)sh378 zebrafish line was used
to visualize macrophages chemotaxis towards injection sites. Zebrafish
were raised and maintained according to standard protocols in UK
Home Office-approved facilities in The Bateson Centre aquaria at the
University of Sheffield under AWERB (Animal Welfare and Ethical
Review Bodies). Eggs were obtained from pairs of adult fish by natural
spawning and raised at 28.5 °C in tank water. All animal experiments
described in the present study were conducted on the Project Licence
P1A4A7A5E held by Professor Stephen Renshaw at the University of
Sheffield. 2.12. In vitro nanobiotics uptake by THP-1 cells THP-1 cells were plated on glass coverslips, infected with a mCherry
fluorescent reporter strain of M. tuberculosis H37Rv ΔleuD ΔpanCD
(Bleupan) [19] using a MOI of 10:1, and treated for 1 h with Nano INH
Cou-6 (to a final concentration of 50 nM Cou-6), rinsed with PBS, fixed
with 4% formaldehyde, rinsed with water and then mounted with
ProLong Gold antifade containing DAPI (Invitrogen). Images were ac-
quired on a Zeiss 780 confocal microscope (Plan-Apochromat × 63/
1.40 Oil-immersion lens) and analysed with Zen 2010 (Carl Zeiss) and
Fiji (open source). 2.11. In vitro mycobacterial infection assays At 4 h post-infection, either free antibiotics or nanobiotics of known
concentrations were intravenously administered to embryos. Groups of
infected/treated embryos were then transferred into 6-well plates and
incubated at 28.5 °C. To determine efficiency of nanodrugs vs free drugs
on infection outcomes, embryos were collected at 3 days post infection/
treatment and imaged for both granuloma quantification (defined at
least 10 infected cells) and bacterial burdens analysis as Fluorescent
Pixel Count (FPC) by fluorescence microscopy. Primary monocyte-derived human macrophages, generated as de-
scribed [18], from healthy consented subjects (Regional Ethics ap-
proval: REC: 14/EE/1187 IRAS: 161095), or THP-1 cells (ATCC) were
differentiated by treatment with either 100 ng/mL MCSF or 5 ng/mL
PMA 48 h before infection, inoculated with M. tuberculosis H37Rv ΔleuD
ΔpanCD (Bleupan) [19], grown as described [20], using a multiplicity
of infection (MOI) of 10:1 for 2 h at 37 °C, washed with PBS, and in-
cubated with either DMEM media, supplemented with 10% FCS, 0.4%
L-leucine, 0.1% calcium pantothenate (untreated control), Nano Blank
in media (negative control), INH in media (positive control) or Nano
INH in media for 48 h at 37 °C. Two drug concentrations were tested: 10
μM and 100 μM. Cells were lysed by osmotic shock and intracellular
bacteria were counted. In case of primary monocyte-derived human
macrophages, cells were lysed and plated to count colony-forming units
(CFUs). In the case of THP-1 cells, a validated luminescent reporter
strain of M. tuberculosis H37Rv ΔleuD ΔpanCD (Bleupan) [19] encoding
the Vibrio luxAB gene was used for infection and luminescence was
measured as described [18] after cells lysis. Correlation between CFUs
and luminescence was established before experiments. Experiments
were carried out in sextuplicate. 2.10. Flow cytometry study of nanobiotic uptake by peripheral blood cells Peripheral blood from healthy volunteers (Regional Ethics approval:
REC: 14/EE/1187 IRAS: 161095) was centrifuged at 500 g and 21 °C for
10 min. The supernatant (plasma) was discarded and the pellet was
resuspended and incubated at 37 °C for 30 min. Nano INH Cou-6 were
added to the cells (to a nanobiotic final concentration of 0.05 mg/mL)
and the mixture was incubated at 37 °C for 30 min, followed by 15 min
at 4 °C. Clinical grade polyclonal human IgG (Vivaglobin ®) was added
to the cells and incubated at 4 °C for 5 min to block Fc receptors. Cells 118 Journal of Controlled Release 314 (2019) 116–124 I.L. Batalha, et al. density at 600 nm (OD600) of 1 in PBST and mixed with phenol red. density at 600 nm (OD600) of 1 in PBST and mixed with phenol red. were stained with the antibodies above for 30 min at 4 °C. Red blood
cells were lysed using BD FACS™lysing solution (BD Biosciences), fixed
with BD Cell Fix and transferred to Corning™Falcon™test tube with cell
strainer snap cap. Samples were analysed using an BD LSRFortessa™cell
analyzer (BD Biosciences). Data were processed using FlowJo® 10.5.0
software (FlowJo LLC). y
Microinjections of 2 nL of bacterial suspensions of known con-
centration (containing around 150 mycobacteria) were carried out di-
rectly into the caudal vein in 30 hpf embryos previously dechorionated
and 0.02% w/v tricaine-anesthetized. The inoculum size was checked
by injection of 2 nL in sterile PBST and plated on 7H10OADC agar. Infected embryos were then transferred into plates and incubated at
28.5 °C. 2.15. Macrophages recruitment observation Macrophage mobilization towards nanobiotic-injected sites was
elicited through injection of Cou-6-labelled Nano Blank into the muscle
compartment
of
3
days
post-fertilization
transgenic
larvae
Tg
(mpeg1:mCherry-CAAX)sh378 [21]. Leucocytes chemotaxis was visua-
lized and imaged at 1 and 4 h post-injection using confocal microscopy. 2.16. Epifluorescence, confocal microscopy and imaging Epifluorescence microscopy was performed using a Leica MZ10 F
stereomicroscope
(Leica
Microsystems,
Germany)
equipped
with
GXCAM-U3 Series 5 M P (GT Vision) camera. Confocal microscopy was
performed using a Leica TCS-SPE confocal DMi8 inverted microscope
(Leica Microsystems, Germany) using a HC FL PLAB 10x/0.40, 20x or
40x objective lenses and captured using a Hammamatsu ORCA-Flash
4.0 camera (Hammamatsu, Japan). 2.17. Statistical analysis All data are expressed as mean ± SEM. Statistical analysis for
comparing two experimental groups was performed using two-sided
Student’s t-tests. A value of P < 0.05 was considered statistically sig-
nificant. Analyses were performed with Prism 7 (Graph pad Software). Differences are labelled n.s. for not significant, * for P ≤0.05, ** for
P ≤0.01, *** for P ≤0.001 and **** for P ≤0.0001. The sample size
of each experiment was determined to be the minimal necessary for
statistical significance by the common practice in the field. No animals
were excluded from the experiments. 2.14. In vivo mycobacterial infection and treatment M. marinum strain M carrying pTEC27 (Addgene, plasmid 30182)
that express red fluorescent protein (tdTomato) were grown at 28.5 °C
under hygromycin B selection in Middlebrook 7H9 broth medium
supplemented with oleic acid, albumin, dextrose, catalase (OADC) en-
richment and 0.05% Tween 80 (7H9OADC/T). Mid-log-phase cultures of
M. marinum expressing tdTomato were pelleted, washed twice and re-
suspended in PBS Tween (PBST). Mycobacterial suspensions were then
homogenized through a 26-gauge needle and adjusted to an optical 119 I.L. Batalha, et al. Journal of Controlled Release 314 (2019) 116–124 ctrum, are too weak to observe following polymer incorpora-
nds at 3105 cm−1, 1633 cm−1 and 1321 cm−1 generated from
ng and bending vibrations of the NH2 group are present in the
attributed to the aromatic protons of INH. The integration ratio of INH
aromatic protons to polymer chain protons indicated complete con-
jugation of the INH to the polymer. The integration ratio of protons
haracterization of polymer-drug conjugates. a. FTIR spectra of INH (top), Blank Polymer (middle), and INH Polymer (Bottom). b. 1H-NMR spectrum of INH
in DMSO-d6 and peak assignments. Fig. 2. Characterization of polymer-drug conjugates. a. FTIR spectra of INH (top), Blank Polymer (middle), and INH Polymer (Bottom). b. 1H-NMR spectrum of INH
Polymer in DMSO-d6 and peak assignments. Fig. 2. Characterization of polymer-drug conjugates. a. FTIR spectra of INH (top), Blank Polymer (middle), and INH Polymer (
Polymer in DMSO-d6 and peak assignments. rug conjugates. a. FTIR spectra of INH (top), Blank Polymer (middle), and INH Polymer (Bottom). b. 1H-NMR spectrum of INH
ments. Fig. 2. Characterization of polymer-drug conjugates. a. FTIR spectra of INH (top), Blank Polymer (middle), and INH Polymer (Bottom). b. 1H-NMR spectrum of INH
Polymer in DMSO-d6 and peak assignments. INH spectrum, are too weak to observe following polymer incorpora-
tion. Bands at 3105 cm−1, 1633 cm−1 and 1321 cm−1 generated from
stretching and bending vibrations of the NH2 group are present in the
INH spectrum, disappear from the INH polymer spectrum, corrobor-
ating the formation of the hydrazone bond. The peaks at 1663 cm−1
(for INH), 1721 cm−1 (for the blank polymer), and 1685 cm−1 (for the
INH polymer) correspond to C]O stretching of several carbonyl
groups. attributed to the aromatic protons of INH. The integration ratio of INH
aromatic protons to polymer chain protons indicated complete con-
jugation of the INH to the polymer. 2.14. In vivo mycobacterial infection and treatment The integration ratio of protons
from the methyl ester end-groups to protons on the repeating unit in-
dicates a polymer chain with an average of 16 and 18 repeating units,
equivalent to a Mn of 4147 g/mol for the blank polymer and a Mn of
6804 g/mol for INH polymer, respectively. The 1H-NMR spectra of the blank (Figure S1; chemical shifts re-
ported in Supplementary Information) and INH (Fig. 2b; chemical
shifts reported in Supplementary Information) polymers in DMSO-d6
confirmed our predicted structure. The spectrum of the blank polymer
shows the presence of peaks at 2.53–2.56 ppm and 3.04–3.07 ppm from
the protons of the ketoglutarate unit and peaks between 1.26 ppm and
1.67 ppm attributed to the aliphatic protons of the octanediol unit. Protons α to the backbone ester group on the octanediol unit produced
peaks at 3.97–4.00 ppm and 4.16–4.19 ppm (Figure S1). The INH
polymer spectrum shows new peaks at 7.69–7.73 and 8.71–8.78 ppm 3.2. Formulation and characterization of nanobiotics and in vitro drug
release Nanoparticles were generated from both blank and INH-conjugated
polymers using single-emulsion solvent evaporation [23], which we
visualized in their native solutions using Cryo-EM [24,25] (Fig. 3a). Nanoparticles formulated from blank polymer were irregular in shape
and presented crystalline visual appearance with regular shaded pat-
terns (consistent with the semi-crystalline nature of many polyesters
[26]) while those made from INH polymer were perfectly spherical 120 I.L. Batalha, et al. Journal of Controlled Release 314 (2019) 116–124 Fig. 3. Characterization of polymeric nanobiotics a. Cryo-EM images show different shape and crystallinity of the Blank (left) and INH (right) nanobiotics (scale bar,
100 nm). b. Size distribution (nm) and zeta potential (mV) of Blank (blue) and INH (red) polymeric nanobiotics formulated by single emulsion solvent evaporation
technique (n = 3). c. INH release from Nano INH at pH 5 (blue), 6 (red), and 7.4 (green), mimicking the acidic conditions of the phagolysosome and physiological
conditions of systemic circulation (n = 3). (For interpretation of the references to colour in this figure legend, the reader is referred to the web version of this article). Fig. 3. Characterization of polymeric nanobiotics a. Cryo-EM images show different shape and crystallinity of the Blank (left) and INH (right) nanobiotics (scale bar,
100 nm). b. Size distribution (nm) and zeta potential (mV) of Blank (blue) and INH (red) polymeric nanobiotics formulated by single emulsion solvent evaporation
technique (n = 3). c. INH release from Nano INH at pH 5 (blue), 6 (red), and 7.4 (green), mimicking the acidic conditions of the phagolysosome and physiological
conditions of systemic circulation (n = 3). (For interpretation of the references to colour in this figure legend, the reader is referred to the web version of this article). Fig. 4. Nanobiotic uptake by phagocytic cells and in vitro efficacy against M. tuberculosis. a. Nanobiotics uptake by white blood cells quantified by fluorescence-
activated cell sorting (FACS). Coumarin 6-labelled Nano INH are preferentially uptaken by phagocytic cells, namely monocytes (CD14+) and neutrophils (CD15+),
ather than lymphocytes, such as T cells (CD3+) and B cells (CD19+), likely reflecting their large size distribution. b. Confocal microscopy images of differentiated
THP-1 cells infected with a mCherry fluorescent reporter strain of M. tuberculosis H37Rv ΔleuD ΔpanCD (red) and treated with Cou-6-labelled Nano INH (green) (scale
bar, 10 μm) c. Differentiated THP-1 cells were infected with a luminescent reporter strain M. 3.2. Formulation and characterization of nanobiotics and in vitro drug
release tuberculosis H37Rv ΔleuD ΔpanCD, treated with 100 μM INH either as a
ree drug or as nanodrug and intracellular M. tuberculosis was assessed 48 h post-infection by relative luminescence units (RLUs). Untreated cells and cells treated
with drug-free nanobiotics (Nano Blank) were used as negative controls. Results are presented in terms of RLUs normalized to untreated cells (Mean ± SEM, n = 6). For interpretation of the references to colour in this figure legend, the reader is referred to the web version of this article). Fig. 4. Nanobiotic uptake by phagocytic cells and in vitro efficacy against M. tuberculosis. a. Nanobiotics uptake by white blood cells quantified by fluorescence-
activated cell sorting (FACS). Coumarin 6-labelled Nano INH are preferentially uptaken by phagocytic cells, namely monocytes (CD14+) and neutrophils (CD15+),
rather than lymphocytes, such as T cells (CD3+) and B cells (CD19+), likely reflecting their large size distribution. b. Confocal microscopy images of differentiated
THP-1 cells infected with a mCherry fluorescent reporter strain of M. tuberculosis H37Rv ΔleuD ΔpanCD (red) and treated with Cou-6-labelled Nano INH (green) (scale
bar, 10 μm) c. Differentiated THP-1 cells were infected with a luminescent reporter strain M. tuberculosis H37Rv ΔleuD ΔpanCD, treated with 100 μM INH either as a
free drug or as nanodrug and intracellular M. tuberculosis was assessed 48 h post-infection by relative luminescence units (RLUs). Untreated cells and cells treated
with drug-free nanobiotics (Nano Blank) were used as negative controls. Results are presented in terms of RLUs normalized to untreated cells (Mean ± SEM, n = 6). (For interpretation of the references to colour in this figure legend, the reader is referred to the web version of this article). 392 ± 75 nm), had similar polydispersity index (0.321 compared to
0.344), and had less negative zeta potentials (-20 ± 3 mV compared to
-31 ± 5 mV; consistent with the presence of basic functional groups). INH loading in the nanobiotics was 25 ± 5% wt (g INH/g nanobiotic),
higher than 16% wt reported for INH-chitosan conjugates [13] and 30
times higher than the 0.8% wt (7.78 mg/g) reported for other INH (reportedly favouring uptake by phagocytes [27]), densely packed, and
presented amorphous visual appearance, possibly due to an increase in
disordered regions due to the presence of the bulky pyridyl ring after
drug conjugation [28]. Using Dynamic Light Scattering (Fig. 3.4. Pre-clinical studies in a zebrafish larval model of mycobacterial
infection We proceeded to examine the fate and activity of nanobiotics in vivo
by
exploiting
the
optical
transparency
of
zebrafish
larvae. Nanoparticles were rapidly engulfed by macrophages following in-
tramuscular injection (Fig. 5a) and, in fish infected with M. marinum (a
pathogenic mycobacterial species closely related to Mtb), nanoparticles
were taken up by over 70% of all infected macrophages (Fig. 5b&c). By
3 days post infection, we observed delivery of nanoparticles, pre-
sumably by macrophages, to both granuloma and extracellular myco-
bacterial cords (Fig. 5d). Funding This work was supported by the Rosetrees Trust Interdisciplinary
Prize 2015 (ILB, MS, MEW, RAF), the Wellcome Trust awards 107032/
Z/15/Z
(RAF)
and
10/H0305/55
(MS),
the
NIHR
Cambridge
Biomedical Research Centre Award (RAF), the MRC AMR Theme
award MR/N02995X/1 (SAR, AB) and Marie-Curie IF CFZEBRA 751977
(AB). We next explored the potential application of nanobiotics for syn-
chronous delivery of multiple drugs and successfully encapsulated CFZ
within INH-nanobiotics (Nano INH & CFZ), with a drug loading of
22 ± 1% wt (g CFZ/g nanobiotic), and these remained structurally
stable in solution for 9 months at room temperature (Supplementary
Figure S3).i Author contributions I.L.B. designed and performed the experiments, and analysed the
data. A.B. and S.A.R. were responsible for the in vivo testing in zebrafish
model. C.P. assisted with bacterial killing assays. C.K. assisted with
fluorescence-activated cell sorting. K.S. and P.C.H. performed cryo-EM
imaging. M.S., M.M.O., S.K. and S.M. started the project and performed
initial experiments. I.L.B., R.A.F. and M.E.W. wrote the paper, with
contributions from all co-authors. R.A.F. and M.E.W. supervised the
project. Zebrafish larvae were then infected with M. marinum and, 4 h later,
treated with either free drug (INH alone or INH with CFZ) or injected
with nanobiotics (Nano INH, Nano INH & CFZ, or blank nanoparticles). At 3 days post infection, both INH- and INH & CFZ- nanobiotics, but not
the equivalent concentration of free drugs, were able to significantly
reduce bacterial burden and granuloma number in M. marinum-infected
fish compared to controls (Fig. 6a-c). 3.2. Formulation and characterization of nanobiotics and in vitro drug
release 3b), we
found that, compared to blank nanoparticles, INH nanobiotics were
slightly
smaller
(Z-average
of
284 ± 11 nm
compared
to 121 I.L. Batalha, et al. Journal of Controlled Release 314 (2019) 116–124 conjugated polyesters, such as PLGA [15]. In recent years, zebrafish has been recognised as a useful vertebrate
animal model, particularly due to its low cost, ease of manipulation and
optical transparency, which allows non-invasive and real-time mon-
itoring using imaging tools of host-pathogen interactions at a cellular
level in a live animal [37]. Despite mammal models being evolutionary
more similar to humans, zebrafish and human genomes present high
homology (71% of human protein-encoding genes and 82% of disease
related genes have zebrafish orthologues), with functional domains of
proteins being almost identical in both species [38]. The zebrafish-M. marinum model also presents pharmacological similarities (i.e. similar
effect of drugs) and homologous immune responses to humans, in-
cluding robust granuloma formation [35,36]. This model is not, of
course, intended to replace mammalian infection models, but is rather
an unique and powerful tool for phenotypic screenings and to study
pathophysiological events. We next examined the drug release properties of INH nanobiotics in
vitro (Fig. 3c). As expected from the hydrolytic cleavage of the hy-
drazone bond, greater and more rapid INH release was observed during
incubation at acidic pH levels that replicate those reported for Mtb-
containing phagosomes [29] and the inside of Mtb granulomas [29,30]. 3.3. In vitro evaluation of INH nanobiotics in Mtb-infected human primary
macrophages and THP-1 cells We next examined the drug release properties of INH nanobiotics in
vitro (Fig. 3c). As expected from the hydrolytic cleavage of the hy-
drazone bond, greater and more rapid INH release was observed during
incubation at acidic pH levels that replicate those reported for Mtb-
containing phagosomes [29] and the inside of Mtb granulomas [29,30]. 3.3. In vitro evaluation of INH nanobiotics in Mtb-infected human primary
macrophages and THP-1 cells We hypothesized that nanobiotics would be successfully targeted to
infecting Mtb since, following internalization by phagocytes, they
would be targeted directly to intracellular mycobacteria, through en-
dosomal fusion [31], and delivered to extracellular mycobacteria in
granuloma or cords through the frequent trafficking of macrophages
and neutrophils to these sites [32–34]. Declaration of Competing Interest i
Due to the wide-ranging tools and strains available, the mouse in-
fection model has been the most extensively studied in Mtb research. However, the main disadvantage of this model is the inability of mice to
effectively replicate human pathologies, such as the caseous granuloma
formation. Instead, mice form diffuse and noncaseating lesions, likely
due to the fact that Mtb is not a natural pathogen of mice. Other
mammalian models, such as guinea pigs and rabbits, which produce
necrotic granulomas and more closely resemble the human Mtb pa-
thology, are not as amenable for transgenic and knockout line pro-
duction. The primate infection model (e.g. macaques) is perhaps the
most clinically relevant, but it is limited by high costs and ethical re-
strictions [35,36]. 4. Conclusions To investigate uptake by human cells, we incubated peripheral
blood samples from healthy subjects with fluorescently-labelled INH
nanobiotics. Using flow cytometry, we observed rapid high-level ac-
cumulation in all monocytes and neutrophils, as well as less efficient
uptake by B and T cells (Fig. 4a). We also monitored internalization by
Mtb-infected cells using confocal microscopy (Fig. 4b; Supplementary
Movie). Nanobiotics were avidly taken up by macrophage cell lines and
trafficked to internal compartments including mycobacteria-containing
phagosomes. We then confirmed that INH nanobiotics were active
against intracellular Mtb, showing equivalent potency to free INH when
added to infected THP-1 cells and primary human macrophages
(Fig. 4c; Supplementary Figure S2). We report a smart multi-drug delivery vehicle, which allows the
simultaneous incorporation of both hydrophilic and hydrophobic drugs
at high concentrations and their targeted delivery to both intracellular
and granuloma-resident mycobacteria in vivo. The main advantage of
this system is the synthetic simplicity and versatility. The drug is di-
rectly conjugated to the polymer without the need for any further
chemical modifications. The drug-polymer bond is acid-labile, allowing
site-specific drug release, and the polymer itself is hydrolysable facil-
itating excretion. Polymer size can be tuned without affecting the high
drug loading capacity, since there is one drug conjugation site per
monomeric unit of polymer. With the slow development of new anti-
biotics, tunable polymeric nanobiotics have the potential to deliver
more effective and more tolerable combination chemotherapy using
existing drugs for Mtb and other infectious diseases. Acknowledgments We are grateful to Dr Rishi Matadeen (Cambridge Cryo-EM
Pharmaceutical Consortium) and Dr Burak Caliskan (Nanoscience
Centre, University of Cambridge), for their assistance with Cryo-EM
imaging and FTIR respectively, and Drs Jonathan Howe and Jerome
Boulanger (MRC Laboratory of Molecular Biology, Cambridge) for
training in light microscopy and image analysis. We thank: the NMR
facility, Department of Chemistry University of Cambridge; the ZF 122 Journal of Controlled Release 314 (2019) 116–124 I.L. Batalha, et al. ig. 5. In vivo testing of multi-drug nanobiotics in a M. marinum-infected zebrafish larval model. a. Confocal microscopy images showing nanobiotics-induced
acrophage mobilization in vivo. Suspension of coumarin 6-labelled Nano Blank (green) was injected into the muscle of 3 dpf Tg(mpeg1:mCherryCAAX)sh378
ebrafish line harbouring red macrophages. Macrophage chemotaxis towards injection site has been monitored at 1 and 4 h post injection (scale bar, 20 μm). b. uantification and c. Confocal imaging of coumarin 6-labelled Nano Blank (green) uptake by M. marinum-infected macrophages (red) after 4 h post infection (scale
ar, 1 μm). d. Confocal imaging showing the repartition and accumulation of coumarin 6-labelled Nano Blank (green) into a M. marinum (red)-granuloma (left, scale
ar, 5 μm) and a mycobacterial cord structure (right, scale bar, 5 μm). (For interpretation of the references to colour in this figure legend, the reader is referred to the
eb version of this article). Fig. 5. In vivo testing of multi-drug nanobiotics in a M. marinum-infected zebrafish larval model. a. Confocal microscopy images showing nanobiotics-induced
macrophage mobilization in vivo. Suspension of coumarin 6-labelled Nano Blank (green) was injected into the muscle of 3 dpf Tg(mpeg1:mCherryCAAX)sh378
zebrafish line harbouring red macrophages. Macrophage chemotaxis towards injection site has been monitored at 1 and 4 h post injection (scale bar, 20 μm). b. Quantification and c. Confocal imaging of coumarin 6-labelled Nano Blank (green) uptake by M. marinum-infected macrophages (red) after 4 h post infection (scale
bar, 1 μm). d. Confocal imaging showing the repartition and accumulation of coumarin 6-labelled Nano Blank (green) into a M. marinum (red)-granuloma (left, scale
bar, 5 μm) and a mycobacterial cord structure (right, scale bar, 5 μm). (For interpretation of the references to colour in this figure legend, the reader is referred to the
web version of this article). infected with fluorescently-labelled M. marinum. a. representative images (scale bar, 200 μm). b. rom 2 independent experiments; n = 21). c. Quantification of granuloma number at 3dpi. Acknowledgments Results
= 19) Fig. 6. Effect of nanobiotics at 3 days post infection on zebrafish infected with fluorescently-labelled M. marinum. a. representative images (scale bar, 200 μm). b. quantification of bacterial load (results plotted as mean ± SEM from 2 independent experiments; n = 21). c. Quantification of granuloma number at 3dpi. Results
are plotted as mean ± SEM from 2 independent experiments (n = 19). Fig. 6. Effect of nanobiotics at 3 days post infection on zebrafish infected with fluorescently-labelled M. marinum. a. representative images (scale bar, 200 μm). b. quantification of bacterial load (results plotted as mean ± SEM from 2 independent experiments; n = 21). c. Quantification of granuloma number at 3dpi. Results
are plotted as mean ± SEM from 2 independent experiments (n = 19). 123 I.L. Batalha, et al. Journal of Controlled Release 314 (2019) 116–124 Journal of Controlled Release 314 (2019) 116–124 facility at The Bateson Centre, and Dr Phil Elks (Bateson Centre,
University of Sheffield) for giving access to his injection platform and
microscopy facility. Biomaterials 52 (2015) 417–425. Biomaterials 52 (2015) 417–425. [16] A. Zumla, P. Nahid, S.T. Cole, Nat. Rev. Drug Discov. 12 (2013) 388. [17] J.P. Sarathy, F. Zuccotto, H. Hsinpin, L. Sandberg, L.E. Via, G.A. Marriner,
T. Masquelin, P. Wyatt, P. Ray, Vr. Dartois, ACS Infect. Dis. 2 (2016) 552–563. [17] J.P. Sarathy, F. Zuccotto, H. Hsinpin, L. Sandberg, L.E. Via, G.A. Marriner,
T. Masquelin, P. Wyatt, P. Ray, Vr. Dartois, ACS Infect. Dis. 2 (2016) 552–563. [18] B. Kampmann, P.O. Gaora, V.A. Snewin, M.P. Gares, D.B. Young, M. Levin, J. Infect. Dis. 182 (3) (2000) 895–901. References [22] V. Ferraresi-Curotto, G.A. Echeverría, O.E. Piro, R. Pis-Diez, A.C. González-Baró, J. Mol. Struct. 1133 (2017) 436–447. [1] Global Tuberculosis Report 2018, World Health Organization, Geneva, 2018
Licence: CC BY-NC-SA 3.0 IGO. [23] C.P. Reis, R.J. Neufeld, A.J. Ribeiro, F. Veiga, Nanomedicine 2 (2006) 8–21. [24] P. Renz, M. Kokkinopoulou, K. Landfester, I. Lieberwirth, Macromol. Chem. Phys. 217 (2016) 1879–1885. [2] E. Nathanson, P. Nunn, M. Uplekar, K. Floyd, E. Jaramillo, K. Lönnroth, D. Weil,
M. Raviglione, N. Engl. J. Med. 363 (2010) 1050–1058. [25] P.L. Stewart, Wiley Interdiscip. Rev. Nanomed. Nanobiotechnol. 9 (2017) e1417. [
]
k di
i
l
l
(
) [25] P.L. Stewart, Wiley Interdiscip. Rev. Nanomed. Nanobiotechnol. 9 (2017) e1417. [26] H.K. Makadia, S.J. Siegel, Polymers 3 (2011) 1377–1397. [3] J. Costa-Gouveia, J.A. Ainsa, P. Brodin, A. Lucia, Drug Discov. Today 22 (2017)
600–607. [26] H.K. Makadia, S.J. Siegel, Polymers 3 (2011) 1377–1397. [26] H.K. Makadia, S.J. Siegel, Polymers 3 (2011) 1377–1397. [27] J.A. Champion, S. Mitragotri, Proc. Natl. Acad. Sci. U. S. A. 103 (2006) 4930–49 [4] M.E. Fox, F.C. Szoka, J.M. Fréchet, Acc. Chem. Res. 42 (2009) 1141–1151.f [4] M.E. Fox, F.C. Szoka, J.M. Fréchet, Acc. Chem. Res. 42 (2009) 1141–1151. [5] I. Ekladious, Y.L. Colson, M.W. Grinstaff, Nat. Rev. Drug Discov. (2018) 1. [28] G. Pertici, The effect of molecular structure on the properties of biomedical poly-
mers, Durability and Reliability of Medical Polymers, Elsevier, 2012, pp. 30–48. [5] I. Ekladious, Y.L. Colson, M.W. Grinstaff, Nat. Rev. Drug Discov. (2018) 1. f
[6] F. Danhier, E. Ansorena, J.M. Silva, R. Coco, A. Le Breton, V. Préat, J. Control. Release 161 (2012) 505–522. [29] T. Yuan, N.S. Sampson, Chem. Rev. 118 (2018) 1887–1916. [30] R.R. Kempker, M.T. Heinrichs, K. Nikolaishvili, I. Sabulua, N. Bablishvili, [30] R.R. Kempker, M.T. Heinrichs, K. Nikolaishvili, I. Sabulua, N. Bablishvili,
S. Gogishvili, Z. Avaliani, N. Tukvadze, B. Little, A. Bernheim, Antimicrob. Agents
Chemother. 61 (2017) e00226–00217. [7] K. Xu, Z.C. Liang, X. Ding, H. Hu, S. Liu, M. Nurmik, S. Bi, F. Hu, Z. Ji, J. Ren, Adv
Healthcare Mater. 7 (2018) 1700509. S. Gogishvili, Z. Avaliani, N. Tukvadze, B. Little, A. Bernheim, Antimicrob. Agents
Chemother. 61 (2017) e00226–00217. [8] J. Bhattacharyya, I. Weitzhandler, S.B. Ho, J.R. McDaniel, X. Li, L. Tang, J. Liu,
M. Dewhirst, A. Chilkoti, Adv. Funct. Mater. 27 (2017) 1605421. [31] I. Canton, G. Battaglia, Chem. Soc. Rev. 41 (2012) 2718–2739 [32] C.T. Appendix A. Supplementary data [19] S.L. Sampson, C.C. Dascher, V.K. Sambandamurthy, R.G. Russell, W.R. Jacobs,
B.R. Bloom, M.K. Hondalus, Infect. Immun. 72 (2004) 3031–3037. Supplementary material related to this article can be found, in the
online version, at doi:https://doi.org/10.1016/j.jconrel.2019.10.009. [20] C.B. Péan, M. Schiebler, S.W.S. Tan, J.A. Sharrock, K. Kierdorf, K.P. Brown,
M.C. Maserumule, S. Menezes, M. Pilátová, K. Bronda, P. Guermonprez,
B.M. Stramer, R. Andres Floto, M.S. Dionne, Nat. Commun. 8 (2017) 14642. M.C. Maserumule, S. Menezes, M. Pilátová, K. Bronda, P. Guermonprez,
B.M. Stramer, R. Andres Floto, M.S. Dionne, Nat. Commun. 8 (2017) 14642. [21] A. Bojarczuk, K.A. Miller, R. Hotham, A. Lewis, N.V. Ogryzko, A.A. Kamuyango,
H. Frost, R.H. Gibson, E. Stillman, R.C.J.S. May, Sci. Rep. 6 (2016) 21489. References Yang, C.J. Cambier, J.M. Davis, C.J. Hall, P.S. Crosier, L. Ramakrishnan, Cell
Host Microbe 12 (2012) 301–312. [9] J. Ritsema, E. Herschberg, S. Borgos, C. Løvmo, R. Schmid, Y. Te [9] J. Ritsema, E. Herschberg, S. Borgos, C. Løvmo, R. Schmid, Y. Te Welsch
G. Storm, C.F. van Nostrum, Int. J. Pharm. 548 (2018) 730–739. Host Microbe 12 (2012) 301–312. [33] C.H. Liu, H. Liu, B. Ge, Cell. Mol. Immunol. 14 (2017) 963 [10] C. Thedrattanawong, C. Manaspon, N. Nasongkla, J. Drug Deliv. Sci. Technol. (2018). [34] S. Ehrt, D. Schnappinger, K.Y. Rhee, Nat. Rev. Microbiol. 16 ( [34] S. Ehrt, D. Schnappinger, K.Y. Rhee, Nat. Rev. Microbiol. 16 (2018) 496–507. [35] G. Bouz, N. Al Hasawi, Crit. Rev. Microbiol. 44 (2018) 779–792 [11] S. Guo, Y. Nakagawa, A. Barhoumi, W. Wang, C. Zhan, R. Tong, C. Santamaria,
D.S. Kohane, J. Am. Chem. Soc. 138 (2016) 6127–6130. [36] M.R. Cronan, D.M. Tobin, Dis. Models Mech. 7 (2014) 777–784. [37] A. Bernut, J.-L. Herrmann, D. Ordway, L. Kremer, Front. Cell. Infect. Microbiol. 7
(2017) 100. [12] S. Sharma, Infect. Genet. Evol. 4 (2004) 167. [13] A.S. Berezin, Y.A. Skorik, Carbohydr. Polym. 127 (2015) 309–3 [38] D.G. Howe, Y.M. Bradford, A. Eagle, D. Fashena, K. Frazer, P. Kalita, P. Mani,
R. Martin, S.T. Moxon, H. Paddock, Nucleic Acids Res. 45 (2016) D758–D768. [14] M.L. Manca, R. Cassano, D. Valenti, S. Trombino, T. Ferrarelli, N. Picci, A.M. Fadda,
M. Manconi, J. Pharm. Pharmacol. 65 (2013) 1302–1311. [14] M.L. Manca, R. Cassano, D. Valenti, S. Trombino, T. Ferrarelli, N M. Manconi, J. Pharm. Pharmacol. 65 (2013) 1302–1311. [15] D. Huang, D. Li, T. Wang, H. Shen, P. Zhao, B. Liu, Y. You, Y. Ma, F. Yang, D. Wu, [15] D. Huang, D. Li, T. Wang, H. Shen, P. Zhao, B. Liu, Y. You, Y. M 124
|
https://openalex.org/W3003728570
|
https://link.springer.com/content/pdf/10.1007/s11661-020-05648-w.pdf
|
English
| null |
Size Effects of Hardness and Strain Rate Sensitivity in Amorphous Silicon Measured by Nanoindentation
|
Metallurgical and materials transactions. A, Physical metallurgy and materials science
| 2,020
|
cc-by
| 6,073
|
I.
INTRODUCTION a
deeper
understanding
between
processing-struc-
ture-property-performance relationship for a-Si, much
recent research interest has focused on characterizing
hardness,
Young’s
modulus,
and
pressure-induced
phase transformations measured, i.e., by nanoindenta-
tion.[3,5–7] This measurement method is particularly
dedicated and successfully applied for studying of
mechanical properties of thin films.[8–11] For example,
Kiran et al. performed high-temperature nanoindenta-
tion on pure ion-implanted amorphous silicon.[5] They
shown that unrelaxed a-Si deforms entirely via plastic
flow, whereas a clear transition in the mode of defor-
mation is observed in an annealed a-Si. Furthermore,
Haberl et al. studied the deformation behavior of a-Si
using spherical nanoindentation[7] and again observed
phase transformations only in the annealed sample. It
should also be noted that the phase transformation was
different for different unloading rates. Furthermore, the
issue of indentation-induced phase transformations in
silicon has been widely studied.[12–15] Unfortunately, the
problem of dependence of mechanical properties, i.e.,
hardness of a-Si on indentation depth or strain rate, can
rarely be found in the literature. THE use of thin films in many branches of modern
industry, i.e., in photonics, micro- and nanoelectronics,
and
micro-
and
nano-electromechanical
systems
(MEMS and NEMS), has generated a strong interest
in their mechanical properties. The requirements for
mechanical reliability and durability become more
crucial with increasing the complexity of the systems. Amorphous silicon (a-Si) is one of the most common
materials used for manufacturing of MEMS.[1,2] In the
literature, one can find evidence that processing condi-
tions like deposition temperature, pressure, and flow
rate of gases and processes like etching, chemical
treatments, or ion implantation are known to affect
the microstructural features as well as the distribution of
volume and surface defects in this material.[3–5] To gain DARIUSZ M. JARZA˛ BEK is with the Institute of Fundamental
Technological Research, Polish Academy of Sciences, 02-106 Warsaw,
Poland and also with the Institute of Fundamental Technological
Research, Polish Academy of Sciences, Pawinskiego 5B, 02-106
Warsaw,
Poland. Contact
e-mail:
djarz@ippt.pan.pl
MICHAŁ
MILCZAREK and SZYMON NOSEWICZ are with the Institute of
Fundamental Technological Research, Polish Academy of Sciences. PIOTR BAZARNIK is with the Warsaw University of Technology,
00-661 Warsaw, Poland. HELMUT SCHIFT is with the Paul Scherrer
Institut, 5232 Villigen PSI, Switzerland. i
b
i
d S
b One of the parameters that indicate the dynamic
behavior of the material is the strain rate sensitivity
(SRS). DARIUSZ M. JARZA˛ BEK, MICHAŁ MILCZAREK, SZYMON NOSEWICZ,
PIOTR BAZARNIK, and HELMUT SCHIFT In this work, dynamic mechanical properties of amorphous silicon and scale effects were
investigated by the means of nanoindentation. An amorphous silicon sample was prepared by
plasma-enhanced chemical vapor deposition (PECVD). Next, two sets of the samples were
investigated: as-deposited and annealed in 500 C for 1 hour. A three-sided pyramidal diamond
Berkovich’s indenter was used for the nanoindentation tests. In order to determine the strain
rate sensitivity (SRS), indentations with different loading rates were performed: 0.1, 1, 10,
100 mN/min. Size effects were studied by application of maximum indentation loads in the
range from 1 up to 5 mN (penetrating up to approximately one-third of the amorphous layer). The value of hardness was determined by the Oliver–Pharr method. An increase of hardness
with decrease of the indentation depth was observed for both samples. Furthermore, the
significant dependence of hardness on the strain rate has been reported. Finally, for the
annealed samples at low strain rates a characteristic ‘‘elbow’’ during unloading was observed on
the force-indentation depth curves. It could be attributed to the transformation of (b-Sn)-Si to
the PI (pressure-induced) a-Si end phase. https://doi.org/10.1007/s11661-020-05648-w
The Author(s) 2020 https://doi.org/10.1007/s11661-020-05648-w
The Author(s) 2020 https://doi.org/10.1007/s11661-020-05648-w
The Author(s) 2020 DARIUSZ M. JARZA˛ BEK is with the Institute of Fundamental
Technological Research, Polish Academy of Sciences, 02-106 Warsaw,
Poland and also with the Institute of Fundamental Technological
Research, Polish Academy of Sciences, Pawinskiego 5B, 02-106
Warsaw,
Poland.
Contact
e-mail:
djarz@ippt.pan.pl
MICHAŁ
MILCZAREK and SZYMON NOSEWICZ are with the Institute of
Fundamental Technological Research, Polish Academy of Sciences.
PIOTR BAZARNIK is with the Warsaw University of Technology,
00-661 Warsaw, Poland. HELMUT SCHIFT is with the Paul Scherrer
Institut, 5232 Villigen PSI, Switzerland.
M
i t
b
itt d S
t
b
24 2019 I.
INTRODUCTION In the literature, one can find many papers about
measuring
SRS
of
different
materials
with
split
Hopkinson bar.[16,17] On the other hand, thin films
cannot be investigated with this technique. Fortunately, VOLUME 51A, APRIL 2020—1625 METALLURGICAL AND MATERIALS TRANSACTIONS A NB5000 microscope. The samples do not differ in
structure. Moreover, to confirm the presence of amor-
phous regions selected area electron diffraction (SAED)
was used.[26] Figure 1(a) shows SAED image taken from
interfacial regions confirmed the presence of amorphous
layer (blurred diffraction rings) on the crystalline h001i
Si substrate (diffraction spots with a 001 zone axis). the nanoindentation can also be useful for SRS deter-
mination. This technique offers hardness measurement
with constant strain rates at a small scale. It should be
mentioned that classical nanoindentation is not dedi-
cated to dynamic measurements but the most advanced
nanoindenters are
equipped
with
high-performance
electronics, which allows to apply loads and collect data
with high speeds. (
p
)
Due to the fact that the thickness of the coatings
could vary in different positions at the wafer, it was
determined according to the following procedure. After
the indentations with small forces, an indentation test
with a relatively high force equal to 30 mN was
performed in a distance of about 20 lm. Since nanocrys-
talline silicon is brittle, a part of the coating was split off
and flown free (Figure 1(b)). Next, the created structure
was measured by an atomic force microscope (Nanosurf
Flex AFM) (Figure 1(c)). The thickness of the coating
was determined as a difference between the surface of a
nanocrystalline film and the surface of single-crystal
silicon. An example profile is shown in Figure 1(d). We
have taken into consideration only the results obtained
in the center of the wafer where the coating thickness
was equal to 510 ± 5 nm (the average obtained from all
the experiments). g
p
It is generally accepted that the values of SRS are
strongly size-dependent: for face-centered-cubic (fcc)
metals, SRS normally increases with decreasing grain
sizes, whereas the opposite holds for body-center-cubic
(bcc) structures.[18] In contrast, amorphous, metallic
glasses exhibit high strength, wear, and corrosion
resistance,[19] but their deformation is shear band
controlled; hence, they are generally brittle and their
SRS is close to zero.[20,21] On the other hand, a-Si is
different from amorphous metallic glasses due to the
highly directional covalent bonding that can make
deformation processes more complex. I.
INTRODUCTION Understanding
the pressure-induced deformation in a-Si is not only
important from the scientific point of view but also it is
also relevant for a number of applications, i.e., photo-
voltaic cells and MEMS.[22,23] Unfortunately, there are
many ‘forms’ of a-Si and its structural ordering,
morphology, and impurity content vary considerably
depending on its formation method and thermal history. Therefore, in this paper we perform nanoindentation
tests with different strain rates and depths on two
samples: as-deposited and annealed, in order to study
dynamic properties and size effects in probably the most
popular a-Si films prepared by PECVD.[24,25] B. Nanoindentation The nanoindentation of the samples was performed
on an Ultra Nanoindentation Tester (UNHT Anton
Paar). The device exhibits an extremely low thermal
drift[27] which is of particular importance in long, strain
rate sensitivity measurements.[28] It is due to the fact that
the principle of the UNHT is based on the idea of using
two independent vertical axes: one axis is dedicated to
the indentation measurement itself and the other axis is
used for active top referencing (Figure 2). The system
can be considered as two nanoindentation systems,
which are intimately linked: one responsible for active
top referencing of the sample surface and the other for
performing the indentation. Hence, all thermal dilata-
tion of the sample and the frame are simply disregarded
because of the fact that the displacement of the indenter
is measured with respect to the reference, load of which
is independently controlled—the reference is ‘‘floating’’
on the sample surface. A. Samples Preparation In this work, mechanical properties of amorphous
silicon were investigated by the means of nanoindenta-
tion. An amorphous silicon sample was prepared at the
Paul Scherrer Institute by PECVD (plasma-enhanced
chemical vapor deposition) in the following way. As
substrate a blank 100 mm Si h100i wafer was used. A
chamber and the substrate were precleaned with N2O
for 3 minutes with a flow of 710 sccm, pressure of 1000
mTorr, temperature of 250 C, and with HF power of
20 W. Then a deposition gas was introduced to the
chamber, as a 5 pct of SiF4 in N2O, and a flow stabilized
at the level of 500 sccm with a chamber pressure of 1000
mTorr and temperature of 250 C. Deposition was done
with HF power of 20 W and lasted for 10 minutes. As
the last step, the whole chamber was purged with N2
gas. Before experiments the wafer was cut into smaller
specimens. Some of them were annealed before nanoin-
dentation in 500 C for 1 hour, others were investigated
as-deposited. Nanoindentation was performed by the nanoindenta-
tion tester equipped with a three-sided pyramidal
diamond Berkovich’s indenter. In order to determine
SRS, indentations with different loading rates were
performed: 0.1, 1, 10, 100 mN/min. Size effects were
studied by application of maximum indentation loads in
the range from 1 up to 5 mN (penetrating up to
approximately one-third of the amorphous layer[7]). The value of hardness was determined by the Oliver–
Pharr method. It is a standard procedure for determin-
ing hardness from the nanoindentation load–displace-
ment
curves. The
projected contact
area
between
indenter tip and sample can be estimated by the depth
of impression obtained from a load–displacement curve. The hardness is determined by dividing the load by the
area of contact. On the other hand, one can determine
the projected contact area by direct measurement. It is p
The samples were studied to obtain microstructural
properties with the use of transmission electron micro-
scope (TEM) Jeol 1200, operating under accelerating
voltage of 120 kV. The samples were cut using the
focussed ion beam (FIB) technique, with a Hitachi METALLURGICAL AND MATERIALS TRANSACTIONS A 1626—VOLUME 51A, APRIL 2020 Fig. 1—(a) SAED image taken from this region showing blurred rings stemming from the amorphous layer. A. Samples Preparation (b) SEM micrograph of a spot
where the a-Si coating on top of the silicon substrate was split off; (c) AFM topography plot and (d) profile of the defect indicating the thickness
of the coating. Fig. 1—(a) SAED image taken from this region showing blurred rings stemming from the amorphous layer. (b) SEM micrograph of a spot
where the a-Si coating on top of the silicon substrate was split off; (c) AFM topography plot and (d) profile of the defect indicating the thickness
of the coating. recommended in the case of high pile-ups or significant
thermal drift (Figure 3(a)). We have applied the latter
approach to verify the results obtained by the Oliver–
Pharr method and to confirm the high thermal stability
of our nanoindenter. The indentation imprints were
measured by AFM (Figure 3(b)). The real projected
contact area was then evaluated by applying the
algorithms
available
in
Gwyddion
software
(Figure 3(c)). Fig. 2—A scheme of the nanoindentation device used in this study. III.
RESULTS The exemplary load–displacement curves for two
amorphous
silicon
samples—as-deposited
and
annealed—are shown in Figure 5. The curves were
obtained at various loading rates starting from 100
down to 0.1 mN/min from samples loaded to 5 mN. In
the case of as-deposited sample (Figure 5(a)), both
loading and unloading portions are smooth without any
discontinuities suggesting homogeneous deformation. The quantitative data obtained from the above curves
and curves obtained for 1 mN max load are given in
Table I. We observe that the maximum penetration
depth hmax decreases with the loading rate in both cases. Similarly, the residual depth hp also decreases with
increasing loading rate. The indentation modulus also decreases with the
applied load/indentation depth. However, the decrease
is significantly stronger for the annealed sample. For
each loading rate, for 1 mN max load the value of the
indentation modulus is higher for the annealed sample. On the other hand, for 5 mN max load it is usually
higher for the as-deposited sample. Similarly to the
hardness, the indentation modulus decreases also with
the decrease of the loading rate. The only exception
occurs for the lowest loading rate (0.1 mN/min). A
significant jump of the indentation modulus has been
observed for both investigated samples. g
g
In contrast, the indentation curves obtained from the
annealed a-Si sample under similar experimental condi-
tions shown in Figure 5(b) are significantly different. Firstly, the sample is slightly softer. Secondly, the
unloading curves obtained from indentation with the
loading rates equal to 1 and 0.1 mN/min exhibit a
so-called elbow behavior. The information from the
load–displacement curves is quantitatively confirmed by
the data presented in Table II. Similarly to as-deposited
sample, hmax decreases with the loading rate for both 5
and 1 mN max load. However, hp for both cases start to
decrease for loading rates equal to 1 and 0.1 mN/min. This suggests different modes of elastic recovery or
phase transformation on unloading. Furthermore, one can determine the indentation
strain rate according to Eq. [3] and then plot hardness
vs indentation strain rate as it is shown in Figure 7(a). It
should be noted that when the indentation strain rate is
plotted on a logarithmic scale, the linear relation
between hardness and logarithm of strain rate can be
postulated. The coefficient of determination is equal to
0.94 and 0.89 for the as-deposited and annealed samples,
respectively. C. Determination of Strain Rate Sensitivity Strain rate sensitivity (SRS) provides information
about the influence of strain rate on material properties
such as hardness. In the case of bulk materials, the
tensile tests are used for SRS determination. It is then
defined as the change in the yield stress or tensile
strength r
r0 divided by the change in the uniaxial strain
rate
_e
_e0 at a constant temperature using the following
equation: Fig. 2—A scheme of the nanoindentation device used in this study. m ¼
@ ln r
r0
@ ln _e
_e0
:
½1 mind ¼ @ ln H
@ ln _ei
;
½2 mind ¼ @ ln H
@ ln _ei
;
½2 ½1 ½2 where H is the hardness and _ei is the indentation strain
rate, which according to Reference 29 On the other hand, for nanoindentation, SRS is deter-
mined as follows[18]: METALLURGICAL AND MATERIALS TRANSACTIONS A
VOLUME 51A, APRIL 2020—1627 NSACTIONS A
VOLUME 51A, APRIL 2020—1627 METALLURGICAL AND MATERIALS TRANSACTIONS A VOLUME 51A, APRIL 2020—1627 _ei ¼
_P
2P :
½3 The results of hardness and indentation modulus
measurement of the two samples are shown in Figure 6. The hardness in almost every case is higher for the
as-deposited sample. There are just two exceptions but
the difference is within the measurement error. It is also
apparent that for both samples hardness significantly
depends on the loading rate. In both cases, the highest
hardness values are obtained for the highest loading
rate, which is equal to 100 mN/min. The evidence
suggests the occurrence of a size effect—the hardness
depends on applied load/indentation depth.[30–32] This
effect is clearly visible for the highest loading rate but is
less significant in the case of lower loading rates. For
example, in the case of the as-deposited sample the
difference between measured hardness values obtained
for 1 and 5 mN max load is equal to 3200 MPa when the
loading rate is 100 mN/min and to 1000 MPa when the
loading rate is 0.1 mN/min. A similar effect is observed
for the annealed sample. ½3 In this work, we have plotted ln H vs ln _ei and performed
the linear fit. SRS is then equal to the slope of the fitted
straight line (Figure 4). III.
RESULTS Furthermore, in Figure 7(b) a plot of SRS
vs max load is shown. In general, SRS is higher for the
as-deposited sample and it is independent of max load
for higher loads. However, a significant increase of SRS Fig. 3—(a) Schematic of the indentation of a three-sided pyramidal
tip into the surface. (b) AFM topography plot of the indentation
and (c) processed micrograph. respectively. Furthermore, in Figure 7(b) a plot of SRS
vs max load is shown. In general, SRS is higher for the
as-deposited sample and it is independent of max load
for higher loads. However, a significant increase of SRS
Fig. 3—(a) Schematic of the indentation of a three-sided pyramidal
tip into the surface. (b) AFM topography plot of the indentation
and (c) processed micrograph. Fig. 4—An example ln H vs ln _e plot for SRS determination. 1628—VOLUME 51A, APRIL 2020
METALLURGICAL AND MATERIALS TRANSACTIONS A Fig. 4—An example ln H vs ln _e plot for SRS determination. METALLURGICAL AND MATERIALS TRANSACTIONS A Fig. 3—(a) Schematic of the indentation of a three-sided pyramidal
tip into the surface. (b) AFM topography plot of the indentation
and (c) processed micrograph. Fig. 4—An example ln H vs ln _e plot for SRS determination. METALLURGICAL AND MATERIALS TRANSACTIONS A 1628—VOLUME 51A, APRIL 2020 Fig. 5—Indentation curves. Table I. The Data Obtained from the P–h Curves for 1 and 5 mN Max Load (As-deposited Sample)
Loading Rate (mN/min)
As-deposited Sample
Max Load: 1 mN
Max Load: 5 mN
hmax (nm)
hp (nm)
hmax (nm)
hp (nm)
100
60 ± 1
21 ± 3
161 ± 4
79 ± 5
10
65 ± 2
26 ± 2
168 ± 4
84 ± 6
1
66 ± 2
27 ± 3
169 ± 4
86 ± 8
0.1
67 ± 2
29 ± 1
174 ± 4
93 ± 6
Table II. The Data Obtained from the P–h Curves for 1 and 5 mN Max Load (Annealed Sample)
Loading Rate [mN/min]
Annealed Sample
Max Load: 1 mN
Max Load: 5 mN
hmax
(nm)
hp
(nm)
hmax
(nm)
hp
(nm)
100
61 ± 1
25 ± 3
164 ± 3
80 ± 2 Fig. 5—Indentation curves. Fig. 5—Indentation curves. Table I. The Data Obtained from the P–h Curves for 1 and 5 mN Max Load (As-deposited Sample) Table I. III.
RESULTS The Data Obtained from the P–h Curves for 1 and 5 mN Max Load (As-deposited Sample)
Loading Rate (mN/min)
As-deposited Sample
Max Load: 1 mN
Max Load: 5 mN
hmax (nm)
hp (nm)
hmax (nm)
hp (nm)
100
60 ± 1
21 ± 3
161 ± 4
79 ± 5
10
65 ± 2
26 ± 2
168 ± 4
84 ± 6
1
66 ± 2
27 ± 3
169 ± 4
86 ± 8
0.1
67 ± 2
29 ± 1
174 ± 4
93 ± 6 Table II. The Data Obtained from the P–h Curves for 1 and 5 mN Max Load (Annealed Sample)
Loading Rate [mN/min]
Annealed Sample
Max Load: 1 mN
Max Load: 5 mN
hmax
(nm)
hp
(nm)
hmax
(nm)
hp
(nm)
100
61 ± 1
25 ± 3
164 ± 3
80 ± 2
10
65 ± 1
31 ± 2
174 ± 2
89 ± 3
1
66 ± 1
27 ± 1
175 ± 1
87 ± 4
0.1
67 ± 4
20 ± 2
178 ± 3
83 ± 3
The bolded values indicate the decrease of the residual depth, which suggests different mode of elastic recovery or phase transformation on
unloading. Table II. The Data Obtained from the P–h Curves for 1 and 5 mN Max Load (Annealed Sample) is observed for max load lower than 2 mN. Moreover, in
this case, SRS values for both samples are very similar. AFM measurements of the indentation imprints made
with 5 mN max load are shown in Figure 8. It should be
noted that in order to highlight the pile-ups we have
presented the AFM measurements in two forms. The is observed for max load lower than 2 mN. Moreover, in
this case, SRS values for both samples are very similar. is observed for max load lower than 2 mN. Moreover, in
this case, SRS values for both samples are very similar. upper images show raw AFM images, whereas in the
lower images, in order to show the pile-up patterns in
more detail, the measured imprint was artificially
truncated at the bottom by the thresholding operation. The thresholding is useful especially for suppressing the
deep, high load imprints, as the absolute value of their AFM measurements of the indentation imprints made
with 5 mN max load are shown in Figure 8. III.
RESULTS It should be
noted that in order to highlight the pile-ups we have
presented the AFM measurements in two forms. The VOLUME 51A, APRIL 2020—1629 METALLURGICAL AND MATERIALS TRANSACTIONS A ordinates is much higher than that of the surrounding
pile-ups. The AFM results show that, firstly, pile-ups are
clearly visible around the imprints made in the as-de-
posited sample and they are not observed in the
annealed sample. It indicates that due to annealing the
internal stress caused by the samples preparation
method has been significantly reduced. The same effect
(but not so clearly visible due to samples roughness) has
been observed for imprints made with lower max load. Furthermore, it should be noted that apparently, the
pile-ups are higher for lower loading rates. The average
height of the highest pile-up for 0.1 mN/min is equal to
3.0 ± 0.5 nm whereas for 100 mN/min to 2.4 ± 0.6 nm. Nevertheless, the pile-ups are very small and they should
not
affect
the
hardness
determination
from load–displacement curves according to the Oliver–Pharr
method. This assumption has been confirmed by the
comparison of the hardness determined by direct mea-
surements of the projected contact area. The discrep-
ancy between the Oliver–Pharr method and the direct
measurement was in every case lower than 4 pct which
also confirms the high thermal stability of the used
nanoindentation tester. Annealing has also increased samples roughness. Parameter Ra for the as-deposited sample was equal to
0.3 ± 0.1 nm whereas for the annealed sample to
1.0 ± 0.2. 0—VOLUME 51A, APRIL 2020
METALLURGICAL AND MAT IV.
DISCUSSION Furthermore, we tentatively attribute the
sudden increase of the indentation modulus for the
lowest loading rate to the above discussed phase
transformation. However, this sudden change occurs
for both samples. It is less significant for the as-de-
posited one, which can suggest that although the ‘elbow’
is not observed, partial phase transformation occurs in
this sample. It is probable that the observation of
changes in the indentation modulus is a more sensitive
way to detect phase transformations in the amorphous
silicon. Nevertheless, further studies are required to
confirm this assumptions. To date, the effect of loading rates on the mechanical
behavior of a-Si under the indentation tip has not been
studied in detail. In Figure 6(a), it is shown that the
hardness significantly depends on the loading rate and it
also depends on the maximal indentation load. On the
other hand, high coefficients of determination estimated
for a postulated linear relation between hardness and
logarithm of the strain rate (Figure 7(a)) suggest that the
size effect is apparent. It is probably the effect of plastic
flow under the indenter, which can be described by any
equation in which stress r is related to a strain rate _c
,i.e., Herschel–Bulkley equation: r ¼ r0 þ C_cb;
½4 ½4 where r0 is the quasi-static flow stress and C and b are
constants. p
It should be noted that Haberl et al.[7] loaded and
unloaded their a-Si sample annealed in 450 C with the
loading rate equal to 60 and 120 mN/min and during
unloading they observed a clear discontinuity called a
‘‘pop-out’’ instead of the elbow. Furthermore, Kiran
et al.[5] reported that elbows are formed for higher and
pop-outs for lower unloading rates. We tentatively
attribute these discrepancies to the differences in the
investigated samples. In the above-mentioned papers,
a-Si
prepared
by
self-ion implantation is
studied,
whereas in the current study we have prepared a-Si by
PECVD. The microstructure of the samples prepared by
these two methods may be significantly different. In
general, a-Si prepared by CVD methods usually con-
tains voids,[34] whereas ion implantation has been
recognized as a successful technique to produce pure
and voidless amorphous layers.[35] In this paper, we have
chosen to study PECVD a-Si due to its high importance
for
microelectronic
and
MEMS
industries. IV.
DISCUSSION In this paper, we study size effects of hardness and
strain rate sensitivity of the as-deposited and annealed
PECVD amorphous silicon by nanoindentation. The
increase of hmax with a decrease of loading rate suggests
that as-deposited a-Si exhibits significant plastic flow
under the nanoindentation tip. The load–displacement
curves shown in Figure 5(a) support this statement. There are no discontinuities on either the loading or
unloading sections of the curves. Similar behavior was
previously reported by Haberl et al.[7] In that work, the
authors performed nanoindentation of amorphous sili-
con samples with a spherical indenter and they investi-
gate the cross section of imprints with a transmission
electron microscope. They did not observe any phase
transformations under the indenter tip in their as-de-
posited sample. In contrast, the nanoindentation curves
taken
from
annealed
a-Si
sample,
as
shown
in
Figure 5(b), exhibit a so-called elbow behavior on
unloading for lower loading rates. The occurrence of
this discontinuity has been reported previously by Kiran
et al.[5] or Bradby et al.[33] and attributed to the
transformation of (b-Sn)-Si to the PI (pressure-induced)
a-Si end phase. Therefore, the presence of this distinctive
feature in the unloading curve is a reliable indicator that Fig. 6—The
results
of
hardness
and
indentation
modulus
measurements. Fig. 6—The
results
of
hardness
and
indentation
modulus
measurements. Fig. 7—(a) Plot of hardness vs indentation strain rate in which one can postulate the linear relation; (b) a plot of SRS vs max load. Fig. 7—(a) Plot of hardness vs indentation strain rate in which one can postulate the linear relation; (b) a plot of SRS vs max load. 1630—VOLUME 51A, APRIL 2020
METALLURGICAL AND MATERIALS TRANSACT METALLURGICAL AND MATERIALS TRANSACTIONS A 1630—VOLUME 51A, APRIL 2020 Fig. 8—AFM topography of the nanoindentation imprints and the featured pile-up patterns. Fig. 8—AFM topography of the nanoindentation imprints and the featured pile-up patterns. Furthermore, the observed differences in the behavior
of a-Si under the indentation tip may also be caused also
by different loading rates. metallic (b-Sn)-Si formed on loading—a pressure-in-
duced phase transition had occurred for lower loading
rates. These statements are also supported by the observed
changes in the indentation modulus. The decrease of its
measured value with the decrease of the loading rate can
be explained by the plastic flow which occurs under the
indenter. REFERENCES 1. R.M.R. Pinto, P. Brito, V. Chu, and J.P. Conde: J. Microelec-
tromech. Syst., 2019, vol. 28, pp. 390–400. 2. J. Gaspar, A. Gualdino, B. Lemke, O. Paul, V. Chu, and
J.P. Conde: J. Appl. Phys., 2012, vol. 112, art. no. 024906. pp
y ,
,
,
3. B. Haberl, M. Guthrie, D.J. Sprouster, J.S. Williams, and
J.E. Bradby: J. Appl. Crystallogr., 2013, vol. 46, pp. 758–68. 4. Y.B. Gerbig, C.A. Michaels, J.E. Bradby, B. Haberl, and
R.F. Cook: Phys. Rev. B, 2015, vol. 92, art. no. 214110. 5. M. Kiran, B. Haberl, J.S. Williams, and J.E. Bradby: J. Appl. Phys., 2014, vol. 115, art. no. 113511. y
6. B. Haberl, A.C.Y. Liu, J.E. Bradby, S. Ruffell, J.S. Williams, and
P. Munroe: Phys. Rev. B, 2009, vol. 79, art. no. 155209. P. Munroe: Phys. Rev. B, 2009, vol. 79, art. no. 155209. 7. B. Haberl, J.E. Bradby, M.V. Swain, J.S. Williams, and P. Mun-
roe: Appl. Phys. Lett., 2004, vol. 85, pp. 5559–61. roe: Appl. Phys. Lett., 2004, vol. 85, pp. 5559–6 8. Z. Wu, J. Thou, W. Chen, J. Shen, C. Lv, and Y.Y. Qi: Ceram. Int., 2015, vol. 41, pp. 11632–36. 9. Z.W. Chen and N. Brandon: Ceram. Int., 2016, vol. 42,
pp. 8316–24. 10. D. Naujoks et al.: Acta Mater., 2017, vol. 138, pp. 100–10. 11. S. Kauffmann-Weiss,
S. Hahn,
C. Weigelt,
L. Schultz,
M.F.X. Wagner, and S. Fahler: Acta Mater., 2017, vol. 132,
pp. 255–63. 12. J.E. Bradby, J.S. Williams, J. Wong-Leung, M.V. Swain, and
P. Munroe: J. Mater. Res., 2001, vol. 16, pp. 1500–507. 13. S. Wong et al.: J. Appl. Phys., 2019, vol. 126, art. no. 105901. 14. J.I. Jang, M.J. Lance, S.Q. Wen, T.Y. Tsui, and G.M. Pharr: Acta
Mater., 2005, vol. 53, pp. 1759–70. Furthermore, we have also shown that the hardness of
PECVD a-Si significantly depends on the loading rate
and on the maximal indentation load. These results
presented in this paper may be important especially for
MEMS/NEMS industry in which this material is often
used. Nevertheless, an important question for further
studies is to explain the increase of SRS for small
indentation depths and to study PECVD a-Si with much
higher strain rates. 15. T. Juliano, Y. Gogotsi, and V. Domnich: J. Mater. Res., 2003,
vol. 18, pp. 1192–1201. 16. Y.M. Wei, Z.G. Lu, K.H. Hu, X.Y. Li, and P.J. Li: Ceram. Int.,
2018, vol. 44, pp. 11702–10. 17. D. IV.
DISCUSSION This assumption is to some extent supported by the
results shown in Figure 7(b), where SRS vs max load for
both samples is plotted. The constant SRS for higher
loads confirms that in this range the apparent hardness
size effect is caused by strain rate-dependent plastic flow. On the other hand, for the smallest loads, SRS signif-
icantly increases. This result is encouraging and should
be validated and further explored for example with the
use of nanoindentation strain rate jump test.[36] Unfor-
tunately, we do not have a facility to perform such
measurements. It should be also noted that SRS is
slightly smaller for the annealed sample. According to
Haberl et al.,[7] it may be explained by the fact that
annealing of a-Si reduced the density of ‘‘defects’’
(broken bonds, vacancies, and interstitials-like defects). The defects facilitate the deformation through localized
shear flow; hence, the as-deposited sample, with more
defects, should be characterized by the higher SRS. VOLUME 51A, APRIL 2020—1631 METALLURGICAL AND MATERIALS TRANSACTIONS A OPEN ACCESS This article is licensed under a Creative Commons
Attribution 4.0 International License, which permits
use, sharing, adaptation, distribution and reproduction
in any medium or format, as long as you give appro-
priate credit to the original author(s) and the source,
provide a link to the Creative Commons licence, and
indicate if changes were made. The images or other
third party material in this article are included in the
article’s Creative Commons licence, unless indicated
otherwise in a credit line to the material. If material is
not included in the article’s Creative Commons licence
and your intended use is not permitted by statutory
regulation or exceeds the permitted use, you will need
to
obtain
permission
directly
from
the
copyright
holder. To view a copy of this licence, visit http://creat
ivecommons.org/licenses/by/4.0/. ACKNOWLEDGMENTS Finally, in a few previous papers, it was shown that
annealing of voidless a-Si prepared by ion implantation
leads to higher hardness due to above-mentioned
reduction of the density of defects. In this research, we
have observed the opposite effect. Hardness of the
annealed samples is always slightly lower, independently
from the experimental conditions. Furthermore, the
hardness of the as-deposited sample measured for lower
max loads is significantly higher than the hardness
reported for a-Si produced by ion implantation.[5,7,37] It
is apparent that the behavior of PECVD a-Si is
significantly different than the a-Si produced by ion
implantation. It is, therefore, speculated that the higher
hardness is caused by higher internal stress introduced
to the sample during the manufacturing process. The
annealing reduces the internal stress and, therefore,
reduces hardness. High internal stress is responsible for
pile-up creation during nanoindentation. In Figure 8, it
is shown that the pile-ups form in the as-deposited
sample and are missing in the annealed sample, which is
in good agreement with the proposed explanation of the
differences in hardness. The present research was partially supported by the
Polish National Science Centre, Grant 2015/19/D/ST8/
03200. V.
CONCLUSION The measurements of dynamic mechanical properties
of a-Si produced by PECVD show significant differences
in comparison to the results obtained for defect-free
amorphous silicon produced by ion implantation. The
as-deposited PECVD a-Si exhibits significant plastic flow
under the nanoindentation tip, whereas the annealed
PECVD a-Si sample exhibits a so-called elbow behavior
on unloading for lower loading rates, which could be
attributed to the transformation of (b-Sn)-Si to the PI
(pressure-induced) a-Si end phase. In contrast, the results
reported in the literature for the materials made by
self-ion implantation reported that elbows are formed for
higher and pop-outs for lower unloading rates. We
tentatively attribute these discrepancies to the fact that
a-Si prepared by CVD methods usually contains voids,
whereas ion implantation has been recognized as a
successful technique to produce pure and voidless amor-
phous layers. Another difference is the fact that in this
paper we have measured lower hardness for the annealed
sample. This effect is opposite to the behavior of the a-Si
produced by self-ion implantation described in the
literature, in which hardness usually increases after
annealing. In this case, the possible explanation is the
reduction of internal stress after annealing. METALLURGICAL AND MATERIALS TRANSACTIONS A Publisher’s Note
Springer Nature remains neutral with regard to
jurisdictional claims in published maps and institutional affiliations. REFERENCES Ghosh, M. Banda, J.E. John, and D.A. Terrones: Scripta
Mater., 2018, vol. 154, pp. 236–40. pp
18. Q. Wei: J. Mater. Sci., 2007, vol. 42, pp. 1709–27. 19. J. Fornell et al.: Acta Mater., 2010, vol. 58, pp. 6256–66. 20. C. Liu, V. Roddatis, P. Kenesei, and R. Maass: Acta Mater., 2017,
vol. 140, pp. 206–16. 1632—VOLUME 51A, APRIL 2020 METALLURGICAL AND MATERIALS TRANSACTIONS A 21. P.Y. Zhao, J. Li, J. Hwang, and Y.Z. Wang: Acta Mater., 2017,
vol. 134, pp. 104–15. 30. J.G. Swadener, E.P. George, and G.M. Pharr: J. Mech. Phys. Solids, 2002, vol. 50, pp. 681–94. ,
,
, pp
31. S.J. Bull, T.F. Page, and E.H. Yoffe: Philos. Mag. Lett., 1989,
vol. 59, pp. 281–88. pp
22. J.L. Wang, L.X. Wu, X. Chen, W.J. Zhuo, and G.F. Wang:
Sensors Actuators A, 2018, vol. 276, pp. 11–16. pp
32. D.E. Stegall and A.A. Elmustafa: Metall. Mater. Trans. A, 2018,
vol. 49A, pp. 4649–58. 23. D.A. Antartis, R.N. Mott, and I. Chasiotis: Mater. Des., 2018,
vol. 144, pp. 182–91. pp
33. J.E. Bradby, J.S. Williams, J. Wong-Leung, M.V. Swain, and
P. Munroe: Appl. Phys. Lett., 2000, vol. 77, pp. 3749–51. 24. H.K. Malik, S. Juneja, and S. Kumar: J. Theor. Appl. Phys., 2019,
vol. 13, pp. 107–13. 34. S. Saitoh, T. Sugii, H. Ishiwara, and S. Furukawa: Jpn. J. Appl. Phys., 1981, vol. 20, pp. L130–L132. 25. A. Soman, U. Nsofor, U. Das, T.Y. Gu, and S. Hegedus: ACS
A
l M
I
f
2019
l 11
16181 16190 25. A. Soman, U. Nsofor, U. Das, T.Y. Gu, and S. Hegedus: ACS
Appl. Mater. Interfaces, 2019, vol. 11, pp. 16181–16190. . Saitoh, T. Sugii, H. Ishiwara, and S. Furukawa: Jpn. J Appl. Mater. Interfaces, 2019, vol. 11, pp. 16181–16190 y
35. S. Roorda et al.: Phys. Rev. B, 1991, vol. 44, pp. 3702–25. 26. M. Chmielewski et al.: Int. J. Refract Metal Hard Mater., 2016,
vol. 59, pp. 78–86. 36. V. Maier, K. Durst, J. Mueller, B. Backes, H.W. H 36. V. Maier, K. Durst, J. Mueller, B. Backes, H.W. Hoppel, an
M Goken: J Mater Res
2011 vol 26 pp 1421 30 36. V. Maier, K. Durst, J. Mueller, B. Backes, H.W. Hoppel, and
M. Goken: J. Mater. Res., 2011, vol. 26, pp. 1421–30. ,
,
,
,
pp ,
M. Goken: J. Mater. Res., 2011, vol. 26, pp. 1421–30. 27. J. Nohava, N.X. Randall, and N. VOLUME 51A, APRIL 2020—1633 METALLURGICAL AND MATERIALS TRANSACTIONS A REFERENCES Conte: J. Mater. Res., 2009,
vol. 24, pp. 873–82. M. Goken: J. Mater. Res., 2011, vol. 26, pp. 1421–30. 37. S.K. Bhuyan et al.: MRS Commun., 2012, vol. 2, pp. 9–12. 28. Y. Liu, J. Hay, H. Wang, and X. Zhang: Scripta Mater., 2014,
vol. 77, pp. 5–8. Publisher’s Note
Springer Nature remains neutral with regard to
jurisdictional claims in published maps and institutional affiliations. 29. D. Pan, A. Inoue, T. Sakurai, and M.W. Chen: Proc. Natl. Acad. Sci. USA, 2008, vol. 105, pp. 14769–772. METALLURGICAL AND MATERIALS TRANSACTIONS A VOLUME 51A, APRIL 2020—1633
|
https://openalex.org/W2617617875
|
https://discovery.ucl.ac.uk/id/eprint/1557151/1/Avery_2017_Physiol._Meas._38_1116.pdf
|
English
| null |
Reproducible 3D printed head tanks for electrical impedance tomography with realistic shape and conductivity distribution
|
Physiological measurement
| 2,017
|
cc-by
| 8,485
|
Reproducible 3D printed head tanks for electrical impedance tomography with realistic shape
and conductivity distribution Reproducible 3D printed head tanks for electrical impedance tomography with realistic shape
and conductivity distribution IP Address: 128.41.35.111
This content was downloaded on 28/06/2017 at 11:49 Please note that terms and conditions apply. You may also be interested in: Correcting electrode modelling errors in EIT on realistic 3D head models
Markus Jehl, James Avery, Emma Malone et al. A comparison of two EIT systems suitable for imaging impedance changes in epilepsy
L Fabrizi, A McEwan, T Oh et al. A comparison of two EIT systems suitable for imaging impedance changes in epilepsy
L Fabrizi, A McEwan, T Oh et al. Comparison of frequency difference reconstruction algorithms for the detection of acute stroke
using EIT in a realistic head-shaped tank
B Packham, H Koo, A Romsauerova et al. B Packham, H Koo, A Romsauerova et al. Modelling for imaging neuronal depolarization by electrical and magnetic detection impedance
tomography
O Gilad, L Horesh and D S Holder
The effect of layers in imaging brain function using electrical impedance tomograghy
A D Liston, R H Bayford and D S Holder
A novel multi-frequency electrical impedance tomography spectral imaging algorithm for early stroke
detection
Lin Yang, Canhua Xu, Meng Dai et al. EIT of human brain function
A T Tidswell, A Gibson, R H Bayford et al. Correction of electrode modelling errors in multi-frequency EIT imaging
Markus Jehl and David Holder Institute of Physics and Engineering in Medicine Institute of Physics and Engineering in Medicine Physiological Measurement
https://doi.org/10.1088/1361-6579/aa6586 Physiological Measurement Physiol. Meas. 38 (2017) 1116–1131 https://doi.org/10.1088/1361-6579/aa6586 James Avery1, Kirill Aristovich, Barney Low
and David Holder University College London, London WC1E 6BT, United Kingdom University College London, London WC1E 6BT, United Kingdom E mail: j.avery@ucl.ac.uk
Received 11 January 2017, revised 16 February 2017
Accepted for publication 9 March 2017
Published 22 May 2017 Received 11 January 2017, revised 16 February 2017
Accepted for publication 9 March 2017
Published 22 May 2017 1 Author to whom any correspondence should be addressed. 1116
Original content from this work may be used under the terms of the Creative
Commons Attribution 3.0 licence. Any further distribution of this work must maintain
attribution to the author(s) and the title of the work, journal citation and DOI.
1 Author to whom any correspondence should be addressed.
1361-6579/17/061116+16$33.00 © 2017 Institute of Physics and Engineering in Medicine Printed in the UK 1. Introduction Electrical impedance tomography (EIT) is a medical imaging technique which produces
images of the internal electrical impedance of a subject from multiple measurements using
surface electrodes. EIT of the brain has many potential applications including stroke type
differentiation (Romsauerova et al 2006), localising epilepsy (Fabrizi et al 2006), and moni-
toring brain injuries such as intracranial and intraventricular haemorrhage, cerebral oedema,
and ischaemia (Tang and Sadleir 2011, Dai et al 2013, Manwaring et al 2013, Fu et al 2014). Studies in phantoms represent a bridge between idealised computer simulations and
clinical measurements in which instrumentation noise is present but errors in modelling the
electric fields are minimised. Constructing a representative head phantom poses a particular
challenge, given the irregular geometry of the head and the complexity of the internal tissues. The conductivity of the skull is inhomogeneous due to variations in the layered structure
throughout (Law 1993, Akhtari et al 2002). These layers result in a low, spatially variable
conductivity, which is highly anisotropic, with tangential:normal conductivity ratios of 10:1
when modelled as a homogenous layer (Wolters et al 2006) or 1.4:1 when considered as three
separate layers (Sadleir and Argibay 2007). Previously, these complexities have necessitated considerable simplifications when creat-
ing phantoms for EIT. Cylindrical, hemispherical and spherical tanks are common as they are
quick to achieve and do not require finite element models (FEMs) as geometrically detailed as
those of the cranial cavity (Liston et al 2004, Tang et al 2010, Ahn et al 2011, Sperandio et al
2012). To improve upon these models, a more realistic tank incorporating a real skull was cre-
ated by Tidswell et al (2003). Subsequently, the prevalence of rapid prototyping technology
has enabled phantoms to replicate the morphology of the head with a high precision, either by
3D printing the tanks directly (Tang and Sadleir 2011), or more commonly, printing moulds
used for casting (Collier et al 2012, Li et al 2014, Mobashsher and Abbosh 2014). Creating a volume conductor model of the skull mimicking the spatial variation of the elec-
trical properties has yet to be achieved in the literature. Layers of carefully chosen concentra-
tions of saline, gel or plaster of paris have been used to create a single shell of accurate, but
spatially invariant conductivity (Liston et al 2004, Collier et al 2012, Sperandio et al 2012). (Some figures may appear in colour only in the online journal) (Some figures may appear in colour only in the online journal) Abstract Objective. Electrical impedance tomography (EIT) has many promising
applications in brain injury monitoring. To evaluate both instrumentation
and reconstruction algorithms, experiments are first performed in head tanks. Existing methods, whilst accurate, produce a discontinuous conductivity, and are
often made by hand, making it hard for other researchers to replicate. Approach. We have developed a method for constructing head tanks directly in a 3D printer. Conductivity was controlled through perforations in the skull surface, which
allow for saline to pass through. Varying the diameter of the holes allowed for
the conductivity to be controlled with 3% error for the target conductivity range. Taking CT and MRI segmentations as a basis, this method was employed to
create an adult tank with a continuous conductivity distribution, and a neonatal
tank with fontanelles. Main results. Using 3D scanning a geometric accuracy
of 0.21 mm was recorded, equal to that of the precision of the 3D printer used. Differences of 6.1% ± 6.4% (n = 11 in 4 tanks) compared to simulations were
recorded in c. 800 boundary voltages. This may be attributed to the morphology
of the skulls increasing tortuosity effects and hole misalignment. Despite
significant differences in errors between three repetitions of the neonatal tank,
images of a realistic perturbation could still be reconstructed with different tanks
used for the baseline and perturbation datasets. Significance. These phantoms
can be reproduced by any researcher with access to a ‘hobbyist’ 3D printer in a Original content from this work may be used under the terms of the Creative
Commons Attribution 3.0 licence. Any further distribution of this work must maintain
attribution to the author(s) and the title of the work, journal citation and DOI. 1116 1361-6579/17/061116+16$33.00 © 2017 Institute of Physics and Engineering in Medicine Printed in the UK Physiol. Meas. 38 (2017) 1116 J Avery et al matter of days. All design files have been released using an open source license
to encourage reproduction and modification. 1. Introduction To address this invariance, Li et al (2014) created a geometrically accurate phantom with eight
separate bone structures, each with a different, but fixed conductivity. Tidswell et al (2003)
incorporated a real skull in their phantom, which may have replicated the spatial variability of
the skull structure, the dead tissue was not representative of the conductivity itself. There has
been less focus in the literature on phantoms in neonatal head EIT; the most detailed phantom
to date, was geometrically representative but did not include the skull (Tang and Sadleir 2011). As well as lacking the true smooth spatial distribution of the skull, these phantoms also
have practical disadvantages. The agar or gypsum used in the skull shells degrades over time
in contact with air or water, making them unsuitable for sporadic use over the long term. At least some elements are also constructed by hand, which makes them difficult for other 1117 Physiol. Meas. 38 (2017) 1116 J Avery et al researchers to replicate. Further, even small changes in boundary shape can result in signifi-
cant modelling errors (Nissinen et al 2010), so results may not be repeatable even in cases
where the method is sufficiently described. researchers to replicate. Further, even small changes in boundary shape can result in signifi-
cant modelling errors (Nissinen et al 2010), so results may not be repeatable even in cases
where the method is sufficiently described. The purpose of this study was to develop a method of creating head tanks for EIT which
are readily reproducible, whilst allowing for smoothly varying conductivity across the skull. It employed rapid prototyping and extruded plastic to create both the tank and skull directly,
for both and adult and neonatal head. Perforations were made in the skulls in order to simulate
their realistic conductivities. This approach reduces the variation during construction and thus
offers practical advantages to any research group wishing to utilise reproducible faithful tanks
for EIT evaluation. As the tanks were constructed entirely with insulating plastic, a method of controlling the
conductivity through surface perforations was first validated in small tubular tanks. Using CT
and MRI segmentations as reference, both a neonatal and adult tank were constructed with
the method developed herein. To assess the accuracy of the construction, the adult tank was
compared to the ideal CAD model using high resolution optical 3D scanning. 1. Introduction Simulations of
the voltages expected when using the same CAD model were then compared to those obtained
experimentally. Finally, the repeatability of the method was evaluated in three neonatal phan-
toms, comparing the experimental voltages and reconstructed images in each case. 2.1. Modelling conductivity ank design
geometry for the phantoms was obtained from segmentations of a combination of CT
MRI for the adult head, and a CT for the neonatal head. The resolution of these scans
<1 mm in all 3 dimensions, providing sufficient resolution to obtain segmentations of
calp, skull, dura, grey and white matter and CSF in the MRI (Jehl et al 2014), and the
and skull in the CT. Smoothing of segmentation artefacts was first performed in MeshLab
noni et al 2008) before trimming of extraneous features and further artefact removal in
CAD (www.freecadweb.org/). This smoothed geometry could then be converted to the
parametric format traditionally used in CAD, figure 2. Figure 1. (a) Left: tubular phantom with two sections of varying length and resistive
surface inserts, right: 2 mm holes with 6 mm spacing and 1 mm with 5 mm spacing
(b) relationship between effective resistivity ρeff and combined cross sectional area of
surface perforations, Asaline for three different hole spacing. Figure 1. (a) Left: tubular phantom with two sections of varying length and resistive
surface inserts, right: 2 mm holes with 6 mm spacing and 1 mm with 5 mm spacing
(b) relationship between effective resistivity ρeff and combined cross sectional area of
surface perforations, Asaline for three different hole spacing. Figure 1. (a) Left: tubular phantom with two sections of varying length and resistive
surface inserts, right: 2 mm holes with 6 mm spacing and 1 mm with 5 mm spacing
(b) relationship between effective resistivity ρeff and combined cross sectional area of
surface perforations, Asaline for three different hole spacing. Figure 1. (a) Left: tubular phantom with two sections of varying length and resistive
surface inserts, right: 2 mm holes with 6 mm spacing and 1 mm with 5 mm spacing
(b) relationship between effective resistivity ρeff and combined cross sectional area of
surface perforations, Asaline for three different hole spacing. the 3D printer, resulting in errors >20%. Therefore, in subsequent phantom designs, a hole
spacing of 5 mm was used and diameters adjusted to stay within this range. 2.1. Modelling conductivity To maximise the reproducibility of the phantom, the entirety of tank and skull, with the excep-
tion of the electrodes themselves, were constructed directly by the 3D printer. This precluded
methods commonly used to create multi shell phantoms, namely casting plaster of paris
(Liston et al 2004, Li et al 2014) or agar and volume conductive film (Sperandio et al 2012). The technique employed in these tanks was of a solid printable skull with variable diameter
surface perforations, which allow saline, and thus current, to pass through. The effective resis-
tivity can be approximated through a simple estimation of the area ratio of saline to insulating
plastic, equation (1). By smoothly altering the density or diameter of the holes the effective
conductivity can vary smoothly across the phantom, as opposed to a fixed value per segment ρ
ρ
≈A
A
. eff
total
saline
saline
ρ
ρ
≈A
A
. eff
total
saline
saline
(1) (1) To validate this model experimentally, a square tube phantom with two electrodes was
created, figure 1(a), between which grids of differing hole density and diameter could be
inserted. The tank was filled with 0.2% saline, and the three components formed a water tight
seal. Inserts with hole spacing of 5, 6 and 8 mm were created with hole diameters 1–3 mm in
0.5 mm increments, covering the range which could reliably be 3D printed. The impedance
was measured with a Hewlett-Packard HP4284A impedance analyser at 1 kHz. Corrections
for the electrode contact impedance were calculated using measurements with three different
tank lengths and with no insert present. The theoretical effective resistance was calculated for the cross sectional surface area equiv-
alent to that occupied by the combined aperture cross sectional area Asaline, figure 1(b). In the
tank, the target resistivity could only be matched with <3% error using a hole spacing of 5 mm
and diameters between >1 mm and <2.5 mm, equivalent to a ratio of Asaline between 0.05 and
0.13. This was likely because diameters outside this range were not accurately reproduced by 1118 Physiol. Meas. 38 (2017) 1116 J Avery et al D printer, resulting in errors >20%. Therefore, in subsequent phantom designs, a hole
ng of 5 mm was used and diameters adjusted to stay within this range. 2.2. Tank design This plane also 1120 J Avery et al Physiol. Meas. 38 (2017) 1116 Figure 4. Design of skull models. (a) skull model was split along the same plane as
figure 3, (b) the model was trimmed and walls vertically extended 1 cm to match those
of the tank, (c) a support base added to position the skull in the tank, (d) cylinders were
created normal to the surface, centred at evenly distributed points 5 mm apart, (e) the
cylinders were subtracted from the tank volume giving the final model, (f ) skull then
suspended in the tank. Figure 4. Design of skull models. (a) skull model was split along the same plane as
figure 3, (b) the model was trimmed and walls vertically extended 1 cm to match those
of the tank, (c) a support base added to position the skull in the tank, (d) cylinders were
created normal to the surface, centred at evenly distributed points 5 mm apart, (e) the
cylinders were subtracted from the tank volume giving the final model, (f ) skull then
suspended in the tank. defines the height of the saline within the tank, and thus the extent of the volume used for
mesh generation. In both tanks, 32 electrodes and a ground electrode were positioned based on the distribu-
tions proposed by Tidswell et al (2001b), which includes 21 positions from the EEG 10–20
system (Jasper 1958), six from the extended 20–20 and four additional positions. These addi-
tional positions were modified in these tanks to the nearest locations in either the 10–10 or
10–5 extensions (Oostenveld and Praamstra 2001). Recesses were then created at each of
these points to ensure that when placed, the electrodes were flush against the tank walls. To
create these recesses, models of the electrodes expanded by 0.2 mm were aligned normal to
the surface at each of these locations and then subtracted from the tank volume. To complete
the models, support legs and a base were then added. 2.2. Tank design The geometry for the phantoms was obtained from segmentations of a combination of CT
and MRI for the adult head, and a CT for the neonatal head. The resolution of these scans
were <1 mm in all 3 dimensions, providing sufficient resolution to obtain segmentations of
the scalp, skull, dura, grey and white matter and CSF in the MRI (Jehl et al 2014), and the
scalp and skull in the CT. Smoothing of segmentation artefacts was first performed in MeshLab
(Cignoni et al 2008) before trimming of extraneous features and further artefact removal in
FreeCAD (www.freecadweb.org/). This smoothed geometry could then be converted to the
solid parametric format traditionally used in CAD, figure 2. 1119 Physiol. Meas. 38 (2017) 1116 J Avery et al Figure 2. Workflow to create solid models from a CT segmentation for the design of
the neonatal tank. Figure 2. Workflow to create solid models from a CT segmentation for the design of
the neonatal tank. Figure 3. Design of tank models. (a) scalp duplicated and scaled by 10% to create the
outer tank surface, (b) inner surface removed and shell split below inion-nasion line, (c)
the model was trimmed and the walls extended 10 mm, (d) electrodes with clearance
were arranged in modified EEG 10–20 positions, (e) the clearance electrodes removed
from the tank volume, (f ) supports added and model was finalised. Figure 3. Design of tank models. (a) scalp duplicated and scaled by 10% to create the
outer tank surface, (b) inner surface removed and shell split below inion-nasion line, (c)
the model was trimmed and the walls extended 10 mm, (d) electrodes with clearance
were arranged in modified EEG 10–20 positions, (e) the clearance electrodes removed
from the tank volume, (f ) supports added and model was finalised. To create the tanks, figure 3, the models were split along a plane parallel to that described
by the inion-nasion line, the offset chosen to include the entirety of the brain cavity yet remove
as much of the neck and facial structures as possible. This is a common simplification in brain
EIT, as structures like the ear canal and eyeballs are challenging to model even in simula-
tion, and their effect on EIT measurements still require investigation (Vauhkonen et al 1999,
Tidswell et al 2003, Liston et al 2004, Sperandio et al 2012, Li et al 2014). 2.3. Skull design with realistic conductivity distribution xy
0
max
max
(2) (2) As the spatial distribution is not as well understood in the literature, the conductivity for the
neonatal skull phantom was fixed at a value of 0.03 S m−1 as suggested by Pant et al (2011). As the spatial distribution is not as well understood in the literature, the conductivity for the
neonatal skull phantom was fixed at a value of 0.03 S m−1 as suggested by Pant et al (2011). 2.3.2. Spatially varying conductivity phantoms. The holes were positioned on the surface of
the skulls in an evenly spaced 5 mm distribution. The diameter of each hole was then calcu-
lated based upon a combination of equations (1) and (2), assuming 0.2% saline concentration. Cylinders of the required diameter, aligned to the normal of the surface were generated at each
point. The cylinders were then removed from the skull model using the same Boolean subtrac-
tion to create the electrode recesses. 2.4. Construction The phantoms were created from polylactic acid (PLA) using a MakerBot Replicator 2
Desktop 3D printer (MakerBot, Brooklyn, NY), a common ‘hobbyist’ fused deposition mod-
elling (FDM) 3D printer. The rated accuracy of the printer is 0.1 mm in the X and Y dimensions
and user controlled within a range of 0.1–0.4 mm in the Z direction. All models were printed
entirely solid to ensure they are water tight, the Z layer height was set to 0.2 mm to reduce the
build time to less than 48 h. Support material artefacts and burrs were removed through sand-
ing and blasting with compressed air. The electrodes were 316 stainless steel, chosen for its
high corrosion resistance, and sealed in place with silicon adhesive. 2.3. Skull design with realistic conductivity distribution To create the skull phantoms, the model was cut along the same plane parallel to the inion—
nasion line, figure 4, leaving only the bone encasing the brain and the resultant surfaces were
extended to match the height of the tank walls. Support structures were added so the skull
could be suspended in place inside the tank, aligned through corresponding concentric holes
in both pieces. In the neonatal tank, it was necessary to split the skull into two parts along
the lambdoid suture between the parietal and occipital bones. Each half was then positioned
separately inside the tank, forming a water tight seal once correctly aligned. 1121 Physiol. Meas. 38 (2017) 1116 J Avery et al 2.3.1. Conductivity distributions. The conductivity for the adult skull was based upon the
results from Tang et al (2008) at 1kHz, and spatial distribution obtained from Li et al (2014). Given the resolution of the 3D printer and the surface perforation method, the sutures between
bone types were neglected. 2.3.1. Conductivity distributions. The conductivity for the adult skull was based upon the
results from Tang et al (2008) at 1kHz, and spatial distribution obtained from Li et al (2014). Given the resolution of the 3D printer and the surface perforation method, the sutures between
bone types were neglected. The sphenoid and temporal bones at the side of the head are of the same ‘quasi-compact’
bone type, σ = 0.005 S m−1. The frontal and occipital bones at the front and back of the head are
both ‘standard tri-layer’, with σ = 0.0126 S m−1. The conductivity at the top of the head—the
parietal bone, is of the ‘quasi-tri-layer’ type, σ = 0.0069 S m−1 (Tang et al 2008). Thus, start-
ing from the top of the head, the conductivity increases by 87% towards the front and back
of the head, and decreases by 32% towards the sides. Taking the top of the skull as the origin
with σ = 0.0069
0
S m−1, the conductivity was modelled as a linear combination of an increase
towards the front and back (x direction) of the skull and a decrease in conductivity towards the
sides (y direction), equation (2) σ
σ
=
+
−
⎛
⎝
⎜
⎞
⎠
⎟
x
x
y
y
1
0.87
0.32
. 2.6. Comparison to EIT simulations The results were thus
expressed as a mean plus standard deviation calculated on 819 and 775 voltage measurements
for the adult and neonatal tanks respectively. Results below, unless stated otherwise, were
expressed as mean and standard deviation. 2.5. Geometry testing—deviation analysis Determining the accuracy of the resultant geometry is made difficult by the irregular morph
ology of the head. Previous techniques compare distances between key points on the tank mea-
sured by hand. The error is then expressed as a percentage of the distance in either the initial CT
(Silva et al 2008) or a cast used in a previous manufacturing step (Li et al 2014). In these cases,
mean errors ranging from 0.6 mm or 0.75% to 1.07 mm or 2.67% have been found. These methods have two drawbacks: measurements are not compared to the same model
used in simulations and they do not adequately capture the spatial distribution of the error
across the head. The method proposed in this study is ‘deviation analysis’, a common
reverse-engineering technique (Várady et al 1997), wherein surfaces of the CAD model are
compared point-by-point to those obtained through a high-precision scan. The adult tank was 1122 Physiol. Meas. 38 (2017) 1116 J Avery et al scanned using a Creaform Go!Scan20 3D optical scanner (Creaform Inc. Québec, Canada),
to provide a mesh of the surface, with an accuracy of ⩽0.2 mm and >2 million elements. Scanning targets were placed upon the model to enable registration and alignment with the
CAD model. To simplify the scanning procedure, only the inner surface of the tank was con-
sidered, and the electrodes were removed from the analysis due to artefacts arising from the
reflective surfaces. 2.6. Comparison to EIT simulations 2.6.1. Data collection. Data were collected in both tanks using the UCL ‘ScouseTom’ EIT
system (Avery et al 2017). For the adult tank experiments, the injection electrode protocol
proposed by Malone et al (2014) was used, with 31 injection pairs chosen to maximise the
number of independent measurements. A protocol with 32 injections was used for the neonatal
tank measurements, which maximised the number of injections from opposites sides of the
head. In all experiments, the current injected was 200 μA amplitude at 1.4 kHz, with 75 ms per
injection pair and results averaged over 20 complete frames. The tanks were filled with 0.2%
saline with a conductivity of 0.4 S m−1, to represent the scalp. Experiments were performed
first without the skull present, and then with the skull in place. 2.6.2. Data analysis. The boundary voltages collected in each experiment were compared
to simulations in ≈4 million element FEMs generated from the CAD models using the par-
allel EIT solver (PEITS) (Jehl et al 2014). The error was calculated as the absolute differ-
ence between the experimental and simulated voltages for each measurement. The data were
summarised both as a mean and standard deviation in absolute voltage and as a percentage
of the simulated voltage. Measurement channels with a simulated voltage less than 0.1 mV
were removed from this analysis, based on the threshold described by Packham et al (2012),
to avoid overestimation of the error from noise on small channels. The results were thus
expressed as a mean plus standard deviation calculated on 819 and 775 voltage measurements
for the adult and neonatal tanks respectively. Results below, unless stated otherwise, were
expressed as mean and standard deviation. 2.6.2. Data analysis. The boundary voltages collected in each experiment were compared
to simulations in ≈4 million element FEMs generated from the CAD models using the par-
allel EIT solver (PEITS) (Jehl et al 2014). The error was calculated as the absolute differ-
ence between the experimental and simulated voltages for each measurement. The data were
summarised both as a mean and standard deviation in absolute voltage and as a percentage
of the simulated voltage. Measurement channels with a simulated voltage less than 0.1 mV
were removed from this analysis, based on the threshold described by Packham et al (2012),
to avoid overestimation of the error from noise on small channels. 2.7. Reproducibility of method Alongside the first neonatal tank, labelled ‘tank 1’, two further neonatal tanks and skulls were
constructed, labelled tank 2 and 3, and the data collection procedure repeated in each one. To
represent the expected intermittent usage of the tanks, each dataset was collected starting from
an empty tank with the EIT system disconnected each time. These voltages were analysed
using the same methods as the previous experiments. EIT images were reconstructed of the conductivity changes resulting from the insertion
of a 20 mm diameter sponge of approximately 10% contrast located in the anterior position. Similar perturbations have been used previously to represent the change expected resulting
from cell swelling in epilepsy (Tidswell et al 2001b, Fabrizi et al 2009). The perturbation
dataset was collected in tank 1 only, with the baseline from all three tanks used as a reference
for three separate images. Reconstructions were performed using the method described by
Aristovich et al (2014), using Zeroth order Tikhonov regularisation with noise based correc-
tion and hexahedral meshes of c. 100 000 elements. 1123 J Avery et al Physiol. Meas. 38 (2017) 1116 3 R
lt
Figure 5. Deviation analysis comparing 3D scan of adult tank to target CAD model. Figure 6. Comparison of voltages (a) in adult head tank R2 = 0.997 and (b) in adult
head tank with skull R2 = 0.997. Figure 5. Deviation analysis comparing 3D scan of adult tank to target CAD model. Figure 5. Deviation analysis comparing 3D scan of adult tank to target CAD model. Figure 5. Deviation analysis comparing 3D scan of adult tank to target CAD model. Figure 5. Deviation analysis comparing 3D scan of adult tank to target CAD model. Figure 6. Comparison of voltages (a) in adult head tank R2 = 0.997 and (b) in adult
head tank with skull R2 = 0.997. Figure 6. Comparison of voltages (a) in adult head tank R2 = 0.997 and (b) in adult
head tank with skull R2 = 0.997. 3.1. Geometry analysis The deviation in the adult tank was 0.21 mm ± 0.101 mm. The error was distributed evenly
across the tank, figure 5, with the exception of occasional printing artefacts at the rim of tank
and at the bottom of tank, equivalent to the top of the head. 3.2. Comparison to simulation 3.2.1. Adult tank. Qualitatively, the simulated and experimental voltages in the adult tank
were in good agreement, figure 6(a), with a correlation coefficient of R2 = 0.997. The error 1124 Physiol. Meas. 38 (2017) 1116 J Avery et al Figure 7. Comparison of voltages (a) in neonatal head tank R2 = 0.999 and (b) in
neonate head tank with skull R2 = 0.994. Figure 7. Comparison of voltages (a) in neonatal head tank R2 = 0.999 and (b) in
neonate head tank with skull R2 = 0.994. Figure 8. Absolute percentage errors in voltages compared to simulation for both adult
(Ad) and neonatal (Nn) tanks with and without skull. Figure 8. Absolute percentage errors in voltages compared to simulation for both adult
(Ad) and neonatal (Nn) tanks with and without skull. across all measurements was
±
0.147
0.099 mV. Injecting between electrode positions FP1
and P04 produced the greatest error of 0.251 mV. 3.2.2. Adult tank with skull. The inclusion of the skull had no effect upon the correlation
between the simulated and experimental voltages R2 = 0.997, figure 6(b). However, the over-
all error increased to
±
0.215
0.229 mV, and some channels had significantly larger errors than
the average. These were all found either when injecting between FP2 and 0Z, or measuring at
T4, with a mean error of 0.513 and 0.439 mV respectively. 3.2.3. Neonatal tank. In the neonatal tank, figure 7(a), the simulated and experimental volt
ages, were also in good correlation R2 = 0.999, with an error of
±
0.079
0.062 mV. The largest
errors were found when injecting current between the electrodes at TP8 and AF3, with an error
of 0.187 mV. 1125 Physiol. Meas. 38 (2017) 1116 J Avery et al J Avery et al Figure 9. Comparison of voltages recorded in (a) three separate neonatal tanks, and
(b) three separate neonatal tanks with skull, averaged over three repeats each. Figure 9. Comparison of voltages recorded in (a) three separate neonatal tanks, and
(b) three separate neonatal tanks with skull, averaged over three repeats each. Figure 10. Perturbation of 10% contrast (a) ideal location, recorded in phantom 1
reconstructed with (b) phantom 1 as baseline (c) phantom 2 as baseline (d) phantom
3 as baseline. Figure 10. 3.3. Reproducibility The recorded voltages for all three neonatal tanks were consistent, figure 9(a), with R2 = 0.999,
R2 = 0.998 and R2 = 0.998. The mean error across all measurements and repeats was sig-
nificantly less for tank 3 P < 0.05, with 5.06% compared to 6.2 and 6.3% for tanks 1 and 2
respectively. As is clear from figure 9(b), the error with three different skulls was both less
correlated with simulation R2 = 0.994, R2 = 0.995 and R2 = 0.993 respectively, and the mea-
surements with largest errors were different for each phantom. The skull phantom in tank 2
had a mean error of 5.3% which was significantly less P < 0.05 than the 7.63% error in tank
1 and 7.1% error in tank 3. Despite these differences in baseline voltages between the phantoms, the perturbation was
successfully reconstructed in every case, including both mismatched datasets, figure 10. The
distance between ideal and reconstructed centre of mass for the matched datasets, figure 10(b),
was 2.9 mm or 2.6% of the tank width. This increased to 3.2 and 3.4 mm or 2.9 and 3.0% when
tank 2 and tank 3 were used as baseline recordings respectively. There were also increased
artefactual positive changes and distortion in the shape of the reconstructed perturbations with
mismatched phantoms. 3.2. Comparison to simulation Perturbation of 10% contrast (a) ideal location, recorded in phantom 1
reconstructed with (b) phantom 1 as baseline (c) phantom 2 as baseline (d) phantom
3 as baseline. 1126 Physiol. Meas. 38 (2017) 1116 J Avery et al 3.2.4. Neonatal tank with skull. The addition of the skull decreased the correlation with the
simulated voltages, figure 7(b), to R2 = 0.994. This was also reflected in the increased error of
±
0.185
0.198 mV. The largest errors were found when injecting between AF8 and FCz, with
an error of 0.569 mV. 3.2.5. Summary. The mean error expressed as a percentage, figure 8, in the adult tank was
4.87% without the skull and 4.50% including the skull. The error in the neonatal tank was
2.4% and 5.14% for measurements without and with the skull respectively. In both cases the
inclusion of the skull significantly changed the mean error P < 0.05, decreasing by 0.37% for
the adult tank, and increasing error by 1.74% for the neonatal tank. 4. Discussion This was not the
case for the neonatal tank, in which the error doubled both as an absolute voltage from 0.079
mV to 0.185 mV, and as a percentage, from 2.4% to 5.14%. In both tanks the errors were greatest when injecting current from opposite sides of the
head, from the anterior to the posterior electrodes. In both tanks, this corresponds to the areas
with the greatest curvature. In which case, the normals of the skull and nearby scalp surfaces
may not be aligned, increasing the errors arising from tortuosity effects. This is especially
true of the parts of the orbital plate included in the neonatal tank, which in some cases is
perpendicular to the nearest tank surface. Given the anisotropy of the conductivity using sur-
face perforations, the approximation only holds when the holes are aligned with the injecting
electrodes. Defining the axis of the holes in the skull perpendicular to the nearest scalp surface
as opposed to the skull could reduce this error. Further, the electrodes with the greatest error
are also closest to the top of the tank, and are thus more vulnerable to changes in saline level
between experiments. Assessment of the tanks with respect to the literature is difficult as the error is often not
quantified to the same extent as in this study. The most detailed study using plaster of paris
showed there was no statistical difference with conductivity distributions in the skull, and the
error in transimpedance, while not quantified by the authors, appears to be 1–5% in six meas-
urements (Li et al 2014). This is approximately the interquartile range of the errors in the adult
skull measurements in figure 8, suggesting the errors are similar in the two tanks. The error
across all recordings in all tanks was 6.1% ± 6.4% (n = 11 in 4 tanks). The reconstructions in
figure 10 suggest the effect of these modelling errors is likely minimised when subtracting the
reference dataset in time difference imaging. Absolute or multi-frequency imaging, without
this subtraction, are much more sensitive to these modelling errors, but the acceptable range is
not yet quantified in the literature. However, a study in a similar adult head tank demonstrated
that these errors are large enough to mask any real conductivity changes (Jehl et al 2015). 4. Discussion Using 3D scanning allowed for direct comparison with the mesh used in EIT reconstructions,
and overcame the limitations of methods implemented previously in the literature. A consis-
tent error of 0.21 mm was found across the tank, which is close to the rated accuracy of both
the 3D printer and 3D scanner used. Printing artefacts were largely confined to areas which
would not affect the experimental results; however some were still present at the top of the
head. Thus the accuracy could be improved through the use of a more accurate rapid prototyp-
ing technology, and better post processing methods. Despite these, the mean error is similar to
that found using distances measured by hand (Silva et al 2008, Li et al 2014). The accuracy of the method is further demonstrated in the comparison of simulated and
experimental voltages, which are all in clear agreement (R2 > 0.994) for all injections and
measurements. Without the skull, the measurements with greatest error occur at electrodes
with the greatest surface curvature, where the electrodes may not be perfectly aligned with,
or flush against, the tank wall. These effects were less pronounced in the neonatal tank and as
such had reduced error, both as a voltage and as a percentage: 0.079 mV and 2.4% compared
to 0.147 mV and 4.87%. The larger percentage errors, >8%, all correspond to the channels
with the lowest boundary voltages <1 mV, and are thus more susceptible to measurement
noise and modelling errors. This explains why counter-intuitively, the mean error decreased 1127 Physiol. Meas. 38 (2017) 1116 J Avery et al in the adult tank with the inclusion of the skull, from 4.87% to 4.5%, whilst the voltage error
increased from 0.147 mV to 0.215 mV. The inclusion of the skull approximately doubled the
mean voltage, resulting in fewer low standing potential, high error channels. This was not the
case for the neonatal tank, in which the error doubled both as an absolute voltage from 0.079
mV to 0.185 mV, and as a percentage, from 2.4% to 5.14%. in the adult tank with the inclusion of the skull, from 4.87% to 4.5%, whilst the voltage error
increased from 0.147 mV to 0.215 mV. The inclusion of the skull approximately doubled the
mean voltage, resulting in fewer low standing potential, high error channels. 4. Discussion Skull conductivity has been shown to
correlate with ratio of diploe thickness (Tang et al 2008). Segmentation of the diploe layer
could thus be used to estimate a per element conductivity in the FEM, and the hole diameter
recalculated using these values. recalculation. In this case, multi-frequency studies over a wide frequency range could require
a multitude of different skulls, which increases the chances of errors arising from printing
artefacts or misalignment. Finally, the spatial variation of the adult skull conductivity is cur
rently too simplified to represent the true distribution. Skull conductivity has been shown to
correlate with ratio of diploe thickness (Tang et al 2008). Segmentation of the diploe layer
could thus be used to estimate a per element conductivity in the FEM, and the hole diameter
recalculated using these values. Acknowledgment The authors would like to acknowledge Mrs Tugba Demiray who collaborated on earlier
designs of the adult and tubular phantoms. These were then superseded by those constructed
for the experiments detailed in this manuscript. This work was supported by EPSRC grant EP/
M506448/1. 5. Conclusion Previous head tanks utilised rapid prototyping to create moulds for casting conductive mat
erials. Whilst this may enable accurate representation of the geometry of the head tissues, this
indirect use does not take full advantage of reproducibility of the technology. In this study, a
novel construction method was described wherein the tank and skull are constructed entirely
by 3D printing, and the effective skull conductivity is controlled through surface perforations. This technique was shown to have an error of approximately 3% across the target conductivity
range in test phantoms. Two phantoms were created: an adult tank with a smoothly varying
skull conductivity, and a neonatal tank with fontanelles. 3D scanning measured a geometric
error of 0.21 mm, equal to the rated accuracy of the 3D printer used. Curvature and tortuos-
ity effects increased the error to ≈5% in the full tank models. Despite significantly differ-
ent errors in three neonatal tanks, reconstructions of biologically representative perturbations
were successful with mismatched reference and perturbation datasets. These phantoms can be replicated by any researcher with access to a ‘hobbyist’ standard
3D printer, with minimal construction by hand. The models and software to create these tanks
have been released on an open-source license (see appendix), to encourage reproduction and
modification. 4. Discussion g
g
y
y
g
The significant differences in the errors between repetitions of the tank are likely to be caused
by printing artefacts. It was not always possible to fully remove the support material from the
electrode recesses in the tank, leading to slight misalignment of electrodes when they were
inserted. Similarly, printing the required number of small diameter holes in thin shells was on
the limits of what was capable with the printer. Consequently, the diameter of all c. 3000 holes
did not meet the rated accuracy in every instance. As with the geometric accuracy, both of these
issues would be improved through the use of a higher resolution rapid prototyping technology. p
g
g
p
p
yp g
gy
The results show that creating a head tank directly via rapid prototyping yields a phantom
that is both quick to build—with our setup, both tanks could be completed within a week—
and provides stable results over time and between copies. However, given the material is
purely insulating, replicating the conductivity of the skull is not without compromise. The
conductivity of the skull phantom is highly anisotropic, as it is close to a pure insulator in
the tangential direction. This is a less realistic representation of the true conductivity than
the isotropic layers used in existing methods. In both current injection protocols chosen, the
direction of injected current was normal to the skull surface, where errors arising from this
anisotropy are minimised. It is likely that if an ‘adjacent pair’ or similar protocol was chosen,
the error would increase. As the whole tank is filled with a single saline concentration, the
scalp and brain volumes have the same conductivity. Whilst this a common simplification
(Tidswell et al 2001a, Li et al 2014), it does not reflect the true tissue properties. Further, each
of the skulls were calculated based upon empirical values from the literature at 1 kHz (Tang
et al 2008, Pant et al 2011), and thus experiments at higher frequencies would benefit from 1128 Physiol. Meas. 38 (2017) 1116 J Avery et al recalculation. In this case, multi-frequency studies over a wide frequency range could require
a multitude of different skulls, which increases the chances of errors arising from printing
artefacts or misalignment. Finally, the spatial variation of the adult skull conductivity is cur
rently too simplified to represent the true distribution. Appendix. Hardware and software resources All CAD models and software used in creating both tanks are available at https://github.com/
EIT-team/Tanks, released under a GNU General Public License v3.0. Replication, contrib
utions and distribution are welcomed. Akhtari M et al 2002 Conductivities of three-layer live human skull Brain Topogr. 14 151–67
Aristovich K Y, dos Santos G S, Packham B C and Holder D S 2014 A method for reconstructing
tomographic images of evoked neural activity with electrical impedance tomography using
intracranial planar arrays Physiol. Meas. 35 1095–109 Ahn S, Oh T I, Jun S C, Seo J K and Woo E J 2011 Validation of weighted frequency-difference EIT
using a three-dimensional hemisphere model and phantom Physiol. Meas. 32 1663–80 References 10 370–5 Jehl M, Avery J, Malone E, Holder D and Betcke T 2015 Correcting electrode modelling errors in EIT
on realistic 3D head models Physiol. Meas. 36 2423–42 on realistic 3D head models Physiol. Meas. 36 2423–42 Jehl M, Dedner A, Betcke T, Aristovich K, Klofkorn R and Holder D 2014 A fast parallel solver for the
forward problem in electrical impedance tomography IEEE Trans. Bio-Med. Eng. 9294 1–13 Law S K 1993 Thickness and resistivity variations over the upper surface of the human skull Brain Topogr. 6 99–109 Li J B et al 2014 A new head phantom with realistic shape and spatially varying skull resistivity
distribution IEEE Trans. Bio-Med. Eng. 61 254–63 Liston A D, Bayford R H and Holder D S 2004 The effect of layers in imaging brain function using
electrical impedance tomograghy Physiol. Meas. 25 143–58 Malone E, Jehl M, Arridge S, Betcke T and Holder D 2014 Stroke type differentiation using spectrally
constrained multifrequency EIT: evaluation of feasibility in a realistic head model Physiol. Meas. 35 1051–66 Manwaring P K, Moodie K L, Hartov A, Manwaring K H and Halter R J 2013 Intracranial electrical
impedance tomography: a method of continuous monitoring in an animal model of head trauma
Anesthesia Analgesia 117 866–75 Mobashsher A T and Abbosh A M 2014 Three-dimensional human head phantom with realistic electrical
properties and anatomy IEEE Antennas Wireless Propag Lett 13 1401 4 Mobashsher A T and Abbosh A M 2014 Three-dimensional human head phantom with realis
properties and anatomy IEEE Antennas Wireless Propag. Lett. 13 1401–4 Nissinen A, Kolehmainen V and Kaipio J P 2010 Compensation of errors due to incorrect model
geometry in electrical impedance tomography J. Phys.: Conf. Ser. 224 012050i Oostenveld R and Praamstra P 2001 The five percent electrode system for high-resolution EEG and ERP
measurements Clin. Neurophysiol. 112 713–9 Oostenveld R and Praamstra P 2001 The five percent
measurements Clin. Neurophysiol. 112 713–9 Packham B, Koo H, Romsauerova A, Ahn S, McEwan A, Jun S and Holder D 2012 Comparison of
frequency difference reconstruction algorithms for the detection of acute stroke using EIT in a
realistic head-shaped tank Physiol. Meas. 33 767–86 Pant S, Te T, Tucker A and Sadleir R R J 2011 The conductivity of neonatal piglet skulls Physiol. Meas. References Ahn S, Oh T I, Jun S C, Seo J K and Woo E J 2011 Validation of weighted frequency-difference EIT
using a three-dimensional hemisphere model and phantom Physiol. Meas. 32 1663–80 Akhtari M et al 2002 Conductivities of three layer live human skull Brain Topogr. 14 151 67
Aristovich K Y, dos Santos G S, Packham B C and Holder D S 2014 A method for reconstructing
tomographic images of evoked neural activity with electrical impedance tomography using
intracranial planar arrays Physiol. Meas. 35 1095–109 1129 Physiol. Meas. 38 (2017) 1116 J Avery et al Avery J, Dowrick T, Faulkner M, Goren N and Holder D 2017 A versatile and reproducible multi-
frequency electrical impedance tomography system Sensors 17 280 q
y
p
g p y y
Cignoni P, Callieri M, Corsini M, Dellepiane M, Ganovelli F and Ranzuglia G 2008 MeshLab: an open-
source mesh processing tool 6th Eurographics Italian Chapter Conf. pp 129–36 Collier T J, Kynor D B, Bieszczad J, Audette W E, Kobylarz E J and Diamond S G 2012 Creation
of a human head phantom for testing of electroencephalography equipment and techniques
IEEE Trans. Bio-Med. Eng. 59 2628–34 Dai M, Li B, Hu S, Xu C, Yang B, Li J, Fu F, Fei Z and Dong X 2013 In vivo imaging of twist drill
drainage for subdural hematoma: a clinical feasibility study on electrical impedance tomography
for measuring intracranial bleeding in humans PloS One 8 e55020 Fabrizi L, McEwan A, Oh T, Woo E J and Holder D S 2009 A comparison of two EIT systems suitable
for imaging impedance changes in epilepsy Physiol. Meas. 30 S103–20 Fabrizi L, Sparkes M, Horesh L, Perez-Juste Abascal J F, McEwan A, Bayford R H, Elwes R, Binnie C D
and Holder D S 2006 Factors limiting the application of electrical impedance tomography for
identification of regional conductivity changes using scalp electrodes during epileptic seizures in
humans Physiol. Meas. 27 S163–74 y
Fu F et al 2014 Use of electrical impedance tomography to monitor regional cerebral edema during
clinical dehydration treatment PloS One 9 e113202 y
Jasper H H 1958 Report of the committee on methods of clinical examination in electroencephalography
Electroencephalogr. Clin. Neurophysiol. Suppl. 10 370–5 Jasper H H 1958 Report of the committee on methods of clinical examinatio
Electroencephalogr. Clin. Neurophysiol. Suppl. References 32 1275–83 Romsauerova A, McEwan A, Horesh L, Yerworth R, Bayford R H and Holder D S 2006 Multi-frequency
electrical impedance tomography (EIT) of the adult human head: initial findings in brain tumours,
arteriovenous malformations and chronic stroke, development of an analysis method and calibration
Physiol. Meas. 27 S147 Sadleir R and Argibay A 2007 Modeling skull electrical properties Ann. Biomed. Eng. 35 1699–712 g
y
g
p
p
g
Silva D N, Gerhardt de Oliveira M, Meurer E, Meurer M I, Lopes da Silva J V and Santa-Bárbara A
2008 Dimensional error in selective laser sintering and 3D-printing of models for craniomaxillary
anatomy reconstruction J. Cranio-Maxillo-Facial Surg. 36 443–9 va D N, Gerhardt de Oliveira M, Meurer E, Meurer M I, Lopes da Silva J V and Santa-Bárbara A
2008 Dimensional error in selective laser sintering and 3D-printing of models for craniomaxillary
i
J C
i
M
ill
F
i l S
36 443 9 Sperandio M, Guermandi M and Guerrieri R 2012 A four-shell diffusion phantom of the head for
electrical impedance tomography IEEE Trans. Bio-Med. Eng. 59 383–9 1130 Physiol. Meas. 38 (2017) 1116 J Avery et al Tang C, You F, Cheng G, Gao D, Fu F, Yang G and Dong X 2008 Correlation between structure and
resistivity variations of the live human skull IEEE Trans. Bio-Med. Eng. 55 2286–92i y
g
Tang T and Sadleir R J 2011 Quantification of intraventricular hemorrhage with electrical impedance
tomography using a spherical model Physiol. Meas. 32 811–21 Tang T and Sadleir R J 2011 Quantification of intraventricular hemor
tomography using a spherical model Physiol. Meas. 32 811–21 mography using a spherical model Physiol. Meas. 32 811–2 Tang T, Oh S and Sadleir R 2010 A robust current pattern for the detection of intraventricular hemorrhage
in neonates using electrical impedance tomography Ann. Biomed. Eng. 38 2733–47 in neonates using electrical impedance tomography Ann. B Tidswell A T, Bagshaw A P, Holder D S, Yerworth R J, Eadie L, Murray S, Morgan L and Bayford R H
2003 A comparison of headnet electrode arrays for electrical impedance tomography of the human
head Physiol. Meas. 24 527–44 Tidswell A T, Gibson A, Bayford R H and Holder D S 2001a Validation of a 3D reconstruction algorithm
for EIT of human brain function in a realistic head-shaped tank Physiol. Meas. References 22 177–85 Tidswell T, Gibson A, Bayford R and Holder D 2001b Three-dimensional electrical impedance
tomography of human brain activity NeuroImage 13 283–94 Várady T, Martin R R and Cox J 1997 Reverse engineering of geometric models—an introduction
Comput.-Aided Des. 29 255–68 Vauhkonen P, Vauhkonen M, Savolainen T and Kaipio J 1999 Three-dimensional electrical impedance
tomography based on the complete electrode model IEEE Trans. Biomed. Eng. 46 1150–60l Wolters C H, Anwander A, Tricoche X, Weinstein D, Koch M A and MacLeod R S 2006 Influence of
tissue conductivity anisotropy on EEG/MEG field and return current computation in a realistic
head model: a simulation and visualization study using high-resolution finite element modeling
NeuroImage 30 813–26 1131
|
https://openalex.org/W2142960697
|
https://europepmc.org/articles/pmc2822241?pdf=render
|
English
| null |
Influence of<i>Salvia miltiorrhizae</i>on the Mesenteric Lymph Node of Rats with Severe Acute Pancreatitis or Obstructive Jaundice
|
Mediators of inflammation
| 2,009
|
cc-by
| 5,648
|
Zhang Xiping,1 Zhang Jie,2 Ye Shuyun,2 Wang Qili,2 Feng Guanghua,1 and Pan Yan2
1Department of General Surgery, Hangzhou First People’s Hospital, Zhejiang Province, Hangzhou 310006, China
2The First College of Clinical Medicine, Zhejiang Traditional Chinese Medicine University, Zhejiang Province,
Hangzhou 310053, China Zhang Xiping,1 Zhang Jie,2 Ye Shuyun,2 Wang Qili,2 Feng Guanghua,1 and Pan Yan2
1Department of General Surgery, Hangzhou First People’s Hospital, Zhejiang Province, Hangzhou 310006, China
2The First College of Clinical Medicine, Zhejiang Traditional Chinese Medicine University, Zhejiang Province,
Hangzhou 310053, China Zhang Xiping,1 Zhang Jie,2 Ye Shuyun,2 Wang Qili,2 Feng Guanghua,1 and Pan Yan2
1Department of General Surgery, Hangzhou First People’s Hospital, Zhejiang Province, Hangzhou 310006, China
2The First College of Clinical Medicine, Zhejiang Traditional Chinese Medicine University, Zhejiang Province,
Hangzhou 310053, China Correspondence should be addressed to Feng Guanghua, zxp99688@sina.com Received 24 October 2009; Accepted 21 December 2009 Recommended by Vera L. Petricevich Objective. To observe the effect of salvia miltiorrhizae injection on inflammatory mediator levels and mesenteric lymph nodes in
severe acute pancreatitis (SAP) and obstructive jaundice (OJ) rats and explore the protective mechanism of salvia miltiorrhizae on
the lymph nodes of these rats. Methods. A total of 288 rats were used in SAP-associated and OJ-associated experiments. The rats
were randomly divided into sham-operated group, model control group, and treated group. At various time points after operation,
the pathological changes in mesenteric lymph nodes of rats in each group were observed, respectively. Results. The pathological
severity scores in lymph nodes of SAP rats in treated group were significantly lower than those in model control group (P < .05)
while the pathological changes in lymph nodes of OJ rats in treated group also showed varying degrees of mitigation. Conclusion. Salvia miltiorrhizae can exert protective effects on the lymph nodes of SAP or OJ rats via a mechanism that is associated with
reducing the contents of inflammatory mediators in blood. Copyright © 2009 Zhang Xiping et al. This is an open access article distributed under the Creative Commons Attribution License,
which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited. Copyright © 2009 Zhang Xiping et al. This is an open access article distributed under the Creative Commons Attribution License,
which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited. 1. Introduction the remission duration of abdominal pain, as well as the
duration of hospitalization [9, 10]. Severe acute pancreatitis (SAP) and obstructive jaundice
(OJ), characterized by multiple complications and high
mortality rate, have become one of the important medical
problems that need to be overcome over the years. The
pathogenesis of two different diseases is very complicated. In
recent years, it has been found that the main contributing
factors to the occurrence of multiple organ dysfunction in
SAP and OJ include the massive release of various inflam-
matory mediators, the dysregulation of the equilibrium
between proinflammatory and antiinflammatory effects, and
the disturbance of immune function [1–3]. Moreover, salvia miltiorrhizae can also protect intestinal
mucosa, reduce the translocation of bacteria and endotoxins,
enhance the immunity, improve operative safety, and shorten
the recovery duration in OJ patients [11]. In our previous
reports, we have detected the serum contents of some inflam-
matory mediators in SAP and OJ rats (including endotoxin,
PLA2, and TNF-α) [10, 12–14] and found them significantly
increased in model control groups of rats. In this study, we
observed the pathological changes in the mesenteric lymph
nodes of SAP and OJ rats, explored the protective effect
and mechanism of salvia miltiorrhizae injection on lymph
nodes, and further clarified the therapeutic significance of
inflammatory mediators in SAP and OJ. This study was
closely related to our serial previous reports. Therefore, it is important to inhibit the excessive release
of inflammatory mediators [4, 5] and restore the function
of the immune system [6] in the therapy of SAP or OJ. Some clinical studies have shown that salvia miltiorrhizae,
as an auxiliary drug, has some therapeutic effects on SAP
and OJ. Salvia miltiorrhizae is a common Chinese herbal
medicine [7, 8]; it is able to improve the clinical symptoms
in patients with SAP and significantly shorten the recovery
duration of urinary amylase and liver function damage, Hindawi Publishing Corporation
Mediators of Inflammation
Volume 2009, Article ID 675195, 7 pages
doi:10.1155/2009/675195 Hindawi Publishing Corporation
Mediators of Inflammation
Volume 2009, Article ID 675195, 7 pages
doi:10.1155/2009/675195 Hindawi Publishing Corporation
Mediators of Inflammation
Volume 2009, Article ID 675195, 7 pages
doi:10.1155/2009/675195 2.2. Methods Continuous infusion of physiological saline solution using a
microinjection pump was then maintained until the end of 2.2.4. Specimens Collection. At the corresponding time
points after operation, SAP or OJ rats were anesthetized with
2.5% sodium pentobarbital and mercifully killed. The tissue
specimens of mesenteric lymph node were collected and then
observed the pathological changes as well as pathological
severity scores, respectively. At 12 hours after operation in
SAP experiment, two rats were randomly selected from each
group, and the ultrastructural changes in the mesenteric
lymph node were observed under an electron microscope. 2.2.5. Statistical Analysis. The compiled data were first input
into the Excel sheet and then read into SPSS15.0 for further
analysis. Normal data were expressed as means (standard
deviation) while nonnormal data were expressed as medians
(interquartile range). Analysis of variance and pairwise
comparisons were used for normal data, whereas nonnormal
data were subjected to nonparametric test, among which
Kruskal-Wallis H test was used for pairwise comparisons
and Mann-Whitney U test for multiple comparisons. Yates’
chi-square test (χ2) was used for intergroup comparisons of
mortality rates. 2. Materials and Methods 2.1. Materials. 288 healthy male SD rats of clean grade,
weighing between 270 and 330 g, were provided by the 2 Mediators of Inflammation 2 the 3-hour, 6-hour, and 12-hour observation periods in the
corresponding groups. Laboratory Animal Research Center of the Zhejiang Uni-
versity of Traditional Chinese Medicine (Hangzhou, China);
sodium taurocholate and sodium pentobarbital were pur-
chased from Sigma Corporation Ltd. (St. Louis, MO, USA);
salvia miltiorrhizae injection (each 10 mL vial contains active
components equivalent to 15 g of the original medicine) was
purchased from Chiatai Qingchunbao Pharmaceutical Co.,
Ltd. (Hangzhou, China). 2.2.3. Preparation of OJ Models and Associated Ther-
apeutic Regimen. After rats were anesthetized with an
intraperitoneal injection of 2.5% sodium pentobarbital
(0.2 ml/100 g), the abdominal cavity was opened to identify
and dissociate common bile duct along the hepatoduodenal
ligament. For rats in the model control group and the
treated group, the proximal end of common bile duct was
double-ligated with surgical threads, common bile duct
was cut off, and a layered suture of the abdominal wall
was performed to close the abdominal cavity. For rats in
the sham-operated groups, common bile duct was only
dissociated but not ligated, and a layered suture of the
abdominal wall was also performed to close the abdominal
cavity. An intraperitoneal injection of salvia miltiorrhizae
injection at a dose of 0.2 ml/100 g/d was given to rats in
the treated group while equal volume of physiological saline
solution was used in the sham-operated and the model
control groups [10–16]. Injection was maintained until the
end of the 7-day, 14-day, 21-day, and 28-day observation
periods in the corresponding groups. 2.2. Methods 2.2.1. Animal Grouping. 108 rats were used for SAP-
associated experiments and randomly divided into sham-
operated, model controlgroup, and treated group (n = 36),
which were further randomly subdivided into 3-hour, 6-
hour, and 12-hour groups (n
= 12) according to time
points after operation; another 180 rats were utilized for OJ-
associated experiments and randomly divided into sham-
operated group, model control group, and treated group
(n = 60), which were further randomly subdivided into 7-
day, 14-day, 21-day, and 28-day groups (n = 15) according
to time duration after operation. 2.2.2. Preparation of SAP Models and Associated Therapeutic
Regimen. The rats were anesthetized with an intraperitoneal
injection of 2.5% sodium pentobarbital (0.2 ml/100 g). Under aseptic conditions, the thigh skin was cut open to
expose femoral vein and a transfusion passage was estab-
lished, through which continuous infusion was maintained
using a microinfusion pump (1 ml/h/100 g). Subsequently,
a median abdominal wall incision was made to expose
duodenal papilla, and a No. 5 syringe needle was used to
prick a small hole in the mesenteric avascular area. The
epidural catheter was first inserted into duodenal cavity
via the hole and then placed into the bile-pancreatic duct
towards the direction of papilla. The catheter head was tem-
porarily clamped using a microvascular clamp, and another
microvascular clamp was used to occlude the common bile
duct at the confluence of hepatic ducts to prevent a backflow
of injected drugs into the liver. After connecting the epidural
catheter end with the transfusion converter, 3.5% sodium
taurocholate (0.1 ml/100 g) was transfused at a flow rate
of 0.2 ml/min using a microinjection pump (produced by
Zhejiang University, Hangzhou, China). After completing
the transfusion, microvascular forceps and epidural catheter
were maintained for further 4 minutes and then removed. A checkup was then conducted to see whether bile leakage
was present. After suturing the hole in the lateral wall of the
duodenum, the abdominal cavity was closed conventionally. Sham-operated group was performed just by moving the
pancreas and duodenum after opening the abdominal cavity. Fifteen minutes after successful operation, a single dose of
salvia miltiorrhizae injection (0.4 ml/100 g body weight) was
given via femoral vein to rats in the treated group while
equal volume of physiological saline solution was used in
the sham-operated and the model control groups [10–16]. 3.1. SAP-Associated Experiments 3.1.1. Pathological Changes in Lymph Nodes 3.1.1. Pathological Changes in Lymph Nodes (1) Sham-Operated Group. Grossly, the morphology of
lymph nodes was normal. Under light microscopy, the
morphology and structure of lymph nodes were roughly
normal. The expansion of lymph sinuses, the hyperplasia of
sinus cells as well as the filtration of few lymphocytes, and
eosinophilic granulocytes in the sinuses were observed in Mediators of Inflammation 3 Figure 3: Treated group: 12 hours in SAP rat expansion of lymph
sinuses, hyperplasia of sinus cells, and infiltration of small part of
some lymphocytes, HE × 200. Figure 1: Model control group: 12 hours in SAP rat focal necrosis
of germinal center of lymph node, HE × 400. Figure 3: Treated group: 12 hours in SAP rat expansion of lymph
sinuses, hyperplasia of sinus cells, and infiltration of small part of
some lymphocytes, HE × 200. Figure 1: Model control group: 12 hours in SAP rat focal necrosis
of germinal center of lymph node, HE × 400. Figure 2: Model control group: 12 hours in SAP rat expansion of
lymph sinuses, hyperplasia of sinus cells, and obvious infiltration of
many inflammatory cells, HE × 200. Figure 4: Model control group: 28 days in OJ rat hyperplasia of
sinus cells, infiltration of a little eosinophile granulocytes, and a
little hemorrhage, HE × 200. Figure 2: Model control group: 12 hours in SAP rat expansion of
lymph sinuses, hyperplasia of sinus cells, and obvious infiltration of
many inflammatory cells, HE × 200. Figure 4: Model control group: 28 days in OJ rat hyperplasia of
sinus cells, infiltration of a little eosinophile granulocytes, and a
little hemorrhage, HE × 200. some rats. Under electron microscopy, the morphology and
structure of lymph nodes were roughly normal. sham-operated group (P < .05); at 3 and 6 hours after
operation, the pathological severity scores in treated group
were significantly lower than those in model control group
(P < .05); see Table 1. (2) Model Control Group. Grossly, the morphology of
lymph nodes was roughly normal. Under light microscopy,
slight enlargement and congestion of interstitial capillaries
were observed. Under electron microscopy, mitochondrial
swelling, disappearance and vacuolation, as well as lympho-
cyte necrosis and apoptosis were seen; see Figures 1 and 2. 4. Discussion Grossly, the pathological changes showed
no marked difference on 7 days after operation when
compared with those in model control group; on 14 days
after operation, lymph nodes became yellow but showed no
marked difference compared with those in model control
group; on 21 and 28 days after operation, the pathological
changes showed no marked difference compared with those
in model control group. Under light microscopy, no marked
difference in pathological changes was observed among each
time points after operation; at all time points after operation,
the boundary of the follicular germinal centers in lymph
nodes was clear; the enlargement of the follicular germinal
centers and the hyperplasia of sinus cells were seen in the
majority of rats; and few rats showed no obvious pathological
changes in lymph nodes; see Figure 5. (3) Treated Group. Grossly, the pathological changes showed
no marked difference on 7 days after operation when
compared with those in model control group; on 14 days
after operation, lymph nodes became yellow but showed no
marked difference compared with those in model control
group; on 21 and 28 days after operation, the pathological
changes showed no marked difference compared with those
in model control group. Under light microscopy, no marked
difference in pathological changes was observed among each
time points after operation; at all time points after operation,
the boundary of the follicular germinal centers in lymph
nodes was clear; the enlargement of the follicular germinal
centers and the hyperplasia of sinus cells were seen in the
majority of rats; and few rats showed no obvious pathological
changes in lymph nodes; see Figure 5. In this study, we found that, on 14, 21, and 28 days
after operation, the pathological severity scores in the lymph
nodes of OJ rats in model control group were significantly
higher than those in sham-operated group (P
<
.05),
indicating that OJ can induce pathological injury in the
lymph nodes of rats. Under electron microscopy, the necrosis
and apoptosis of lymphocytes as well as the swelling of
mitochondrial cristae were observed in the lymph nodes of
SAP rats in model control group, suggesting that SAP can
also induce pathological injury in the lymph nodes of rats. Since both SAP and OJ can induce lymph node injury, they
may threaten the stability of the body’s immune function. 3.2.2. Comparison of the Pathological Severity Scores in Lymph
Nodes. 3.2.1. Pathological Changes in Lymph Nodes Compare to sham-operated group, ∗∗∗P < .001; compare to model control group, +P < .05. Table 1: Comparison of pathological score of lymph node in SAP groups (M(QR)). Figure 5: Treated group: 28 days in OJ rat expansion of lymph
sinuses, hyperplasia of sinus cells, and infiltration of inflammatory
cells, HE × 200. .05); at all time points after operation, the pathological
severity scores in treated group were significantly higher
than those in sham-operated group (P < .05), and no
significant difference was noted between treated group and
model control group (P > .05); see Table 2. 3.2.1. Pathological Changes in Lymph Nodes (1) Sham-Operated Group. Grossly, the morphology of
lymph nodes was normal. Under light microscopy, no
marked difference in pathological changes was observed
among each time points after operation; the morphology
and structure of lymph nodes were roughly normal; the
enlargement of the follicular germinal centers and the
hyperplasia of sinus cells were seen in few rats. (3) Treated Group. Grossly, the morphology of lymph nodes
was roughly normal. Under light microscopy, the patholog-
ical changes were slightly mitigated when compared with
those in model control group. Under electron microscopy,
the swelling, disappearance, and vacuolation of few mito-
chondria were seen; see Figure 3. (2) Model Control Group. Grossly, lymph nodes became
yellow in half of the rats on 7 days after operation and in
the majority of rats on 14, 21, and 28 days after operation;
the texture of lymph nodes showed no alterations at all
time points after operation. Under light microscopy, no
marked difference in pathological changes was observed
among each time points after operation; on 7 days after
operation, the enlargement of the follicular germinal centers (2) Model Control Group. Grossly, lymph nodes became
yellow in half of the rats on 7 days after operation and in
the majority of rats on 14, 21, and 28 days after operation;
the texture of lymph nodes showed no alterations at all
time points after operation. Under light microscopy, no
marked difference in pathological changes was observed
among each time points after operation; on 7 days after
operation, the enlargement of the follicular germinal centers 3.1.2. Comparison of the Pathological Severity Scores in Lymph
Nodes. The pathological score of lymph nodes was con-
ducted according to the standard reported in the literature
[17]. At all time points after operation, no significant
difference in the pathological severity scores in lymph nodes
was noted between sham-operated group and model control
group (P > .05); at 6 hours after operation, the pathological
score in treated group was significantly higher than that in Mediators of Inflammation 4 Table 1: Comparison of pathological score of lymph node in SAP groups (M(QR)). Indexes
Sham-operated group
Model control group
Treated group
3 hours
6 hours
12 hours
3 hours
6 hours
12 hours
3 hours
6 hours
12 hours
Pathological
score
2.00
2.00
2.00
2.00
2.00
2.00
1.00+
1.00∗∗∗+
1.00
(1.00)
(0.00)
(0.50)
(1.00)
(0.00)
(1.00)
(2.00)
(1.50)
(1.00)
Note. 4. Discussion Although the pathogenesis of SAP and OJ is very com-
plicated, their ultimate outcomes are always local and
systemic inflammatory response [18] that will eventually
induce MODS (multiple organ dysfunction syndrome) and
MOF (multiple organ failure). Besides being involved in the
infections and inflammatory damage secondary to SAP or
OJ, the translocation of intestinal bacteria is also closely
related with the decline in the defense function of mesen-
teric lymph nodes [19–22]. Therefore, it is of important
significance to protect mesenteric lymph nodes in the
therapy of SAP and OJ. Salvia miltiorrhizae has advantages
of lower cost, extensive pharmacological effects, and fewer
side effects, and it is able to prevent calcium overload,
scavenge oxygen free radicals, protect against inflammation,
and improve microcirculation. Some animal experimental
results have shown [23–25] that salvia miltiorrhizae is able
to significantly reduce the excessive levels of bilirubin and
endotoxin, antagonize SAP- or OJ-induced increase in the
intestinal permeability and intestinal bacterial translocation,
effectively inhibit the production of inflammatory mediators,
and thereby protect multiple organs. The results of our
previous study showed that the survival rates of SAP and
OJ rats in treated group were higher than those in model
control group [10–16], suggesting that salvia miltiorrhizae
has some therapeutic effects on SAP or OJ rats. However,
no statistical difference in the survival rates of rats was
noted between treated group and model control group. We
surmise that this may be because the sample size is too
small. Figure 5: Treated group: 28 days in OJ rat expansion of lymph
sinuses, hyperplasia of sinus cells, and infiltration of inflammatory
cells, HE × 200. and the hyperplasia of sinus cells were seen in the majority
of rats, and few rats showed no obvious pathological changes
in lymph nodes; on 14 days after operation, the enlargement
of the follicular germinal centers and the hyperplasia of
sinus cells were seen; on 21 and 28 days after operation,
the enlargement of the follicular germinal centers and the
hyperplasia of sinus cells were seen in the majority of rats,
and spotty necrosis could be seen in the mantle zone and
germinal centers; see Figure 4. (3) Treated Group. 4. Discussion The pathological score of lymph nodes was con-
ducted according to the standard reported in the literature
[17]. On 14, 21, and 28 days after operation, the pathological
severity scores in lymph nodes in model control group were
significantly higher than those in sham-operated group (P < Mediators of Inflammation 5 Table 2: Comparison of pathological score of lymph node in OJ groups (M(QR)). Indexes
Sham-operated group
Model control group
Treated group
7 days
14 days
21 days
28 days
7 days
14 days
21 days
28 days
7 days
14 days
21 days
28 days
Pathological
score
0.00
0.00
0.00
0.00
1.00
1.00∗
1.00∗
1.50∗
1.00∗
2.00∗
1.00∗
1.00∗
(2.00)
(1.00)
(1.00)
(1.00)
(2.00)
(1.00)
(2.00)
(1.00)
(1.00)
(1.50)
(0.00)
(1.00)
Note. Compare to sham-operated group, ∗P < .05. Table 2: Comparison of pathological score of lymph node in OJ groups (M(QR)). effect on lymph nodes mainly through the following two
aspects. Thus, it is of important significance to protect lymph nodes
with drugs. Endotoxin, as one of the most important inflammatory
mediators that are involved in the pathogenesis of SAP and
OJ, may play important roles in the initiation of SIRS/MOD
and the aggravation of AP [26]. It can directly stimulate the
Kupffer cells to release inflammatory mediators, including
oxygen free radicals, TNF-α, IL-6, and IL-8, and thereby
aggravate the body’s inflammatory response [27, 28]. TNF-
α is the most important factor that mediates the toxic effects
of endotoxin. The excessive release of TNF-α may induce a
variety of pathological injuries. PLA2 is a Ca2+-dependent
enzyme [29]. In SAP and OJ, PLA2 is activated and massively
released into the blood. After reaching various organs via the
blood circulation, PLA2 can destroy the lipid membrane of
cells and cause direct cytotoxicity. Some studies showed that,
in SAP complicated with MODS, PLA2 could induce severe
damage and dysfunction of lung, liver, kidney and heart
[30] while its antagonists could improve the pathological
changes in SAP [31]. The results of our previous study
showed that the contents of inflammatory mediators such
as endotoxin, TNF-α and PLA2 in model control group
were significantly higher than those in sham-operated group
[10, 12–14], suggesting that inflammatory mediators were
massively released in SAP and OJ. Those results indicated
that salvia miltiorrhizae was able to significantly reduce
the levels of endotoxin, TNF-α and PLA2 in SAP and OJ
rats. 4. Discussion This may be because salvia miltiorrhizae can effectively
inhibit the release of endotoxin and thereby reduce the
production of other inflammatory mediators such as TNF-
α, IL-6, and IL-8. Concomitant with the decline in the
levels of inflammatory mediators, at 3 and 6 hours after
operation, the pathological scores in the lymph nodes of
SAP rats in treated group were significantly lower than
those in model control group (P
<
.05). Under light
microscopy, we found that the boundary of the follicular
germinal centers in lymph nodes of OJ rats in treated group
was clear, and few lymph nodes revealed no abnormality. Compared to model control group, the pathological changes
in treated group were improved. Therefore, we believe
that inflammatory mediator levels are correlated with the
severity of pathological changes in lymph nodes in SAP
and OJ rats. In other words, inflammatory mediators can
induce pathological damage in the lymph nodes of rats,
whereas salvia miltiorrhizae can exert protective effects on
the lymph nodes of rats through reducing the levels of
inflammatory mediators. Some studies have proved that
salvia miltiorrhizae can antagonize endotoxin-induced effects
[32]. We think that salvia miltiorrhizae exerts its protective Antiinflammatory Effect. The inflammatory mediators pro-
duced in SAP and OJ can cause intestinal mucosal dam-
age. Both endotoxin and TNF-α can directly induce local
ischemia and increased permeability of intestinal mucosa,
mediate the massive production of cytokines and pro-
inflammatory response, and thereby induce the necrosis and
apoptosis of cells as well as multiple organ dysfunction
[33–37]. Through inhibiting the production of endotoxin
and blocking the pathways through which inflammatory
cytokines are produced, salvia miltiorrhizae is able to effec-
tively modulate the gut microenvironment, protect mesen-
teric lymph nodes, and restore intestinal barrier function. Our series of studies, which have been or will be reported in
other papers, have proved that salvia miltiorrhizae is able to
protect intestinal mucosa, providing a further basis for this
hypothesis. Improving the Body’s Immune Function. It has been pointed
out in some studies [38] that endotoxin can affect the
body’s immune function through inducing the production
of inflammatory mediators. The decline in immune func-
tion can impair the body’s antiinfection ability, increase
the translocation of intestinal bacteria and endotoxins,
and aggravate inflammatory response and tissue damage. References [1] J. Granger and D. Remick, “Acute pancreatitis: models,
markers, and mediators,” Shock, vol. 24, supplement 1, pp. 45–
51, 2005. [17] X. P. Zhang, H. M. Xu, Y. Y. Jiang, et al., “Influence of
dexamethasone on meseneric lymph node of rats with severe
acute pancreatitis,” World Journal of Gastroenterology, vol. 14,
no. 22, pp. 3511–3517, 2008. [2] Z. Ling, Z. Xiping, Q. Fengmei, Y. Ping, and C. Qihui,
“Protective effects of Salvia miltiorrhizae on multiple organs
of rats with obstructive jaundice,” Mediators of Inflammation,
vol. 2009, Article ID 602935, 9 pages, 2009. [3] A. N. Kimmings, S. J. H. van Deventer, H. Obertop, E. A. J. Rauws, K. Huibregtse, and D. J. Gouma, “Endotoxin,
cytokines, and endotoxin binding proteins in obstructive
jaundice and after preoperative biliary drainage,” Gut, vol. 46,
no. 5, pp. 725–731, 2000. [18] J. Norman, “The role of cytokine in the pathogenesis of acute
pancreatitis,” American Journal of Surgery, vol. 175, no. 1, pp. 76–83, 1998. [19] X. P. Zhang, J. Zhang, Q. L. Song, and H. Q. Chen,
“Mechanism of acute pancreatitis complicated with injury of
intestinal mucosa barrier,” Journal of Zhejiang University—
Science B, vol. 8, no. 12, pp. 888–895, 2007. [4] A. Kingsnorth, “Role of cytokines and their inhibitors in acute
pancreatitis,” Gut, vol. 40, no. 1, pp. 1–4, 1997. [20] P. Garside, O. Millington, and K. M. Smith, “The anatomy
of mucosal immune responses,” The New York Academy of
Sciences, vol. 1029, pp. 9–15, 2004. [5] F. Kimura, M. Miyazaki, T. Suwa, et al., “Anti-inflammatory
response in patients with obstructive jaundice caused by bil-
iary malignancy,” Journal of Gastroenterology and Hepatology,
vol. 16, no. 4, pp. 467–472, 2001. [21] Y. Harari, N. W. Weisbrodt, and F. G. Moody, “Ileal mucosal
response to bacterial toxin challenge,” The Journal of Trauma,
vol. 49, no. 2, pp. 306–313, 2000. [6] C. Shi, X. Zhao, A. Lagergren, M. Sigvardsson, X. Wang, and
R. Andersson, “Immune status and inflammatory response
differ locally and systemically in severe acute pancreatitis,”
Scandinavian Journal of Gastroenterology, vol. 41, no. 4, pp. 472–480, 2006. [22] H. Kiyono, M.-N. Kweon, T. Hiroi, and I. Takahashi, “The
mucosal immune system: from specialized immune defense
to inflammation and allergy,” Acta Odontologica Scandinavica,
vol. 59, no. 3, pp. 145–153, 2001. [7] R. Zhang, X. Zhang, J. Mediators of Inflammation 6 animal study reported, and all rats have not been abused and
executive merciful killing when the observing time in this
study was over was conducted. [12] Z. Xiping, L. Chuyang, Z. Jie, R. Yuefang, and M. Meili,
“Protection of Salvia miltiorrhizae to the spleen and thymus
of rats with severe acute pancreatitis or obstructive jaundice,”
Mediators of Inflammation, vol. 2009, Article ID 186136, 14
pages, 2009. Acknowledgments [13] X.-P. Zhang, G.-H. Feng, J. Zhang, et al., “Protective effects
of Salvia miltiorrhizae on the hearts of rats with severe
acute pancreatits or obstructive jaundice,” Journal of Zhejiang
University—Science B, vol. 10, no. 3, pp. 193–202, 2009. This work was supported by technological foundation
project of Traditional Chinese Medicine Science of Zhejiang
province (no. 2003C130; no. 2004C142), foundation project
for medical science and technology of Zhejiang province
(no. 2003B134), grave foundation project for technological
and development of Hangzhou (no. 2003123B19), inten-
sive foundation project for technology of Hangzhou (no. 2004Z006), foundation project for medical science and
technology of Hangzhou (no. 2003A004) and foundation
project for technology of Hangzhou (no. 2005224). [14] Z. Xiping, W. Dijiong, L. Jianfeng, et al., “Effects of salvia
miltiorrhizae on ICAM-1, TLR4, NF-κB and Bax proteins
expression in multiple organs of rats with severe acute
pancreatitis or obstructive jaundice,” Inflammation, vol. 32,
no. 4, pp. 218–232, 2009. [15] X. Zhang, L. Chen, J. Zhang, et al., “Effect of salvia miltior-
rhizae on apoptosis and NF-κB p65 expression in the liver
of rats with severe acute pancreatitis or obstructive jaundice,”
Journal of Gastroenterology and Hepatology, vol. 24, no. 5, pp. 841–852, 2009. [16] Z. Xiping, C. Yang, W. Dijiong, et al., “Effects of Salvia
miltiorrhiza on intercellular adhesion molecule 1 protein
expression in the lungs of rats with severe acute pancreatitis
or obstructive jaundice,” Pancreas, vol. 38, no. 3, pp. 309–317,
2009. 4. Discussion Therefore, suppressing the production of inflammatory
mediators, such as endotoxin, TNF-α, and PLA2, can block
inflammatory mediator-induced damage to the immune
defense system, which is important for stabilizing the body’s
immune system. Since salvia miltiorrhizae perhaps has this
effect, it is worthy to deeply study its effect on the immune
function in the future. In short, salvia miltiorrhizae is able to protect the mesen-
teric lymph nodes of SAP and OJ rats through reducing
the levels of inflammatory mediators in blood. This effect
plays a positive role in stabilizing the body’s immune func-
tion, protecting intestinal mucosa barrier, and reducing the
translocation of bacteria. As the pharmacological effects of
salvia miltiorrhizae are further clarified, salvia miltiorrhizae
will be more widely used to treat SAP and OJ. Note. We claimed that this paper was original and would not
have any financial interest in a company or its competitor
and that all authors meet standard for authorship. We abided
by the ethics in this animal experiment study. The ethics
committee approval of our hospital was secured for the Mediators of Inflammation References Zhang, et al., “Efficacy and mechanism
of Salvia miltiorrhizae injection in the treatment of rats with
severe acute pancreatitis,” Inflammation, vol. 32, no. 2, pp. 109–119, 2009. [23] R. Xia and X. Chen, “Effects of Danshen injection on the
malignant obstructive jaundice in the SD rat model,” Journal
of Huazhong University of Science and Technology, vol. 26, no. 6, pp. 686–689, 2006. [8] Z. Xiping, F. Jun, W. Chengjun, et al., “Effect of salvia
miltiorrhizae on pulmonary apoptosis of rats with severe acute
pancreatitis or obstructive jaundice,” Inflammation, vol. 32,
no. 5, pp. 287–295, 2009. [24] X. P. Zhang, Z. J. Li, and D. R. Liu, “Progress in research into
the mechanism of Radix salviae miltiorrhizae in treatment
of acute pancreatitis,” Hepatobiliary & Pancreatic Diseases
International, vol. 5, no. 4, pp. 501–504, 2006. [9] Z. Z. Qin, Z. L. Hua, and Y. S. Zhou, “Efficacy of compound
Danshen in the treatment of acute pancreatitis,” Modern
Journal of Integrated Traditional Chinese and Western Medicine,
vol. 16, no. 2, pp. 194–195, 2007 (Chinese). [25] X. P. Zhang and D. R. Liu, “Advances in research on
the mechanism of salvia miltiorrhizae in the treatment of
obstructive jaundice,” Chinese Journal of Surgery of Integrated
Traditional and Western Medicine, vol. 1, no. 1, pp. 69–70, 2006
(Chinese). [10] X. Zhang, D. Liu, D. Wu, et al., “Effect of salvia miltiorrhizae
on the expressions of TLR4 protein in the liver of rats with SAP
or OJ,” Inflammation, vol. 32, no. 3, pp. 151–162, 2009. [26] G. Zhang, X. Yang, Z. Liu, X. H. Zhou, Z. D. Zhang,
and X. L. Chen, “The mechanisms underlying the effects
of endotoxemia in the development of lung injury in acute
pancreatitis,” China Journal of Modern Medicine, vol. 10, no. 7, pp. 40–41, 2000 (Chinese). [11] Y. N. Yuan, H. X. Shi, and A. H. Chen, “Efficacy of compound
Danshen injection for auxiliary treatment of obstructive
jaundice,” Lishizhen Medicine and Materia Medica Research,
vol. 17, no. 8, p. 1414, 2006 (Chinese). 7 Mediators of Inflammation 7 [27] W. X. Li, F. Wang, D. W. Han, R. L. Xu, and Y. C. Zhao,
“Correlation between intestinal endotoxemia and TNF-α, IL-
1β, NO and MDA in severe acute pancreatitis,” Journal of
Shanxi Medical University, vol. 37, no. 9, pp. 911–914, 2006
(Chinese). [28] W. X. Li, F. Wang, D. W. Han, R. L. Xu, and Y. C. References Zhao, “Effects
of intestinal endotoxemia on pathological changes of pancreas
of severe acute pancreatitis,” Shanxi Medical Journal, vol. 35,
no. 12, pp. 1083–1085, 2006 (Chinese). [29] J. H. Wu, G. G. Xu, Y. H. Guo, Q. Miao, and W. Chen,
“Changes of phospholipase A2 in the patients with acute
pancreatitis and therapeutic effect of verapamil,” Chinese
Journal of Critical Care Medicine, vol. 27, no. 2, pp. 103–105,
2007 (Chinese). [30] E. Q. Mao, J. C. Wang, T. Q. Han, C. L. Zhang, and S. Q. Zhang, “Roles of phospholipase A2 in the development of
multiple organ dysfunction in acute pancreatitis,” Journal of
Shanghai Second Medical University, vol. 16, no. 2, pp. 135–
136, 1996 (Chinese). [31] Y. Tomita, K. Kuwabara, S. Furue, et al., “Effect of a selective
inhibitor of secretory phospholipase A2, S-5920/LY315920Na,
on experimental acute pancreatitis in rats,” Journal of Pharma-
cological Sciences, vol. 96, no. 2, pp. 144–154, 2004. [32] Q. K. Hu, W. L. Shen, and Y. G. Luo, “Study on the Anti-
endotoxin effects of salvia miltiorrhiza solution (SMS) in
vitro,” Journal of Shantou University Medical College, vol. 13,
no. 3, pp. 16–17, 2000 (Chinese). [33] R. Andersson and X. D. Wang, “Gut barrier dysfunction in
experimental acute pancreatitis,” Annals of the Academy of
Medicine Singapore, vol. 28, no. 1, pp. 141–146, 1999. [34] S. H. Rahman, B. J. Ammori, J. Holmfield, et al., “Intestinal
hypoperfusion contributes to gut barrier failure in severe acute
pancreatitis,” Journal of Gastrointestinal Surgery, vol. 7, no. 1,
pp. 26–36, 2003. [35] T. Yasuda, Y. Takeyama, T. Ueda, et al., “Breakdown of
intestinal mucosa via accelerated apoptosis increases intestinal
permeability in experimental severe acute pancreatitis,” Jour-
nal of Surgical Research, vol. 135, no. 1, pp. 18–26, 2006. [36] K. Nagpal, V. R. Minocha, V. Agrawal, and S. Kapur, “Evalu-
ation of intestinal mucosal permeability function in patients
with acute pancreatitis,” American Journal of Surgery, vol. 192,
no. 1, pp. 24–28, 2006. [37] S. F. Assimakopoulos, C. D. Scopa, and C. E. Vagianos, “Patho-
physiology of increased intestinal permeability in obstructive
jaundice,” World Journal of Gastroenterology, vol. 13, no. 48,
pp. 6458–6464, 2007. [38] J. W. Greve, D. J. Gouma, P. B. Soeters, and W. A. Buurman,
“Suppression of cellular immunity in obstructive jundice is
caused by endotoxins: a study with germ-free rats,” Gastroen-
terology, vol. 98, no. 2, pp. 478–485, 1990.
|
https://openalex.org/W2764089487
|
https://pure.uva.nl/ws/files/30028018/Heinen_et_al_2018_Ecography.pdf
|
English
| null |
Extinction‐driven changes in frugivore communities on oceanic islands
|
Ecography
| 2,017
|
cc-by
| 12,383
|
Citation for published version (APA):
Heinen, J. H., van Loon, E. E., Hansen, D. M., & Kissling, W. D. (2018). Extinction-driven
changes in frugivore communities on oceanic islands. Ecography, 41(8), 1245–1255.
https://doi.org/10.1111/ecog.03462 xtinction-driven changes in frugivore communities on oceanic islands Citation for published version (APA):
Heinen, J. H., van Loon, E. E., Hansen, D. M., & Kissling, W. D. (2018). Extinction-driven
changes in frugivore communities on oceanic islands. Ecography, 41(8), 1245–1255. https://doi.org/10.1111/ecog.03462 General rights
It is not permitted to download or to forward/distribute the text or part of it without the consent of the author(s)
and/or copyright holder(s), other than for strictly personal, individual use, unless the work is under an open
content license (like Creative Commons). UvA-DARE (Digital Academic Repository) Extinction-driven changes in frugivore communities on oceanic islands
Heinen, J.H.; van Loon, E.E.; Hansen, D.M.; Kissling, W.D. DOI
10.1111/ecog.03462
Publication date
2018
Document Version
Final published version
Published in
Ecography
License
CC BY
Link to publication
Citation for published version (APA):
Heinen, J. H., van Loon, E. E., Hansen, D. M., & Kissling, W. D. (2018). Extinction-driv
changes in frugivore communities on oceanic islands. Ecography, 41(8), 1245–1255. https://doi.org/10.1111/ecog.03462 Extinction-driven changes in frugivore communities
on oceanic islands Julia H. Heinen, E. Emiel van Loon, Dennis M. Hansen and W. Daniel Kissling Julia H. Heinen, E. Emiel van Loon, Dennis M. Hansen and W. Daniel Kissling J. H. Heinen (http://orcid.org/0000-0001-8384-4273), E. E. van Loon and W. D. Kissling (http://orcid.org/0000-0002-7274-6755) (wdkissling@gmail. com), Inst. for Biodiversity and Ecosystem Dynamics (IBED), Univ. of Amsterdam, Amsterdam, the Netherlands. – D. M. Hansen, Dept of Evolutionary
Biology and Environmental Studies, Univ. of Zurich, Zurich, Switzerland. Ecography
Subject Editor: Alison Boyer
Editor-in-Chief: Miguel Araújo
Accepted 26 September 2017
41: 1245–1255, 2018
doi: 10.1111/ecog.03462 Ecography
Subject Editor: Alison Boyer
Editor-in-Chief: Miguel Araújo
Accepted 26 September 2017
41: 1245–1255, 2018
doi: 10.1111/ecog.03462 Global change and human expansion have resulted in many species extinctions
worldwide, but the geographic variation and determinants of extinction risk in
particular guilds still remain little explored. Here, we quantified insular extinctions
of frugivorous vertebrates (including birds, mammals and reptiles) across 74 tropical
and subtropical oceanic islands within 20 archipelagos worldwide and investigated
extinction in relation to island characteristics (island area, isolation, elevation and
climate) and species’ functional traits (body mass, diet and ability to fly). Out of the
74 islands, 33 islands (45%) have records of frugivore extinctions, with one third
(mean: 34%, range: 2–100%) of the pre-extinction frugivore community being
lost. Geographic areas with more than 50% loss of pre-extinction species richness
include islands in the Pacific (within Hawaii, Cook Islands and Tonga Islands) and
the Indian Ocean (Mascarenes, Seychelles). The proportion of species richness lost
from original pre-extinction communities is highest on small and isolated islands,
increases with island elevation, but is unrelated to temperature or precipitation. Large and flightless species had higher extinction probability than small or volant
species. Across islands with extinction events, a pronounced downsizing of the
frugivore community is observed, with a strong extinction-driven reduction of
mean body mass (mean: 37%, range: –18–100%) and maximum body mass (mean:
51%, range: 0–100%). The results document a substantial trophic downgrading of
frugivore communities on oceanic islands worldwide, with a non-random pattern
in relation to geography, island characteristics and species’ functional traits. This
implies severe consequences for ecosystem processes that depend on mutualistic
plant–animal interactions, including ecosystem dynamics that result from the
dispersal of large-seeded plants by large-bodied frugivores. This is an open access article under the terms of the Creative Commons
Attribution License, which permits use, distribution and reproduction in any
medium, provided the original work is properly cited.
© 2017 The Authors. This is an Online Open article ECOGRAPHY ECOGRAPHY Disclaimer/Complaints regulations Disclaimer/Complaints regulations
If you believe that digital publication of certain material infringes any of your rights or (privacy) interests, please
let the Library know, stating your reasons. In case of a legitimate complaint, the Library will make the material
inaccessible and/or remove it from the website. Please Ask the Library: https://uba.uva.nl/en/contact, or a letter
to: Library of the University of Amsterdam, Secretariat, Singel 425, 1012 WP Amsterdam, The Netherlands. You
will be contacted as soon as possible. UvA-DARE is a service provided by the library of the University of Amsterdam (https://dare.uva.nl) Download date:24 Oct 2024 h
© 2017 The Authors. This is an Online Open article Introduction It
makes island biota particularly vulnerable to global change, e.g. related to human colonization, habitat degradation, hunting
and invasive species (Whittaker and Fernandez-Palacios 2007). As a consequence, many species have gone extinct on islands
worldwide, especially on oceanic islands (Blackburn et al. 2004,
Steadman 2006, Hansen and Galetti 2009, Duncan et al. 2013, Boyer and Jetz 2014, Whittaker et al. 2017). The loss
of these species from island ecosystems also means a loss of
interactions among species, such as those between plants and
frugivores (Meehan et al. 2002, Kaiser-Bunbury et al. 2010,
Pérez-Méndez et al. 2015, 2016). Plant–frugivore interactions are important for the resil Global biodiversity is decreasing at an alarming rate
(Ceballos et al. 2017). Extinction rates of species are cur
rently at least 100 times higher than the ‘background’ rate
from the fossil record (Ceballos et al. 2015) and many plants
and animals show extirpations and severe declines in species
abundance (Butchart et al. 2010, Dirzo et al. 2014). Extinction
is a selective process because both extrinsic and intrinsic fac
tors result in a non-random loss of species (Blackburn et al. 2004, Cardillo et al. 2005, Boyer and Jetz 2010, 2014). For
instance, extrinsic factors such as small area size, low rain
fall and high elevation can increase species extinction risk on
islands (MacArthur and Wilson 1967, Blackburn et al. 2004,
Duncan et al. 2013). On the other hand, intrinsic factors
such as species traits can determine extinction risk, e.g. large-
bodied, flightless or specialized species might be more prone
to extinction than small-bodied, volant or generalist species
(Cardillo et al. 2005, Estes et al. 2011, Duncan et al. 2013,
Eskildsen et al. 2015). These non-random extinctions can
have severe consequences for ecosystem processes, e.g. when
extinction-driven body-size reduction results in the downsiz
ing and trophic downgrading of communities (Galetti et al. 2013, Pérez-Méndez et al. 2015, 2016, Svenning et al. 2016). l
d
d
h
l
h
d p
Here, we quantify extinction-driven changes of insular
frugivore communities on 74 islands within 20 archipela
gos worldwide. We focus on oceanic islands in the tropics
and subtropics because 1) seed dispersal by vertebrates is
most prevalent in the (sub)tropics (Kissling et al. 2009), and
2) oceanic islands are of particular conservation concern
due to their unique biota, high levels of endemism as well
as high rates of biodiversity loss (Blackburn et al. 2004,
Kier et al. Introduction 2009, Kaiser-Bunbury et al. 2010, Whittaker et al. 2017). We determine the magnitude of frugivore extinctions
(including birds, mammals and reptiles) and explore how
it varies with extrinsic and intrinsic factors, i.e. with island
characteristics (island area, isolation, elevation and climate)
and species’ functional traits (body mass, diet and ability to
fly). First, we ask to what extent insular frugivore extinctions
vary geographically, expecting that islands in the Pacific are
most strongly affected because they have at least two waves
of anthropogenic extinctions of the native fauna, one related
to pre-European colonization about 1600 yr ago (i.e. sea
faring Polynesians) and one after European contact in the
18th century (Steadman 2006, Boyer 2008, Duncan et al. 2013, Boyer and Jetz 2014). Second, we test whether the
proportional extinction of frugivores could be explained
by key island characteristics, especially by small island area,
large isolation, low rainfall and high elevation (MacArthur
and Wilson 1967, Blackburn et al. 2004, Whittaker and
Fernandez-Palacios 2007, Duncan et al. 2013). Third, we
quantify whether extinction probability can be predicted
with functional traits of frugivores, e.g. that large and flight
less species are most prone to extinctions (Cardillo et al. 2005, Hansen and Galetti 2009, Boyer and Jetz 2010, 2014). Fourth, we ask how extinction affects the frequency distribu
tions of body masses within insular frugivore communities,
expecting an ecological downsizing and body-size reduction
at the community level that results from the predominant
loss of large-bodied species (Estes et al. 2011, Dirzo et al. 2014, Pérez-Méndez et al. 2015). Islands tend to have lower species richness, more endemic
species, and more pronounced extinction rates than mainland
ecosystems (Whittaker and Fernandez-Palacios 2007, Hansen
and Galetti 2009, Kier et al. 2009, Whittaker et al. 2017). This
is partly due to their isolation and small size which accom
modates a limited number of species and facilitates the evo
lution of unique species interactions (MacArthur and Wilson
1967, Kaiser-Bunbury et al. 2010, Lomolino et al. 2010). It
makes island biota particularly vulnerable to global change, e.g. related to human colonization, habitat degradation, hunting
and invasive species (Whittaker and Fernandez-Palacios 2007). As a consequence, many species have gone extinct on islands
worldwide, especially on oceanic islands (Blackburn et al. 2004,
Steadman 2006, Hansen and Galetti 2009, Duncan et al. 2013, Boyer and Jetz 2014, Whittaker et al. 2017). Introduction The loss
of these species from island ecosystems also means a loss of
interactions among species, such as those between plants and
frugivores (Meehan et al. 2002, Kaiser-Bunbury et al. 2010,
Pérez-Méndez et al. 2015, 2016). Plant–frugivore interactions are important for the resil
ience of ecosystems to global change because frugivores
disperse the seeds of many plants and thereby contribute
to reproductive success, colonization ability and genetic
structure of plant populations (Nathan and Muller-Landau
2000, Nathan 2006, Pérez-Méndez et al. 2015, 2016). This
is particularly true in the tropics and subtropics where many
vertebrates include fleshy fruits in their diet and up to 90
percent of all shrubs and trees depend on animal-mediated
seed dispersal (Fleming et al. 1987, Fleming and Kress 2013). Specialized frugivores (i.e. obligate fruit eaters) are strictly
tropical/subtropical in their distribution (Kissling et al. 2009) and some specific groups of birds (e.g. fruit pigeons)
and volant mammals (e.g. fruit bats) are of key importance Extinction-driven changes in frugivore communities
on oceanic islands We suggest that targeted
conservation and rewilding efforts on islands are needed to halt the defaunation of
large and non-volant seed dispersers and to restore frugivore communities and key
ecological interactions. www.ecography.org 1245 Introduction for long-distance dispersal of plants, e.g. for the colonization
of oceanic islands (Whittaker and Jones 1994, Nathan 2006). Moreover, large-bodied frugivorous birds such as pigeons or
flightless birds such as the dodo (Meehan et al. 2002, Hansen
and Galetti 2009, Wotton and Kelly 2012) as well as reptiles
such as lizards and tortoises are especially relevant for seed
dispersal on islands (Olesen and Valido 2003, Blake et al. 2012, Hansen and Traveset 2012). Hence, on islands all three
vertebrate groups (birds, mammals and reptiles) are impor
tant seed dispersers, and the particular traits of frugivores
(e.g. flight ability and body mass) might play an important
role for seed dispersal. Global biodiversity is decreasing at an alarming rate
(Ceballos et al. 2017). Extinction rates of species are cur
rently at least 100 times higher than the ‘background’ rate
from the fossil record (Ceballos et al. 2015) and many plants
and animals show extirpations and severe declines in species
abundance (Butchart et al. 2010, Dirzo et al. 2014). Extinction
is a selective process because both extrinsic and intrinsic fac
tors result in a non-random loss of species (Blackburn et al. 2004, Cardillo et al. 2005, Boyer and Jetz 2010, 2014). For
instance, extrinsic factors such as small area size, low rain
fall and high elevation can increase species extinction risk on
islands (MacArthur and Wilson 1967, Blackburn et al. 2004,
Duncan et al. 2013). On the other hand, intrinsic factors
such as species traits can determine extinction risk, e.g. large-
bodied, flightless or specialized species might be more prone
to extinction than small-bodied, volant or generalist species
(Cardillo et al. 2005, Estes et al. 2011, Duncan et al. 2013,
Eskildsen et al. 2015). These non-random extinctions can
have severe consequences for ecosystem processes, e.g. when
extinction-driven body-size reduction results in the downsiz
ing and trophic downgrading of communities (Galetti et al. 2013, Pérez-Méndez et al. 2015, 2016, Svenning et al. 2016). Islands tend to have lower species richness, more endemic
species, and more pronounced extinction rates than mainland
ecosystems (Whittaker and Fernandez-Palacios 2007, Hansen
and Galetti 2009, Kier et al. 2009, Whittaker et al. 2017). This
is partly due to their isolation and small size which accom
modates a limited number of species and facilitates the evo
lution of unique species interactions (MacArthur and Wilson
1967, Kaiser-Bunbury et al. 2010, Lomolino et al. 2010). Island characteristics For each island, we defined the spatial extent using island
polygons from the Global Administrative Areas (GADM)
database ver. 2.8 ( www.gadm.org/ ). We extracted
data on island characteristics including area (km2), isola
tion distance from the nearest mainland (km), maximum
elevation (m), annual precipitation (mm), surrounding
landmass proportion (log10 proportion) and annual mean
temperature (°C). These data were obtained from a stan
dardized dataset on environmental characterization of the
world’s islands (Weigelt et al. 2013). As measures of isola
tion we included both distance to the nearest mainland
(DIST) and surrounding landmass proportion (SLMP)
because they represent different aspects of isolation
(Weigelt and Kreft 2013). DIST represents the shortest
distance from an island to the nearest mainland and is typ
ically used in studies of island biogeography (MacArthur
and Wilson 1967, Whittaker and Fernandez-Palacios
2007, Whittaker et al. 2017). SLMP characterizes the
amount of land area surrounding the target island and has
been shown to be a key predictor of insular plant diversity
(Weigelt and Kreft 2013). The islands were distributed across the Pacific, Atlantic
and Indian Ocean. We included resident species but excluded
non-native and vagrant species as well as migrants and non-
breeding visitors. Occurrence data was compiled from a
comprehensive literature search including scientific articles,
books, monographs, inventory lists, fossil descriptions and
other sources (Supplementary material Appendix 1). Litera
ture was searched systematically by using keywords such as
‘extinct’ or ‘frugivore’ and with specific names of island or
species. Functional traits For all extant and extinct frugivores, we collected infor
mation about three functional traits that are important
for seed dispersal: body mass (continuous, in g), amount
of fruit in the diet (categorical: low, medium or high) and
ability to fly (categorical: volant vs non-volant). These trait
data were obtained using a range of different methods,
including calculations, available databases, estimates, genus
means, primary literature, morphology or surrogate spe
cies. A summary of methods for the three functional traits is
provided in Table 1. We generally included all species that have fruit in their
diet. For extant species this information was derived from
books and published databases (Kissling et al. 2009, 2014,
Del Hoyo et al. 2011), complemented with systematic
literature searches for particular species (Supplementary
material Appendix 1). For extinct species, frugivory was
assessed with diet descriptions from the literature, often
based on morphological and dental analyses. In some cases,
frugivory had to be estimated from the closest taxonomic
relative. A detailed overview of the diet collection and
literature sources is provided in Supplementary material
Appendix 1. Body mass data were mostly obtained from existing data
bases (Table 1), including birds (Dunning 2008), mammals
(Sandom et al. 2013) and reptiles (Meiri 2010). Additionally,
the primary literature was searched for body mass values
or to estimate body mass from literature (Supplementary
material Appendix 1). For about a quarter of the species,
body mass data had to be approximated by the genus mean
(i.e. mean value across species within the same genus) or Table 1. Percent use of methods to collect information on traits (body mass, amount of fruit in diet, flight ability) for extant and extinct fru
givorous vertebrates (birds, mammals, reptiles) on islands worldwide. Percentages reflect how often a particular method has been used to
derive information on a particular trait across all included vertebrates (n=387 species). derive information on a particular trait across all included vertebrates (n = 387 species). Frugivore occurrences on islands We collected species occurrences of extant and extinct
frugivorous vertebrates (incl. birds, mammals and reptiles) 1246 for 74 tropical and subtropical oceanic islands within 20
archipelagos worldwide. We refer to the frugivore assemblage
on an island as ‘community’. Our focus was on tropical or
subtropical regions where seed dispersal by frugivorous ver
tebrates is most pronounced (Howe and Smallwood 1982,
Kissling et al. 2009) and conservation concern about the
loss of frugivores particularly high (Kaiser-Bunbury et al. 2010). We further focused on oceanic islands because they
have formed over oceanic plates with phases of emergence,
development, and submergence, and have never been con
nected to continental landmasses (Whittaker and Fernandez-
Palacios 2007, Whittaker et al. 2017). Oceanic islands are
more isolated than continental ones (Weigelt et al. 2013)
and consequently have high levels of endemism (Kier et al. 2009) with a large number of documented extinctions since
human contact (Blackburn et al. 2004, Steadman 2006,
Duncan et al. 2013, Whittaker et al. 2017). We also included
four granitic islands in the Seychelles which are considered
para-oceanic islands, i.e. comparatively small islands of a
continental origin that behave like oceanic ones in terms of
distance to the mainland and a long-term isolation (Kaiser-
Bunbury et al. 2010). Island characteristics Functional traits Following Nakagawa and
Schielzeth (2013), we calculated the marginal R2 to quantify
the variance explained by fixed factors, and the conditional
R2 to quantify the variance explained by both fixed and
random factors. pp
y
pp
The amount of fruit in the diet of each of the species was
predominantly derived from existing databases (Kissling et al. 2009, 2014), but also from the primary literature, genus
means or surrogate species (Table 1). Because the diet of
extinct species could not be quantified at the same level of
detail as the diet of extant species, we used three categories
to harmonize diet estimates across species and vertebrate
classes: low ( 30%), medium (40–60%) or high (70–
100%) amounts of fruit in the diet. For birds, we used the
proportion of fruits in the diet (Kissling et al. 2009) relative
to other dietary components mentioned in the literature (Del
Hoyo et al. 2011). For mammals, we used the MammalDIET
database (Kissling et al. 2014) and assigned data from rank 1,
rank 2 and rank 3 as high, medium and low amounts of fruit,
respectively. For reptiles as well as mammal and bird species
without diet information from existing diet databases, addi
tional primary literature, books and websites (Supplementary
material Appendix 1) were used to assign the most likely diet
category. In case species-specific diet descriptions were not
available, the diet of close taxonomic relative(s) or the genus
mean from existing databases (Kissling et al. 2009, 2014) was
used as a surrogate. hl a do
acto s. In a third step, we used a GLMM with a binomial error
distribution and a logit link function to test the relationship
between extinction status and functional traits (body mass,
amount of fruit in the diet, and ability to fly). We used the
binary extinction status (extinct = 1, extant = 0) for a spe
cific species on a specific island as response variable (i.e. island populations). This included a total of 1185 extinct
and extant species occurrences (extinct: 122, extant: 1063)
because the same species might be extinct and/or present
on different islands. This way of implementation does not
account for the independence of species, but we addition
ally tested species-level analysis by calculating the proportion
of islands where each species went extinct. Functional traits Method
Explanation
Body mass Diet Flight ability
Calculation
Calculated using equations for morphometrics (only for body mass)
9
0
0
Database
Directly obtained from available databases (body mass and diet)
60
58
0
Estimate
Estimates supported by literature (body mass and diet)
4
19
0
Genus mean Mean value across species within the same genus (all traits)
13
7
1
Literature
Directly obtained from individual articles (all traits)
3
14
5
Morphology
Evaluated on the basis of morphological characteristics (only flight ability)
0
0
94
Surrogate
Surrogate data from taxonomically related, morphologically similar species (body mass and diet)
11
2
0
Total %
100
100
100 1247 extinction probability and functional traits, and 4) ecological
downsizing, i.e. community-level body-size reduction due to
extinctions.ii by surrogate species (i.e. using data from taxonomically
related and morphologically similar species) because species-
specific body mass values were missing. For some species
(e.g. many reptiles), body mass was calculated using pub
lished morphometric data together with allometric equations
(Pough 1980) using family-specific slopes (Meiri 2010). For
tortoises, body mass was calculated with morphometric data
from 182 Aldabra giant tortoises (Aldabrachelys gigantea)
(Supplementary material Appendix 2).h In a first step, we quantified the magnitude of extinction
on each island by calculating the species richness of the pre-
extinction community (sum of extant and extinct species
richness) as well as their proportional extinction (number
of extinct species divided by the pre-extinction species
richness). In a second step, we assessed the relationship between
proportional extinction and six island characteristics (area,
distance from mainland, maximum elevation, annual pre
cipitation, surrounding landmass proportion, and mean
temperature) using a generalized linear mixed effects model
(GLMM) with a binomial error distribution and a logit link
function, following the guidelines of Zuur et al. (2009). Proportional extinction was used as the response variable
and pre-extinction species richness as a weight. The weight
defines the variance structure of the GLMM by dividing
the scale parameter of the model with the weight values
for each island. The six island characteristics were used as
predictor variables. They were not strongly correlated with
each other (all Spearman rank correlations r 0.6), scaled
with a Z-transformation, and included as fixed effects in
the GLMM. We further accounted for the nestedness of
islands within archipelagos by using archipelago (n = 20) as
a random effect for the intercept. Functional traits This alternative
implementation (using species rather than island popula
tions as the unit of analysis) yielded similar results, so we
only present the analysis at the level of island populations. In
this analysis, both body mass and amount of fruit in the diet
were used as predictor variables. Body mass was scaled with
a Z-transformation. To correct for taxonomic nestedness of
species, we used taxonomic class (bird, mammal or reptile)
as a random effect for the intercept in the GLMM. We fur
ther included ability to fly (yes = 1, no = 0) as a random effect
within each taxonomic class. g
The ability to fly was recorded as a binomial trait (volant
or non-volant), mostly evaluated on the basis of morpho
logical characteristics as well as descriptions from books and
the primary literature (Table 1; Supplementary material
Appendix 1). All reptiles and rodents were assigned to the
non-volant category and all fruit bats to the volant category. For birds, the ability to fly was evaluated separately for each
species by considering their morphology (e.g. with the help
of images). Additional literature was searched for extinct bird
species, for bird species which belong to a taxonomic group
with several known flightless species (e.g. rails, pigeons, mega
podes), or when wings were absent or small. g
An overview of all trait data sources is provided in
Supplementary material Appendix 1. Moreover, the species-
specific trait data, the methods used, and specific sources for
each species are available as a dataset from the Dryad Digital
Repository. Data deposition mean and maximum body mass for all islands with recorded
extinction events (n = 33). Changes in mean body mass (g)
were calculated by subtracting the post-extinction mean
(using extant species) from the pre-extinction mean (using
extant and extinct species). We then divided the change in
body mass by the pre-extinction mean body mass to obtain
the proportional change in body mass. The change in
maximum body mass was calculated in a similar way. Body
mass frequency distributions were visualized with horizontal
density strips and data were log-transformed before
calculations. Data available from the Dryad Digital Repository: http://
dx.doi.org/10.5061/dryad.s522m (Heinen et al. 2017). Analysis Four aspects of extinction-driven changes in insular frugi
vore communities were investigated: 1) geographic varia
tion in the extent of insular frugivore extinctions, 2) the
relationship between the proportion of frugivore extinctions
and key island characteristics, 3) the relationship between In a fourth step, we assessed ecological downsizing by
quantifying extinction-driven changes in community-level 1248 Magnitude and geographical distribution of insular
frugivore extinctions The 74 studied islands covered a size range of 5–5000 km2,
and included oceanic islands in all three major oceans (Fig. 1). Of those islands, 33 (45%) showed 1 recorded frugivore
extinction (Fig. 1, Supplementary material Appendix 3). For those 33 islands, the proportion of species richness lost
from the original frugivore communities was 34% (median:
18%), ranging from 2% on the Solomon islands Choiseul
and Malaita to 100% on the Hawaiian islands Ohau and
Lanai. Insular frugivore extinctions varied geographi
cally, with Hawaii, the Cook Islands, Tonga Islands, the
Mascarenes and the Seychelles being those archipelagos that
are most affected (all containing islands with proportional
extinctions 0.5; Fig. 1). Hence, Pacific islands show the
highest proportions of frugivore extinctions worldwide, but
high proportional extinctions were additionally recorded
in parts of the Indian Ocean (Mascarenes and Seychelles). All statistical analyses were done with the R software ver. 3.2.5 (R Development Core Team). For quantifying the mag
nitude of extinctions (step 1) we used R packages ‘ggplot2’
and ‘ggmap’. For the GLMMs (steps 2 and 3), we used the R
packages ‘lme4’ and ‘lmerTest’. We further employed a step
wise model selection based on Akaike information criterion
corrected for small sample sizes (AICc) using the R package
‘MuMIn’. GLMMs with a ΔAICc 4 compared to the best
model were rejected (Burnham and Anderson 2002). For the
first GLMM analysis (step 2), this resulted in eight GLMMs
with similar support. We therefore used model averaging
(R function ‘model.avg()’) to obtain a final model. For the
second GLMM analysis (step 3), no model averaging was
needed because only one GLMM had ΔAICc 4. Finally,
for visualizing changes in body mass frequency distributions
(step 4) the R package ‘denstrip’ was used. Figure 1. Geographic variation of insular frugivore extinctions across 74 islands worldwide. Pre-extinction species richness of frugivores is
illustrated by circle size (using quantile classification), with larger circles showing islands with more species. Proportional extinction reflects
the relative number of species lost from the original frugivore communities, illustrated with colors ranging from blue (no extinctions) to
orange and red (high proportional extinctions). Only native fruit-eating vertebrate species (birds, mammals, reptiles) are included. The circles are plotted for the mass centroid of each island using the World Geodetic System 1984 (WGS 84) as the geographic coordinate
system. Figure 1. Geographic variation of insular frugivore extinctions across 74 islands worldwide. Traits and extinction probability The best GLMM to explain extinction probability of 1063
extant and 122 extinct occurrences of 378 unique frugivo
rous species was a model that included body mass as a fixed
effect and ability to fly (volant or non-volant) and taxonomic
class (bird, mammal, reptile) as random effects (Fig. 3). This
model was better (ΔAICc 4) than GLMMs using diet
or no fixed effects (Supplementary material Appendix 5) (a)
Relative importance
(b)
AREA
(c)
DIST
(d)
ELEV
Coefficients (scaled)
Proportional extinction
Proportional extinction
Proportional extinction
Distance to mainland (km)
Elevation (m)
Island characteristics
Area (km2)
−1.5
−1.0
−0.5
0.0
0.5
1.0
1.5
**
**
*
n.s. n.s. n.s. AREA
DIST
ELEV
PREC SLMP
TEMP
●
●
●
●
●
●
●
●●
●●●
●
●
●
●
●
●
●
●
●
●
●●
●
●
0
1000
2000
3000
4000
5000
0.0
0.2
0.4
0.6
0.8
1.0
●●●●●●●●●●●●
●
●
●
●●●
●
●
●
●●
●●●●●●●●●●●●●●●
●●●●●●●●
●●
●●
●
●
●
●
●
●
●
●
●●●
●●
●
●
●
●
●
●
●
●
●
●
●
0
1000
2000
3000
4000
5000
0.0
0.2
0.4
0.6
0.8
1.0
●
●
●●
●●●●
●●
●●●●
●●●●
●●●
●●●●
●●●●
●●●●
●●●
●●●●
●●●●●●
●●●●●
●●●
●●●
●●●●●●●
●●●
●●
●●●●●
●●●●●●
●●●●●
●●●
●●
●●
●●
●●
●●
●●
●●●●●●●
●●●
●●●●
●●●
●●●
●●●
●●●
●●●●
●●●
●●●
●●
●●
●●
●●
●
●●●
●●
●
●
●
●
●
●
●
●
●●●●
●●●
●●
●
●
●
●
●
●
●
●
●
●
●
Mean model fit
Azores
Bismarck Archipelago
Bonin Islands
Canary Islands
Cape Verde
Comoros
Cook Islands
Fernando de Noronha
Fiji Islands
Galapagos Islands
Hawaii
Lord Howe Islands
Madeira Archipelago
Mascarene Islands
New Caledonia
Revillagigedo Islands
Seychelles
Solomon Islands
Tonga Islands
Vanuatu
●
●●
●●●●●●●●●●●●●●●●●●●
●
●●
●
●●●
●
●
●
●
●●
●
●
●
●
●
●●
●
●
●
●
●
●
●
●
●
●●
●
●
●
0
500
1000
1500
2000
2500
3000
0.0
0.2
0.4
0.6
0.8
1.0
Figure 2. The effect of island characteristics on frugivore extinctions across 74 islands worldwide. (a) Relative importance of predictor vari
ables as indicated by the strength of coefficients (grey = statistically significant predictors, white = not significant). (b–d) Relationship
between proportional extinction and island area, distance to mainland, and maximum elevation, respectively. Traits and extinction probability AREA
DIST
ELEV
PREC SLMP
TEMP (b)
AREA
Proportional extinction
Area (km2)
●
●
●
●
●
●
●
●●
●●●
●
●
●
●
●
●
●
●
●
●
●●
●
●
0
1000
2000
3000
4000
5000
0.0
0.2
0.4
0.6
0.8
1.0
●●●●●●●●●●●●
●
●
●
●●●
●
Mean model fit
Azores
Bismarck Archipelago
Bonin Islands
Canary Islands
Cape Verde
Comoros
Cook Islands
Fernando de Noronha
Fiji Islands
Galapagos Islands (b)
AREA
Proportional extinction
Area (km2)
●
●
●
●
●
●
●
●●
●●●
●
●
●
●
●
●
●
●
●
●
●●
●
●
0
1000
2000
3000
4000
5000
0.0
0.2
0.4
0.6
0.8
1.0
●●●●●●●●●●●●
●
●
●
●●●
● (a) (c)
DIST
Proportional extinction
Distance to mainland (km)
●
●
●●
●●●●●●●●●●●●●●●
●●●●●●●●
●●
●●
●
●
●
●
●
●
●
●
●●●
●●
●
●
●
●
●
●
●
●
●
●
●
0
1000
2000
3000
4000
5000
0.0
0.2
0.4
0.6
0.8
1.0
●
●
●●
●●●●
●●
●●●●
●●●●
●●●
●●●●
●●●●
●●●●
●●●
●●●●
●●●●●●
●●●●●
●●●
●●●
●●●●●●●
●●●
●●
●●●●●
●●●●●●
●●●●●
●●●
●●
●●
●●
●●
●●
●●
●●●●●●●
●●●
●●●●
●●●
●●●
●●●
●●●
●●●●
●●●
●●●
●●
●●
●●
●●
●
●●●
●●
●
●
●
●
●
●
●
●
●●●●
●●●
●●
●
●
●
●
●
●
●
●
●
●
● (d)
ELEV
Proportional extinction
Elevation (m)
●
●●
●●●●●●●●●●●●●●●●●●●
●
●●
●
●●●
●
●
●
●
●●
●
●
●
●
●
●●
●
●
●
●
●
●
●
●
●
●●
●
●
●
0
500
1000
1500
2000
2500
3000
0.0
0.2
0.4
0.6
0.8
1.0 (d) (c) Elevation (m) Elevation (m) Distance to mainland (km) Figure 2. The effect of island characteristics on frugivore extinctions across 74 islands worldwide. (a) Relative importance of predictor vari
ables as indicated by the strength of coefficients (grey = statistically significant predictors, white = not significant). (b–d) Relationship
between proportional extinction and island area, distance to mainland, and maximum elevation, respectively. Results were obtained from
model averaging of eight generalized linear mixed effects models with similar support (Supplementary material Appendix 4). The propor
tion of extinct frugivores (birds, mammals, reptiles) was used as response variable and six island characteristics as predictors (scaled before
analysis). Magnitude and geographical distribution of insular
frugivore extinctions Pre-extinction species richness of frugivores is
illustrated by circle size (using quantile classification), with larger circles showing islands with more species. Proportional extinction reflects
the relative number of species lost from the original frugivore communities, illustrated with colors ranging from blue (no extinctions) to
orange and red (high proportional extinctions). Only native fruit-eating vertebrate species (birds, mammals, reptiles) are included. The circles are plotted for the mass centroid of each island using the World Geodetic System 1984 (WGS 84) as the geographic coordinate
system. Figure 1. Geographic variation of insular frugivore extinctions across 74 islands worldwide. Pre-extinction species richness of frugivores is
illustrated by circle size (using quantile classification), with larger circles showing islands with more species. Proportional extinction reflects
the relative number of species lost from the original frugivore communities, illustrated with colors ranging from blue (no extinctions) to
orange and red (high proportional extinctions). Only native fruit-eating vertebrate species (birds, mammals, reptiles) are included. The circles are plotted for the mass centroid of each island using the World Geodetic System 1984 (WGS 84) as the geographic coordinate
system. 1249 Across all islands (n = 74), proportional extinction was only
weakly (and negatively) correlated with pre-extinction species
richness (Spearman rank: r = –0.11), but for the subset of
islands with extinction events (n = 33) this correlation was
strong (Spearman rank: r = –0.70), indicating that species-
rich islands tend to have lower proportional extinctions than
species-poor islands. Across all islands (n = 74), proportional extinction was only
weakly (and negatively) correlated with pre-extinction species
richness (Spearman rank: r = –0.11), but for the subset of
islands with extinction events (n = 33) this correlation was
strong (Spearman rank: r = –0.70), indicating that species-
rich islands tend to have lower proportional extinctions than
species-poor islands. factors such as surrounding landmass proportion, tem
perature and precipitation did not play an important role
in explaining proportional extinction. Hence, the hypoth
esis that small island area, large isolation, and high eleva
tion could explain the magnitude of insular extinctions was
supported, but climate variables did not play an impor
tant role. Furthermore, variance partitioning showed that
variation among archipelagos dominates: the marginal R2
values (fixed factors only) for the eight models with similar
support (Supplementary material Appendix 4) were around
0.25, while the conditional R2 values (both fixed and random
factors) were around 0.65. Extinction and island characteristics Model averaging of eight GLMMs with similar support
(Supplementary material Appendix 4) revealed that three
island characteristics were particularly important to explain
proportional extinction: island area, distance to mainland,
and elevation (Fig. 2). The most important factor was
distance to mainland (Fig. 2a), with proportional extinction
increasing the further away islands are from the mainland
(Fig. 2c). Island area was the second most important fac
tor (Fig. 2a), with small islands having large proportional
extinction (Fig. 2b). Maximum elevation of islands also
showed a statistically significant effect, with low elevation
islands having less proportional extinction (Fig. 2d). Other Traits and extinction probability Results were obtained from
model averaging of eight generalized linear mixed effects models with similar support (Supplementary material Appendix 4). The propor
tion of extinct frugivores (birds, mammals, reptiles) was used as response variable and six island characteristics as predictors (scaled before
analysis). The model was weighted with pre-extinction species richness. Archipelago was used as a random effect on the intercept (reflected
with colored lines in b–d). Abbreviations of island characteristics: AREA = island area (km2); DIST = distance to mainland (km);
ELEV = maximum elevation (m); PREC = annual precipitation (mm yr–1); SLMP = surrounding landmass proportion (log10 proportion);
TEMP = annual mean temperature (°C). (a)
Relative importance
Coefficients (scaled)
Island characteristics
−1.5
−1.0
−0.5
0.0
0.5
1.0
1.5
**
**
*
n.s. n.s. n.s. Traits and extinction probability 0.0
0.2
0.4
0.6
0.8
1.0
Body mass (g)
Extinction probability
0
10
100
1000
10 000
100 000
1
1
1
1
1
1
1
1
1
1
1
1
11
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
111
1
1
1
1
1 1
111
1
11
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1 1
1
1
1
1 1
1
1
1
1
1
0
0
0
0
0 0 0
0
0
0
0
0
00
0
0
0
0
0
0
0 0 0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0 0
0
0
0
0
0
0
0
0
0
0
0
0 00
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0 00
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
00
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0 0
0
0
0
0
0
0
0
0 0
0
0 0
0
0
0 0
0 0
0
00
0
0
0
0
0
0
0
0
0
00
0
00
0
0
0
0
0
0
0
0
0
0
0
0
0
0 0
0 0
0
0
0
0
0
0
000
00
0
0
0
0
0
0
0
00
0
0
0
0
0
0
0
0
00 00 0
0 0
0
0
0
0
0
0
00
0
00
00
0
0
0
0
0 0
0 00
0
0
0
0
0 0
0 0
0
0 0 0
0
0
0
0
0
0
0
0
0
0
00
0
0
00
0
0
0
0
0
0
0
00
0
0
0
0
0
0 0
0
0000
0
0
0
0 0
0
0
0
0 0
0
0
0
0
0
00
0
0
0
0
0
00
0
0
0
0
0
000
0
0
0
0
00
0
0
0
0
0
0
0
0
0
0
0
0
00
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0 0
0
0 0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0 0
0
0
00
0
0
0
0
0
0
0
0 0
0
0 0
0
0
0
0
0
0
0
0
0
0 0
0
0
0
0
0
0 0
0 0
0
00
0
0
0
0
0
0
0
0
0
0
0
0
0
0 0
0
0
0
0
0
0
0
0 000
0
0
00
0
0
0
0
0
0
0
0
0
0
0 0
0
0
0
0
0
0
0
0
0
0
0
0
0 0
0
00 00
0
0
0
0
0
000
0
0
0
0
0
0
0
0
0
0
0
0 0
0
0
0
0
0
0
0 0 0
0
0
0
00 000
0
0
0
0
00 0
0
0 0
0
0
0
0
0
00
0 0
0
0
0
0
0
0
0
0
0
0
00
0
0
0
0
0 000
0
0
0
0
00
0
0
0
0
0
0
0
0
0
0
00
0 0
0
0
0
0
0
0
0
0
0
0 0
0
00
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0 00
0
0
0
0
00
0
0
0
0
0
0
0
0
0
0
0
0
00
0
0
0
00
00
0
000
0
0
0
0
0
0
0
0
0
0 0
0
0
0
0
0
0
0
00
0
0
0
0
0
0
0 0
0
0
0
0
00
0
00
0
0
0
0
0
0
0
0
0
0 0
0
0
0
0
0
00
0
000
0
0
0
0
0
0
0
0
0
0 0
0
0
0
0
0
0
0
00
00
0 0
0
0
00
0
0
0
0
0
0
0
0
0
0
0 0
0
0
0
00
0
000
0
0
0
0
0
0
0
0
0
0
00
0
0
0
0
0
0
0
0
0
0
0
0
0
00
0 0
0
0
0
0
0
0
0
0
0
0
0
00
0
00
0
0
0
0
0
0
0
0
0
0
0 0
0
0
0
0
00
0
0 0
0
0
0
0
0
000
0
0 0
0
0
0
0
00
0
0 0
0
0
0
0
0
00
0
0 0
0
0
0
0
0
000
0
0 0
0
0
0
0
0
00
0
0 0
0
0
0
0
0
00
0
0 0
0
0
0
0
0
000
0
0
0 0
0
0
0
0
0
000
0
1
1
1
1
1
1
1
1
1 1
1
1
1
1
11
1
1 1
1
1
1
1
1
1
1
1
0
0
0
0 0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0 0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
00
0
00
00
00
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
Birds (non−volant)
Birds (volant)
Mammals (non−volant)
Mammals (volant)
Reptiles (non−volant) Figure 3. Traits and extinction probability The model was weighted with pre-extinction species richness. Archipelago was used as a random effect on the intercept (reflected
with colored lines in b–d). Abbreviations of island characteristics: AREA = island area (km2); DIST = distance to mainland (km);
ELEV = maximum elevation (m); PREC = annual precipitation (mm yr–1); SLMP = surrounding landmass proportion (log10 proportion);
TEMP = annual mean temperature (°C). Traits and extinction probability 1250 0.0
0.2
0.4
0.6
0.8
1.0
Body mass (g)
Extinction probability
0
10
100
1000
10 000
100 000
1
1
1
1
1
1
1
1
1
1
1
1
11
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
111
1
1
1
1
1 1
111
1
11
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1 1
1
1
1
1 1
1
1
1
1
1
0
0
0
0
0 0 0
0
0
0
0
0
00
0
0
0
0
0
0
0 0 0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0 0
0
0
0
0
0
0
0
0
0
0
0
0 00
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0 00
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
00
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0 0
0
0
0
0
0
0
0
0 0
0
0 0
0
0
0 0
0 0
0
00
0
0
0
0
0
0
0
0
0
00
0
00
0
0
0
0
0
0
0
0
0
0
0
0
0
0 0
0 0
0
0
0
0
0
0
000
00
0
0
0
0
0
0
0
00
0
0
0
0
0
0
0
0
00 00 0
0 0
0
0
0
0
0
0
00
0
00
00
0
0
0
0
0 0
0 00
0
0
0
0
0 0
0 0
0
0 0 0
0
0
0
0
0
0
0
0
0
0
00
0
0
00
0
0
0
0
0
0
0
00
0
0
0
0
0
0 0
0
0000
0
0
0
0 0
0
0
0
0 0
0
0
0
0
0
00
0
0
0
0
0
00
0
0
0
0
0
000
0
0
0
0
00
0
0
0
0
0
0
0
0
0
0
0
0
00
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0 0
0
0 0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0 0
0
0
00
0
0
0
0
0
0
0
0 0
0
0 0
0
0
0
0
0
0
0
0
0
0 0
0
0
0
0
0
0 0
0 0
0
00
0
0
0
0
0
0
0
0
0
0
0
0
0
0 0
0
0
0
0
0
0
0
0 000
0
0
00
0
0
0
0
0
0
0
0
0
0
0 0
0
0
0
0
0
0
0
0
0
0
0
0
0 0
0
00 00
0
0
0
0
0
000
0
0
0
0
0
0
0
0
0
0
0
0 0
0
0
0
0
0
0
0 0 0
0
0
0
00 000
0
0
0
0
00 0
0
0 0
0
0
0
0
0
00
0 0
0
0
0
0
0
0
0
0
0
0
00
0
0
0
0
0 000
0
0
0
0
00
0
0
0
0
0
0
0
0
0
0
00
0 0
0
0
0
0
0
0
0
0
0
0 0
0
00
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0 00
0
0
0
0
00
0
0
0
0
0
0
0
0
0
0
0
0
00
0
0
0
00
00
0
000
0
0
0
0
0
0
0
0
0
0 0
0
0
0
0
0
0
0
00
0
0
0
0
0
0
0 0
0
0
0
0
00
0
00
0
0
0
0
0
0
0
0
0
0 0
0
0
0
0
0
00
0
000
0
0
0
0
0
0
0
0
0
0 0
0
0
0
0
0
0
0
00
00
0 0
0
0
00
0
0
0
0
0
0
0
0
0
0
0 0
0
0
0
00
0
000
0
0
0
0
0
0
0
0
0
0
00
0
0
0
0
0
0
0
0
0
0
0
0
0
00
0 0
0
0
0
0
0
0
0
0
0
0
0
00
0
00
0
0
0
0
0
0
0
0
0
0
0 0
0
0
0
0
00
0
0 0
0
0
0
0
0
000
0
0 0
0
0
0
0
00
0
0 0
0
0
0
0
0
00
0
0 0
0
0
0
0
0
000
0
0 0
0
0
0
0
0
00
0
0 0
0
0
0
0
0
00
0
0 0
0
0
0
0
0
000
0
0
0 0
0
0
0
0
0
000
0
1
1
1
1
1
1
1
1
1 1
1
1
1
1
11
1
1 1
1
1
1
1
1
1
1
1
0
0
0
0 0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0 0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
00
0
00
00
00
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
Birds (non−volant)
Birds (volant)
Mammals (non−volant)
Mammals (volant)
Reptiles (non−volant)
Figure 3. Traits and extinction probability Extinction probability of frugivorous birds, mammals and reptiles in relation to body mass and ability to fly. For all taxa, the
probability of extinction increases with large body mass (logarithmic x-axis, in g), but extinction is higher for non-volant birds and mam
mals compared to volant species. Results are obtained from a binomial generalized linear mixed effects model (GLMM) with body mass as
a fixed effect and the ability to fly (volant, non-volant) and taxonomic class (bird, mammal, reptile) as random effects (compare Supplemen
tary material Appendix 5). The GLMM was based on 1185 species occurrences of 387 unique species on 74 islands. The body masses of
extinct species are plotted as ‘1’ and those of extant species as ‘0’, with black indicating non-volant species and grey representing volant
species. Downsizing and changes in community-level body mass For most of the 33 islands with recorded extinction events,
the loss of species resulted in a downsizing of the frugivore
community (Fig. 4). On average, extinctions decreased com
munity-level mean body mass by one third (mean: 37%,
range: –18–100%) relative to the pre-extinction commu
nity. The two Hawaiian islands Oahu and Lanai showed a
complete loss of their native frugivore community (Fig. 4). In contrast, three islands in the Pacific (Rarotonga of Cook
Islands, Ambrym of Vanuatu, and Santa Cruz of Galapagos)
surprisingly showed a slight increase in community-level body
mass because the extinct frugivores were smaller in body size
than the community mean (Fig. 4). Nevertheless, all other
30 islands showed a decrease in community-level body mass
indicating ecological downsizing (Fig. 4). Across islands, the
largest species ( 1.2 kg) that were lost were reptiles and
birds (Fig. 4). These included tortoises (order Testudines)
such as species in the genera Chelonoidis, Cylindraspis and
Meiolania, as well as scaled reptiles (order Squamata) such
as giant lizards and skinks, e.g. the goliath Tenerife lizard
Gallotia goliath and the Mauritian giant skink Leiolopisma
mauritiana. For birds, this included flightless ducks (order
Anseriformes), pigeons (order Columbiformes) and wildfowl
(order Galliformes) such as the giant Sylviornis neocaledoniae Traits and extinction probability Extinction probability of frugivorous birds, mammals and reptiles in relation to body mass and ability to fly. For all taxa, the
probability of extinction increases with large body mass (logarithmic x-axis, in g), but extinction is higher for non-volant birds and mam
mals compared to volant species. Results are obtained from a binomial generalized linear mixed effects model (GLMM) with body mass as
a fixed effect and the ability to fly (volant, non-volant) and taxonomic class (bird, mammal, reptile) as random effects (compare Supplemen
tary material Appendix 5). The GLMM was based on 1185 species occurrences of 387 unique species on 74 islands. The body masses of
extinct species are plotted as ‘1’ and those of extant species as ‘0’, with black indicating non-volant species and grey representing volant
species. from New Caledonia, the Rodrigues solitaire Pezophaps
solitaria, the dodo Raphus cucullatus from Mauritius, and a
large goose-like duck (Thambetochen chauliodous) from the
Hawaiian Islands. and revealed that extinction probability is particularly high
for large-bodied species (Fig. 3). Moreover, for birds and
mammals it showed that extinction probability depends on
the ability to fly, with non-volant species having a higher
probability of going extinct than volant species (Fig. 3). This
supported the hypothesis that large and flightless species are
most prone to extinctions. Compared to the mean body mass changes at the
community-level, the changes in the largest (heaviest) spe
cies were often pronounced. Across the 33 islands, the maxi
mum body mass of post-extinction frugivore communities
was severely reduced (mean = 51%, range: 0–100%) when
compared to the pre-extinction communities. On about one
third (11 of 33) of the islands, the maximum body mass did
not change (i.e. the largest species is still present). However,
most of them still showed an ecological downsizing, i.e. a
decrease in community-level mean body mass (due to the
loss of species with a body mass larger than the pre-extinction
community-level mean). Overall, our results document a
pervasive ecological downsizing and body-size reduction of
insular frugivore communities. However, they also provide
hope because some large-bodied species are still left, at least
on some islands. Discussion Our study provides a first comprehensive overview of
frugivore extinctions on oceanic islands worldwide. The
results show that losses of frugivorous species have been
reported on almost half of the 74 studied islands, and that
on average about one third of the original community has
disappeared. We further show that both extrinsic factors (e.g. physical conditions such as island size, isolation and archi
pelago membership) and intrinsic traits (e.g. large body size
and inability to fly) are key predictors of extinction risk, and 1251 Figure 4. Extinction-driven downsizing of frugivore communities. Changes in community-level body mass of frugivores (birds, mammals,
reptiles) are shown for 33 islands on which frugivore extinctions have occurred. For each island, the body mass frequency distribution before
extinction is indicated with blue density strips (dark blue = high density of species with a particular body mass, light blue = low density). The pre-extinction mean body mass (dark blue dash) and the post-extinction mean body mass (red dash) are illustrated for each community. The body masses of extinct bird species (circles), mammals (squares) and reptiles (triangles) are indicated for each community. Islands are
sorted according to the amount of change in mean community-level body mass (grey arrow) indicating the magnitude of downsizing of
frugivore communities. Figure 4. Extinction-driven downsizing of frugivore communities. Changes in community-level body mass of frugivores (birds, mammals,
reptiles) are shown for 33 islands on which frugivore extinctions have occurred. For each island, the body mass frequency distribution before
extinction is indicated with blue density strips (dark blue = high density of species with a particular body mass, light blue = low density). The pre-extinction mean body mass (dark blue dash) and the post-extinction mean body mass (red dash) are illustrated for each community. The body masses of extinct bird species (circles), mammals (squares) and reptiles (triangles) are indicated for each community. Islands are
sorted according to the amount of change in mean community-level body mass (grey arrow) indicating the magnitude of downsizing of
frugivore communities. role for ecological functioning varies geographically among
archipelagos. that a pervasive ecological downsizing of insular frugivore
communities has occurred due to the predominant loss of
large-bodied vertebrates. We suspect that this has serious
consequences for insular plant populations and ecosystems. Many islands have lost large proportions of their origi
nal frugivore communities, with some islands (e.g. Discussion within
Hawaii, the Cook Islands, Tonga Islands, the Mascarenes and
the Seychelles) having lost 50% of their pre-extinction fru
givore richness. Especially the Pacific region – following waves
of extinctions after both pre-European and European coloni
zation – has seen one of the largest extinction events in the
Holocene (Steadman 2006, Boyer 2008, Duncan et al. 2013). Our results further show high proportional extinctions in
other parts of the world such as the Indian Ocean (Mascarenes
and Seychelles). For instance, in the Mascarenes at least 30
frugivores became extinct (18 birds, 2 mammals, 10 reptiles),
including the dodo Raphus cucullatus, the Mauritian giant
skink Leiolopisma mauritiana and giant tortoises Cylindraspis
spp. This suggests a dramatic loss of ecological function rela
tive to pre-human native ecosystems (Rijsdijk et al. 2009,
Kaiser-Bunbury et al. 2010) and severe consequences for
insular seed-dispersal networks (Schleuning et al. 2014). Magnitude and geographical distribution of insular
frugivore extinctions Previous community-level studies of insular extinctions
across multiple archipelagos have usually focused on birds
(Blackburn et al. 2004, Duncan et al. 2013, Boyer and Jetz
2014). Instead, our analysis summarizes insular frugivore
extinctions across all three vertebrate classes (birds, mam
mals and reptiles). Although birds represent the majority of
frugivore species on islands, reptiles such as lizards (Olesen
and Valido 2003) and tortoises (Hansen and Galetti 2009)
as well as mammals such as frugivorous bats (Whittaker
and Jones 1994) also play a key role for animal-mediated
seed dispersal on islands (Kaiser-Bunbury et al. 2010). Within some archipelagos (e.g. Galapagos, the Mascarenes,
Comoros, Canary Islands and Solomon Islands), reptiles
or mammals are relatively more diverse than in other
archipelagos (Supplementary material Appendix 6). This
suggests that the relative importance of particular verte
brate classes for animal-mediated seed dispersal and their Downsizing and changes in community-level body mass Our results revealed a substantial ecological downsizing of
many insular frugivore communities worldwide. Across the 33
islands that have recorded extinction events, post-extinction
changes in body size were predominantly character
ized by a community-wide reduction (mean: 37%; range:
–18–100%), often losing the largest frugivore. Such eco
logical downsizing can entail the loss of unique ecological
functions (Boyer and Jetz 2014) and negatively affect seed
dispersal, gene flow and recruitment of plants (Meehan et al. 2002, Pérez-Méndez et al. 2015, 2016). h ,
,
)
The actual consequences of frugivore extinctions for plant
regeneration and ecosystem dynamics on islands have yet
been little studied. Large-bodied frugivores are dispropor
tionately important as dispersers because they can swallow
or handle large fruits and deposit seeds far away from the
parent plant (Blake et al. 2012, Wotton and Kelly 2012). For two large-seeded New Zealand tree species, field experi
ments suggest that dispersal failure reduces plant regen
eration through effects on seed predation, germination
and seedling survival (Wotton and Kelly 2011). In the
Canarian archipelago, the extinction of large-bodied lizards
(Gallotia, Lacertidae) has led to a strong reduction of seed
ling recruitment (Pérez-Méndez et al. 2015) and gene flow
(Pérez-Méndez et al. 2016) in plants that rely exclusively on
lizards for seed dispersal. On Galapagos, the presence of giant
tortoises (genus Geochelone) facilitates a greater recruitment
of an arboreal cactus which is an important keystone resource
for many other animals (Gibbs et al. 2008). These examples
highlight the important consequences of insular frugivore
extinctions for trophic interactions, but more studies are
clearly needed to better quantify the effects of ecological
downsizing on plant regeneration and ecosystem dynamics,
especially on islands (Hansen and Galetti 2009). Precipitation, temperature and surrounding landmass
proportion did not show a statistically significant effect on
proportional extinction of frugivores. Precipitation and tem
perature might play a minor role in our study because all
islands are located within the (sub-)tropics and the oceanic
climate around islands might have a buffering effect (Cronk
1997, Kissling et al. 2012). Surrounding landmass propor
tion, a good predictor of plant species richness on islands
(Weigelt and Kreft 2013), might have been unimportant in
our GLMMs because we used archipelago as a random effect,
which probably captures most of the spatial variation in land
masses around islands. Traits and extinction probability Extinction probability of frugivores was strongly linked
to species’ traits. As previously observed for insular bird
communities (Steadman 2006, Boyer 2008, Duncan et al. 2013, Boyer and Jetz 2014), large and flightless species
suffered higher rates of extinction than did small-bodied
and volant species. This substantiates that the loss of species
is not a random but a selective process, where specific
intrinsic traits lead to high vulnerability (McKinney 1997,
Cardillo et al. 2005, Eskildsen et al. 2015). Large-bodied
frugivores are particularly prone to extinction because their
space use and resource requirements increase with increasing
body size (Haskell et al. 2002), and hence large frugivores
are much more sensitive to small areas than small frugivo
res are (Sodhi et al. 2004, Boyer and Jetz 2010). The loss
of large-bodied vertebrates is ubiquitous across the globe
and can have far-reaching consequences for ecosystem pro
cesses (Estes et al. 2011), including negative effects on plant
regeneration and seedling recruitment (Meehan et al. 2002, Extinction and island characteristics Our analysis of island characteristics revealed that distance to
the mainland, island area and maximum elevation were the 1252 Galetti et al. 2013, Pérez-Méndez et al. 2015, 2016). Flightless
species are extinction prone because they are vulnerable to
hunting, habitat loss and exotic predatory mammals which
were often established on islands after European coloniza
tion (Blackburn et al. 2004, Steadman 2006, Boyer 2008,
Duncan et al. 2013). three strongest correlates of proportional extinction. Propor
tional extinction was highest on isolated and small islands and
increased with island elevation. Larger areas allow more spe
cies to coexist than smaller areas (Arrhenius 1921), e.g. due
to reduced extinction rates (MacArthur and Wilson 1967). Moreover, because resources are limited on small islands,
human-induced habitat loss, hunting and invasive species
can quickly affect the whole island (Whittaker and Fernan
dez-Palacios 2007). For the positive effect of distance to the
mainland, we suggest that island isolation often causes exag
gerated ecological release, resulting in evolutionary modifi
cations such as gigantism and flightlessness (MacArthur and
Wilson 1967, Lomolino et al. 2010). Such traits can make
species particularly prone to extinction. The positive effect of
maximum elevation (as similarily found in Blackburn et al. 2004) is unexpected and might be explained by species being
restricted to small areas of suitable habitat along the eleva
tional gradient, making them more vulnerable to extinction
if this is lost. Downsizing and changes in community-level body mass Moreover, it is important to note
that variation among archipelagos dominated in explaining
proportional extinction compared to the effects of single
island characteristics. This suggests that additional factors
at the archipelago level – such as human arrival time – are
required to more fully explain insular extinctions worldwide. Conclusion and declines. – Proc. Natl Acad. Sci. USA 114: E6089–E6096. C
k Q C B 1997 I l
d
bili
di
i
i Our study synthesizes current knowledge of extant and extinct
frugivorous vertebrate species and documents widespread
extinctions and pervasive ecological downsizing on many
islands. Two key conclusions can be drawn. From a scien
tific point of view, our findings encourage predictions of
extinction risks by demonstrating that large and flightless
species on small and isolated islands are most prone to
extinction. The models that were built in this study pro
vide an important building block for forecasting the likely
effects of future global change on biodiversity because both
extrinsic factors (e.g. island characteristics) and intrinsic
traits (e.g. body size and inability to fly) emerge as key pre
dictors of extinction risk. From a conservation perspective,
our results suggest that targeted conservation and rewilding
efforts on islands are urgently needed to halt the defauna
tion of large and non-volant seed dispersers and the tro
phic downgrading of insular frugivore communities. This
requires empirical studies on the ecological consequences
of insular frugivore extinctions and how to restore func
tional (mega)faunas and extinct interactions within island
ecosystems. and declines. Proc. Natl Acad. Sci. USA 114: E6089 E6096. Cronk, Q. C. B. 1997. Islands: stability, diversity, conservation. – Biodivers. Conserv. 6: 477–493. Del Hoyo, J. et al. 2011. Handbook of the birds of the world
volume 1–16. – Lynx Edicions. y
Dirzo, R. et al. 2014. Defaunation in the Anthropocene. – Science
345: 401–406. Duncan, R. P. et al. 2013. Magnitude and variation of prehistoric
bird extinctions in the Pacific. – Proc. Natl Acad. Sci. USA 110:
6436–6441. Dunning, J. B. 2008. Handbook of avian body masses. – CRC
Press. Eskildsen, A. et al. 2015. Ecological specialization matters:
long-term trends in butterfly species richness and assemblage
composition depend on multiple functional traits. – Divers. Distrib. 21: 792–802. Estes, J. A. et al. 2011. Trophic downgrading of planet Earth. – Science 333: 301–306.h Fleming, T. H. and Kress, W. J. 2013. The ornaments of life:
coevolution and conservation in the tropics. – Chicago Univ. Press. Fleming, T. H. et al. 1987. Patterns of tropical vertebrate frugivore
diversity. – Annu. Rev. Ecol. Syst. 18: 91–109. Galetti, M. et al. 2013. Functional extinction of birds drives rapid
evolutionary changes in seed size. – Science 340: 1086–1090.h Gibbs, J. P. et al. 2008. References Arrhenius, O. 1921. Species and area. – J. Ecol. 9: 95–99. Arrhenius, O. 1921. Species and area. – J. Ecol. 9: 95–99. Blackburn, T. M. et al. 2004. Avian extinction and mammalian
introductions on oceanic islands. – Science 305: 1955–1958. Blackburn, T. M. et al. 2004. Avian extinction and mammalian
introductions on oceanic islands. – Science 305: 1955–1958. Blake, S. et al. 2012. Seed dispersal by Galápagos tortoises. – J. Biogeogr. 39: 1961–1972. g
g
Boyer, A. G. 2008. Extinction patterns in the avifauna of the
Hawaiian islands. – Divers. Distrib. 14: 509–517.i Boyer, A. G. and Jetz, W. 2010. Biogeography of body size in Pacific
island birds. – Ecography 33: 369–379. Boyer, A. G. and Jetz, W. 2014. Extinctions and the loss of ecological
function in island bird communities. – Global Ecol. Biogeogr. 23: 679–688. Burnham, K. P. and Anderson, D. R. 2002. Model selection and
multimodel inference – a practical information-theoretic
approach. – Springer. pp
p
g
Butchart, S. H. M. et al. 2010. Global biodiversity: indicators of
recent declines. – Science 328: 1164–1168. Cardillo, M. et al. 2005. Multiple causes of high extinction risk in
large mammal species. – Science 309: 1239–1241. large mammal species. – Science 309: 1239–124 Ceballos, G. et al. 2015. Accelerated modern human–induced spe
cies losses: entering the sixth mass extinction. – Sci. Adv. 1:
e1400253. Ceballos, G. et al. 2017. Biological annihilation via the ongoing
sixth mass extinction signaled by vertebrate population losses
d d li
P
N tl A d S i USA 114 E6089 E6096 Implications for conservation and restoration Our results provide several implications for the conservation
and restoration of island ecosystems. While we acknowledge
that the (sub-)fossil record is incomplete and can be biased
towards particular regions, taxa or size classes (Steadman
2006, Boyer 2008, Duncan et al. 2013, Kemp and Hadly
2015), our data provide a first comparative baseline for pre
serving and restoring plant–frugivore interactions on many
islands. The underlying dataset (available from the Dryad
Digital Repository) allows to identify particular frugivores
and their ecological function (e.g. approximated by traits such
as body size) in the context of pre-extinction communities. 1253 References This can facilitate the restoration of native seed dispersal
interactions, e.g. by increasing populations of functionally
unique species or by replacing extinct species with taxon sub
stitutes (Kaiser-Bunbury et al. 2010). The latter approach,
rewilding of ecosystems by introducing extant species as
functional replacements or ecological analogues for extinct
native species, is controversial but gaining momentum given
the need to remedy defaunation and restore trophic cascades
in the Anthropocene (Svenning et al. 2016), including on
islands (Kaiser-Bunbury et al. 2010). Conservation practitio
ners are starting to employ large and giant tortoises (family
Testudinidae) in island ecosystems such as the Galapagos,
the Mascarenes and the Seychelles to replace extinct tortoises
with extant analogues (Hansen et al. 2010). This can improve
seed germination and establishment of endangered large-
fruited plant species (Griffiths et al. 2011). However, imple
mentation of rewilding projects to restore plant–frugivore
interactions on islands is still rare (Kaiser-Bunbury et al. 2010), and empirical evidence as well as rigorous scientific
studies on trophic rewilding are often lacking (Svenning et al. 2016). Acknowledgements – We thank Donald R. Drake and Kevin C. Burns
for constructive feedback on an earlier version of this manuscript,
Renske Onstein, James Campbell and Boris Berkhout for support
with statistical analyses, and Richard Baxter and Joliene Wennink
for providing data. We are also grateful to Robert Whittacker, José
María Fernández-Palacios, Kenneth Rijsdijk, Henri Hooghiemstra
and Chris Clements for discussion.
Funding – Financial support was provided by a Univ. of Amsterdam
starting grant to WDK. Acknowledgements – We thank Donald R. Drake and Kevin C. Burns
for constructive feedback on an earlier version of this manuscript,
Renske Onstein, James Campbell and Boris Berkhout for support
with statistical analyses, and Richard Baxter and Joliene Wennink
for providing data. We are also grateful to Robert Whittacker, José
María Fernández-Palacios, Kenneth Rijsdijk, Henri Hooghiemstra
and Chris Clements for discussion.
Funding – Financial support was provided by a Univ. of Amsterdam
starting grant to WDK. Funding – Financial support was provided by a Univ. of Amsterdam
starting grant to WDK. Conclusion The role of endangered species reintroduction
in ecosystem restoration: tortoise–cactus interactions on
Española Island, Galápagos. – Restor. Ecol. 16: 88–93.fi g
Griffiths, C. J. et al. 2011. Resurrecting extinct interactions with
extant substitutes. – Curr. Biol. 21: 762–765.h Hansen, D. M. and Galetti, M. 2009. The forgotten megafauna. – Science 324: 42–43. Hansen, D. M. and Traveset, A. 2012. An overview and introduction
to the special issue on seed dispersal on islands. – J. Biogeogr. 39: 1935–1937. 1254 Olesen, J. M. and Valido, A. 2003. Lizards as pollinators and seed
dispersers: an island phenomenon. – Trends Ecol. Evol. 18:
177–181. Hansen, D. M. et al. 2010. Ecological history and latent conserva
tion potential: large and giant tortoises as a model for taxon
substitutions. – Ecography 33: 272–284. Pérez-Méndez, N. et al. 2015. Downsized mutualisms: consequences
of seed dispersers’ body-size reduction for early plant recruit
ment. – Perspect. Plant Ecol. Evol. Syst. 17: 151–159.h Haskell, J. P. et al. 2002. Fractal geometry predicts varying body
size scaling relationships for mammal and bird home ranges. – Nature 418: 527–530. Heinen, J. H. et al. 2017. Data from: Extinction-driven changes in
frugivore communities on oceanic islands. – Dryad Digital
Repository, http://dx.doi.org/10.5061/dryad.s522m . Pérez-Méndez, N. et al. 2016. The signatures of Anthropocene
defaunation: cascading effects of the seed dispersal collapse. – Sci. Rep. 6: 24820.h p
Pough, F. H. 1980. The advantages of ectothermy for tetrapods. – Am. Nat. 115: 92–112. p
y
p
g
y
Howe, H. F. and Smallwood, J. 1982. Ecology of seed dispersal. – Annu. Rev. Ecol. Syst. 13: 201–228. Kaiser-Bunbury, C. N. et al. 2010. Conservation and restoration
of plant–animal mutualisms on oceanic islands. – Perspect. Plant Ecol. Evol. Syst. 12: 131–143. Rijsdijk, K. F. et al. 2009. Mid-Holocene vertebrate bone
Concentration-Lagerstätte on oceanic island Mauritius provides
a window into the ecosystem of the dodo (Raphus cucullatus). – Quat. Sci. Rev. 28: 14–24. y
Kemp, M. E. and Hadly, E. A. 2015. Extinction biases in Quaternary
Caribbean lizards. – Global Ecol. Biogeogr. 24: 1281–1289. Sandom, C. et al. 2013. Mammal predator and prey species richness
are strongly linked at macroscales. – Ecology 94: 1112–1122. Kier, G. et al. 2009. A global assessment of endemism and species
richness across island and mainland regions. – Proc. Natl Acad. Sci. USA 106: 9322–9327.h Schleuning, M. et al. 2014. At a loss for birds: insularity increases
asymmetry in seed-dispersal networks. Conclusion – Global Ecol. Biogeogr. 23: 385–394. Kissling, W. D. et al. 2009. The global distribution of frugivory in
birds. – Global Ecol. Biogeogr. 18: 150–162. Sodhi, N. S. et al. 2004. Avian extinctions from tropical and
subtropical forests. – Annu. Rev. Ecol. Evol. Syst. 35: 323–345. Sodhi, N. S. et al. 2004. Avian extinctions from tropical and
subtropical forests. – Annu. Rev. Ecol. Evol. Syst. 35: 323–345. Steadman, D. W. 2006. Extinction and biogeography of tropical
Pacific birds. – Univ. of Chicago Press. Kissling, W. D. et al. 2012. Quaternary and pre-Quaternary
historical legacies in the global distribution of a major tropical
plant lineage. – Global Ecol. Biogeogr. 21: 909–921. Steadman, D. W. 2006. Extinction and biogeography of tropical
Pacific birds. – Univ. of Chicago Press. Kissling, W. D. et al. 2014. Establishing macroecological trait
datasets: digitalization, extrapolation, and validation of diet
preferences in terrestrial mammals worldwide. – Ecol. Evol. 4:
2913–2930. Svenning, J.-C. et al. 2016. Science for a wilder Anthropocene:
synthesis and future directions for trophic rewilding research. – Proc. Natl Acad. Sci. USA 113: 898–906. Weigelt, P. and Kreft, H. 2013. Quantifying island isolation
– insights from global patterns of insular plant species richness. – Ecography 36: 417–429. Lomolino, M. V. et al. 2010. Biogeography. – Sinauer Associates.h MacArthur, R. H. and Wilson, E. O. 1967. The theory of island
biogeography. – Princeton Univ. Press. g p y
Weigelt, P. et al. 2013. Bioclimatic and physical characterization
of the world’s islands. – Proc. Natl Acad. Sci. USA 110:
15307–15312.h g
g
McKinney, M. L. 1997. Extinction vulnerability and selectivity:
combining ecological and paleontological views. – Annu. Rev. Ecol. Syst. 28: 495–516. Whittaker, R. J. and Jones, S. H. 1994. The role of frugivorous bats
and birds in the rebuilding of a tropical forest ecosystem, Kraka
tau, Indonesia. – J. Biogeogr. 21: 245–258. y
Meehan, H. J. et al. 2002. Potential disruptions to seed dispersal
mutualisms in Tonga, western Polynesia. – J. Biogeogr. 29:
695–712. Whittaker, R. J. and Fernandez-Palacios, J. M. 2007. Island
biogeography. – Oxford Univ. Press. Meiri, S. 2010. Length–weight allometries in lizards. – J. Zool. 281:
218–226. g
g
Whittaker, R. J. et al. 2017. Island biogeography: taking the long
view of nature’s laboratories. – Science 357: eaam8326. Nakagawa, S. and Schielzeth, H. 2013. A general and simple
method for obtaining R2 from generalized linear mixed-effects
models. – Methods Ecol. Evol. 4: 133–142. Supplementary material (Appendix ECOG-03462 at www.
ecography.org/appendix/ecog-03462 ). Appendix 1–6. Conclusion Wotton, D. M. and Kelly, D. 2011. Frugivore loss limits recruitment
of large-seeded trees. – Proc. R. Soc. B 278: 3345–3354. Nathan, R. 2006. Long-distance dispersal of plants. – Science 313:
786–788. g
Wotton, D. M. and Kelly, D. 2012. Do larger frugivores move seeds
further? Body size, seed dispersal distance, and a case study of
a large, sedentary pigeon. – J. Biogeogr. 39: 1973–1983.f Nathan, R. and Muller-Landau, H. C. 2000. Spatial patterns of
seed dispersal, their determinants and consequences for
recruitment. – Trends Ecol. Evol. 15: 278–285. g
p g
g
g
Zuur, A. F. et al. 2009. Mixed effects models and extensions in
ecology with R. – Springer. 1255
|
https://openalex.org/W2921257751
|
http://gala.gre.ac.uk/id/eprint/23244/7/23244%20MARAGIANNIS_Diversity_and_Inclusivity_in_the_Age_of_Wearables_%28OA%29_2019.pdf
|
English
| null |
Diversity and Inclusivity in the Age of Wearables: A Buzzword, a Myth, an Uncertain Reality
|
Body, space & technology journal
| 2,019
|
cc-by
| 6,209
|
How to Cite: Maragiannis, A and Ashford, R 2019 Diversity and Inclusivity
in the Age of Wearables: A Buzzword, a Myth, an Uncertain Reality. Body,
Space & Technology, 18(1), pp. 198–214. DOI: https://doi.org/10.16995/
bst.320 Published: 12 March 2019 Published: 12 March 2019 Peer Review: This article has been peer reviewed through the double-blind process of Body, Space & Technology, which
is a journal published by the Open Library of Humanities. Open Access: Body, Space & Technology is a peer-reviewed open access journal. Copyright: © 2019 The Author(s). This is an open-access article distributed under the terms of the Creative
Commons Attribution 4.0 International License (CC-BY 4.0), which permits unrestricted use,
distribution, and reproduction in any medium, provided the original author and source are credited. See http://creativecommons.org/licenses/by/4.0/. Diversity and Inclusivity in the Age of
Wearables: A Buzzword, a Myth, an
Uncertain Reality Anastasios Maragiannis1 and Rain Ashford2
1 School of Design, University of Greenwich, London, UK
2 Department of Computing, Goldsmiths, University of London, UK
Corresponding author: Dr. Anastasios Maragiannis (a.maragiannis@gre.ac.uk) ‘Wearable Technology’ is a buzzword of our contemporary era. It could be
argued there are few examples of aesthetically pleasing devices that are
designed to meet our needs and/or our consumer desires. However, do we
focus on design and aesthetics of technology as a holistic action with the
capacity to simultaneously engage conceptual and practical shifts that make
our society a place with no boundaries? To design inclusively is to engage
the user deeply throughout the design process, sharing our practices and
amalgamating people’s unique knowledge as technological interventions. Design diversity and inclusion seems to be used interchangeably with two
other terminologies, a) Universal Design and b) Design for All. The terms
have a parallel purpose but their origin and use is distinguished in various
parts of the world. For example, Inclusive Design is used within Europe
and goes beyond age, ethnicity, gender, sex, and disabilities to focus on
other excluded groups to deliver mainstream solutions. Inspired by the
limited understanding and choices around aesthetics and personalisation
in wearables, this article discusses how we use technology to empower
individuals in a variety of contexts; to improve our way of living in the
world, through a number of contextual resources and practice-research,
which were devised and conducted to address women’s concerns and
preferences on wearable technologies. Digital Preservation: The Open Library of Humanities and all its journals are digitally preserved in the CLOCKSS scholarly
archive service. Anastasios Maragiannis and Rain Ashford, ‘Diversity
and Inclusivity in the Age of Wearables: A Buzzword, a
Myth, an Uncertain Reality’ (2019) 18(1) Body, Space &
Technology. DOI: https://doi.org/10.16995/bst.320 Anastasios Maragiannis and Rain Ashford, ‘Diversity
and Inclusivity in the Age of Wearables: A Buzzword, a
Myth, an Uncertain Reality’ (2019) 18(1) Body, Space &
Technology. DOI: https://doi.org/10.16995/bst.320 Keywords: Wearables; design; technology; gender; diversity; inclusivity ‘Wearable Technology’ is a buzzword of our contemporary era. The wearable
devices currently on offer vary in functionality, including: activity tracking, medical
monitoring, mobile connectivity and more. However, it could be argued that there
are few examples of aesthetically pleasing devices that are designed to meet both ‘Wearable Technology’ is a buzzword of our contemporary era. The wearable
devices currently on offer vary in functionality, including: activity tracking, medical
monitoring, mobile connectivity and more. However, it could be argued that there
are few examples of aesthetically pleasing devices that are designed to meet both Maragiannis and Ashford: Diversity and Inclusivity in the Age of Wearables 199 199 our needs and our consumer desires. The question then, is whether we focus on the
design aspect and aesthetics of technology or, whether we should make utility the
focal point and centre of our attention instead. Aesthetics can often be a matter of
taste; subjective, personal and therefore diverse. In contrast, utility is more inclusive
in that, design in this context encompasses the needs of a wide range of people and
must appeal to the growing needs of an increasingly global culture. Nevertheless,
in the design process the terms diversity and inclusivity are not necessarily a
contradiction in terms. To design inclusively is to engage the user deeply throughout
the design process, sharing practices and amalgamating people’s unique knowledge
as technological interventions; it is therefore inclusive, precisely by including the
diverse. Design diversity and inclusion seem to be used interchangeably with two
other terms, a) Universal Design and b) Design for All. This article looks to investigate
artistic practices that engage with the idea of diversity and inclusivity in digital
technologies and particularly, the gendered body which, is in keeping with the scope
and focus of BST Journals’ publications. One of the marginalised groups that have been under-represented regarding
choices and design in technology are women. Technology and tools when aimed
at this group have often embodied stereotypes that could be seen as patronising
and are often superficially marketed, for example, through the use of stereotypical
colour schemes such as, ‘pinking’ (Schroeder, 2010). Also by being accompanied
by condescending accoutrements or texts such as, recipes (Laird, 2010). Keywords: Wearables; design; technology; gender; diversity; inclusivity In 2011,
the Fitbit Ultra clip-on activity tracker came in a choice of a black enclosure with
a pink (named ‘plum’ by the manufacturers) or blue underside, which upheld an
obvious western gender stereotype and marketing ploy of pink as feminine and blue
as masculine. Inspired by the limited understanding and choices around aesthetics
and personalisation in wearables, this article discusses how we use technology to
empower individuals and in particular women. This article will engage with this
question in a variety of contexts and through a number of contextual resources
and practice-research studies, which have been devised and conducted to address
women’s concerns and preferences on wearable technologies. What then, are the
implications of diversity and inclusivity in wearable technology for women? Maragiannis and Ashford: Diversity and Inclusivity in the Age of Wearables 200 Dominant Stereotypes and Other Myths yp
y
Culturally constructed gender stereotypes begin in childhood and can have a
substantial influence on children’s self-concepts (Witts, 1997). For instance, toys can
communicate and reinforce gender-based stereotypes. Toys for boys are often highly
manipulative and/or electronic (Caleb 2000; Sanders, 1997). Whereas, toys for girls
tend to be designed to include interpersonal interaction, like dolls which encourage
the development of social skills and relationships instead (Caleb, 2000). Researchers
such as, Sanders, Koch and Urso (1997) have long asserted that girls are not exposed
to toys that encourage scientific, mathematical and technological thinking and
are therefore less likely to develop any interest in related subjects. These cultural
stereotypes are communicated and reinforced from a young age and consequently
often influence and inform the school curriculum developing gender identities in
adulthood (Fighting Stereotype Stigma). This then informs the technological design
process and the features which will be marketed to either men or women. Does this
lead to a more complex design process? Or, does it in fact do the very opposite i.e. promote reproductions of gender pigeonholing perpetuating the myths of gender
stereotypes? A recent (2017) international research exhibition named Diversity and Inclusivity
by Design [d+iD Research Hub] in London (Figure 1), demonstrated artworks
which explored how design can empower individuals in a variety of contexts; to
improve our way of living in the world; the implications when it comes to wearables
and inclusion; and how we can enable designers when they design with, and for
others. The selection of artworks showcased in this exhibition demonstrated design
thinking through multidisciplinary approaches that positively impact our social and
interdisciplinary landscapes. These works explored the deeper meaning of co-design
methodological processes through the theme of gender in wearable design, with a
focus on “objects” that are designed to engage people from diverse and marginalised
groups. Gender biases and stereotypes are consistently being refuted in contemporary
society. For example, the performance artist Viktoria Modesta challenges a
stereotypical image of being disabled by using wearable technology, in the form 201 Maragiannis and Ashford: Diversity and Inclusivity in the Age of Wearables Figure 1: The diversity and Inclusivity by Design International Exhibition, London
(Maragiannis, 2016). Figure 1: The diversity and Inclusivity by Design International Exhibition, London
(Maragiannis, 2016). of stylised and bespoke prosthetics to embolden and accentuate her image as a
powerful woman (Saner, 2014). Dominant Stereotypes and Other Myths Emotive wearables research, conducted in London
in 2014, investigates how wearable technology can be used to create nonverbal
communication and explores how physiological data drawn from the body can be
visualised and broadcast. This research focuses on the preferences and concerns
of women in various age groups. The collected data indicated that each group
had different expectations and requirements from wearables and this was akin to
daily choices made in respect to clothing such as, form, colour, shape, texture and
prominence of the object. The feedback also included preferences regarding where
on the body devices be worn. This was influenced by new variables determined by
the purpose of the wearable and also how exposed it made the women feel in terms
of broadcasting and visualising personal data (Ashford, 2018). A large-scale Danish
study looking at female interaction with electronic products, which was based on the
observation of a male gender bias in tech devices, found that various factors had an
effect on motivation and how women operated certain technologies. The outcomes
of this research addressed how to make electronic products more relevant, beneficial
and appealing to female users, including design aspects and aesthetics. Significantly,
according to the authors of this research, the idea of gender-focused design appears
to be in conflict with the political idea of gender equality. They claim: “The idea
of equality and equal rights for men and women sometimes prevents people from Maragiannis and Ashford: Diversity and Inclusivity in the Age of Wearables 202 tapping the potential that lies in an analysis of and a focus on gender differences”
(Schroeder, 2010). The study did not aim to dispel gender stereotypes and myths
but rather, provide practical guidelines for designing products by avoiding gender
stereotypes and working with different female personas and attitudes towards
technology based on statistic clusters. However, regardless of intention the study does
not entirely manage to avoid gender stereotypes altogether. Rather, it differentiates
between male approaches as detail orientated and female sensemaking through
coherence; claiming that female approaches are socially orientated and more
emotional. It therefore, reproduces some of the cultural stereotypes that it tries to
avoid. At the same time, it is important to note that research in technology and
emotive or, affective responses is gaining traction. Dominant Stereotypes and Other Myths Applications now exist that are operable through our wearables and other
technologies that allow us to reconnect with emotional qualities that can mirror
our human essence in prescient terms of artificial responses. These artifacts are now
linked with our body through portable intelligent devices including smartphones
that drive, send and process data from wearables. We use these technologies in the
hope of better communication; the data extracted from these technologies can aid
communication by rethinking and reshaping our physical and social interactions. These apps include non-gender specific tools such as EEG headsets which connect and
share physiological data with various devices such as computers and smartphones. For marketing and crowd feedback online apps such as, CrowdEmotion are used for
tracking “attention, facial coding to understand engagement, and implicit testing to
quantify memorability” via eye-tracking (CrowdEmotion, 2018). Emotive wearables,
such as pendants, garments and accessories which track and broadcast physiological
data associated with emotions and mental states such as the EEG Visualising Pendant
(2012) (Ashford, 2018) (Figure 2) may process data in a non gender-specific way, but
require some interpretational skills depending on the situation and wearer. Diversity and Design Choices: Aesthetics and Technology To help understand how design for current wearables has evolved, it is necessary to
look back at the history of these devices. Depending on how far back we go, early
wearable technology can be traced back to the thirteenth century when eyeglasses 203 Maragiannis and Ashford: Diversity and Inclusivity in the Age of Wearables Figure 2: The EEG Visualising Pendant, an example of an emotive wearable by Rain
Ashford (2012). Figure 2: The EEG Visualising Pendant, an example of an emotive wearable by Rain
Ashford (2012). were invented. Three centuries later, one of the earliest portable watches, designed
to be worn around the neck and named the Nuremburg Egg was created. Since
the twentieth century leaps and bounds have been made in the evolution of these
technologies. For example, head-mounted augmented reality (AR) wearables can
be traced back to devices such as the 1961 Philco display for remote viewing of
surveillance cameras (Mann, 2015; Comeau and Bryan, 1961). In the early 1990s,
wearable technology was not necessarily created to be aesthetically pleasing. The
priority of many of those working on early wearables was to find ways to achieve
a particular function or functionalities from the hardware that was available to
them, such as hacking desktop computers and gaming devices and what resulted
was often large, obtrusive and complex (Dvorak, 2007). Artefacts that were starting
to evolve in research labs and universities were akin to research prototypes and
not commercially viable. Yet, pioneering researchers such as Steve Mann and Thad were invented. Three centuries later, one of the earliest portable watches, designed
to be worn around the neck and named the Nuremburg Egg was created. Since
the twentieth century leaps and bounds have been made in the evolution of these
technologies. For example, head-mounted augmented reality (AR) wearables can
be traced back to devices such as the 1961 Philco display for remote viewing of
surveillance cameras (Mann, 2015; Comeau and Bryan, 1961). In the early 1990s,
wearable technology was not necessarily created to be aesthetically pleasing. The
priority of many of those working on early wearables was to find ways to achieve
a particular function or functionalities from the hardware that was available to
them, such as hacking desktop computers and gaming devices and what resulted
was often large, obtrusive and complex (Dvorak, 2007). Diversity and Design Choices: Aesthetics and Technology Artefacts that were starting
to evolve in research labs and universities were akin to research prototypes and
not commercially viable. Yet, pioneering researchers such as Steve Mann and Thad Maragiannis and Ashford: Diversity and Inclusivity in the Age of Wearables 204 Starner illustrate that wearable computing was gradually changing the paradigms
of human-computer interaction. More lightweight, portable and wearable artifacts
were being championed from the mid-1990s’ (Figure 3). A common trend was
emerging amongst independent inventors: “a personal computer should be worn,
much as eyeglasses or clothing is worn, to provide access to computing power at all
times. These new machines are now mature enough to provide personal, portable,
augmented realities” (Starner et al. 1995). By 1995, in Europe, Philips Electronics had launched its Vision of the Future initiative,
which included investigating user needs in technology (Ryan, 2014). Philips followed
this with a book of aspirational ideas for lifestyles focused on wearable technology. New
Nomads: An Exploration of Wearable Electronics (2001) featured clothing for work and
leisure with embedded technology intended to make the workload or leisure time of
the wearer easier or more enjoyable. The garments were envisioned to work seamlessly
with technology including mobile communications, displays and personal audio
(Philips, 2000). Ideas concerning both the use and aesthetics of wearables has changed
dramatically over the years. The use of wearables has been transformed, for example,
eyeglasses as a medical necessity become a fashion artifact. However, some of these
endeavours to produce both useful and aesthetically pleasing wearables have failed. Figure 3: Steve Mann: Evolution of Wearcomp (CC BY-SA 3.0) (2004) https://
commons.wikimedia.org/wiki/File:Wearcompevolution.jpg
(accessed
30/12/2018). Figure 3: Steve Mann: Evolution of Wearcomp (CC BY-SA 3.0) (2004) https://
commons.wikimedia.org/wiki/File:Wearcompevolution.jpg
(accessed
30/12/2018). Maragiannis and Ashford: Diversity and Inclusivity in the Age of Wearables 205 Despite best efforts of putting wearables on women on the catwalk and in
magazines, it has not paid off for all wearables. For example Google’s Glass was
promoted as a must-have accessory. It was sported by fashion bloggers, critics and
celebrities in 2013, and featured in a 13-page spread in Vogue, as well as, appearing
on the catwalk in the established US designer Diane Von Fürstenberg’s Spring 2013
show (Klein, 2013). The non-gendered but prominently geeky-looking headset
attracted controversy over privacy and issues of exclusivity, leading to wearers being
called “glassholes”. Diversity and Design Choices: Aesthetics and Technology The Glass project was shelved in 2015 before it became publicly
available or had the chance to be taken up or tested properly by women or any other
groups. The device later found its place in industry (and various workplaces) where
aesthetics and issues around it breaching privacy were less problematic (Levy, 2017). This example illustrates how a wearable product can be hyped and prominently
marketed to women via the catwalk but extenuating issues, in this case social and
privacy issues, can alter the path of the eventual design and development of a device. Wearing electronics on the body makes a statement about the wearer in terms
of their relationship with technology but also how they want to be perceived by
others. It is comparable to making a choice about personal style; wearing a particular
garment, or a piece of jewellery and/or any other accessory. It is therefore, essential
for designers and manufacturers alike to consider that ‘one size does not fit all’ and
to consider giving wearers more options for customisation and consider allowing for
more personalisation through bespoke designs. More recently, a combination of factors such as, the rise of the Maker movement,
which embraces collaboration-based learning and demystification of technologies,
stimulated the development of smaller and more easily obtained electronics
components, sensors and actuators, which has helped encourage interest in wearables. In particular, technology such as, the sewable Lilypad Arduino microcontroller,
which was intended to inspire girls and female designers to investigate the design of
wearables creation and spark an interest in STEM subjects and lead to the design of
technology products better aimed towards women’s requirements (Buechley, 2006). Importantly, media and public interest in wearable devices began to rise when
self-tracking became popular due to the Quantified Self movement. The movement Maragiannis and Ashford: Diversity and Inclusivity in the Age of Wearables 206 began in California in 2008 and inspired a plethora of inventors, start-ups and
industry experts to take an interest in fitness and wellbeing devices through self-
tracking (Wolf, 2016). In 2011, the aforementioned Fitbit activity tracker was one of
the first devices that tapped into and commercialised this interest. It first appeared
on the market as a small and unobtrusive device that clipped onto the body and
came in limited colour choices with a design that was functional and minimal. Diversity and Design Choices: Aesthetics and Technology At the
same time, Fitbits were not overtly fashionable items that reflected a personalised
look or, a style choice. As the popularity of these devices increased, the availability of
similar wearables for fitness and wellbeing grew including Nike’s FuelBand and the
Jawbone UP, consequently stimulating discussion around the design and aesthetics
of these artifacts. Wearables have faced a multitude of challenges over the years. They
have developed from heavy, clunky forms and poor battery life to more streamlined
devices. In the past it might have been assumed that design and aesthetics would
not be such an issue (in comparison to the functionality of the technology itself), but
disappointment with regard to design values were already reported in technology
media from 2015. Often wearables were discussed in terms of ‘ugly’ aesthetics,
which put into question women’s desirability to purchase such devices, especially
from established technology brands racing to get products such as, smartwatches
out before competitors (Arthur, 2014). If these early commercial devices were not
engaging users through their aesthetics, then what are the other factors that make
them compelling? When Fitbits and similar devices burst onto the commercial scene
they brought with them a certain social cachet, due to wearables being a desirable
up and coming technology to possess. These fitness and wellbeing devices connect
users through gamification, using various visualisations of accumulated data and are
therefore social; they contain social media features that encourage the sharing of data
between friends as competitive motivation to increase exercise and ultimately, the
usage of these devices. These compelling features required the user to agree to terms
and conditions releasing the data for upload and sharing to third-party companies. Users of fitness trackers willingly parted with and shared with others private
information gathered by their devices. Due to popularity, Fitbit quickly expanded
their range of devices, which included more prominently placed wristbands on the Users of fitness trackers willingly parted with and shared with others private
information gathered by their devices. Due to popularity, Fitbit quickly expanded
their range of devices, which included more prominently placed wristbands on the Maragiannis and Ashford: Diversity and Inclusivity in the Age of Wearables 207 body and more colour choices soon followed. Diversity and Design Choices: Aesthetics and Technology These rubberised objects still drew
comments on design issues; Fitbit responded by partnering with designers such as
Vera Wang and Tory Burch to create bracelets for their devices, which gave activity
trackers the look of bespoke designer jewellery rather than objects to be hidden
(Mahajan, 2016). Although this suggested that there is an audience for fashion
orientated wearables that look personal and are chosen to accompany an outfit,
there still seems to be a bridge to be crossed where design and technology have hit
the right note with consumers (Wissinger, 2017). Reasons for this might be because
although younger generations of western women may have more opportunities to
use technology through schooling, gaming and mobile communications, generations
with disposable income may not; hence, some women still find their personal and
domestic technology intimidating, suggesting accessibility to be key (Charara, 2015). In recent years, the Internet of Things, aimed to link us to all the “stuff” we are
wearing, everything we need to work with and for us (Medaglia, 2010). Additionally,
we start to see the evolving connection of our bodies directly to the machine and
virtual “others”; in particular, by opening up instantaneous encompassing haptic
senses to wearable robots and avatars that we intentionally create or choose to wear:
“the internet of bodies” (BDS, 2017). Looking further into design choices, issues
and preferences for women, emotive wearables research into aesthetics in a 2014
study by Rain Ashford, discussed earlier in this article, also brought up the issue
about where on the body women would want to wear devices. The study found
that for many this was dependent on whether they were comfortable sharing their
data in public, thereby making body areas important to the design. Displays facing
outwards suited those who were comfortable with their data being visible, whereas
displays facing inwards was essential for those who just wanted to keep their data
for their own viewing. Form factors of wearable technology mentioned by women
(as preference) included jewellery such as, wrist worn devices, badges, pendants,
earrings and brooches but also, devices that are embedded into garment cuffs, lapels
and lining which further reflected the discussion on issues of privacy in regard of
visibility of emotive data and also when sharing was appropriate. Diversity and Design Choices: Aesthetics and Technology Having devices that Maragiannis and Ashford: Diversity and Inclusivity in the Age of Wearables 208 were customisable and personalised to fit usage on different occasions appropriately,
such as, work and special occasions made the prospect of using an emotive wearable
more viable for everyday use. A 2013 study into cross-cultural societal perceptions
of gesture interactions with a wearable e-textile interface, found that American and
South Korean women and men’s attitudes steered towards favouring the wrist and
forearm for placement on the body, but reported some unease with performing
touch gestures on the upper body of a female actor (Profita, 2013). Activity tracking:
barriers, workarounds and customisation, examined the abandonment of popular
activity trackers and discovered reasons for unhappiness with their device including
the desire for better functionality and aesthetics (Harrison et al. 2015). Investigations
found that the physical design was important and could be a barrier to usage. Research into emotive wearables found that when asked about the functionality
of such personal devices, potential users requested the ability to customise modes
for different settings, which included the aesthetics of how data was displayed. This
would allow for the display of certain kinds of emotive data when appropriate or
give the user the ability to personalise how data is visualised, so that data appeared
to be scrambled or could not be read for privacy or aesthetic reasons (Ashford, 2018). Hence, when discussing aesthetics it is not only the exterior of the object which is
to be considered but also, how the information is displayed to the consumer. Utility
and aesthetics are thereby, not mutually exclusive issues for wearable technology. Rather, the one tends to inform and depend on the other, as well as, how consumers
will take to the product. Uncertain reality: Data, Women and Privacy Who is willing to wear the wearables of the future? Designs that encompass bespoke
and personalised aspects have been found to be attractive, but the devices themselves
need to provide technology that is useful or compelling to women, and also be
accessible in terms of design and functionality. For example, the notion of security
wearables that track or include a panic button for women might sound like a useful
application for wearable technology but brings with it an assumption that women Maragiannis and Ashford: Diversity and Inclusivity in the Age of Wearables 209 alone are seen as “victims-in-the-making” (Wissinger, 2017). These are unwanted
identities that add to the discussion of what women want from wearables and how
a wearable can change how they are seen by others or, how women themselves feel
whilst wearing them. The design and aesthetics of wearables for women is a problematic area because
we need to consider that one form, one set of aesthetics, or functionalities does not
work for everyone. As mentioned previously, in regard to the barriers for accepting
fitness devices, an artifact that is unattractive is not likely to be used by women and
may be hidden, or shunned. This is something that various studies have discovered
through focus groups and field tests and demonstrated by looking at possible
audiences for wearable technology. The studies conducted (and mentioned above),
returned valuable feedback about the use of technology in everyday life scenarios
such as work and socialising. This included inquiring about aesthetic qualities and
functionality. Designing for privacy is an important matter for women, so wearables
need to consider carefully how any data is broadcast. For those who feel comfortable
about sharing their physiological data this is not such a problem, but for those who
are worried about, for example, how physiological data concerned with emotive
states might be recorded or interpreted by others, it was worrisome enough to be a
barrier to usage and so would require measures in place that would reassure them. Conclusion No two people have the same taste in clothing or fashion worn on the body so it
should be assumed that no two people will have the same taste in wearables. This
was made apparent in user studies where particular attention was paid to the
requirements and concerns of women. The studies indicated that when women were
asked what they required from wearable technology, in terms of aesthetics, they
emphasised a requirement for wearables to have elements of personalisation and
to be customised to suit different situations and styles. One size or style does not fit
all, which was reflected by those who said they were comfortable with conspicuous
or brightly lit wearables on the body, compared to others who said they would
prefer their wearables to be discrete or with muted or toned down visualisations Maragiannis and Ashford: Diversity and Inclusivity in the Age of Wearables 210 of data. In terms of wearables in everyday interactions, prominence is a factor
and so design elements to enhance or, make technology more in keeping with
the rest of one’s dress sense is desirable. Lessons should be learned from the early
commercialisation of wearables where complaints about large, ugly, clunky and very
masculine wearables, such as smartwatches were common. A very important factor
in the development of wearables is the social and cultural connotations of these
devices, how they then portray women and whether women are ready to accept the
attention that these devices might bring to them, this includes privacy and what
data these devices might give away to companies and those sharing information
via social media or directly viewing these devices. What is coming next though
for wearables? Through practice-based research we could amplify, diversify, and
mobilise this under-investigated aspect of practice and discourse; where design acts
inclusively regardless of disability, gender, ethnicity, vulnerability, language or age;
where design can empower individuals in a variety of contexts; and improve our
way of living in the world. Looking to the future, how will women use wearables
to enhance their day-to-day lives and empower themselves? Will it be through the
idea of using technology to further longevity, or physical perfection as transhumans? Conclusion That is, as feminist Haraway discussed in 1990 through the eyes of a chimera in her
Cyborg Manifesto (Haraway, 1990) which, is still relevant today as we try find our own
ways of expressing ourselves and personal meaning through the use of technology. Competing Interests The authors have no competing interests to declare. Dr Anastasios Maragiannis A designer, academic and researcher, Dr Anastasios Maragiannis has devoted
his career to exploring and understanding the fundamentals of design diversity
and inclusion, within the context of the digital landscape and a multidisciplinary
approach to design thinking. Anastasios is currently the Deputy Head School of
Design at the University of Greenwich, London and a Fellow of the Royal Society of
the Arts. He is the director of the d+iD – diversity + inclusivity by Design research Maragiannis and Ashford: Diversity and Inclusivity in the Age of Wearables 211 unit which explores how design acts inclusively regardless of disability, gender,
ethnicity, vulnerability, language and age. He is also the Design Critique Lead for
TypeThursday London with a great interest on how computational technologies
impact on the existing principles and forms of design (readability and legibility). Anastasios presented in a number of international conferences and he has been
showing his work in various places including the London Design Festival, and the
V&A museum, London. Dr Rain Ashford Rain Ashford is a creative technologist, designer, lecturer and consultant working
in the field of wearable technology. Her research investigates how wearables can
be used to create new forms of nonverbal communication using physiological
and environmental data. Rain was selected for a Convocation Trust Student
Entrepreneur Award (2014), and was a finalist in EPSRC’s UK ICT Pioneers
Competition (2015). Her Baroesque Barometric Skirt was featured in New
Scientist Magazine. In 2015 Rain was part of the organising committee for the
Critical Wearables Research Lab at LCF, UAL London, She has presented her work
extensively in the UK and internationally in Europe, Asia and USA. References Arthur, R 2014 “Why is wearable technology so damn ugly?” https://www.telegraph. co.uk/women/womens-life/10569007/Why-is-wearable-technology-so-damn-
ugly.html (accessed 20/09/2018). Arthur, R 2014 “Why is wearable technology so damn ugly?” https://www.telegraph. co.uk/women/womens-life/10569007/Why-is-wearable-technology-so-damn-
ugly.html (accessed 20/09/2018). Ashford, R 2018 Responsive and Emotive Wearable Technology: physiological data,
devices and communication (Doctoral dissertation, Goldsmiths, University of
London). Ashford, R 2018 Responsive and Emotive Wearable Technology: physiological data,
devices and communication (Doctoral dissertation, Goldsmiths, University of
London). BDS 2017 http://www.bodydataspace.net/2018/10/we-in-social-tech-accelerating-
women-entrepreneurs/ (accessed 1/10/18). BDS 2017 http://www.bodydataspace.net/2018/10/we-in-social-tech-accelerating-
women-entrepreneurs/ (accessed 1/10/18). Buechley, L 2006 October. A construction kit for electronic textiles. In: 2006 10th
IEEE international symposium on wearable computers, 83–90. IEEE. Caleb, L 2000 Design Technology: Learning how girls learn best. Equity & Excellence,
33(1): 22–25. DOI: https://doi.org/10.1080/1066568000330105 212 Maragiannis and Ashford: Diversity and Inclusivity in the Age of Wearables Charara, S 2015 If you’re not designing for women, get out of the wearable tech
game now. Wareable. Retrieved October 6 from: http://www.wareable.com/
wearable-tech/design-for-women-or-go-home. Comeau, C P and Bryan, J S Nov. 1961 “Headsight television system provides remote
surveillance.” Electronics, 10: 86–90. CrowdEmotion 2018 https://www.crowdemotion.co.uk (accessed 11/09/18). Dvorak, J L 2007 Moving Wearables into the Mainstream: Taming the Borg, 2008
edn, Springer. Fighting Stereotype Stigma https://www.wiley.com/college/psyc/westen240494/
student/mod2/Stereoty.htm (accessed 20/10/2017). Haraway, D J 1990 Cyborg Manifesto, Simians, Cyborgs, and Women: The Reinvention
of Nature, first edition edn, Routledge. Harrison, D, Marshall, P, Bianchi-Berthouze, N and Bird, J 2015 September. Activity tracking: barriers, workarounds and customisation. In: Proceedings of the
2015 ACM International Joint Conference on Pervasive and Ubiquitous Computing,
617–621. ACM. DOI: https://doi.org/10.1145/2750858.2805832 Klein, S 2013 Google Glass and a Futuristic Vision of Fashion – Magazine. Available
at:
http://www.vogue.com/magazine/article/the-final-frontier-google-glass-
and-futuristic-fashion/. Laird, K 2010 ‘Marketer News Hitting The Mark If Women Hold So Much Buying
Power, Why Aren’t Marketers Doing A Better Job Of Communicating To Them?’
Marketing Magazine. Levy, S 2017 ‘Google Glass 2.0 Is a Startling Second Act’. https://www.wired.com/
story/google-glass-2-is-here/ (accessed 29/08/2017). Mann, S 2015 Phenomenal Augmented Reality: Advancing technology for the future
of humanity. IEEE Consumer Electronics Magazine, 4(4): 92–97. DOI: https://doi. org/10.1109/MCE.2015.2463312 Maragiannis 2016 Can non-anthropocentric relationships lead to true intimacy with
technology? Proceedings: full papers. ed./O. Tapio Leiono ISEA 2016: Hong Kong. Medaglia, C M and Serbanati, A 2010 An Overview of Privacy and Security Issues
in the Internet of Things In: Giusto, D, et al. References (eds.), The Internet of Things: 20th Maragiannis and Ashford: Diversity and Inclusivity in the Age of Wearables 213 Thirrhenian Workshop on Digital Communications, 389–395. New York: Springer. DOI: https://doi.org/10.1007/978-1-4419-1674-7_38 Profita, H P, Clawson, J, Gilliland, S, Zeagler, C, Starner, T, Budd, J and Do, E Y-L
2013 Don’t mind me touching my wrist: a case study of interacting with on-body
technology in public. In: ‘Proceedings of the 2013 International Symposium on
Wearable Computers’. 89–96. Zurich, Switzerland: ACM Press. DOI: https://doi. org/10.1145/2493988.2494331 Ryan, S E 2014 Garments of paradise: wearable discourse in the digital age. MIT
Press. DOI: https://doi.org/10.7551/mitpress/8873.001.0001 Sanders, J 1997 Women in technology: Attribution, learned helplessness,
self-esteem, and achievement. Paper presented at the Conference on Women, Girls
and Technology. Tarrytown, New York. Sanders, J, Koch, J and Urso, J 1997 Gender equity right from the start: Instructional
activities for teacher educators in mathematics, science and technology. Mahwah:
Lawrence Erlbaum Associates. Saner, E 2014 “Viktoria Modesta, the world’s first amputee pop star: ‘if you don’t fit
in, then don’t fit in’”. https://www.theguardian.com/music/2014/dec/20/-sp-
amputee-pop-star-viktoria-modesta (accessed 11/09/2018). Schroeder, K 2010 Gender dimensions of product design. In: Gender, Science,
and Technology Expert Group meeting of the United Nations Division for the
Advancement of Women (UN-DAW), 28: 93–200. Starner, T, Mann, S, Rhodes, B, Healey, J, Russell, K B, Levine, J and Pentland,
A 1995 Wearable Computing and Augmented Reality. M.I.T. Media Lab Vision
and Modeling Group Technical Report No. 355, Nov. 1995 Submitted to Presence
special issue on Augmented Reality, Nov. 1995. Wissinger, E 2017 Wearable tech, bodies, and gender. Sociology Compass, 11: e12514. (accessed 13/10/18). DOI: https://doi.org/10.1111/soc4.12514 Witts, S 1997 Parental influence on children’s socialization to gender roles. Adolescence, 32(126): 253–259. 214 Maragiannis and Ashford: Diversity and Inclusivity in the Age of Wearables How to cite this article: Maragiannis, A and Ashford, R 2019 Diversity and Inclusivity
in the Age of Wearables: A Buzzword, a Myth, an Uncertain Reality. Body, Space &
Technology, 18(1), pp. 198–214. DOI: https://doi.org/10.16995/bst.320
Submitted: 29 October 2018 Accepted: 22 January 2019 Published: 12 March 2019
Copyright: © 2019 The Author(s). This is an open-access article distributed under the
terms of the Creative Commons Attribution 4.0 International License (CC-BY 4.0), which
permits unrestricted use, distribution, and reproduction in any medium, provided the
original author and source are credited. See http://creativecommons.org/licenses/by/4.0/. OPEN ACCESS
Body, Space & Technology is a peer-reviewed open
access journal published by Open Library of Humanities. Maragiannis and Ashford: Diversity and Inclusivity in the Age of Wearables How to cite this article: Maragiannis, A and Ashford, R 2019 Diversity and Inclusivity
in the Age of Wearables: A Buzzword, a Myth, an Uncertain Reality. Body, Space &
Technology, 18(1), pp. 198–214. DOI: https://doi.org/10.16995/bst.320
Submitted: 29 October 2018 Accepted: 22 January 2019 Published: 12 March 2019
Copyright: © 2019 The Author(s). This is an open-access article distributed under the
terms of the Creative Commons Attribution 4.0 International License (CC-BY 4.0), which
permits unrestricted use, distribution, and reproduction in any medium, provided the
original author and source are credited. See http://creativecommons.org/licenses/by/4.0/.
OPEN ACCESS
Body, Space & Technology is a peer-reviewed open
access journal published by Open Library of Humanities. References How to cite this article: Maragiannis, A and Ashford, R 2019 Diversity and Inclusivity
in the Age of Wearables: A Buzzword, a Myth, an Uncertain Reality. Body, Space &
Technology, 18(1), pp. 198–214. DOI: https://doi.org/10.16995/bst.320 Submitted: 29 October 2018 Accepted: 22 January 2019 Published: 12 March 2019 Submitted: 29 October 2018 Accepted: 22 January 2019 Published: 12 March 2019 Copyright: © 2019 The Author(s). This is an open-access article distributed under the
terms of the Creative Commons Attribution 4.0 International License (CC-BY 4.0), which
permits unrestricted use, distribution, and reproduction in any medium, provided the
original author and source are credited. See http://creativecommons.org/licenses/by/4.0/. OPEN ACCESS
Body, Space & Technology is a peer-reviewed open
access journal published by Open Library of Humanities.
|
https://openalex.org/W3166627195
|
https://research.rug.nl/files/174703448/The_erythropoietin_receptor_expressed_in_skeletal_muscle_is_essential_for_mitochondrial_biogenesis_and_physiological_exercise.pdf
|
English
| null |
Successful rapid desensitization in a glioma patient with procarbazine‐mediated anaphylaxis
|
Allergy
| 2,021
|
cc-by
| 10,773
|
University of Groningen
Successful rapid desensitization in a glioma patient with procarbazine-mediated anaphylaxis
Hutten, Evelien M.; Lambeck, Annechien J. A.; Dijkstra, Hilda; Nuver, Janine; Oude Elberink,
Hanneke N. G.; van de Ven, Annick A. J. M.
Published in:
Allergy University of Groningen
Successful rapid desensitization in a glioma patient with procarbazine-mediated anaphylaxis
Hutten, Evelien M.; Lambeck, Annechien J. A.; Dijkstra, Hilda; Nuver, Janine; Oude Elberink,
Hanneke N. G.; van de Ven, Annick A. J. M. Published in:
Allergy University of Groningen University of Groningen Successful rapid desensitization in a glioma patient with procarbazine-mediated anaphylaxis
Hutten, Evelien M.; Lambeck, Annechien J. A.; Dijkstra, Hilda; Nuver, Janine; Oude Elberink,
Hanneke N. G.; van de Ven, Annick A. J. M. P bli h d i Successful rapid desensitization in a glioma patient with procarbazine-mediated anaphylaxis
Hutten, Evelien M.; Lambeck, Annechien J. A.; Dijkstra, Hilda; Nuver, Janine; Oude Elberink,
Hanneke N. G.; van de Ven, Annick A. J. M. DOI:
10.1111/all.14698 DOI:
10.1111/all.14698 IMPORTANT NOTE: You are advised to consult the publisher's version (publisher's PDF) if you wish to cite from
it. Please check the document version below. Document Version
Publisher's PDF, also known as Version of record IMPORTANT NOTE: You are advised to consult the publisher's version (publisher's PDF) if you wish to cite from
it. Please check the document version below. D
t V
i Document Version
Publisher's PDF, also known as Version of record Publication date:
2021 Link to publication in University of Groningen/UMCG research database Citation for published version (APA):
Hutten, E. M., Lambeck, A. J. A., Dijkstra, H., Nuver, J., Oude Elberink, H. N. G., & van de Ven, A. A. J. M. (2021). Successful rapid desensitization in a glioma patient with procarbazine-mediated anaphylaxis. Allergy, 76(6), 1932-1933. https://doi.org/10.1111/all.14698 Copyright
Other than for strictly personal use, it is not permitted to download or to forward/distribute the text or part of it without the consent of the
author(s) and/or copyright holder(s), unless the work is under an open content license (like Creative Commons). Citation for published version (APA):
Hutten, E. M., Lambeck, A. J. A., Dijkstra, H., Nuver, J., Oude Elberink, H. N. G., & van de Ven, A. A. J. M.
(2021). Successful rapid desensitization in a glioma patient with procarbazine-mediated anaphylaxis.
Allergy, 76(6), 1932-1933. https://doi.org/10.1111/all.14698 Abstract Erythropoietin (EPO) is a haematopoietic hormone that regulates erythropoiesis, but the EPO-receptor (EpoR) is also
expressed in non-haematopoietic tissues. Stimulation of the EpoR in cardiac and skeletal muscle provides protection from
various forms of pathological stress, but its relevance for normal muscle physiology remains unclear. We aimed to deter-
mine the contribution of the tissue-specific EpoR to exercise-induced remodelling of cardiac and skeletal muscle. Baseline
phenotyping was performed on left ventricle and m. gastrocnemius of mice that only express the EpoR in haematopoietic
tissues (EpoR-tKO). Subsequently, mice were caged in the presence or absence of a running wheel for 4 weeks and exercise
performance, cardiac function and histological and molecular markers for physiological adaptation were assessed. While
gross morphology of both muscles was normal in EpoR-tKO mice, mitochondrial content in skeletal muscle was decreased
by 50%, associated with similar reductions in mitochondrial biogenesis, while mitophagy was unaltered. When subjected
to exercise, EpoR-tKO mice ran slower and covered less distance than wild-type (WT) mice (5.5 ± 0.6 vs. 8.0 ± 0.4 km/day,
p < 0.01). The impaired exercise performance was paralleled by reductions in myocyte growth and angiogenesis in both
muscle types. Our findings indicate that the endogenous EPO-EpoR system controls mitochondrial biogenesis in skeletal
muscle. The reductions in mitochondrial content were associated with reduced exercise capacity in response to voluntary
exercise, supporting a critical role for the extra-haematopoietic EpoR in exercise performance. Keywords Erythropoietin receptor · Cardiac and skeletal muscle · Mitochondrial biogenesis · Exercise performance ·
Exercise-induced physiological adaptation Keywords Erythropoietin receptor · Cardiac and skeletal muscle · Mitochondrial biogenesis · Exercise performance ·
Exercise induced physiological adaptation Exercise-induced physiological adaptation * B. Daan Westenbrink
b.d.westenbrink@umcg.nl The erythropoietin receptor expressed in skeletal muscle is essential
for mitochondrial biogenesis and physiological exercise Kirsten T. Nijholt1 · Laura M. G. Meems1 · Willem P. T. Ruifrok1 · Alexander H. Maass1 · Salva R. Yurista1 ·
Mario G. Pavez‑Giani1 · Belend Mahmoud1 · Anouk H. G. Wolters2 · Dirk J. van Veldhuisen1 · Wiek H. van Gilst1 ·
Herman H. W. Silljé1 · Rudolf A. de Boer1 · B. Daan Westenbrink1 Kirsten T. Nijholt1 · Laura M. G. Meems1 · Willem P. T. Ruifrok1 · Alexander H. Maass1 · Salva R. Yurista1 ·
Mario G. Pavez‑Giani1 · Belend Mahmoud1 · Anouk H. G. Wolters2 · Dirk J. van Veldhuisen1 · Wiek H. van Gilst1 ·
Herman H. W. Silljé1 · Rudolf A. de Boer1 · B. Daan Westenbrink1 Received: 1 February 2021 / Revised: 16 April 2021 / Accepted: 5 May 2021
© The Author(s) 2021 Introduction Erythropoietin (EPO) is a hormone that is secreted by the
adult kidneys and the foetal liver and regulates erythropoie-
sis in the haematopoietic tissues [46, 61]. EPO stimulates
erythropoiesis by binding to a homodimeric EPO-receptor
(EpoR) on erythroid progenitor cells, which promotes their
survival, proliferation and differentiation into erythrocytes
[46, 61]. In certain pathological conditions such as chronic
renal failure, EPO production becomes insufficient and
anaemia ensues [46, 54, 61]. Treatment with recombinant
human EPO (rhEPO) does not only increase haematocrit in
these patients, but also improves quality of life and promotes
exercise capacity [9, 31]. rhEPO is also used to enhance
exercise performance by various endurance athletes without
anaemia [3, 4]. Copyright strictly personal use, it is not permitted to download or to forward/distribute the text or part of it without the consent of th
or copyright holder(s), unless the work is under an open content license (like Creative Commons). The publication may also be distributed here under the terms of Article 25fa of the Dutch Copyright Act, indicated by the “Taverne” license. More information can be found on the University of Groningen website: https://www.rug.nl/library/open-access/self-archiving-pure/taverne-
amendment. Take-down policy
If you believe that this document breaches copyright please contact us providing details, and we will remove access to the work immediately
and investigate your claim. Take-down policy
If you believe that this document breaches copyright please contact us providing details, and we will remove access to the work immediately
and investigate your claim. Take-down policy
If you believe that this document breaches copyright please contact us providing details, and we will remove access to
and investigate your claim. Downloaded from the University of Groningen/UMCG research database (Pure): http://www.rug.nl/research/portal. For technical reasons the
number of authors shown on this cover page is limited to 10 maximum. Downloaded from the University of Groningen/UMCG research database (Pure): http://www.rug.nl/research/portal. For technical reasons the
number of authors shown on this cover page is limited to 10 maximum. Download date: 24-10-2024 Pflügers Archiv - European Journal of Physiology
https://doi.org/10.1007/s00424-021-02577-4 MUSCLE PHYSIOLOGY 2
Department of Cell Biology, University Medical Centre
Groningen, University of Groningen, Groningen,
The Netherlands Key points y p
• Ablation of the erythropoietin receptor (EpoR) in skeletal
muscle results in defective mitochondrial biogenesis. • EpoR deficiency in cardiac and skeletal muscle reduces exercise
performance.i p
• EpoR deficiency disrupts physiological myocyte hypertrophy
and adaptive angiogenesis in cardiac and skeletal muscle. 1
Department of Cardiology, University Medical Centre
Groningen, University of Groningen, HPC AB31, 9700 RB,
P.O. Box 30.001 Groningen, The Netherlands 2
Department of Cell Biology, University Medical Centre
Groningen, University of Groningen, Groningen,
The Netherlands (0123
1 23456789)
3 Pflügers Archiv - European Journal of Physiology The mechanisms underlying the effects of rhEPO on exer-
cise are generally thought to result from the associated increase
in haematocrit, under the assumption that this augments oxy-
gen delivery to aerobic tissues [4, 25–27]. Although this may
partially explain the beneficial effects in patients with severe
anaemia, extra-haematopoietic effects may also contribute to
the improvements in quality of life and exercise performance
[14, 25]. For instance, studies have shown that rhEPO activates
mitochondrial biogenesis and subsequently promotes exer-
cise capacity, prior to an elevation in haematocrit [7]. Also,
enhancement of mitochondrial function was observed in skel-
etal muscle fibres of athletes after rhEPO administration [39]. Indeed, most of the effects of rhEPO on exercise performance
in endurance athletes are not exclusively explained by changes
in haematocrit and the accessory effects could therefore result
from non-haematopoietic effects [2]. all animal experiments followed the protocols from Direc-
tive 2010/63/EU of the European Parliament. Mouse model We employed the well-described EpoR-/-rescued mouse
model in which EpoR-tissue knock-out mice have been
rescued from lethality by a transgene that drives EpoR
expression exclusively in haematopoietic tissues [47]. These mice (EpoR-tissue knock-out; EpoR-tKO) develop
normally and are fertile without a clear (cardiac) pheno-
type, despite the absence of the EPO receptor in all non-
haematopoietic tissues (Fig. 1) [47]. EpoR-tKO mice origi-
nated from the Riken BioResource Research Center (Riken
BRC) repository in Japan. Thereafter, the EpoR-tKO mice
were bred and housed at the central animal facility of the
University of Groningen and back-crossed into our local
C57Bl/6 J genetic background. Age and gender-matched
wild-type (WT) mice were used as controls. Genotyping
of EpoR-tKO mice was performed as described previously
[47]. f
There are several lines of evidence supporting the con-
cept that EPO augments exercise performance by activating
an EpoR subtype in non-haematopoietic tissues [36]. This
EpoR has been discovered in a variety of non-haematopoi-
etic tissues, including cardiac and skeletal muscle, endothe-
lial cells, neurones and kidneys [34, 36, 46]. Evidence for
its role in tissue physiology is based upon studies where
rhEPO protected these tissues from various forms of patho-
logical stress [10, 12, 24, 55, 56], even when administered in
doses that are too low to stimulate erythropoiesis [23]. More
direct evidence for the significance of the extra-haematopoi-
etic EPO-EpoR system comes from studies with a mouse
model that only expresses the EpoR in haematopoietic cells,
but lack EpoR expression in all other tissues (EpoR-tissue
knock-out, EpoR-tKO mice) [47]. While these EpoR-tKO
mice display a normal phenotype with preserved erythro-
poiesis, their non-haematopoietic tissues are more vulner-
able to pathological stress [10, 50]. For instance, cardiac and
skeletal muscle of EpoR-tKO mice are more susceptible to
ischemic or toxic injury and also develop more severe heart
failure when subjected to pressure overload [16, 33, 48]. Exercise performance A cyclometer was attached to the running wheel, which
measured daily exercise parameters, including speed, dis-
tance and time on the running wheel, as described previ-
ously [1]. Experimental model At 8–12 weeks of age, WT and EpoR-tKO male mice
were randomised to the presence or absence of a running
wheel present in their individual cage. The voluntary run-
ning wheel (VWR) was present for the whole duration of
the experiment, which was a period of 4 weeks (referred to
as WT Run or EpoR-tKO Run). Control littermates were
housed similarly but without a running wheel and therefore
remained sedentary for this period (referred to as WT Sed
or EpoR-tKO Sed). An additional group of male mice was
selected for phenotyping of basal mitochondrial parameters. In total, 89 male mice were used for this study. All mice had
ad libitum access to food and water. While this evidence indicates that the EpoR expressed in
muscle is critical for the adaptation to pathological stress,
its relevance for muscle physiology is not well described. We hypothesised that the EpoR expressed in cardiac and
skeletal muscle is critical for the physiological adaptation
to exercise. To test this hypothesis, we subjected EpoR-tKO
mice to a model of voluntary wheel running. Ethical approval Following completion of the 4-week experiment and
2 days prior to sacrifice, cardiac function was deter-
mined by echocardiography (GE Vivid 7 Dimension,
using a 13-MHz probe) as described before [29]. The experimental protocol was accepted by The Ani-
mal Ethical Committee of the University of Groningen
(DEC4585, IvD 1583–02-011, 1583–02-014). In addition, 3 Pflügers Archiv - European Journal of Physiology a
b
d
e
c
Fig. 1 Baseline muscle phenotype of tissue-specific erythropoietin
receptor deficient mice. Phenotyping of tissue-specific erythropoi-
etin receptor deficient (EpoR-tKO) mice. Shown are (a) conceptual
illustration of the mouse model, in which the erythropoietin receptor
(EpoR) is absent in non-haematopoietic tissues of EpoR-tKO mice,
(b) left ventricular (LV) mRNA expression of the EpoR, (n = 2–5/
group), (c) heart weight to body weight ratio in milligrams per
gram (mg/g), (d) cardiomyocyte cross-sectional area in micrometres
squared (μm2) and typical examples, and (e) skeletal myocyte cross-
sectional area (μm2) including typical examples, (n = 11–13/group). WT wild-type mice, EpoR-tKO EpoR-tissue knock-out mice. Data
are presented as mean value ± standard error of the mean (SEM). The
illustration in panel a contains images from Servier Medical Art by
Servier, licensed under a Creative Commons Attribution 3.0 unported
license b
c c a b c a d d e e
squared (μm2) and typical examples and (e) skeletal myocyte cross- e Fig. 1 Baseline muscle phenotype of tissue-specific erythropoietin
receptor deficient mice. Phenotyping of tissue-specific erythropoi-
etin receptor deficient (EpoR-tKO) mice. Shown are (a) conceptual
illustration of the mouse model, in which the erythropoietin receptor
(EpoR) is absent in non-haematopoietic tissues of EpoR-tKO mice,
(b) left ventricular (LV) mRNA expression of the EpoR, (n = 2–5/
group), (c) heart weight to body weight ratio in milligrams per
gram (mg/g), (d) cardiomyocyte cross-sectional area in micrometres squared (μm2) and typical examples, and (e) skeletal myocyte cross-
sectional area (μm2) including typical examples, (n = 11–13/group). WT wild-type mice, EpoR-tKO EpoR-tissue knock-out mice. Data
are presented as mean value ± standard error of the mean (SEM). The
illustration in panel a contains images from Servier Medical Art by
Servier, licensed under a Creative Commons Attribution 3.0 unported
license skeletal muscle were fixated for immunohistochemistry or
snap frozen for molecular analysis. Subsequently, on the day of sacrifice, haemodynamic
parameters were determined with a microtip pres-
sure–volume transducer as described previously (Millar
Instr. Inc., Houston, TX, USA) [37]. Histology After isolation of LV and m. gastrocnemius, cardiac muscle
was cut at the mid-papillary level, and skeletal muscle was
cut at the mid-belly level. Sections were either fixated in 4%
paraformaldehyde and embedded in paraffin after dehydra-
tion steps (Leica TP1020, Germany) or fixated in Tissue-Tec
and subsequently slowly frozen in liquid nitrogen. Following
fixation and embedding processes, sections were cut using a
microtome and were thereafter processed onto slides. Ethical approval All cardiac function
assessments were performed under general anaesthesia
(isoflurane 2%). Capillary density To measure capillary density as a measure of adaptive angio-
genesis, endothelial cells were incubated with the endothe-
lial cell antigen CD31 (purified rat anti-mouse CD31 anti-
body #550,274, BD Biosciences, USA). Prior to incubation
of antibodies, cryosections were fixated in acetone, air-dried
and blocked for endogenous peroxidases. After primary,
secondary (unconjugated rabbit anti-rat IgG antibody, #AI-
4001, Vector Laboratories, USA) and tertiary (anti-rabbit
polymer-HRP, #K4008, EnVision kit, Denmark) antibody
incubation, AEC + solution was applied and counterstained
with Mayer’s hematoxylin. Capillaries were quantified from
at least eight randomly chosen fields at × 40 magnification
(Zeiss KS400, Germany), and adaptive angiogenesis was
presented as capillary to (cardio)myocyte ratio [5]. (Cardio) Myocyte size To determine (cardio) myocyte cross sectional area, depar-
affinised 3-μm thick sections were stained with 4′,6-diami-
dino-2-phenylindole (DAPI) (Vector Laboratories, USA)
and wheat germ agglutinin (WGA) (Sigma-Aldrich, USA),
as described before [60]. Stained transverse sections were
visualised with fluorescence microscopy to generate images
for analysis (Zeiss KS400, Germany). Quantification fields
were chosen at random at × 40 magnification; five fields per
sample were counted and averaged. Organ, body and blood measurements After haemodynamic measurements, blood was drawn and
hearts plus skeletal muscle were collected. After weighing
the hearts, the left ventricle (LV) and the gastrocnemius 1 3 3 Pflügers Archiv - European Journal of Physiology mitochondrial quantity. Analysis of images was performed
in ImageJ (Fiji- ImageJ version of Java 6, USA). Quantitative real‑time polymerase chain reaction To assess molecular analysis at mRNA level, qRT-PCR was
performed as described before [58]. Total RNA from LV
and m. gastrocnemius was isolated using the Trizol isola-
tion method (Invitrogen Corporation, USA). Quantification
of RNA concentrations was performed using Nanodrop
(ThermoFisher, USA). cDNA synthesis was performed with
equal RNA concentrations (500 ng of total RNA) using the
Quantitect Reverse Transcription kit (Qiagen, the Nether-
lands). qRT-PCR was performed using SYBR Green dye
super mix and the running protocol included 3 min at 95 °C,
followed by 35 cycles of 15 s at 95 °C, 30 s at 60 °C, fol-
lowed by dissociation and melting steps (Bio-Rad CFX384,
USA). Data were processed according to the ddCT method,
and were normalised for housekeeping gene 36B4 and con-
trol group WT sedentary. The Online Resource with Sup-
plementary Table S1 contains the primer sequences. Mitochondrial DNA to nuclear DNA ratio (mitochondrial
DNA copy number) To quantify basal mitochondrial DNA copy number as a meas-
ure for mitochondrial biogenesis, qRT-PCR was performed
for genes encoding mitochondrial proteins. DNA was isolated
from LV and m. gastrocnemius using the Nucleospin Tissue
XS kit (Macherey–Nagel, Germany); DNA concentrations
were quantified using Nanodrop (Thermofisher, USA). A total
of 10 μg DNA was used for qRT-PCR, using the standard run-
ning protocol as described above. To determine mitochondrial
DNA copy number, nuclear DNA-encoded mitochondrial pro-
tein hexokinase 2 (HK2) and mitochondrial DNA (mtDNA)-
encoded protein NADH dehydrogenase (ND1) were detected
[41]. mtDNA copy number was calculated using the ddCT
method. HK2 served as reference gene and the WT sedentary
group served as control. Supplementary Table S1 contains the
primer sequences. Electron microscopy To assess basal mitochondrial quantity, large-scale electron
microscopy (EM), known as nanotomy (for nano-anatomy),
was performed. Samples from skeletal muscle were cut
into pieces of approximately 1 mm2 and were fixated in
2% glutaraldehyde and 2% paraformaldehyde solution in
0.1 M sodium cacodylate directly after sacrifice. Following
fixation, samples were further processed with 1% osmium
tetroxide and 1.5% potassium ferrocyanide and, finally, after
dehydration embedded in EPON epoxy resin. Sections of
80 nm were cut using an ultramicrotome, which was fol-
lowed by contrast staining with 4% neodymium acetate for
30 min [19]. Finally, sections were observed using scan-
ning and transmission electron microscopy (STEM) (Zeiss
Supra55, Oberkochen Germany), as described before
[15, 18, 45]. This microscope was connected to an exter-
nal scan generator ATLAS5 (Fibics, Canada) to produce
many high-resolution images that are stitched to generate
a large Nanotomy ‘map’. TIFF files were converted to html
links and uploaded to the website www.nanotomy.org. Of
each nanotomy file, at least six high-resolution images,
selected randomly in the map, were saved for analysis of Western blot To detect molecular changes at protein level, Western blot
was performed [59], specifically to assess levels of the key
regulators of mitochondrial biogenesis and mitophagy; nuclear
respiratory factor 2 (NRF2), outer mitochondrial membrane
protein TOM20 (TOM20), ubiquitin protein 62 (p62) and light
chain 3B (LC3B). Total protein was extracted with RIPA buffer
and fresh sodium vanadate, phosphatases and proteases buff-
ers (Sigma-Aldrich, USA). Quantification of protein concen-
trations was performed using the BCA protein assay (Pierce. No. 232250, ThermoFisher, USA). Equal concentrations of 1 3 3 Pflügers Archiv - European Journal of Physiology protein were boiled and loaded on SDS-PAGE. Semi-dry blot-
ting transferred the proteins onto PVDF membranes. Mem-
branes were incubated with the following primary antibodies:
polyclonal anti-NRF2 (ab31163, Abcam, UK), monoclonal
anti-TOM20 (#42,406, Cell Signalling, USA), monoclonal
anti-sqstm1/p62 antibody (ab56416, Abcam, UK), mono-
clonal anti-LC3B antibody (#2775, Cell Signalling, USA)
and monoclonal anti-GAPDH antibody as a loading control
(10R-G109A, Fitzgerald, USA). This was followed by sec-
ondary antibody incubation, and proteins were detected with
enhanced chemiluminescence (NEL120E001EA, Western
lightening ECL Pro, Perkin Elmer, USA) using ImageQuant
imager. Western blots were quantified in ImageJ and values
were normalised for loading control GAPDH expressions as
well as for the control WT sedentary group. protein were boiled and loaded on SDS-PAGE. Semi-dry blot-
ting transferred the proteins onto PVDF membranes. Mem-
branes were incubated with the following primary antibodies:
polyclonal anti-NRF2 (ab31163, Abcam, UK), monoclonal
anti-TOM20 (#42,406, Cell Signalling, USA), monoclonal
anti-sqstm1/p62 antibody (ab56416, Abcam, UK), mono-
clonal anti-LC3B antibody (#2775, Cell Signalling, USA)
and monoclonal anti-GAPDH antibody as a loading control
(10R-G109A, Fitzgerald, USA). This was followed by sec-
ondary antibody incubation, and proteins were detected with
enhanced chemiluminescence (NEL120E001EA, Western
lightening ECL Pro, Perkin Elmer, USA) using ImageQuant
imager. Western blots were quantified in ImageJ and values
were normalised for loading control GAPDH expressions as
well as for the control WT sedentary group. Basal mitochondrial biogenesis is disrupted
in skeletal but not in cardiac muscle of EpoR‑tKO
mice The generation of new mitochondria, and thereby increas-
ing the mitochondrial content, is known as a process
called mitochondrial biogenesis and is regulated by mul-
tiple signalling pathways and transcription factors [11,
40]. Interestingly, rhEPO has been previously shown to
enhance exercise capacity and stimulate mitochondrial
biogenesis through regulation of markers including eNOS
and PGC-1α [7]. In our model, mitochondrial content
expressed as the mitochondrial to nuclear DNA ratio was
unaltered in cardiac muscle as was the mRNA expres-
sion of molecular markers for mitochondrial biogenesis
(Fig. 2a–b). In skeletal muscle, however, mitochondrial
DNA content was reduced by 50% in EpoR-tKO mice com-
pared to WT mice (Fig. 2c). Similarly, protein levels of the
outer mitochondrial membrane protein (TOM20) were also
significantly lower in skeletal muscle of EpoR-tKO mice
(Fig. 2e–f, Supplementary Fig. S2). These reductions in
mitochondrial content were associated with similar reduc-
tions in the expression of key regulators of mitochondrial
biogenesis, eNOS, SIRT1, PGC-1α and NRF2 (Fig. 2d). In addition, protein levels of NRF2, a key regulator of
mitochondrial biogenesis, were also decreased in EpoR
deficient mice (Fig. 2e–f, Supplementary Fig. S2). To cor-
roborate whether the differences in mitochondrial DNA
content and TOM20 protein levels paralleled with similar
reductions in the actual number of mitochondria, we next
performed electron microscopy (EM) of m. gastrocnemius
muscle. Strikingly, quantification of the number of mito-
chondria in our EM datasets revealed that the number of
interfibrillar mitochondria was reduced in skeletal mus-
cle of EpoR-tKO mice (Fig. 2g, available for complete
datasets). This was accompanied by a trend towards a
decrease in mRNA expression of oxidative muscle fibre
types, which may also reflect a reduction in mitochondrial
abundance (Supplementary Fig. S3a) [40]. Statistical analyses Results are presented as mean value ± standard error of the
mean (SEM). Statistical comparisons between two groups
with normally distributed data were performed with a stu-
dent’s t test; if data was not normally distributed, the non-
parametric Mann Whitney U test was performed. Statistical
comparisons among more than two groups were performed by
one-way ANOVA with Bonferroni post hoc test, if distributed
normally. Otherwise, the non-parametric Kruskal–Wallis test
was used followed by the Mann–Whitney U test. Statistical
significance was stated if the p value was < 0.05, of which all
p values were two-tailed. Statistical analysis was performed in
STATA v11SE software (College Station, TX) and Graph Pad
Prism software (Version 7, USA). Phenotyping of basal cardiac and skeletal muscle
in tissue‑specific erythropoietin receptor‑deficient
(EpoR‑tKO) mice A conceptual illustration of the mouse model is presented
in Fig. 1a. In accordance with this concept, EpoR mRNA
expression was undetectable in cardiac muscle from
EpoR-tKO mice (Fig. 1b), while haematocrit levels were
unaltered, and thus, erythropoiesis was preserved (Sup-
plementary Fig. S1). Cardiac mass and cardiomyocyte
cross-sectional area were comparable between EpoR-tKO
and WT mice (Fig. 1c–d), as was the mRNA expression
of natriuretic peptides and muscle isoforms (Supplemen-
tary Fig. S1). Skeletal myocyte size in m. gastrocnemius
was also comparable between EpoR-tKO and WT mice
(Fig. 1e). Next, we aimed to determine whether the reductions in
mitochondrial content and biogenesis were also associ-
ated with enhanced elimination of mitochondria, through a
process called mitophagy [11]. We therefore assessed p62,
LC3B I and LC3B II at protein level, which are markers
for mitophagy. We did not observe significant differences
in protein levels of mitophagy markers in skeletal muscle
of EpoR-tKO mice compared to WT mice (Supplementary
Fig. S4a–b). Finally, to assess whether mitochondrial per-
formance was altered in EpoR-tKO mice, oxygen consump-
tion in isolated mitochondria was assessed. Surprisingly, the
changes in mitochondrial content were not accompanied by
changes in mitochondrial respiratory capacity (Supplemen-
tary Fig. S3b) [57]. Together, these data indicate that the 1 Pflügers Archiv - European Journal of Physiology a
b
c
d
e
f
g
Fig. 2 Mitochondrial biogenesis in basal cardiac and skeletal muscle. Cardiac and skeletal muscle mitochondrial biogenesis parameters at
baseline are depicted (n = 11–13/group). Shown are (a) left ventric-
ular (LV) mitochondrial to nuclear DNA ratio (mtDNA/nDNA), (b)
LV mRNA expression of markers for mitochondrial biogenesis, (c)
skeletal muscle (SM) mitochondrial to nuclear DNA ratio (mtDNA/
nDNA), (d) SM mRNA expression of markers for mitochondrial
biogenesis, (e) typical examples of SM protein levels of NRF2 and
TOM20 (n = 3/group), (f) quantification of full Western blots of
NRF2 and TOM20 (n = 4–5/group), and (g) typical electron micros-
copy examples of m. gastrocnemius (white arrows indicating mito-
chondria) and quantification of mitochondrial (mito) number per field
of 55μm2; six high power fields were analysed per animal, (n = 2/
group). eNOS endothelial nitric oxide synthase, SIRT1 sirtuin 1,
PGC-1α peroxisome proliferator-activated receptor gamma coactiva-
tor 1-alpha, NRF2 nuclear respiratory factor 2, TOM20 outer mito-
chondrial membrane protein. WT wild-type mice, EpoR-tKO EpoR-
tissue knock-out mice. Data are presented as mean value ± standard
error of the mean (SEM). Phenotyping of basal cardiac and skeletal muscle
in tissue‑specific erythropoietin receptor‑deficient
(EpoR‑tKO) mice WT vs. EpoR-tKO: *p < 0.05, **p < 0.01. Raw data with zoomable files at high resolution are accessible at
www.nanotomy.org a
b
g a b
g a
b
c
d
e
f
g g a g c
e d
f c d c f f e e Fig. 2 Mitochondrial biogenesis in basal cardiac and skeletal muscle. Cardiac and skeletal muscle mitochondrial biogenesis parameters at
baseline are depicted (n = 11–13/group). Shown are (a) left ventric-
ular (LV) mitochondrial to nuclear DNA ratio (mtDNA/nDNA), (b)
LV mRNA expression of markers for mitochondrial biogenesis, (c)
skeletal muscle (SM) mitochondrial to nuclear DNA ratio (mtDNA/
nDNA), (d) SM mRNA expression of markers for mitochondrial
biogenesis, (e) typical examples of SM protein levels of NRF2 and
TOM20 (n = 3/group), (f) quantification of full Western blots of
NRF2 and TOM20 (n = 4–5/group), and (g) typical electron micros-
copy examples of m. gastrocnemius (white arrows indicating mito- chondria) and quantification of mitochondrial (mito) number per field
of 55μm2; six high power fields were analysed per animal, (n = 2/
group). eNOS endothelial nitric oxide synthase, SIRT1 sirtuin 1,
PGC-1α peroxisome proliferator-activated receptor gamma coactiva-
tor 1-alpha, NRF2 nuclear respiratory factor 2, TOM20 outer mito-
chondrial membrane protein. WT wild-type mice, EpoR-tKO EpoR-
tissue knock-out mice. Data are presented as mean value ± standard
error of the mean (SEM). WT vs. EpoR-tKO: *p < 0.05, **p < 0.01. Raw data with zoomable files at high resolution are accessible at
www.nanotomy.org EpoR expressed in skeletal muscle regulates mitochondrial
biogenesis, without affecting mitochondrial oxidative capac-
ity or mitophagy. wheel in their cage (Fig. 3a). With preserved erythropoiesis
(Supplementary Table S2), EpoR-tKO running mice had a
lower average speed compared to WT mice, spent less time
on the running wheel and ran significantly fewer daily and
kilometres (Fig. 3b–d). Consequently, mice in the EpoR-tKO
Run group covered less total distance compared to mice in
the WT Run group (Fig. 3e). Exercise performance is impaired in EpoR‑tKO mice
despite normal haematocrit levels We next sought to determine whether the reductions in mito-
chondrial content in skeletal muscle affected exercise per-
formance. For this purpose, WT and EpoR-tKO mice were
housed individually in the presence or absence of a running 1 3 Pflügers Archiv - European Journal of Physiology Cardiac function To determine if loss of EpoR signalling in the cardiac mus-
cle affected the adaptive changes in cardiac performance
induced by exercise, intracardiac pressure measurements
and echocardiography were performed. Voluntary exercise
resulted in a significant improvement in both contractility
and relaxation of the left ventricle (dP/dt max, dP/dt min,
respectively), as well as a comparable effect in fractional
shortening among both groups (Supplementary Table S3). However, left ventricular end-diastolic pressure (LVEDP),
mean arterial pressure (MAP) and systolic blood pressure
(SBP) were considerably increased in EpoR-tKO mice com-
pared to WT mice after exercise (Fig. 4a, Supplementary
Table S3). Analogous to pathological stress, the cardiac physiologi-
cal response to sustained exercise includes increasing car-
diomyocyte size [44]. Exercise resulted in a 26% increase
in cardiomyocyte cross-sectional area (CSA) in WT mice
(Fig. 4c–d). In contrast, exercise did not result in a consider-
able increase in CSA in EpoR-tKO mice (Fig. 4c–d). Simi-
larly, the increase in the mRNA expression of alpha-actinin 1
(ACTA-1), a molecular marker of muscle growth, was more
pronounced in WT Run than in the EpoR-tKO Run group
(Supplementary Table S4). Furthermore, atrial natriuretic
peptide (ANP) expression was increased and a shift towards
β-myosin heavy chain (β-MHC) isoforms was observed in a
b a
c
d
e
b
Fig. 3 Exercise performance in wild-type and EpoR-tissue knock-
out mice. Exercise parameters during 4-week period of voluntary
wheel running (n = 11–12/group). Shown are (a) experimental design
including voluntary wheel running protocol, (b) average daily run-
ning speed in kilometres per hour (km/h), (c) average running time
in minutes per day (min/day), (d) average daily running distance in
kilometres per day (km/day) and (e) total running distance in kilo-
metres (km) that mice achieved in the 4-week voluntary wheel run-
ning experiment. WT wild-type mice, EpoR-tKO EpoR-tissue knock-
out mice, Run running. Data are presented as mean value ± standard
error of the mean (SEM). WT vs. EpoR-tKO: **p < 0.01. Part of the
illustration in panel a contains images from Servier Medical Art by
Servier, licensed under a Creative Commons Attribution 3.0 unported
license a b b a c
d
e e d e c metres (km) that mice achieved in the 4-week voluntary wheel run-
ning experiment. WT wild-type mice, EpoR-tKO EpoR-tissue knock-
out mice, Run running. Data are presented as mean value ± standard
error of the mean (SEM). WT vs. EpoR-tKO: **p < 0.01. Cardiac hypertrophy Exercise resulted in a significant increase in cardiac weight
in WT mice (WT Sed 155 ± 3.6 vs. WT Run 195 ± 6.6 mg,
p < 0.05) (Fig. 4b). In contrast, exercise did not significantly
increase cardiac mass in EpoR-tKO mice (Fig. 4b). Body
weight was comparable among the experimental groups
(Supplementary Table S2). Accordingly, when cardiac
weight was adjusted for body weight, a similar trend was
observed (Supplementary Table S2). Cardiac function Part of the
illustration in panel a contains images from Servier Medical Art by
Servier, licensed under a Creative Commons Attribution 3.0 unported
license Fig. 3 Exercise performance in wild-type and EpoR-tissue knock-
out mice. Exercise parameters during 4-week period of voluntary
wheel running (n = 11–12/group). Shown are (a) experimental design
including voluntary wheel running protocol, (b) average daily run-
ning speed in kilometres per hour (km/h), (c) average running time
in minutes per day (min/day), (d) average daily running distance in
kilometres per day (km/day) and (e) total running distance in kilo- 1 Pflügers Archiv - European Journal of Physiology Fig. 4 Functional, structural and angiogenic changes in cardiomyo-
cytes: immunohistochemistry analysis. Functional, structural and angi-
ogenic changes in cardiomyocytes after 4 weeks of voluntary wheel
running (n = 10–13/group). Shown are (a) intra cardiac measurement
of left ventricular end-diastolic pressure (LVEDP) in millimetre of
mercury (mmHg) and (b) heart weight (HW) in milligrams (mg). Panel c displays cardiomyocyte cross-sectional area in micrometres-
squared (μm2) and (d) typical examples of cardiomyocyte size. In
panel e, the capillary-to-cardiomyocyte ratio is shown, together with
typical examples of capillary density in panel f. WT wild-type mice,
EpoR-tKO EpoR-tissue knock-out mice, Sed sedentary, Run running. Data are presented as mean value ± standard error of the mean (SEM). Sed vs. Run: *p < 0.05, ***p < 0.001, Sed vs. Sed and Run vs. Run: #
p < 0.05, ## p < 0.01
a
b
c
e
f
d a
b b c
d d d c f f e
f e squared (μm2) and (d) typical examples of cardiomyocyte size. In
panel e, the capillary-to-cardiomyocyte ratio is shown, together with
typical examples of capillary density in panel f. WT wild-type mice,
EpoR-tKO EpoR-tissue knock-out mice, Sed sedentary, Run running. Data are presented as mean value ± standard error of the mean (SEM). Sed vs. Run: *p < 0.05, ***p < 0.001, Sed vs. Sed and Run vs. Run: #
p < 0.05, ## p < 0.01 Fig. 4 Functional, structural and angiogenic changes in cardiomyo-
cytes: immunohistochemistry analysis. Functional, structural and angi-
ogenic changes in cardiomyocytes after 4 weeks of voluntary wheel
running (n = 10–13/group). Shown are (a) intra cardiac measurement
of left ventricular end-diastolic pressure (LVEDP) in millimetre of
mercury (mmHg) and (b) heart weight (HW) in milligrams (mg). Cardiac function Panel c displays cardiomyocyte cross-sectional area in micrometres- the LV of EpoR-tKO Run mice, suggesting that exercise
induced subtle maladaptive remodelling in EpoR-tKO mice
[32] (Supplementary Table S4). significantly reduced in sedentary EpoR-tKO mice compared
to sedentary WT mice (p < 0.01) (Fig. 4e–f). Furthermore,
the normal increase in capillary density in response to exer-
cise that was observed in WT Run mice was abrogated in
the EpoR-tKO Run group (Fig. 4e–f). Similarly, the mRNA
expression of vascular endothelial growth factor α (VEGF-α),
a critical regulator of angiogenesis, was substantially
increased in response to exercise in WT mice, but not altered
by exercise in EpoR-tKO mice (Supplementary Table S4). The angiogenic response in cardiac muscle In response to exercise, cardiomyocyte hypertrophy is paral-
leled by a corresponding expansion of the capillary network
[44]. Cardiac capillary-to-myocyte ratio was slightly, but 1 3 Pflügers Archiv - European Journal of Physiology Discussion The role of the non-haematopoietic EpoR in the adaptive
response to physiological stress remains unexplored. In this
study, we aimed to determine the role of the EpoR in cardiac
and skeletal muscle in the physiological adaptation to exer-
cise. While EpoR-tKO mice were phenotypically normal,
mitochondrial content in skeletal muscle of EpoR-tKO mice
was significantly reduced, and mitochondrial biogenesis was
impaired compared to wild-type mice. EpoR-tKO mice also
displayed reduced exercise performance in response to vol-
untary wheel running, associated with reductions in myo-
cyte hypertrophy and adaptive angiogenesis in both cardiac
and skeletal muscle. Taken together, our findings indicate
that the endogenous EPO-EpoR system is critical for basal
mitochondrial biogenesis in skeletal muscle. We also show
that the associated reductions in mitochondrial content are
accompanied by impaired exercise performance, indicating
that the extra-haematopoietic EpoR is critical for the physi-
ological adaptation to exercise. Myocyte hypertrophy Similar to cardiac muscle, skeletal muscle fibres also
develop myocyte hypertrophy in response to exercise [40]. Exercise resulted in a 32% increase in myocyte CSA in the
WT Run group, but this increase upon exercise was absent
in EpoR-tKO mice (Fig. 5a–b). This pattern was similar for
the mRNA expression of ACTA-1, and can be observed in
Supplementary Table S4. The angiogenic response in the skeletal muscle Physiological myocyte hypertrophy in skeletal muscle is
also paralleled by expansion of the capillary network [40]. To determine whether the EpoR is also critical for adap-
tive angiogenesis in skeletal muscle, capillary density was
evaluated in the m. gastrocnemius. In this skeletal muscle, an
abrogated response in capillary/myocyte ratio and VEGF-α
mRNA expression was also observed in EpoR-tKO mice
(Fig. 5c–d, Supplementary Table S4). Cardiac and skeletal muscle require considerable plas-
ticity to adapt to continuous variations in demand. The
response to endurance exercise training is remarkably Fig. 5 Structural and angiogenic changes in skeletal myocytes: immu-
nohistochemistry analysis. Structural and angiogenic changes in skel-
etal myocytes after 4 weeks of voluntary wheel running (n = 10–13/
group). Shown are (a) skeletal myocyte (SM) cross-sectional area
(CSA) in micrometres-squared (μm2) and (b) typical examples of
SM size. In panel c, the skeletal capillary-to-myocyte ratio is shown,
together with typical examples of capillary density in panel d. WT
wild-type mice, EpoR-tKO EpoR-tissue knock-out mice, Sed seden-
tary, Run running. Data are presented as mean value ± standard error
of the mean (SEM). Sed vs. Run: *p < 0.05, **p < 0.01
a
b
c
d a b b a a d c d c SM size. In panel c, the skeletal capillary-to-myocyte ratio is shown,
together with typical examples of capillary density in panel d. WT
wild-type mice, EpoR-tKO EpoR-tissue knock-out mice, Sed seden-
tary, Run running. Data are presented as mean value ± standard error
of the mean (SEM). Sed vs. Run: *p < 0.05, **p < 0.01 Fig. 5 Structural and angiogenic changes in skeletal myocytes: immu-
nohistochemistry analysis. Structural and angiogenic changes in skel-
etal myocytes after 4 weeks of voluntary wheel running (n = 10–13/
group). Shown are (a) skeletal myocyte (SM) cross-sectional area
(CSA) in micrometres-squared (μm2) and (b) typical examples of 1 1 3 Pflügers Archiv - European Journal of Physiology beta common receptor (βcR) to compose the innate repair
receptor (IRR) [8, 10]. The IRR potentially plays a role in
repair mechanisms in response to injury and inflammation;
however, some controversy has been described regarding the
mechanisms of interaction between these receptor isoforms
[8, 10]. Yet, these differences between the tissue and haema-
topoietic EPO receptor suggest that the tissue receptor can
be targeted specifically and several compounds are currently
under investigation [8, 10]. The angiogenic response in the skeletal muscle The insufficient angio-
genic response could result from loss of EpoR signalling, but
could also be a secondary effect related to reduced oxidative
muscle fibres.i Literature describes some degree of discrepancy regard-
ing the role of the skeletal muscle EpoR. Some studies
observe little to no effects of the EpoR in skeletal muscle
whereas other studies suggest a beneficial role for an EpoR
in skeletal muscle [20–22, 51]. For instance, EPO admin-
istration shows activation of AKT signalling, but this was
accompanied by limited effects on protein synthesis in skel-
etal muscle after exercise [21]. Also, the role of the EpoR
in myogenesis was found to be negligible [22]. However,
findings that describe the limited role for the EpoR in skel-
etal muscle adaptations were based on a single EPO injec-
tion or in vitro EPO treatment and do not specifically target
the EpoR as described above. Contrarily to the above, our
results are in line with previous reports that have indeed
demonstrated the role of the EpoR in basal mitochondrial
muscle metabolism in EpoR-tKO mice [51]. These mecha-
nisms have been attributed to the regulation of AMPK and
PGC-1α, which in turn modulate oxidative muscle fibre
reprogramming [51]. With our study, we now provide the
evidence that the EpoR in skeletal muscle is indeed critical
for basal mitochondrial biogenesis. Our data suggests that
these changes are associated with reduced exercise capacity
and impaired adaptation of cardiac and skeletal muscle to
voluntary exercise. Future investigations should be focussed
on unravelling the possible underlying mechanisms of these
adaptations, such as the regulation of AMPK. Taken together
with our findings, the EpoR in skeletal muscle could be a
novel target to improve skeletal muscle quality in patients
with muscle fatigue, mitochondrial myopathies or heart fail-
ure-associated exercise intolerance [13, 38, 53]. Of note,
ageing EpoR-tKO mice, mostly pronounced in females, also
develop severe obesity and insulin resistance, suggesting that
targeting the tissue-specific EPO receptor might also be an
effective therapy for diabetes [49]. i
The first indication suggesting that a functional EPO
receptor was present in the cardiac and skeletal muscle dates
back almost two decades [6]. While the presence of a func-
tional extra-haematopoietic EPO-receptor was scrutinised at
first, studies using comprehensive analysis have firmly estab-
lished the presence of a functional EPO receptor in muscle
cells [35, 42, 52]. The angiogenic response in the skeletal muscle The insufficient angio-
genic response could result from loss of EpoR signalling, but
could also be a secondary effect related to reduced oxidative
muscle fibres.i similar in cardiac and skeletal muscle, including an increase
in myocyte size, expansion of the capillary network, changes
in muscle fibre isoforms and an increase in mitochondrial
density [40, 43]. While several key regulators of physiologi-
cal hypertrophy have been identified [32], the role of the
endogenous EPO-EpoR system in cardiac and skeletal mus-
cle has not been well described. The tissue-specific EpoR
has however been widely studied in pathological conditions
and findings from these studies suggest that the EpoR exerts
protective effects in response to pathological stress [16, 33,
48]. Current knowledge therefore mainly focusses on the
effects of the tissue-specific EpoR in disease but knowledge
about its physiological role is limited. The sparse evidence
indicating a critical role for this system in the response to
exercise was often confounded by the concomitant changes
in haematocrit. For instance, mutated mice with systemic
EPO-deficiency have reduced exercise performance, asso-
ciated with oxidative and metabolic stress in skeletal mus-
cle and reduced expression of fast oxidative muscle fibres;
yet, these observations were done in the presence of severe
anaemia [30]. rhEPO also improves exercise performance
and induces mitochondrial biogenesis in cardiac and skeletal
muscle of healthy mice through activation of key regula-
tors of physiological hypertrophy, including AKT, eNOS
and PGC-1α [7]. Importantly, these mitochondrial effects
preceded any changes in haematocrit, suggesting that they
were regulated by a tissue-specific receptor. Mitochondrial
respiration was also increased in permeabilised skeletal
muscle fibres obtained from athletes after prolonged treat-
ment with rhEPO [39]. Overexpression of EPO in mice also
activated PGC-1α and increased the proportion of oxidative
muscle fibres, whereas the opposite was observed in our
mouse model [51]. While these studies clearly indicate that
EPO stimulates oxidative metabolism in muscle cells, it is
impossible to exclude the possibility that these results were
confounded by off target effects of EPO. Our study suggests
that the tissue-specific EpoR is a crucial regulator of basal
mitochondrial biogenesis as well as structural and metabolic
adaptations to exercise. Regarding the difference in capillary
density, it should be noted that we cannot reach definitive
conclusions about cause and effect. The angiogenic response in the skeletal muscle This would offer the possibil-
ity to exploit the beneficial effects of rhEPO for exercise
performance without the untoward effects on haematocrit. Whether these isoforms are capable of regulating exercise
performance and exercise-induced muscle growth remains
unknown. similar in cardiac and skeletal muscle, including an increase
in myocyte size, expansion of the capillary network, changes
in muscle fibre isoforms and an increase in mitochondrial
density [40, 43]. While several key regulators of physiologi-
cal hypertrophy have been identified [32], the role of the
endogenous EPO-EpoR system in cardiac and skeletal mus-
cle has not been well described. The tissue-specific EpoR
has however been widely studied in pathological conditions
and findings from these studies suggest that the EpoR exerts
protective effects in response to pathological stress [16, 33,
48]. Current knowledge therefore mainly focusses on the
effects of the tissue-specific EpoR in disease but knowledge
about its physiological role is limited. The sparse evidence
indicating a critical role for this system in the response to
exercise was often confounded by the concomitant changes
in haematocrit. For instance, mutated mice with systemic
EPO-deficiency have reduced exercise performance, asso-
ciated with oxidative and metabolic stress in skeletal mus-
cle and reduced expression of fast oxidative muscle fibres;
yet, these observations were done in the presence of severe
anaemia [30]. rhEPO also improves exercise performance
and induces mitochondrial biogenesis in cardiac and skeletal
muscle of healthy mice through activation of key regula-
tors of physiological hypertrophy, including AKT, eNOS
and PGC-1α [7]. Importantly, these mitochondrial effects
preceded any changes in haematocrit, suggesting that they
were regulated by a tissue-specific receptor. Mitochondrial
respiration was also increased in permeabilised skeletal
muscle fibres obtained from athletes after prolonged treat-
ment with rhEPO [39]. Overexpression of EPO in mice also
activated PGC-1α and increased the proportion of oxidative
muscle fibres, whereas the opposite was observed in our
mouse model [51]. While these studies clearly indicate that
EPO stimulates oxidative metabolism in muscle cells, it is
impossible to exclude the possibility that these results were
confounded by off target effects of EPO. Our study suggests
that the tissue-specific EpoR is a crucial regulator of basal
mitochondrial biogenesis as well as structural and metabolic
adaptations to exercise. Regarding the difference in capillary
density, it should be noted that we cannot reach definitive
conclusions about cause and effect. The angiogenic response in the skeletal muscle The extra-haematopoietic EpoR comprises
a distinct isoform of the EpoR, with a distinct pattern of reg-
ulation, and heterodimerises with other receptors such as the f
The mechanistic pathways underlying the cardiac
effects as a result of the absent EpoR in non-haemat-
opoietic tissues remain unexplored. In our study, we have
observed a rise in left ventricular end diastolic pressure,
together with similar effects in mean arterial pressure as 1 3 3 Pflügers Archiv - European Journal of Physiology well as systolic blood pressure in response to exercise. These functional parameters were accompanied by incre-
ments of ANP expression and a shift towards β-myosin
heavy chain isoforms. Together, these effects describe a
phenotype suiting pathological rather than physiological
adaptation. We may argue that these effects are primarily
due the loss of the EpoR in cardiac tissue; however, these
effects may also be related to secondary effects exerted
upon the heart. Our findings suggest the primary role for
EpoR in response to physiological stress; unfortunately,
it remains questionable whether these effects are in fact
primary. Therefore, future studies are required in order
to determine the specific underlying mechanisms of the
impaired physiological response. well as systolic blood pressure in response to exercise. These functional parameters were accompanied by incre-
ments of ANP expression and a shift towards β-myosin
heavy chain isoforms. Together, these effects describe a
phenotype suiting pathological rather than physiological
adaptation. We may argue that these effects are primarily
due the loss of the EpoR in cardiac tissue; however, these
effects may also be related to secondary effects exerted
upon the heart. Our findings suggest the primary role for
EpoR in response to physiological stress; unfortunately,
it remains questionable whether these effects are in fact
primary. Therefore, future studies are required in order
to determine the specific underlying mechanisms of the
impaired physiological response. Author contribution WPTR, AHM, RAdB and BDW were involved in
conception and design of experiment. KTN, LMGM, WPTR, SRY, MGPG,
BM and AHGW performed data collection, analysis and interpretation. KTN, LMGM, WPTR, AHM, SRY, MGPG, DJvV, WHvG, HHWS, RAdB
and BDW were responsible for the article drafting and revising. All authors
approved the final version of the manuscript, all authors met the criteria for
authorship and all who qualify for authorship are listed. Author contribution WPTR, AHM, RAdB and BDW were involved in
conception and design of experiment. Consent for publication Not applicable. Consent for publication Not applicable. Competing interests The UMCG, which employs several of the
authors, has received research grants and/or fees from AstraZeneca,
Abbott, Bristol-Myers Squibb, Novartis, Novo Nordisk and Roche. Dr. de Boer received speaker fees from Abbott, AstraZeneca, Novartis and
Roche. The other authors do not have any conflict of interest to declare. Open Access This article is licensed under a Creative Commons Attri-
bution 4.0 International License, which permits use, sharing, adapta-
tion, distribution and reproduction in any medium or format, as long
as you give appropriate credit to the original author(s) and the source,
provide a link to the Creative Commons licence, and indicate if changes
were made. The images or other third party material in this article are
included in the article’s Creative Commons licence, unless indicated
otherwise in a credit line to the material. If material is not included in
the article’s Creative Commons licence and your intended use is not
permitted by statutory regulation or exceeds the permitted use, you will
need to obtain permission directly from the copyright holder. To view a
copy of this licence, visit http://creativecommons.org/licenses/by/4.0/. i
In conclusion, this study provides evidence that the
endogenous EPO-EpoR system controls mitochondrial bio-
genesis in skeletal muscle. We also provide evidence that
reduced mitochondrial biogenesis in EpoR-deficient mice
was associated with impaired exercise performance, suggest-
ing that the EpoR expressed in cardiac and skeletal muscle is
critical for physiological exercise and its adaptation. The angiogenic response in the skeletal muscle KTN, LMGM, WPTR, SRY, MGPG,
BM and AHGW performed data collection, analysis and interpretation. KTN, LMGM, WPTR, AHM, SRY, MGPG, DJvV, WHvG, HHWS, RAdB
and BDW were responsible for the article drafting and revising. All authors
approved the final version of the manuscript, all authors met the criteria for
authorship and all who qualify for authorship are listed. Funding This study was funded in part by grants from the Nether-
lands Organization for Research (VENI 016–106-117) and Netherlands
Heart Foundation (2007T046, 2019T064). Part of the work has been
performed in the UMCG Microscopy and Imaging Centre (UMIC),
sponsored by ZonMW 91111.006. Data availability The data from the current study are available from
the corresponding author on reasonable request. Several limitations of our analysis need to be acknowl-
edged. For instance, this is a model of voluntary wheel run-
ning in which the workload on cardiac and skeletal mus-
cle is relatively subtle and the amount of exercise varied
between the groups, whereas forced or exhaustive exercise
may cause more rigorous and equal effects on cardiac and
skeletal muscle. Nevertheless, exercise-induced hypertrophy
in our WT mice was substantial and similar studies also
demonstrated substantial muscle growth induced by volun-
tary wheel running models [1, 17, 28]. Nonetheless, the role
of the EpoR in the adaptation to more strenuous forms of
exercise may also be different. The varied amount of exer-
cise performed, however, can be attributed to our findings of
decreased mitochondrial content in skeletal muscle of EpoR-
tKO mice, resulting in impaired exercise performance and
physiological adaptation. Next to this limitation, it should
also be acknowledged that we compared mitochondrial con-
tent and molecular markers for mitochondrial biogenesis in
our study. From this data, we cannot conclude definitely that
the actual generation of new mitochondria was perturbed. Finally, this is a mouse model in which mice were rescued
from lethality by erythroid lineage-restricted expression of
the EpoR. Therefore, the outcomes cannot be interpreted as
for cardiac and skeletal myocyte-specific KO mice. Declarations Ethics approval The experimental protocol was accepted by The Ani-
mal Ethical Committee of the University of Groningen (DEC4585/
IvD 1583–02-011, 1583–02-014). In addition, all animal experiments
followed the protocols from Directive 2010/63/EU of the European
Parliament. Consent to participate Not applicable. References Carraway MS, Suliman HB, Jones WS, Chen C-W, Babiker A,
Piantadosi CA (2010) Erythropoietin activates mitochondrial bio-
genesis and couples red cell mass to mitochondrial mass in the
heart. Circ Res 106:1722–1730. https://doi.org/10.1161/CIRCR
ESAHA.109.214353 21. Lamon S, Zacharewicz E, Arentson-Lantz E, Della GPA, Ghobrial
L, Gerlinger-Romero F, Garnham A, Paddon-Jones D, Russell
AP (2016) Erythropoietin does not enhance skeletal muscle pro-
tein synthesis following exercise in young and older adults. Front
Physiol 7:292. https://doi.org/10.3389/fphys.2016.00292 8. Cheung Tung Shing KS, Broughton SE, Nero TL, Gillinder K,
Ilsley MD, Ramshaw H, Lopez AF, Griffin MDW, Parker MW,
Perkins AC, Dhagat U (2018) EPO does not promote interaction
between the erythropoietin and beta-common receptors. Sci Rep
8:1–16. https://doi.org/10.1038/s41598-018-29865-x 22. Lamon S, Zacharewicz E, Stephens AN, Russell AP (2014) EPO-
receptor is present in mouse C2C12 and human primary skeletal
muscle cells but EPO does not influence myogenesis. Physiol Rep
2:e00256. https://doi.org/10.1002/phy2.256 23. Lipsic E, Westenbrink BD, van der Meer P, van der Harst P, Voors
AA, van Veldhuisen DJ, Schoemaker RG, van Gilst WH (2008)
Low-dose erythropoietin improves cardiac function in experimen-
tal heart failure without increasing haematocrit. Eur J Heart Fail
10:22–29. https://doi.org/10.1016/j.ejheart.2007.10.008 p
g
9. Cohen RS, Karlin P, Yushak M, Mancini D, Maurer MS (2009)
The effect of erythropoietin on exercise capacity, left ventricular
remodeling, pressure-volume relationships, and quality of life in
older patients with anemia and heart failure with preserved ejec-
tion fraction. Congest Hear Fail 16:96–103. https://doi.org/10.
1111/j.1751-7133.2009.00135.x 24. Lu J, Dai Q, ming, Ma G shan, Zhu Y hong, Chen B, Li B, Yao Y
yu, (2017) Erythropoietin attenuates cardiac dysfunction in rats
by inhibiting endoplasmic reticulum stress-induced diabetic car-
diomyopathy. Cardiovasc Drugs Ther 31:367–379. https://doi.org/
10.1007/s10557-017-6742-1f j
10. Collino M, Thiemermann C, Cerami A, Brines M (2015) Flipping
the molecular switch for innate protection and repair of tissues:
long-lasting effects of a non-erythropoietic small peptide engi-
neered from erythropoietin. Pharmacol Ther 151:32–40. https://
doi.org/10.1016/j.pharmthera.2015.02.005 25. Lundby C, Olsen NV (2011) Effects of recombinant human eryth-
ropoietin in normal humans. J Physiol 589:1265–1271. https://doi.
org/10.1113/jphysiol.2010.195917 11. Dorn GW, Vega RB, Kelly DP (2015) Mitochondrial biogenesis
and dynamics in the developing and diseased heart. Genes Dev
29:1981–1991 26. Lundby C, Thomsen JJ, Boushel R, Koskolou M, Warberg J,
Calbet JAL, Robach P (2007) Erythropoietin treatment elevates
haemoglobin concentration by increasing red cell volume and
depressing plasma volume. J Physiol 578:309–314. https://doi.
org/10.1113/jphysiol.2006.122689 12. References Ghezzi P, Bernaudin M, Bernaudinb M, Bianchi R, Bianchic R,
Blomgren K, Brines M, Campana W, Cavaletti G, Cerami A,
Chopp M, Coleman T, Digicaylioglu M, Ehrenreich H, Erbay-
raktar S, Erbayraktar Z, Gassmann M, Genc S, Gokmen N, Grasso
G, Juul S, Lipton SA, Hand CC, Latini R, Lauria G, Leist M,
Newton SS, Petit E, Probert L, Sfacteria A, Siren A-L, Talan M,
Thiemermann C, Westenbrink D, Yaqoob M, Zhu C (2010) Eryth-
ropoietin: not just about erythropoiesis. Lancet (London, England)
375:2142. https://doi.org/10.1016/S0140-6736(10)60992-0 27. Mancini DM, Katz SD, Lang CC, LaManca J, Hudaihed A, Androne
AS (2003) Effect of erythropoietin on exercise capacity in patients
with moderate to severe chronic heart failure. Circulation 107:294–
299. https://doi.org/10.1161/01.CIR.0000044914.42696.6A 28. Manzanares G, Brito-Da-Silva G, Gandra PG (2019) Voluntary
wheel running: patterns and physiological effects in mice. Brazil-
ian J Med Biol Res 52https://doi.org/10.1590/1414-431X20187830 13. Handschin C, Chin S, Li P, Liu F, Maratos-Flier E, LeBrasseur
NK, Yan Z, Spiegelman BM (2007) Skeletal muscle fiber-type
switching, exercise intolerance, and myopathy in PGC-1α mus-
cle-specific knock-out animals. J Biol Chem 282:30014–30021. https://doi.org/10.1074/jbc.M704817200 29. Meems LMG, Cannon MV, Mahmud H, Voors AA, van Gilst
WH, Silljé HHW, Ruifrok WPT, de Boer RA (2012) The vitamin
D receptor activator paricalcitol prevents fibrosis and diastolic
dysfunction in a murine model of pressure overload. J Steroid
Biochem Mol Biol 132:282–289. https://doi.org/10.1016/j.jsbmb.
2012.06.004 14. Hardeman M, Alexy T, Brouwer B, Connes P, Jung F, Kuipers
H, Baskurt OK (2014) EPO or PlacEPO? Science versus practi-
cal experience: panel discussion on efficacy of erythropoetin in
improving performance. Biorheology 51:83–90. https://doi.org/
10.3233/BIR-140655 30. Mille-Hamard L, Billat VL, Henry E, Bonnamy B, Joly F, Benech
P, Barrey E (2012) Skeletal muscle alterations and exercise per-
formance decrease in erythropoietin-deficient mice: a compara-
tive study. BMC Med Genomics 5:29. https://doi.org/10.1186/
1755-8794-5-29 15. Hoes MF, Grote Beverborg N, Kijlstra JD, Kuipers J, Swinkels
DW, Giepmans BNG, Rodenburg RJ, van Veldhuisen DJ, de Boer
RA, van der Meer P (2018) Iron deficiency impairs contractility of
human cardiomyocytes through decreased mitochondrial function. Eur J Heart Fail 20:910–919. https://doi.org/10.1002/ejhf.1154 31. Muirhead N, Bargman J, Burgess E, Jindal KK, Levin A, Nolin L,
Parfrey P (1995) Evidence-based recommendations for the clinical
use of recombinant human erythropoietin. Am J Kidney Dis 26:S1-24 32. Nakamura M, Sadoshima J (2018) Mechanisms of physiological
and pathological cardiac hypertrophy. Nat Rev Cardiol 15:387–
407. https://doi.org/10.1038/s41569-018-0007-y 16. References 1. Allen DL, Harrison BC, Maass A, Bell ML, Byrnes WC, Lein-
wand LA (2001) Cardiac and skeletal muscle adaptations to vol-
untary wheel running in the mouse. J Appl Physiol 90:1900–1908. https://doi.org/10.1152/jappl.2001.90.5.1900f Supplementary Information The online version contains supplemen-
tary material available at https://doi.org/10.1007/s00424-021-02577-4. p
g
j pp
2. Annaheim S, Jacob M, Krafft A, Breymann C, Rehm M, Boutellier
U (2016) RhEPO improves time to exhaustion by non-hematopoi-
etic factors in humans. Eur J Appl Physiol 116:623–633. https://
doi.org/10.1007/s00421-015-3322-6 p
g
j pp
2. Annaheim S, Jacob M, Krafft A, Breymann C, Rehm M, Boutellier
U (2016) RhEPO improves time to exhaustion by non-hematopoi-
etic factors in humans. Eur J Appl Physiol 116:623–633. https://
doi.org/10.1007/s00421-015-3322-6 Acknowledgements We thank Ben N.G. Giepmans (UMCG) for
advice on nanotomy. Animal and/or laboratory techniques were per-
formed with the help of Bianca Meijeringh, Linda van Genne, Inge
Vreeswijk-Baudoin, Silke Oberdorf-Maass, Janny Takens, Martin Dok-
ter and Marloes Schouten. g
3. Atkinson TS, Kahn MJ (2020) Blood doping: then and now. A nar-
rative review of the history, science and efficacy of blood doping
in elite sport. Blood Rev 39:100632 1 3 Pflügers Archiv - European Journal of Physiology in the mouse. Am J Physiol - Hear Circ Physiol 289:H455-65. https://doi.org/10.1152/ajpheart.00085.2005 p
g
jp
18. Kuipers J, de Boer P, Giepmans BNG (2015) Scanning EM of
non-heavy metal stained biosamples: large-field of view, high con-
trast and highly efficient immunolabeling. Exp Cell Res 337:202–
207. https://doi.org/10.1016/j.yexcr.2015.07.012 5. de Boer RA, Pinto YM, Suurmeijer AJH, Pokharel S, Scholtens E,
Humler M, Saavedra JM, Boomsma F, van Gilst WH, van Veldhu-
isen DJ (2003) Increased expression of cardiac angiotensin II type
1 (AT(1)) receptors decreases myocardial microvessel density after
experimental myocardial infarction. Cardiovasc Res 57:434–442 19. Kuipers J, Giepmans BNG (2020) Neodymium as an alternative
contrast for uranium in electron microscopy. Histochem Cell Biol
153:271–277. https://doi.org/10.1007/s00418-020-01846-0 6. Calvillo L, Latini R, Kajstura J, Leri A, Anversa P, Ghezzi P, Salio
M, Cerami A, Brines M (2003) Recombinant human erythropoi-
etin protects the myocardium from ischemia-reperfusion injury
and promotes beneficial remodeling. Proc Natl Acad Sci U S A
100:4802–4806. https://doi.org/10.1073/pnas.0630444100 20. Lamon S, Russell AP (2013) The role and regulation of erythropoietin
(EPO) and its receptor in skeletal muscle: how much do we really
know? Front Physiol 4:176. https://doi.org/10.3389/fphys.2013.00176 7. 4. Birzniece V (2015) Doping in sport: effects, harm and misconcep-
tions. Intern Med J 45:239–248. https://doi.org/10.1111/imj.12629 References Pin F, Busquets S, Toledo M, Camperi A, Lopez-Soriano FJ, Costelli
P, Argilés JM, Penna F (2015) Combination of exercise training and
erythropoietin prevents cancer-induced muscle alterations. Onco-
target 6:43202–43215. https://doi.org/10.18632/oncotarget.6439 54. Westenbrink BD, De Boer RA, Voors AA, Van Gilst WH, Van
Veldhuisen DJ (2008) Anemia in chronic heart failure: etiology
and treatment options. Curr Opin Cardiol 23:141–147 39. Plenge U, Belhage B, Guadalupe-Grau A, Andersen PR, Lundby C,
Dela F, Stride N, Pott FC, Helge JW, Boushel R (2012) Erythropoi-
etin treatment enhances muscle mitochondrial capacity in humans. Front Physiol 3:50. https://doi.org/10.3389/fphys.2012.00050i 55. Westenbrink BD, Oeseburg H, Kleijn L, van der Harst P, Belonje
AMS, Voors AA, Schoemaker RG, de Boer RA, van Veldhu-
isen DJ, van Gilst WH (2008) Erythropoietin stimulates normal
endothelial progenitor cell-mediated endothelial turnover, but
attributes to neovascularization only in the presence of local
ischemia. Cardiovasc drugs Ther 22:265–274. https://doi.org/10.
1007/s10557-008-6094-y 40. Qaisar R, Bhaskaran S, Van Remmen H (2016) Muscle fiber type
diversification during exercise and regeneration. Free Radic Biol
Med 98:56–67. https://doi.org/10.1016/j.freeradbiomed.2016.03.
025 41. Quiros PM, Goyal A, Jha P, Auwerx J (2017) Analysis of mtDNA/
nDNA ratio in mice. Curr Protoc Mouse Biol 7:47–54. https://doi.
org/10.1002/cpmo.21 56. Westenbrink BD, Ruifrok W-PT, Voors AA, Tilton RG, van Veld-
huisen DJ, Schoemaker RG, van Gilst WH, de Boer RA (2010)
Vascular endothelial growth factor is crucial for erythropoietin-
induced improvement of cardiac function in heart failure. Cardio-
vasc Res 87:30–39. https://doi.org/10.1093/cvr/cvq041 42. Rundqvist H, Rullman E, Sundberg CJ, Fischer H, Eisleitner K,
Ståhlberg M, Sundblad P, Jansson E, Gustafsson T (2009) Activa-
tion of the erythropoietin receptor in human skeletal muscle. Eur
J Endocrinol 161:427–434. https://doi.org/10.1530/EJE-09-0342 43. Seo DY, Kwak HB, Kim AH, Park SH, Heo JW, Kim HK, Ko JR,
Lee SJ, Bang HS, Sim JW, Kim M, Han J (2019) Cardiac adapta-
tion to exercise training in health and disease. Pflugers Arch Eur. J. Physiol 472:155–168 57. Yu H, Tigchelaar W, Koonen DPY, Patel HH, de Boer RA, van
Gilst WH, Westenbrink BD, Silljé HHW (2013) AKIP1 expres-
sion modulates mitochondrial function in rat neonatal cardiomyo-
cytes. PLoS ONE 8:e80815. https://doi.org/10.1371/journal.pone.
0080815 44. Shimizu I, Minamino T (2016) Physiological and pathological
cardiac hypertrophy. J Mol Cell Cardiol 97:245–262. https://doi.
org/10.1016/j.yjmcc.2016.06.001 58. References Jia Y, Suzuki N, Yamamoto M, Gassmann M, Noguchi CT (2012)
Endogenous erythropoietin signaling facilitates skeletal muscle
repair and recovery following pharmacologically induced damage. FASEB J 26:2847–2858. https://doi.org/10.1096/fj.11-196618 33. Nakano M, Satoh K, Fukumoto Y, Ito Y, Kagaya Y, Ishii N,
Sugamura K, Shimokawa H (2007) Important role of erythropoi-
etin receptor to promote VEGF expression and angiogenesis in j
17. Konhilas JP, Widegren U, Allen DL, Paul AC, Cleary A, Lein-
wand LA (2005) Loaded wheel running and muscle adaptation 1 3 Pflügers Archiv - European Journal of Physiology peripheral ischemia in mice. Circ Res 100:662–669. https://doi.
org/10.1161/01.RES.0000260179.43672.fe proopiomelanocortin production. Nat Commun 2:520. https://
doi.org/10.1038/ncomms1526 g
34. Noguchi CT, Wang L, Rogers HM, Teng R, Jia Y (2008) Survival
and proliferative roles of erythropoietin beyond the erythroid line-
age. Expert Rev Mol Med 10:e36. https://doi.org/10.1017/S1462
399408000860 50. Wang L, Di L, Noguchi CT (2014) Erythropoietin, a novel ver-
satile player regulating energy metabolism beyond the erythroid
system. Int J Biol Sci 10:921–939. https://doi.org/10.7150/ijbs.
9518 51. Wang L, Jia Y, Rogers H, Suzuki N, Gassmann M, Wang Q,
McPherron AC, Kopp JB, Yamamoto M, Noguchi CT (2013)
Erythropoietin contributes to slow oxidative muscle fiber specifi-
cation via PGC-1α and AMPK activation. Int J Biochem Cell Biol
45:1155–1164. https://doi.org/10.1016/j.biocel.2013.03.007 35. Ogilvie M, Yu X, Nicolas-Metral V, Pulido SM, Liu C, Ruegg
UT, Noguchi CT (2000) Erythropoietin stimulates proliferation
and interferes with differentiation of myoblasts. J Biol Chem
275:39754–39761. https://doi.org/10.1074/jbc.M004999200 p
g
j
36. Ostrowski D, Heinrich R (2018) Alternative erythropoietin recep-
tors in the nervous system. J Clin Med 7:24. https://doi.org/10.
3390/jcm7020024 p
g
j
52. Wang L, Jia Y, Rogers H, Wu YP, Huang S, Noguchi CT (2012)
GATA-binding protein 4 (GATA-4) and T-cell acute leukemia
1 (TAL1) regulate myogenic differentiation and erythropoi-
etin response via cross-talk with Sirtuin1 (Sirt1). J Biol Chem
287:30157–30169. https://doi.org/10.1074/jbc.M112.376640 j
37. Pacher P, Nagayama T, Mukhopadhyay P, Bátkai S, Kass DA
(2008) Measurement of cardiac function using pressure-volume
conductance catheter technique in mice and rats. Nat Protoc
3:1422–1434. https://doi.org/10.1038/nprot.2008.138 53. Weiss K, Schär M, Panjrath GS, Zhang Y, Sharma K, Bottomley
PA, Golozar A, Steinberg A, Gerstenblith G, Russell SD, Weiss
RG (2017) Fatigability, exercise intolerance, and abnormal skel-
etal muscle energetics in heart failure. Circ Heart Fail 10:e004129. https://doi.org/10.1161/CIRCHEARTFAILURE.117.004129 38. Publisher’s note Springer Nature remains neutral with regard to
jurisdictional claims in published maps and institutional affiliations. References Yu H, Tigchelaar W, Lu B, van Gilst W, de Boer R, Westenbrink
B, Silljé H (2013) AKIP1, a cardiac hypertrophy induced protein
that stimulates cardiomyocyte growth via the Akt pathway. Int J
Mol Sci 14:21378–21393. https://doi.org/10.3390/ijms141121378 45. Sokol E, Kramer D, Diercks GFH, Kuipers J, Jonkman MF, Pas
HH, Giepmans BNG (2015) Large-scale electron microscopy
maps of patient skin and mucosa provide insight into pathogenesis
of blistering diseases. J Invest Dermatol 135:1763–1770. https://
doi.org/10.1038/jid.2015.109 59. Yurista SR, Matsuura TR, Silljé HHW, Nijholt KT, McDaid KS,
Shewale SV, Leone TC, Newman JC, Verdin E, van Veldhuisen
DJ, de Boer RA, Kelly DP, Westenbrink BD (2020) Ketone ester
treatment improves cardiac function and reduces pathologic
remodeling in preclinical models of heart failure. Circ Hear Fail. https://doi.org/10.1161/circheartfailure.120.007684 46. Suresh S, Rajvanshi PK, Noguchi CT (2020) The Many facets of
erythropoietin physiologic and metabolic response. Front Physiol
10:1534 47. Suzuki N, Ohneda O, Takahashi S, Higuchi M, Mukai HY,
Nakahata T, Imagawa S, Yamamoto M (2002) Erythroid-specific
expression of the erythropoietin receptor rescued its null mutant
mice from lethality. Blood 100:2279–2288. https://doi.org/10.
1182/blood-2002-01-0124 60. Yurista SR, Silljé HHW, Nijholt KT, Dokter MM, van Veldhuisen
DJ, de Boer RA, Westenbrink BD (2020) Factor Xa inhibition
with apixaban does not influence cardiac remodelling in rats with
heart failure after myocardial infarction. Cardiovasc Drugs Ther. https://doi.org/10.1007/s10557-020-06999-7 61. Zhang Y, Wang L, Dey S, Alnaeeli M, Suresh S, Rogers H,
Teng R, Noguchi CT (2014) Erythropoietin action in stress
response, tissue maintenance and metabolism. Int J Mol Sci
15:10296–10333 48. Tada H, Kagaya Y, Takeda M, Ohta J, Asaumi Y, Satoh K, Ito K,
Karibe A, Shirato K, Minegishi N, Shimokawa H (2006) Endog-
enous erythropoietin system in non-hematopoietic lineage cells
plays a protective role in myocardial ischemia/reperfusion. Car-
diovasc Res 71:466–477. https://doi.org/10.1016/j.cardiores.2006.
05.010 Publisher’s note Springer Nature remains neutral with regard to
jurisdictional claims in published maps and institutional affiliations. Publisher’s note Springer Nature remains neutral with regard to
jurisdictional claims in published maps and institutional affiliations. 49. Teng R, Gavrilova O, Suzuki N, Chanturiya T, Schimel D,
Hugendubler L, Mammen S, Yver DR, Cushman SW, Mueller
E, Yamamoto M, Hsu LL, Noguchi CT (2011) Disrupted eryth-
ropoietin signalling promotes obesity and alters hypothalamus 1 3
|
https://openalex.org/W4366205688
|
https://bmcsurg.biomedcentral.com/counter/pdf/10.1186/s12893-023-01987-8
|
English
| null |
First assistant experience in total laparoscopic pancreaticoduodenectomy: accelerating the learning curve for an operator
|
BMC surgery
| 2,023
|
cc-by
| 6,875
|
RESEARCH Open Access © The Author(s) 2023. Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use,
sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and
the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this
article are included in the article’s Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included
in the article’s Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will
need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by/4.0/. The
Creative Commons Public Domain Dedication waiver (http://creativecommons.org/publicdomain/zero/1.0/) applies to the data made available
in this article, unless otherwise stated in a credit line to the data. Abstract Objective Compare and analyze clinical data of total laparoscopic pancreaticoduodenectomy (TLPD) cases for
surgeons with / without first assistant experience (FAE) in TLPD. Probe influence of FAE in TLPD on the learning curve
for an operator. Methods The clinical data of 239 patients, that underwent TLPD performed by two surgeons between January 2017
and January 2022) in our department, were consecutively collected and divided into two groups (A and B). Group A
cases were operated by Surgeon A, with FAE of 57 TLPDs in our department prior to initial TLPD as an operator. Group
B cases were operated by Surgeon B with no FAE of TLPD. Cumulative sum (CUSUM) method developed learning
curves. Clinical data and both surgeons’ learning curves were statistically compared between both groups. Results Between both groups, no statistically significant variations were observed for pre-operative health
conditions. Reduced surgical duration, blood loss and transfusion volume during surgery, together with reductions
in major post–operative complication rates and reduced hospital/ICU stays were identified within Group A, having
statistically significant variations. The technical plateau phases of the learning curves were approximately 25–41 cases
and 35–51 cases, for Surgeon A and Surgeon B, respectively. Conclusion FAE in TLPD can accelerate the learning curve of TLPD for an operator, with safer surgical procedures and
enhanced post–operative recovery. Keywords FAE, Total laparoscopic pancreaticoduodenectomy, Learning curve, CUSUM Dongrui Li1, Chengxu Du1, Wenbin Wang1, Jiansheng Zhang1 and Jianhua Liu1* Dongrui Li1, Chengxu Du1, Wenbin Wang1, Jiansheng Zhang1 and Jianhua Liu1* BMC Surgery BMC Surgery Li et al. BMC Surgery (2023) 23:92
https://doi.org/10.1186/s12893-023-01987-8 Introduction in large-volume centers [1, 2]. Abundant experience of
OPD, excelling laparoscopic surgical skills and learning
curve are required to ensure the feasibility and safety of
LPD [3–6]. To better adopt LPD in medical centers, there
have been many studies concerning the learning curve
for LPD, using differing sample sizes. Three phases (ini
tial learning period, technical plateau period and tech
nical proficiency period), with two peaks, were typically
revealed by analyses of such learning curves. Studies on
over 50 cases suggested that the operator could achieve
technical competence in TLPD after successfully accom
plishing 40–50 TLPD cases [7–18]. However, there are Presently, total laparoscopic pancreaticoduodenectomy
(TLPD) is a technically demanding and well-established
procedure, with similar or enhanced short-term out
comes and long-term survival rates in comparison to
open laparoscopic pancreaticoduodenectomy (OPD) *Correspondence:
Jianhua Liu
dr.ljh@outlook.com
1Department of Hepatobiliary Surgery, the Second Hospital of Hebei
Medical University, 215 Heping West Road, Shijiazhuang 050000, Hebei,
China *Correspondence:
Jianhua Liu
dr.ljh@outlook.com
1Department of Hepatobiliary Surgery, the Second Hospital of Hebei
Medical University, 215 Heping West Road, Shijiazhuang 050000, Hebei,
China TLPD procedure protocol outlineh The cases were divided into Group A (the first 127
consecutive TLPD cases performed by Surgeon A) and
Group B (the first 112 consecutive TLPD cases per
formed by Surgeon B) based upon differing operators. Both surgeons were second-generation TLPD operators
in our department. Prior to the initial TLPD case as an
operator, Surgeon A had participated in 57 TLPD cases
in our department as First Assistant, while surgeon B
had no FAE. Both surgeons had a minimum of 6-years’
experience and were competent of fundamental laparo
tomic and laparoscopic hepatobiliary-pancreatic surgi
cal procedures, including cholecystectomy, bile duct
exploration, chole-jejunostomy, hepatectomy, distal pan
createctomy. Furthermore, both surgeons had a FAE of
approximately 30 OPD cases, though had rare experience
as an OPD operator. Pre-operative data included general
conditions (age, sex, BMI and American Society of Anes
thesiologists (ASA) score), initial symptoms (jaundice,
abdominal pain and fever), co-morbidities (coronary
heart disease (CHD), hypertension (HT), diabetes melli
tus (DM), pancreatitis, hepatitis and surgical procedure
history), blood test results (CA19-9, CA12-5 and direct
bilirubin) and pre-operative treatments alleviating jaun
dice (percutaneous trans-hepatic cholangial drainage
(PTCD) and endoscopic naso-biliary drainage (ENBD)). Intra-operative data included pancreatic texture, size of
main pancreatic duct, blood loss, intra-operative trans
fusion, transfusion volume, operation time, pancreato
jejunostomy time, choledochojejunostomy time and
gastrojejunostomy time. Post-operative data included The patient was lying in a supine position with legs
placed apart, whereby the operator and assistant were
positioned on the right side and left side of the patient,
respectively, while laparoscope handler stood between
the patient’s legs. Five trocars were inserted in a “U”
shape. Theses consisted of one 12 mm trocar at the lower
edge of the umbilicus, as the laparoscope port, while one
5 mm trocar, one 12 mm trocar, one 5 mm trocar and one
10 mm trocar were situated at the left and right anterior
axillary line below costal margin, left and right midclavic
ular line above the umbilicus, respectively. The abdomi
nal cavity was filled with carbon dioxide gas at a pressure
of 12-14mmHg. Resection: (1) Dissect gastrocolic ligament and mobi
lize duodenum through Kocher maneuver. (2) Dissect,
ligate and transect the right gastroepiploic and pancre
aticoduodenal inferior vessels. (3) Transect the distal
stomach 2–3 cm from the pylorus. (4) Expose the jeju
num through the Riolan avascular area (on the left to the
SMV) and transect the jejunum 15–20 cm distal from
the Treitz ligament. © The Author(s) 2023. Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use,
sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and
the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this
article are included in the article’s Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included
in the article’s Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will
need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by/4.0/. The
Creative Commons Public Domain Dedication waiver (http://creativecommons.org/publicdomain/zero/1.0/) applies to the data made available
in this article, unless otherwise stated in a credit line to the data. Page 2 of 10 Page 2 of 10 Li et al. BMC Surgery (2023) 23:92 Li et al. BMC Surgery (2023) 23:92 few comparative studies focusing on how to safely and
efficiently accelerate the learning curve for TLPD within
an operator. This study comparatively analyzed clinical
data of TLPD cases and related parameters for learning
curves in surgeons with / without First Assistant Experi
ence (FAE), aiming at elucidating the influence of FAE in
TLPD on the learning curve for an operator. few comparative studies focusing on how to safely and
efficiently accelerate the learning curve for TLPD within
an operator. This study comparatively analyzed clinical
data of TLPD cases and related parameters for learning
curves in surgeons with / without First Assistant Experi
ence (FAE), aiming at elucidating the influence of FAE in
TLPD on the learning curve for an operator. post-operative complications (post-operative pancreatic
fistula, delayed gastric emptying, post-pancreatectomy
hemorrhage and abdominal infection), Clavien-Dindo
classification of surgical complications, pathology results
such as tumor locations (duodenum, ampulla, distal bile
duct and pancreatic head), resection margin status and
number of harvested lymph nodes, re-operation, hospi
talization time, ICU stay time and mortality event within
hospital.i Methods and materials
Clinical data collection Post-operative pancreatic fistula (POPF) [19], post-
pancreatectomy hemorrhage (PPH) [20], delayed gastric
emptying [21] and Clavien-Dindo classification of surgi
cal complications [22] were defined according to known
consensus and guidelines. Abdominal infection was
defined as post-operative fever and increased level of
white blood cells (> 10*10^9/L), with exclusion of infec
tions in other organs. Operation time was defined as the
duration from the first incision up to final closure; anas
tomosis time was defined as the duration from finishing
time of the former procedure to the completing time for
anastomosis procedure; resection time was deemed as
the operation time without the anastomosis time. Clinical data from 239 TLPD cases - subsequently per
formed by two junior attending surgeons (Surgeon A and
B) under identical mentorship throughout the entire sur
gical procedure, in our department between January 2017
and January 2022 - were collected. All surgeons employed
identical laparoscope platform (Olympus™ 3D laparos
copy), with identical assistant / laparoscope handler. All 239 cases were clearly diagnosed as a tumor(< 2 cm)
within pancreatic head, distal bile duct, duodenal or
ampulla -identified through imaging, though excluding
the possibility of inflammatory mass presence. Inclusion
criteria were: (1) standard TLPD with no hand-assisting
or transversion to OPD; (2) no metastasis in other organs
or artery / vein invasion.hi TLPD procedure protocol outlineh (5) Complete lymphadenectomy of
hepatoduodenal ligament and transect the gastroduo
denal artery. (6) Totally mobilize the jejunum and duo
denum from left to right, to expose major vasculatures. (7) Create a tunnel between the pancreatic neck and the
superior mesenteric vein (SMV) or portal vein (PV), at
the inferior border of the pancreas. (8) Transect pan
creatic neck. (9) Complete cholecystectomy and tran
sect common hepatic duct. (10) Expose inferior vena
cava (IVC) and left renal vein through Kocher maneuver. Li et al. BMC Surgery (2023) 23:92 Li et al. BMC Surgery (2023) 23:92 Page 3 of 10 Cases were ordered chronologically, based upon TLPD
duration. CUSUM for operation time was determined as: Separate the uncinate process from SMV. 11. Complete
lymphadenectomy, including lymph node stations 5, 6, 8,
12, 13a, 13b, 14a, 14b, 17a and 17b. 12. Samples placed
in a retrieval bag and extracted through a 5-cm upper
abdominal incision. Separate the uncinate process from SMV. 11. Complete
lymphadenectomy, including lymph node stations 5, 6, 8,
12, 13a, 13b, 14a, 14b, 17a and 17b. 12. Samples placed
in a retrieval bag and extracted through a 5-cm upper
abdominal incision. CUSUMOT =
n
i=1 (xi −µ) Reconstruction: (1) Pancreatojejunostomy: a two-
layer
duct-to-mucosa
anastomosis. (2)
Choledo
chojejunostomy: an end-to-side anastomosis with
approximately 10 cm distance distal to anastomosis for
pancreatojejunostomy. (3) Gastrojejunostomy: antecolic
gastrojejunostomy 40–45 cm downstream from choledo
chojejunostomy location. where xi is the operation time of case i and µ is the
mean operation time. CUSUM of pancreatojejunos
tomy time (CUSUMPJT), choledochojejunostomy time
(CUSUMCJT) and gastrojejunostomy time (CUSUMGJT)
were also determined through the method applied as for
CUSUMOT. Differing phases were defined according to
CUSUMOT learning curve. The principles of uncinate-process-first, novel-artery-
first and inferoposterior duodenal approaches were fol
lowed during such TLPD procedures. Ultrasonic Shears
and linear staplers were used in such procedures, while
vessels were ligated with hemlock clips.h All tests were two-tailed, with P values < 0.05 deemed
to confer statistical significance. Statistical analyses were
performed by the SPSS for Windows® statistical package. This study was performed in accordance with relevant
guidelines and regulations, and was approved by the
Research Ethics Committee of the Second Hospital of
Hebei Medical University (No. 2019-R209). The principles of postoperative management for
TLPD were as follows: 1. Postoperative infusion treat
ment include anti-infection, gastric acid suppression,
liver function protection, reduce phlegm and nutritional
support. Statistical characteristics and comparative analysesh There were no statistical variations in pre-operative data. The statistical characteristics and results of such compar
ison tests for pre-operative data are described in Table 1. No Statistically significant variations were observed
across pancreatic texture and main pancreatic duct
(MPD) size between both groups. There were statistical
variations in all remaining intra-operative data between
Group A and Group B. Such variations suggested that
there was reduced blood loss (400mL (median) in Group
A and B accordingly, P < 0.001) and transfusion vol
ume (0mL (median) in Group A and 400mL (median) in
Group B, P < 0.001), lower intra-operative transfusion rate
(44.88% in Group A and 78.57% in Group B, P < 0.001),
and reduced operation time (424.41 ± 91.41 min in Group
A and 449.11 ± 76.29 min in Group B, P < 0.001), pancre
atojejunostomy time (40 min (median) in Group A and
50 min (median) in Group B, P < 0.001), choledochojeju
nostomy time (20 min (median) in Group A and 30 min
(median) in Group B, P < 0.001) and gastrojejunostomy
time (15 min (median) in Group A and 25 min (median)
in Group B, P < 0.001), in Group A - when compared to
Group B - with statistical significance. There were no sta
tistically significant variations in resection time. Among
post-operative data, there were statistical variations in
major post-operative complications rates (11.02% in
Group A and 30.36% in Group B, P < 0.001), hospital
ization time (15 days (median) in Group A and 19days
(median) in Group B, P < 0.001) and ICU stay duration (0
days (median) in Group A and 0 days (median) in Group
B, P < 0.001). Minimized major post-operative com
plications rates, reduced hospital / ICU stay durations TLPD procedure protocol outlineh 2.The gastric tube remained in place until the
patient passed flatus, and a fluid-based diet was admin
istered to the patient once gastric tube was removed. 3. The two drainage tubes were regularly removed sequen
tially, two days post-operation, once there was no obvious
drainage or clear drainage-fluid less than 50mL/day with
out POPF; the drainage tubes remained if the drainage-
fluid was turbid, > 50mL/day or with POPF. 4. Besides
the emergency reoperations, laparoscopic reoperations
for debridement were performed when the patients had
deteriorating abdominal infection and POPF within one
week post-LPD. Statistical analysis Clinical data were statistically described and compared
between Group A and Group B. Normality tests were
performed for quantitative data. Independent-samples’
t tests were performed for parameters such as age, BMI
and operation time. Wilcoxon rank sum tests were per
formed for parameters such as blood test results, pan
creatojejunostomy time, choledochojejunostomy time,
gastrojejunostomy time, intra-operative blood loss and
transfusion volume, ASA score, POPF, hospitalization
time and ICU stay duration. Chi-square tests were per
formed for parameters such as sex, initial symptoms, co-
morbidities, pre-operative treatments, intra-operative
transfusion, complications (except for POPF), Clavien-
Dindo classification, reoperation, pathology results and
mortality event in hospital. The learning curves for both surgeons were developed
respectively with cumulative sum (CUSUM) method,
which was a graphical method detecting data trends. Page 4 of 10 Page 4 of 10 Li et al. BMC Surgery (2023) 23:92 Table 1 Statistical characteristics and results of the comparison
tests of preoperative data
Parameter
Group
A(127
cases)
Group
B(112
cases)
P
General conditions
Age(years)
60.27 ± 9.78
59.88 ± 9.49
0.832#
Sex (male/female)
78/49
61/51
0.277*
BMI(kg/m2)
22.58 ± 2.91
24.28 ± 3.45
0.083#
ASA(n,I/II/III/IV)
2/101/23/1
0/94/18/0
0.402*
Initial symptoms [n (%)]
Jaundice
74(58.27%)
67(59.82%)
0.807*
Abdominal pain
54(42.51%)
49(43.75%)
0.848*
Fever
12(9.44%)
4(3.57%)
0.070*
Comorbidities [n (%)]
CHD
7(5.51%)
10(8.92%)
0.305*
HBP
39(30.70%)
40(35.71%)
0.412*
DM
20(15.75%)
16(14.28%)
0.752*
Pancreatitis history
4(3.15%)
7(6.25%)
0.254*
Hepatitis
8(6.30%)
8(7.14%)
0.928*
Opearation history
13(10.24%)
15(13.39%)
0.449*
Blood tests
CA199(U/mL)a
94.50
68.50
0.315+
CA125(U/mL)a
12.90
14.70
0.323+
Direct bilirubin(mmol/L)a
60.00
64.00
0.773+
Preoperative treatments [n (%)]
Preoperative PTCD
46(36.22%)
36(32.14%)
0.508*
Preoperative ENBD
17(13.38%)
13(11.60%)
0.679*
BMI: body mass index; ASA: American Society of Anesthesiologists CHD:
coronary heart disease; HBP: high blood pressure; DM: diabetes mellitus; PTCD:
percutaneous tranhepatic cholangial drainage; ENBD: endoscopic nasalbiliary
drainage. a: Median of the parameter; #: Independent-samples t tests; *: Chi-square tests;
+: Wilcoxon rank sum tests Table 1 Statistical characteristics and results of the comparison
tests of preoperative data Surgeon A and Surgeon B, respectively. Initial peak
points for CUSUMOT curves, at which operators began
to achieve technical competence for TLPD, were Case
#25 and Case #35 for Surgeon A and Surgeon B, respec
tively. Initial peak points of CUSUMPJT curves (Fig. 2)
for Surgeon A and Surgeon B were approximately at
Case #15 and Case #31, respectively. Initial peak points
of CUSUMCJT curve (Fig. Statistical analysis 3) for Surgeon A and Surgeon
B were approximately at Case #11 and Case #31, respec
tively. Initial peak points of CUSUMGJT curve (Fig. 4) for
Surgeon A and Surgeon B were approximately at Case
#21 and #51, respectively. This suggested that Surgeon A
required to accomplish fewer TLPD cases than Surgeon
B, in order to achieve technical competence for TLPD
surgical procedure. g
p
CUSUMOT curves consisted of three individual phases,
namely, ascending (Cases #1–24 for Surgeon A; Cases
#1–34 for Surgeon B), plateau (Cases #25–41 for Surgeon
A; Cases #35–51 for Surgeon B), and the descending
phase (Cases #42–127 for Surgeon A; Cases #52–112 for
Surgeon B). Intra-surgical / post-surgical datasets were
comparatively analyzed using statistical methods. No sta
tistically significant variations were observed within pan
creatic texture and MPD dimensions, among all differing
phases for both surgeons. Intra-surgical hemorrhage
demonstrated a decreasing trend from Phase 1 to Phase
3 for both surgeons, which were statistically significant. Post-surgical major complication rates, together with
ICU stay durations, demonstrated a decreasing trend
from Phase 1 to Phase 3 for both surgeons, which were
statistically significant. The repeat surgery cases had sta
tistically significant variation among the 3 phases. Statis
tical profiles / dataset outcomes for comparative analyses
across differing LPD learning curve phases (both sur
geons) are depicted within Tables 3 and 4, respectively. together with increased quantities of harvested lymph
nodes were found in Group A. There were no statistically
significant variations in tumor location and resection
margin status between both groups. The reoperation rate
was 5.51% and 12.50%, in Group A and Group B, respec
tively with no statistical variation. Hospital-based mortal
ity rates were 3.15% and 3.57% in Group A and Group B,
accordingly, with no statistical variations. Safer and more
efficient surgical procedures, together with enhanced
post-operative recovery, were demonstrated in Group
A. Statistical characteristics and results for comparative
analyses of intra-/post–operative data are described in
Table 2. together with increased quantities of harvested lymph
nodes were found in Group A. There were no statistically
significant variations in tumor location and resection
margin status between both groups. The reoperation rate
was 5.51% and 12.50%, in Group A and Group B, respec
tively with no statistical variation. Hospital-based mortal
ity rates were 3.15% and 3.57% in Group A and Group B,
accordingly, with no statistical variations. Statistical analysis Safer and more
efficient surgical procedures, together with enhanced
post-operative recovery, were demonstrated in Group
A. Statistical characteristics and results for comparative
analyses of intra-/post–operative data are described in
Table 2. Discussion TLPD has been widely accepted and applied in many
medical centers globally [23]. During the learning / appli
cation process for TLPD, our surgical team embarked
upon the uncinate-process-first, novel-artery-first and
inferior-posterior duodenal approach [24–26] while also
following the principle of ‘no touch’ in every procedure,
from the beginning of such surgery [27]. To overcome
the learning period, this study preferred and recom
mended the surgical options of ‘easy and safe steps first’. Based on our experience of > 700 LPD cases between
2013 and 2020, the key points to achieve LPD proficiency
highly depended upon the technical skills of other lapa
roscopic surgeries, abundant OPD experience and effec
tive teamwork. This study considered hand-assisted LPD,
TLPD, TLPD with total mesopancreas excision (TMpE),
and TLPD with vessel reconstruction, as four technical
marks. Surgical teams in our department gradually and Learning curves with CUSUM methodh Learning curves with CUSUM method
The
learning
curves
of
CUSUMOT,
CUSUMPJT,
CUSUMCJT and CUSUMGJT for Surgeon A and Sur
geon B were developed, respectively. CUSUMOT curves
(Fig. 1) were divided into three phases, of which the sec
ond phases (the technical plateau phases) were observed
approximately at Case # 25–41 and Case #35–51, for Page 5 of 10 Li et al. BMC Surgery (2023) 23:92 Li et al. BMC Surgery Table 2 Statistical characteristics and results of the comparison tests of intra/post-operative data
Parameter
Group A(127 cases)
Group B(112 cases)
P
Intraoperative
Pancreatic texture [n (%)]
0.576*
Soft pancreas
113(88.98%)
97(86.61%)
Not-soft pancreas
14(11.02%)
15(13.39%)
MPD size(cm) a
0.3
0.3
0.651+
Blood loss(mL)a
400
500
< 0.001+
Transfusion [n (%)]
57(44.88%)
88(78.57%)
< 0.001*
Transfusion volume(mL)a
0
400
< 0.001+
Operation time(min)
424.41 ± 91.41
449.11 ± 76.29
0.020#
Pancreatojejunostomy time(min)a
40.00
50.00
< 0.001+
Choledochojejunostomy time(min)a
20.00
30.00
< 0.001+
Gastrojejunostomy time(min)a
15.00
25.00
< 0.001+
Resection time(min)
347.44 ± 84.22
345.94 ± 71.54
0.883#
Complications [n (%)]
Pancreatic fistula (BL/A/B/C)
96(75.59%)/2/11/7
81(72.32%)/7/11/3
0.311+
Delayed gastric emptying
3(2.36%)
6(5.35%)
0.225*
Postoperative hemorrhage
9(7.09%)
15(16.07%)
0.106*
Abdominal infection
13(10.24%)
20(17.86%)
0.088*
Clavien-Dindo classification [n (%)]
Minor(Grade I and II)
Major(Grade III to V)
36(28.35%)
14(11.02%)
45(40.18%)
34(30.36%)
< 0.001*
Reoperation [n (%)]
7(5.51%)
14(12.50%)
0.057*
Death in hospital [n (%)]
4(3.15%)
4(3.57%)
0.856*
Hospital stay (day)a
15
19
< 0.001+
ICU stay (day)a
0
0
< 0.001+
Pathological results
Tumor Location [n (%)]
0.132*
Pancreatic head tumor
42(33.07%)
51(45.53%)
Bile duct tumor
44(34.64%)
28(25.00%)
Duodenal tumor
31(24.41%)
21(18.75%)
Ampulla tumor
10(7.87%)
12(10.71%)
Harvested lymph nodesa
12(10,14)
6(2,8)
< 0.001+
Resection margin (positive/negative)
4(3.15%)/123(96.85%)
7(6.25%)/105(93.75%)
0.254*
a: Median of the parameter; #: Independent-samples t tests; *: Chi-square tests; +: Wilcoxon rank sum tests. were required to reach the turning point respectively -
after which there was significant reduction of blood loss
and operation time [7, 8, 17, 32, 33]. Through CUSUM
methods, the learning curve for LPD revealed that 12–38
out of 57 cases, 21–30 out of 50 cases, 20–25 out of 50
cases, 34–65 cases out of 98 cases, 47 out of 119 cases,
41–100 cases out of 171 cases, 55 cases (for the first-
generation surgeon) out of 500 cases and 40–104 out of
133 cases (a retrospective multicenter analysis of 1029
patients) were required for proficiency in LPD, respec
tively [9, 10, 12–16, 18]. Learning curves with CUSUM methodh In summary, initial peak for
learning curves involving > 50 cases was mainly within
the range of 40–50 cases [7–10, 12–18]. This study con
cluded that sample size was determined according to the
summary of learning curve-related studies. successfully performed hand-assisted LPD and TLPD,
aiming at accomplishing these procedures: TLPD with
total mesopancreas excision (TMpE) and TLPD with ves
sel reconstruction, focusing on the precision and stan
dardization of such procedures. The required training period could be determined
by learning curves for TLPD. Several studies suggested
that a steep LPD learning curve - that impacts patient
outcomes - could be positively affected by appropriate
training, high-volume practice/institution, proficient
mentorship and an experienced multi-disciplinary team
[6–8, 12, 16–18, 28–30]. Differing results for LPD learn
ing curves were presented from previous studies, based
upon differing methods. By statistically analyzing clinical
data related to LPD, the learning curve was presented to
be a steep curve [28, 31]. Through pre-defining the learn
ing periods, 5 out of 12 cases, 10 out of 30 cases, 50 out of
56 cases, 30–60 out of 120 cases, and 47 out of 473 cases Concomitant to the rapid development of laparoscopic
technology, laparoscopic platforms are widely used
within the field of hepatobiliary and pancreatic surgery. Li et al. BMC Surgery (2023) 23:92 Page 6 of 10 Li et al. BMC Surgery Nearly all such procedures can be accomplished through
laparoscopy, within large medical centers. Junior sur
geons have reduced opportunities to commence practice
on open-surgery procedures. Regarding LPD, which is
the one of the most challenging surgical pr
confirmation of how a surgeon with no OP
would be competent enough to perform a T
cial importance. In order to mitigate such
Fig. 2 Cumulative sum graphs for pancreatojejunostomy time (CUSUMPJT) of Surgeon A and Surgeon B
Fig. 1 Cumulative sum graphs for operation time (CUSUMOT) of Surgeon A and Surgeon B Fig. 1 Cumulative sum graphs for operation time (CUSUMOT) of Surgeon A and Surgeon B Fig. 1 Cumulative sum graphs for operation time (CUSUMOT) of Surgeon A and Surgeon B Fig. 1 Cumulative sum graphs for operation time (CUSUMOT) of Surgeon A and Surgeon B Fig. 2 Cumulative sum graphs for pancreatojejunostomy time (CUSUMPJT) of Surgeon A and Surgeon B Fig. Learning curves with CUSUM methodh 2 Cumulative sum graphs for pancreatojejunostomy time (CUSUMPJT) of Surgeon A and Surgeon B the one of the most challenging surgical procedures, the
confirmation of how a surgeon with no OPD experience
would be competent enough to perform a TLPD, has cru
cial importance. In order to mitigate such an issue, this the one of the most challenging surgical procedures, the
confirmation of how a surgeon with no OPD experience
would be competent enough to perform a TLPD, has cru
cial importance. In order to mitigate such an issue, this Nearly all such procedures can be accomplished through
laparoscopy, within large medical centers. Junior sur
geons have reduced opportunities to commence practice
on open-surgery procedures. Regarding LPD, which is Li et al. BMC Surgery (2023) 23:92 Page 7 of 10 Page 7 of 10 Li et al. BMC Surgery study aimed at probing the influence of FAE upon TLPD
learning curve for an operator. This investigation also
aimed to minimize such factor effects, possibly influenc
ing TLPD performance across both groups. These factors
included laparoscopic equipment, surgical tea
characteristics, pancreas consistency and MPD
Both junior surgeons in this study had at lea
working experience, and were competent in fu
Fig. 4 Cumulative sum graphs for gastrojejunostomy time (CUSUMGJT) of Surgeon A and Surgeon B
Fig. 3 Cumulative sum graphs for choledochojejunostomy time (CUSUMCJT) of Surgeon A and Surgeon B Fig. 3 Cumulative sum graphs for choledochojejunostomy time (CUSUMCJT) of Surgeon A and Surgeon B Fig. 3 Cumulative sum graphs for choledochojejunostomy time (CUSUMCJT) of Surgeon A and Surgeon B study aimed at probing the influence of FAE upon TLPD
learning curve for an operator. This investigation also
aimed to minimize such factor effects, possibly influenc
ing TLPD performance across both groups. These factors
included laparoscopic equipment, surgical team, patient
characteristics, pancreas consistency and MPD diameter. Both junior surgeons in this study had at least 6-years’
working experience, and were competent in fundamental
Fig. 4 Cumulative sum graphs for gastrojejunostomy time (CUSUMGJT) of Surgeon A and Surgeon B Fig. 4 Cumulative sum graphs for gastrojejunostomy time (CUSUMGJT) of Surgeon A and Surgeon B included laparoscopic equipment, surgical team, patient
characteristics, pancreas consistency and MPD diameter. Both junior surgeons in this study had at least 6-years’
working experience, and were competent in fundamental study aimed at probing the influence of FAE upon TLPD
learning curve for an operator. Learning curves with CUSUM methodh This investigation also
aimed to minimize such factor effects, possibly influenc
ing TLPD performance across both groups. These factors Page 8 of 10 Li et al. BMC Surgery (2023) 23:92 Li et al. BMC Surgery Table 3 Statistical characteristics and results of the comparison tests of different phases of the LPD learning curve of Surgeon A. a:
Median of the parameter; #: Independent-samples t tests; *: Chi-square tests; +: Kruskal-Wallis H tests
Parameter
Phase 1(Case #1–24)
Phase 2(Case #25–31)
Phase 3(Case #32–127)
P
Intraoperative
Pancreatic texture [n (%)]
0.873*
Soft pancreas
22(91.67%)
6(85.71%)
85(88.54%)
Not-soft pancreas
2(8.33%)
1(14.29%)
11(11.46%)
MPD size(cm) a
0.3
0.3
0.3
0.711+
Blood loss(mL)a
550
300
300
0.011+
Transfusion [n (%)]
14(58.33%)
3(42.86%)
40(41.67%)
0.338*
Transfusion volume(mL)a
200
200
0
0.074+
Complications [n (%)]
Pancreatic fistula (BL/A/B/C)
20(83.33%)/0/3/1
5(71.43%)/0/1/1
71(73.96%)/2/7/5
0.604
Delayed gastric emptying
1(4.17%)
0
2(2.08%)
0.763*
Postoperative hemorrhage
2(8.33%)
2(28.57%)
5(5.21%)
0.020*
Abdominal infection
1(4.17%)
1(14.28%)
11(11.46%)
0.435*
Clavien-Dindo classification [n (%)]
< 0.001*
Minor(Grade I and II)
11(45.83%)
4(57.14%)
21(21.87%)
Major(Grade III to V)
8(33.33%)
1(14.28%)
7(7.29%)
Reoperation [n (%)]
5(20.83%)
0
2(2.08%)
0.001*
Death in hospital [n (%)]
2(8.33%)
0
2(2.08%)
0.146*
Hospital stay (day)a
19
15
10
0.097+
ICU stay (day)a
2
0
0
0.014+ Table 3 Statistical characteristics and results of the comparison tests of different phases of the LPD learning curve of Surgeon A. Learning curves with CUSUM methodh a:
Median of the parameter; #: Independent-samples t tests; *: Chi-square tests; +: Kruskal-Wallis H tests
h
h
h Table 4 Statistical characteristics and results of the comparison tests of different phases of the LPD learning curve of Surgeon B
Parameter
Phase 1(Case #1–34)
Phase 2(Case #35–41)
Phase 3(Case #42–112)
P
Intraoperative
Pancreatic texture [n (%)]
0.958*
Soft pancreas97
29(85.29%)
6(85.71%)
62
Not-soft pancreas15
5(14.71%)
1(14.29%)
9
MPD size(cm) a
0.3
0.3
0.3
0.651+
Blood loss(mL)a
400
400
300
0.025+
Transfusion [n (%)]
28(82.35%)
5(71.43%)
55(77.46%)
0.787
Transfusion volume(mL)a
200
200
0
0.149+
Complications [n (%)]
Pancreatic fistula (BL/A/B/C)
30/1/1/2
4/1/2/0
47/5/8/1
0.227*
Delayed gastric emptying
3(8.82%)
0
3(4.23%)
0.555*
Postoperative hemorrhage
8(23.53%)
2(28.57%)
5(7.04%)
0.035*
Abdominal infection
2(5.89%)
1(14.28%)
17(23.94%)
0.075*
Clavien-Dindo classification [n (%)]
0.013*
Minor(Grade I and II)
21(61.76%)
4(57.14%)
20(28.17%)
Major(Grade III to V)
8(23.53%)
2(28.57%)
24(33.80%)
Reoperation [n (%)]
5(14.71%)
3(42.86%)
6(8.45%)
0.029*
Death in hospital [n (%)]
2(5.89%)
0
2(2.82%)
0.636*
Hospital stay (day)a
20
21
18
0.457+
ICU stay (day)a
2
0
0
0.003+
a: Median of the parameter; #: Independent-samples t tests; *: Chi-square tests; +: Kruskal-Wallis H tests statistically valid comparative analyses in this study sug
gested that enhanced post-operative recovery together
with increased quantities of harvested lymph nodes were
found within TLPD cases performed by a surgeon having
previous FAE for TLPD, when compared to TLPD cases
performed by a surgeon without previous FAE for TLPD. To prevent PPH in patients with deteriorating abdominal statistically valid comparative analyses in this study sug
gested that enhanced post-operative recovery together
with increased quantities of harvested lymph nodes were
found within TLPD cases performed by a surgeon having
previous FAE for TLPD, when compared to TLPD cases
performed by a surgeon without previous FAE for TLPD. To prevent PPH in patients with deteriorating abdominal laparotomic / laparoscopic hepatobiliary surgical proce
dures. Regarding PD, both surgeons had a FAE approxi
mating 30 OPD cases of OPD, respectively, though had
minimal experience as OPD operators. This study aimed
at exploring the influence of FAE in TLPD on the learn
ing curve for an operator, together with technical com
petence for both surgeons could be referenced. The laparotomic / laparoscopic hepatobiliary surgical proce
dures. Regarding PD, both surgeons had a FAE approxi
mating 30 OPD cases of OPD, respectively, though had
minimal experience as OPD operators. Consent for publication
Not Applicable. 21. Wente MN, Bassi C, Dervenis C, et al. Delayed gastric emptying (DGE) after
pancreatic surgery: a suggested definition by the International Study Group
of pancreatic surgery (ISGPS). Surgery. 2007;142(5):761–8. Received: 31 October 2022 / Accepted: 4 April 2023 Received: 31 October 2022 / Accepted: 4 April 2023 Received: 31 October 2022 / Accepted: 4 April 2023 infection, we chose to initiatively perform laparoscopic
debridement within one week post-LPD. Although that
would lead to a slightly high reoperation rate, which was
around 5–10% of the two surgeons, it was much safer for
the patients. TLPD learning curves for both surgeons
showed that the surgeon with FAE required less TLPD
cases(# 25–41 cases) to accomplish prowess, in com
parison to the surgeon with no previous FAE for TLPD(#
35–51 cases), in order to achieve technical competence
for TLPD. Intra-surgery datasets (blood loss) and post-
surgery datasets (post-surgery major complication rates,
ICU durations and repeated surgery) of the three phases
on the learning curve showed a descending trend for both
surgeons, which was consistent with the learning curve. Acknowledgements Acknowledgements
Not Applicable. 13. Huang L, Tian Y, Wu J, et al. The effectiveness, risks and improvement of lapa
roscopic pancreaticoduodenectomy during the learning curve: a propensity
score-matched analysis. Gland Surg. 2020;9(4):985–99. 13. Huang L, Tian Y, Wu J, et al. The effectiveness, risks and improvement of lapa
roscopic pancreaticoduodenectomy during the learning curve: a propensity
score-matched analysis. Gland Surg. 2020;9(4):985–99. Availability of data and materials The datasets used and analyzed during the current study available from the
corresponding author on reasonable request. 17. Wang X, Cai Y, Jiang J, Peng B. Laparoscopic pancreaticoduodenectomy:
outcomes and experience of 550 patients in a single Institution. Ann Surg
Oncol. 2020;27(11):4562–73. 18. Wang M, Peng B, Liu J, et al. Practice patterns and perioperative outcomes of
laparoscopic pancreaticoduodenectomy in China: a retrospective Multicenter
analysis of 1029 patients. Ann Surg. 2021;273(1):145–53. Authors’ contributions DRL, CXD, WBW, JSZ and LJH performed the LPD procedures. DRL, CXD, WBW
and JSZ collected and analyzed the preclinical and clinical data. DRL and CXD
wrote the manuscript, prepared the figures and tables. All authors read and
approved the final manuscript. 14. Kim S, Yoon YS, Han HS, Cho JY, Choi Y, Lee B. Evaluation of a single surgeon’s
learning curve of laparoscopic pancreaticoduodenectomy: risk-adjusted
cumulative summation analysis. Surg Endosc. 2021;35(6):2870–8. 15. Morató O, Poves I, Burdío F, Sánchez-Velázquez P, Duran X, Grande L. Evalu
ation of the learning curve for laparoscopic pancreatoduodenectomy by
CUSUM analyses. Cohort study. Int J Surg. 2020;80:61–7. Funding
Thi
k This work was supported by the Nature Science Foundation of Hebei Province
of China under Grant No. H2022206284. 16. Song KB, Kim SC, Lee W, et al. Laparoscopic pancreaticoduodenectomy for
periampullary tumors: lessons learned from 500 consecutive patients in a
single center. Surg Endosc. 2020;34(3):1343–52. References 1. Zhang Y, Hong D, Zhang C, Hu Z. Total laparoscopic versus robot-assisted
laparoscopic pancreaticoduodenectomy. Biosci Trends. 2018;12(5):484–90. 2. Zhang H, Lan X, Peng B, Li B. Is total laparoscopic pancreaticoduodenec
tomy superior to open procedure? A meta-analysis. World J Gastroenterol. 2019;25(37):5711–31. 3. Asbun HJ, Stauffer JA. Laparoscopic vs open pancreaticoduodenectomy:
overall outcomes and severity of complications using the Accordion Severity
Grading System. J Am Coll Surg. 2012;215(6):810–9. y
4. Stauffer JA, Coppola A, Villacreses D, et al. Laparoscopic versus open pancre
aticoduodenectomy for pancreatic adenocarcinoma: long-term results at a
single institution. Surg Endosc. 2017;31(5):2233–41. g
g
5. Liao CH, Wu YT, Liu YY, et al. Systemic review of the feasibility and advan
tage of minimally invasive pancreaticoduodenectomy. World J Surg. 2016;40(5):1218–25. There were limitations to this study. Firstly, there was a
lack of comparative analyses for long-term results. Sec
ondly, only two surgeons were involved in this study. Additional surgeons with differing technical levels of
laparotomic and laparoscopic hepatobiliary surgery
procedures, together with larger cohort sizes should be
involved in further studies, in order to better-assess the
positive factors for accelerating TLPD learning curve. 6. van Hilst J, de Rooij T, Bosscha K, et al. Laparoscopic versus open pancre
atoduodenectomy for pancreatic or periampullary tumours (LEOPARD-2): a
multicentre, patient-blinded, randomised controlled phase 2/3 trial. Lancet
Gastroenterol Hepatol. 2019;4(3):199–207. 7. Speicher PJ, Nussbaum DP, White RR, et al. Defining the learning curve for
team-based laparoscopic pancreaticoduodenectomy. Ann Surg Oncol. 2014;21(12):4014–9. 7. Speicher PJ, Nussbaum DP, White RR, et al. Defining the learning curve for
team-based laparoscopic pancreaticoduodenectomy. Ann Surg Oncol. 2014;21(12):4014–9. 8. Lu C, Jin W, Mou YP, et al. Analysis of learning curve for laparoscopic pancre
aticoduodenectomy. J Vis Surg. 2016;2:145. 8. Lu C, Jin W, Mou YP, et al. Analysis of learning curve for laparoscopic pancre
aticoduodenectomy. J Vis Surg. 2016;2:145. In conclusion, TLPD was a feasible, safe and effective
surgery when performed by second-generation TLPD
surgeons. FAE in TLPD can accelerate the learning curve
for TLPD within an operator - with safer surgical proce
dures and enhanced post-operative recovery. Training
and studying in large-volume TLPD centers could be a
safer and more efficient method to positively affect the
learning curve and clinical outcomes for such patients. 9. Wang M, Meng L, Cai Y, et al. Learning curve for laparoscopic pancreatico
duodenectomy: a CUSUM Analysis. J Gastrointest Surg. 2016;20(5):924–35. 9. Wang M, Meng L, Cai Y, et al. References Learning curve for laparoscopic pancreatico
duodenectomy: a CUSUM Analysis. J Gastrointest Surg. 2016;20(5):924–35. 10. Nagakawa Y, Nakamura Y, Honda G, et al. Learning curve and surgical fac
tors influencing the surgical outcomes during the initial experience with
laparoscopic pancreaticoduodenectomy. J Hepatobiliary Pancreat Sci. 2018;25(11):498–507. 10. Nagakawa Y, Nakamura Y, Honda G, et al. Learning curve and surgical fac
tors influencing the surgical outcomes during the initial experience with
laparoscopic pancreaticoduodenectomy. J Hepatobiliary Pancreat Sci. 2018;25(11):498–507. 11. van Hilst J, de Rooij T, van den Boezem PB, et al. Laparoscopic pancre
atoduodenectomy with open or laparoscopic reconstruction during the
learning curve: a multicenter propensity score matched study. HPB (Oxford)
2019;21(7):857–64. 11. van Hilst J, de Rooij T, van den Boezem PB, et al. Laparoscopic pancre
atoduodenectomy with open or laparoscopic reconstruction during the
learning curve: a multicenter propensity score matched study. HPB (Oxford). 2019;21(7):857–64. 12. Choi M, Hwang HK, Lee WJ, Kang CM. Total laparoscopic pancreaticoduode
nectomy in patients with periampullary tumors: a learning curve analysis. Surg Endosc. 2021;35(6):2636–44. 12. Choi M, Hwang HK, Lee WJ, Kang CM. Total laparoscopic pancreaticoduode
nectomy in patients with periampullary tumors: a learning curve analysis. Surg Endosc. 2021;35(6):2636–44. Learning curves with CUSUM methodh This study aimed
at exploring the influence of FAE in TLPD on the learn
ing curve for an operator, together with technical com
petence for both surgeons could be referenced. The Li et al. BMC Surgery (2023) 23:92 Li et al. BMC Surgery (2023) 23:92 Page 9 of 10 Page 9 of 10 Ethics approval and consent to participate 19. Bassi C, Marchegiani G, Dervenis C, et al. The 2016 update of the International
Study Group (ISGPS) definition and grading of postoperative pancreatic
fistula: 11 years after. Surgery. 2017;161(3):584–91. This study was performed in accordance with the relevant guidelines and
regulations and was approved by the Research Ethics Committee of the
Second Hospital of Hebei Medical University (No. 2019-R209). Informed
consent was obtained from all the study subjects. 20. Wente MN, Veit JA, Bassi C, et al. Postpancreatectomy hemorrhage (PPH): an
International Study Group of pancreatic surgery (ISGPS) definition. Surgery. 2007;142(1):20–5. Conflict of interest Reduced dissemination of circulating
tumor cells with no-touch isolation surgical technique in patients with
pancreatic cancer. JAMA Surg. 2014;149(5):482–5. Conflict of interest 22. Clavien PA, Barkun J, de Oliveira ML, et al. The Clavien-Dindo classification of
surgical complications: five-year experience. Ann Surg. 2009;250(2):187–96. 22. Clavien PA, Barkun J, de Oliveira ML, et al. The Clavien-Dindo classification of
surgical complications: five-year experience. Ann Surg. 2009;250(2):187–96. l
The authors declare that they have no competing interests. Page 10 of 10 Li et al. BMC Surgery (2023) 23:92 Li et al. BMC Surgery (2023) 23:92 23. Wang M, Cai H, Meng L, et al. Minimally invasive pancreaticoduodenectomy:
a comprehensive review. Int J Surg. 2016;35:139–46. 30. de Rooij T, van Hilst J, Topal B, et al. Outcomes of a Multicenter Training
Program in Laparoscopic pancreatoduodenectomy (LAELAPS-2). Ann Surg. 2019;269(2):344–50. 30. de Rooij T, van Hilst J, Topal B, et al. Outcomes of a Multicenter Training
Program in Laparoscopic pancreatoduodenectomy (LAELAPS-2). Ann Surg. 2019;269(2):344–50. 30. de Rooij T, van Hilst J, Topal B, et al. Outcomes of a Multicenter Training
Program in Laparoscopic pancreatoduodenectomy (LAELAPS-2). Ann Surg. 2019;269(2):344–50. 24. Zhang Y, Sun DL, Chen XM. The uncinate process First Approach in Lapa
roscopic Pancreaticoduodenectomy: a single-institution experience. Surg
Laparosc Endosc Percutan Tech. 2017;27(6):e141–4. 31. Patel B, Leung U, Lee J, et al. Laparoscopic pancreaticoduodenectomy in
Brisbane, Australia: an initial experience. ANZ J Surg. 2018;88(5):E440–4. 31. Patel B, Leung U, Lee J, et al. Laparoscopic pancreaticoduodenectomy in
Brisbane, Australia: an initial experience. ANZ J Surg. 2018;88(5):E440–4. 25. Gao WT, Xi CH, Tu M, et al. [Laparoscopic pancreaticoduodenectomy with
a novel artery first and uncinate process first approach through Treitz liga
ment]. Zhonghua Wai Ke Za Zhi. 2017;55(5):359–63. 32. Kuroki T, Kitasato A, Adachi T, et al. Learning curve for laparoscopic pancreati
coduodenectomy: a single surgeon’s experience with consecutive patients. Hepatogastroenterology. 2014;61(131):838–41. p
g
gy
33. Liao CH, Liu YY, Wang SY, Liu KH, Yeh CN, Yeh TS. The feasibility of laparoscopic
pancreaticoduodenectomy-a stepwise procedure and learning curve. Lan
genbecks Arch Surg. 2017;402(5):853–61. 33. Liao CH, Liu YY, Wang SY, Liu KH, Yeh CN, Yeh TS. The feasibility of laparoscopic
pancreaticoduodenectomy-a stepwise procedure and learning curve. Lan
genbecks Arch Surg. 2017;402(5):853–61. 26. Wang XM, Sun WD, Hu MH, et al. Inferoposterior duodenal approach
for laparoscopic pancreaticoduodenectomy. World J Gastroenterol. 2016;22(6):2142–8. 27. Gall TM, Jacob J, Frampton AE, et al. Reduced dissemination of circulating
tumor cells with no-touch isolation surgical technique in patients with
pancreatic cancer. JAMA Surg. 2014;149(5):482–5. 27. Gall TM, Jacob J, Frampton AE, et al. Publisher’s Note Springer Nature remains neutral with regard to jurisdictional claims in
published maps and institutional affiliations. Springer Nature remains neutral with regard to jurisdictional claims in
published maps and institutional affiliations. Springer Nature remains neutral with regard to jurisdictional claims in
published maps and institutional affiliations. 28. Tsamalaidze L, Stauffer JA. Pancreaticoduodenectomy: minimizing the learn
ing curve. J Vis Surg. 2018;4:64. 28. Tsamalaidze L, Stauffer JA. Pancreaticoduodenectomy: minimizing the learn
ing curve. J Vis Surg. 2018;4:64. 29. Kuroki T, Fujioka H. Training for laparoscopic pancreaticoduodenectomy. Surg
Today. 2019;49(2):103–7. 29. Kuroki T, Fujioka H. Training for laparoscopic pancreaticoduodenectomy. Surg
Today. 2019;49(2):103–7.
|
https://openalex.org/W2120189956
|
https://europepmc.org/articles/pmc3866392?pdf=render
|
English
| null |
Guided Application of Ventricular Catheters (GAVCA) - multicentre study to compare the ventricular catheter position after use of a catheter guide versus freehand application: study protocol for a randomised trail
|
Trials
| 2,013
|
cc-by
| 6,066
|
© 2013 Schaumann and Thomale; licensee BioMed Central Ltd. This is an Open Access article distributed under the terms of
the Creative Commons Attribution License (http://creativecommons.org/licenses/by/2.0), which permits unrestricted use,
distribution, and reproduction in any medium, provided the original work is properly cited. The Creative Commons Public
Domain Dedication waiver (http://creativecommons.org/publicdomain/zero/1.0/) applies to the data made available in this
article, unless otherwise stated. Abstract Background: The standard technique for the placement of ventricular catheters (VC) comprises a high proportion
of malpositioning of the catheter (12.5 to 40%). Technical advances such as neuronavigation or ultrasound have
been shown to increase the accuracy of the procedure. Since these means result in significant technical and time
consuming efforts, they are used for selected cases only. In order to simplify the controlled placement of ventricular
catheters a newly developed smartphone assisted guiding tool has been introduced. In this study the efficacy and
safety of this guiding tool is determined. Methods/design: This study is a multicentre, randomised, controlled trial. A total of 144 patients planned for an
elective shunting procedure will be enrolled throughout 10 study centres within two years. The primary objective
of the trial is to show the superiority of the guided placement in comparison to the standard freehand technique
of ventricular catheter application. Patients will be followed up for 30 days after the operation in regard to
image-based evaluation of the catheter position as well as possible shunt dysfunction and complications. Discussion: The Guided Application of Ventricular Catheters (GAVCA) trial compares the guided catheter
positioning with the standard freehand technique of catheter placement in hydrocephalic patients. If superiority
is shown, the standard technique may be changed with the advantage of a more reliable and safer positioning of
the ventricular catheter with just a slight effort in time and pre-operative planning. egistration: The GAVCA trial is registered at ClinicalTrials.gov under the number NCT01811589. already described [2-7]. Several studies estimate the pro-
portion of inaccuracy in the standard freehand technique
to be in the range of 12.5 to 40% [2,4,8,9]. This freehand
technique is based on the guidance by anatomical land-
marks and the experience of the surgeon. In selected,
more complicated cases technical advances such as neuro-
navigation or ultrasound might be used. However, this
comprises high technical and time consuming efforts within
a rather short procedure [4,10-13]. Seeing this problem, a
more simplified technique is needed to be applied in a wide Guided Application of Ventricular Catheters
(GAVCA) - multicentre study to compare the
ventricular catheter position after use of a catheter
guide versus freehand application: study protocol
for a randomised trail Andreas Schaumann and Ulrich-Wilhelm Thomale* Andreas Schaumann and Ulrich-Wilhelm Thomale* STUDY PROTOCOL Open Access Guided Application of Ventricular Catheters
(GAVCA) - multicentre study to compare the
ventricular catheter position after use of a catheter
guide versus freehand application: study protocol
for a randomised trail TRIALS TRIALS Schaumann and Thomale Trials 2013, 14:428
http://www.trialsjournal.com/content/14/1/428 * Correspondence: uthomale@charite.de
Department of Neurosurgery/Paediatric Neurosurgery, Campus Virchow
Klinikum, Charité – Universitätsmedizin Berlin, Augustenburger Platz 1, 13353
Berlin, Germany Background The transcortical puncture of the ventricles is considered
to be a simple neurosurgical procedure and is performed
as one of the first procedures in neurosurgical training [1]. The ventricular catheter is considered as a lifelong func-
tioning implant in the case of a shunting procedure; there-
fore special focus should be dedicated to the accurate
positioning of the catheter as a correlation between cath-
eter position and risk for proximal shunt obstruction is * Correspondence: uthomale@charite.de
Department of Neurosurgery/Paediatric Neurosurgery, Campus Virchow
Klinikum, Charité – Universitätsmedizin Berlin, Augustenburger Platz 1, 13353
Berlin, Germany Schaumann and Thomale Trials 2013, 14:428
http://www.trialsjournal.com/content/14/1/428 Page 2 of 8 range of procedures. A mechanical catheter-guiding instru-
ment was developed which enables a controlled placement
of the catheters by applying one individual measure to the
patients’ anatomy [14]. The instrument is suitable for a
frontal pre-coronal approach. In earlier measurements
based on MR images by using the foramen of Monro as
a target, the angle towards the tangent of the skull sur-
face was found to be relevant for the correct trajectory
of puncture. Hereby, it could be shown that this angle is
always rectangular in the sagittal plane whereas in the
coronal plane an individual angulation must be measured
respectively, with a mean deviation from rectangular of
1.96° ± 2.75° towards the medial margin up to 9.56° ± 4.14°
towards the lateral margin of the ventricle [14]. That led
to the development of the ventricular catheter guide
which addresses a strict trajectory of a 90° angle to the
sagittal surface of the skull and may be adjusted to an
individual angle to the coronal surface. The measure-
ments for the correct adjustment of the catheter guide
are done on a coronal section of the cranium at a frontal
paramedian entry point. Since regular DICOM viewing
software does not allow an easy measurement of this
parameter, this is facilitated in an iPhone application
software (Version 1.1.1 for iOS6.1, Apple Inc., Cupertino,
USA). Herein, the guiding instrument can virtually be
visualised and placed on the skull of a coronal recon-
structed section of the cranium for easy measurement
of the trajectory’s angle towards the coronal surface of
the skull. range of procedures. A mechanical catheter-guiding instru-
ment was developed which enables a controlled placement
of the catheters by applying one individual measure to the
patients’ anatomy [14]. The instrument is suitable for a
frontal pre-coronal approach. Interventions The procedure of cerebrospinal fluid shunting in hydro-
cephalus patients is an accepted neurosurgical standard. The placement of a ventricular catheter is a surgical skill
that is taught early within neurosurgical training. There-
fore, every neurosurgeon and neurosurgical resident is
applicable to undertake the procedure within the trial as
long as the handling of the catheter guide was taught
within a standardised training. The technical feasibility of the smartphone assisted
guiding tool was verified in a single centre retrospective
study [14]. In comparison to earlier studies on freehand
ventricular placement, an advantage in the first pass
cannulation and positioning was achieved. The aim of the Guided Application of Ventricular
Catheters (GAVCA) study is to show that the ventricular
catheter guiding tool is superior to the standard freehand
technique in terms of correct catheter position within the
ipsilateral ventricle in a randomised multicentre trial. The surgical technique is standardised as follows. The
preparation of the skin flap is performed in order to
expose at least 2 cm in diameter of the precoronal
paramedian skull surface. One bore hole trepanation is
performed at a localisation of 10 to 12.5 cm measured
from nasion and 2 to 3.5 cm measured from midline in
accordance to the individualised software measurements. The dura mater is to be coagulated and opened in order
to enable the catheter puncture without being deviated at
the dura level. The brain surface is punctiformly coagu-
lated to open the leptomeningeal layers. Methods/design
Objectives This study is designed as a prospective, randomized,
two-armed, multicentre trial comparing two surgical pro-
cedures (Figure 1). The primary objective of the study is
to prove the superiority of the guided ventricular catheter
placement (treatment arm 1) versus standard freehand
technique (treatment arm 2) in terms of correct catheter
position in the ipsilateral ventricle. In treatment arm 1, the catheter guide is used with the
individual parameters of coronal angulation to the skull
surface and catheter length measured in a coronal sec-
tion of the cranium with the catheter guide application
software. Background In earlier measurements
based on MR images by using the foramen of Monro as
a target, the angle towards the tangent of the skull sur-
face was found to be relevant for the correct trajectory
of puncture. Hereby, it could be shown that this angle is
always rectangular in the sagittal plane whereas in the
coronal plane an individual angulation must be measured
respectively, with a mean deviation from rectangular of
1.96° ± 2.75° towards the medial margin up to 9.56° ± 4.14°
towards the lateral margin of the ventricle [14]. That led
to the development of the ventricular catheter guide
which addresses a strict trajectory of a 90° angle to the
sagittal surface of the skull and may be adjusted to an
individual angle to the coronal surface. The measure-
ments for the correct adjustment of the catheter guide
are done on a coronal section of the cranium at a frontal
paramedian entry point. Since regular DICOM viewing
software does not allow an easy measurement of this
parameter, this is facilitated in an iPhone application
software (Version 1.1.1 for iOS6.1, Apple Inc., Cupertino,
USA). Herein, the guiding instrument can virtually be
visualised and placed on the skull of a coronal recon-
structed section of the cranium for easy measurement
of the trajectory’s angle towards the coronal surface of
the skull. Figure 1 Study protocol flow chart representing the
randomisation process. Figure 1 Study protocol flow chart representing the
randomisation process. Secondary endpoints – number of cannulation attempts until cerebrospinal
fluid flow sets in – the proportion of correct intraventricular
localisation of the perforated catheter part – early shunt failures – complication rates (adverse events) Follow-up In addition, the postoperative image will be evaluated for
the position of the respective catheter tip in the ipsilateral,
contralateral, third ventricle or in any extraventricular
compartment. Therefore the length of the intraventricular
catheter part will be measured. Compared to preoperative
ventricular width values, the postoperative FOHR and
FOHWR will also be measured. As adverse events in im-
aging, any possible bleeding or air inclusions will be detected. At the day of surgery relevant perioperative parameters
are reported. This includes, for both treatment arms,
the age, sex, the underlying cause for hydrocephalus,
the ventricular width according to the frontal occipital
horn ratio (FOHR) and the frontal occipital width ratio
(FOHWR), the personal and the device-specific expert-
ise level of the operating surgeon, the number of cannu-
lation attempts, the type of ventricular catheter, the
localisation of the entry point, the time of surgery, and
any adverse events. Primary endpoints – the proportion of primary grade I or grade Ib
catheter tip position in the ipsilateral ventricle – the proportion of primary grade I or grade Ib
catheter tip position in the ipsilateral ventricle In treatment arm 2, the catheter placement is under-
taken in freehand technique. Page 3 of 8 Page 3 of 8 Schaumann and Thomale Trials 2013, 14:428
http://www.trialsjournal.com/content/14/1/428 – grade I: ideal catheter position without contact to
the ventricular wall of more than 0.5 cm – grade I: ideal catheter position without contact to
the ventricular wall of more than 0.5 cm In both arms, the catheter is fixed by a rectangular
redirector of the catheter from the bore hole to the sub-
cutaneous tissue, such as a bore hole reservoir or a 90°
connector. – grade Ib: contact to the medial and lateral
ventricular wall in the case the ventricular width
equals the catheter diameter In case of an unsuccessful cannulation, the surgeon
will re-try the puncture in the freehand group. In the
guided catheter group, the surgeon will repeat the use
of the guide in case any mistakes in its application are
detected and may need to be corrected, otherwise a
crossover to the freehand technique will be performed. – grade II: contact to the ventricular wall or the
choroid plexus of more than 0.5 cm – grade III: partly intraventricular position of the
catheter (less than 1.5 cm intraventricular) – grade IV: extraventricular position of the catheter
(less than 0.5 cm intraventricular) Study software
I
d
id In order to identify the ideal trajectory in the coronal
plane an image processing software (https://itunes.apple. com/en/app/thomale-guide-application/id648839804?mt=8
Thomale Guide App, Version 1.1.1 for iOS 6.1, Apple Inc.,
Cupertino, USA) was developed for the iPhone/iPad, which
is also approved in the EU (certificate registration number:
009066 MR2). This software can be downloaded for free
from the Apple app store [14]. The application is used to
measure the angle of the trajectory, the catheter length
and the position of the bore hole relative to midline from
imported images (Figure 4). Alternatively, the individual Study tool The individual angulation of the catheter guide and
the length of the implanted catheter are reported for the
treatment arm as measured in the ventricular guide
application software. After surgery a MRI-scan or CT-
scan as postoperative control is performed to docu-
ment the catheter position and any adverse events, such
as intra-cerebral haemorrhage. The follow-up examina-
tions will be performed on the day of discharge from
hospital and 30 to 40 days after surgery in order to re-
port catheter dysfunction with the need of a re-
operation and serious adverse events (Table 1). The catheter guide is a medical instrument, which is ap-
proved and bears the CE-Mark (Thomale Guide, certifi-
cate registration number: 009066 MR2) It is produced
by the Miethke company (Potsdam, Germany) and distrib-
uted by Aesculap, Inc. (Tuttlingen, Germany). The cath-
eter guide allows an individual angulation in the coronal
plane whereas the angulation in the sagittal plane is in a
strict rectangular direction (Figure 3). Rating of the catheter position The catheter position is rated in a centralised manner in
regard to the 2 cm tip of the catheter (perforated part)
by a neuroradiologist. The neuroradiologist is blinded
for the treatment arm and outcome of the patient and
evaluates the position of the catheter on the postoperative
reconstructed MRI or CT scans, which were performed
within 30 days after surgery (Figure 2). The rating will be
performed using the following protocol [15]. Table 1 Study protocol representing the relevant CRF visits
Method
Visit 0
Visit 1
pre-OP
Visit 2
peri-OP
Visit 3
discharge
Visit 4
end of study
CT catheter-control
screening, consent, inclusion
x
randomisation, patient data
x
surgery
x
perioperative study visit, complications
x
study visit, complications, early shunt failure
x
x
postoperative brain scan (until study visit 4)
x Table 1 Study protocol representing the relevant CRF visits Page 4 of 8 Schaumann and Thomale Trials 2013, 14:428
http://www.trialsjournal.com/content/14/1/428 Figure 2 Catheter position rating scale. Grade I: ideal catheter position with contact to the ventricular wall less than 0.5 cm. Grade Ib: contact
to the medial and lateral ventricular wall in the case of ventricular width equals the catheter diameter. Grade II: contact to the ventricular wall or
the choroid plexus of more than 0.5 cm. Grade III: only partial intraventricular position of the catheter (less than 1.5 cm intraventricular). Grade IV: extraventricular position of the catheter (less than 0.5 cm intraventricular). n 0.5 cm. Grade Ib: contact Figure 2 Catheter position rating scale. Grade I: ideal catheter position with contact to the ventricular wall less than 0.5 cm. Grade Ib: contac
to the medial and lateral ventricular wall in the case of ventricular width equals the catheter diameter. Grade II: contact to the ventricular wall o
the choroid plexus of more than 0.5 cm. Grade III: only partial intraventricular position of the catheter (less than 1.5 cm intraventricular). Grade IV: extraventricular position of the catheter (less than 0.5 cm intraventricular). patient’s parameters can also be obtained by processing
on an adequate DICOM-viewer-software (Centricity,
Gentricity 3.0, General Electrics Inc., Fairfield, USA), if the
user is familiar with that. – signed informed consent Exclusion criteria – previous known uneven bone surface at the site of
the approach Trial population The trial population will be recruited consecutively from
the hydrocephalic patients who are scheduled for a shunt-
ing procedure in the participating medical centres. The par-
ticipation in the trial ends with the final visit at Day 30 after
surgery. Further follow-up visits are not scheduled. A de-
tailed overview of the eligibility criteria are given as follows. – slit ventricles with an fronto-occipital horn width
ratio (FOHWR) <0.05 [17] – participation in other clinical trials with interfering
endpoints – patients unable to give informed consent Sample size The sample size was determined to verify the hypothesis
of superiority of treatment arm 1 (guided catheter place-
ment) in comparison to treatment arm 2 (freehand tech-
nique) in regard to the proportion of primary cannulation – hydrocephalic patient needing a shunting procedure hydrocephalic patient needing a shunting procedure
– fronto-occipital horn ratio (FOHR) <0.5 [16]
positioning ia a mat re brain tiss e trajector – fronto-occipital horn ratio (FOHR) <0.5 [16] – positioning via a mature brain tissue trajectory ull model as used during surgery. A: In the coronal plane, the
sagittal plane, the angulation to the skull surface is fixed at 90° Figure 3 Ventricular catheter guide placed on the surface of the skull model as used during surgery. A: In the coronal plane, the
individual angulation towards the skull surface can be adjusted. B: In the sagittal plane, the angulation to the skull surface is fixed at 90°. Figure 3 Ventricular catheter guide placed on the surface of the skull model as used during surgery. A: In the coronal plane, the
individual angulation towards the skull surface can be adjusted. B: In the sagittal plane, the angulation to the skull surface is fixed at 90°. Schaumann and Thomale Trials 2013, 14:428
http://www.trialsjournal.com/content/14/1/428 Page 5 of 8 Figure 4 Measurements acquired through the smartphone application. A: Coronal MRI section with moderate enlarged ventricles with
virtual placement of the guide on the skull surface. Angulation of catheter trajectory (red dotted line) is measured at 10° deviation from 90°
(light blue dotted line) tilting the guiding tube towards the midline. The catheter length (green dotted line) is measured at 67 mm. B: Coronal
CT reconstruction along body axis with the measurements revealing an angulation of 5.2° deviation from rectangular with the guiding tube tilt
towards the midline and a catheter length of 69 mm. Figure 4 Measurements acquired through the smartphone application. A: Coronal MRI section with moderate enlarged ventricles with
virtual placement of the guide on the skull surface. Angulation of catheter trajectory (red dotted line) is measured at 10° deviation from 90°
(light blue dotted line) tilting the guiding tube towards the midline. The catheter length (green dotted line) is measured at 67 mm. B: Coronal
CT reconstruction along body axis with the measurements revealing an angulation of 5.2° deviation from rectangular with the guiding tube tilt
towards the midline and a catheter length of 69 mm. Statistical analysis
Primary variable As we calculate with a 5% non-responder rate in treat-
ment arm 1, the positive response in treatment arm 1
was reduced to 78.8% according to the weighted average. The difference between the two balanced patient cohorts
is rated on a one-sided Chi-squared test with a signifi-
cance of 2.5% and a power of 80%. Sixty-five patients are
to be enrolled in each treatment arm as calculated with
the software STATA 12.1, (StataCorp, College Station, TX,
USA). Including a 10% dropout rate, the number increases
to 72 patients per treatment arm. The proportion of patients with a primary placement of
grade I or Ib position of the catheter in the ipsilateral
ventricle, as well as associated 95% confidence intervals,
will be reported for each treatment arm. The study null
hypothesis, which is to reject, states that the catheter
guided treatment arm is equal or inferior to the free-
hand technique arm regarding the proportion of correct
catheter positioning in the ipsilateral ventricle. This null
hypothesis has to be disproved to state the superiority
of the catheter guide treatment. Sample size on the postoperative scans by the treatment-blinded out-
come evaluator, which is an experienced neuroradiologist. of grade I or grade Ib in the ipsilateral ventricle. A suc-
cessful catheter positioning of grade I or Ib in the ipsilat-
eral ventricle is regarded as a positive response. Discussion The placement of ventricular catheters is a relatively
easy and straight-forward procedure, for which the neu-
rosurgeons are trained during their first years of residency
[1]. It may be the most often performed procedure in
neurosurgery, since it becomes necessary in patients with
cerebrospinal fluid drainage blockage, as transient external
drainage before larger tumor surgeries, for intracranial pres-
sure monitoring, for example, in severe traumatic brain
injury patients, and in hydrocephalus patients for cere-
brospinal fluid (CSF) shunting [3,5]. In all of these indi-
cations the accuracy of the ventricular catheter placement
will correlate with the proper functioning of the catheter
[2,6]. In the shunted hydrocephalus patients the catheter
is part of a lifelong implant, thus, accurate care must be
delivered for proper placement and long-term functioning
of the implant [2]. This is regularly performed in freehand
technique according to standardised anatomical landmarks. Recent studies have shown that this technique results in
a 12.5% to 40% inaccuracy of catheter placement
[2,4,8,9]. Thus, in selected cases, technical efforts, such
as neuronavigation and ultrasound, were used to en-
hance the proportion of accuracy [3,7,11,12,16]. The
reason for not using it in all cases is that the preparation
time of technical advances is too long in relation to the
procedure itself and the availability in each centre for these
procedures are limited [12]. Another aspect is a limited
awareness of neurosurgeons concerning the failure rate
of accurate positioning of the catheter in a rather simple
procedure. This holds especially true since the failure of
a catheter which is inaccurately placed will happen
within an interval when the correlation to the initial
procedure cannot be drawn anymore. A survey among
neurosurgeons about the use of technical efforts for the
accurate placement of ventricular catheter disclosed
that 30% did not want to enhance the freehand method
while roughly 60% would use technical advances if they
would not prolong the surgery by 5 or 10 minutes [10]. With these conditions, we recently introduced a mech-
anical instrument for guiding ventricular catheter place-
ment. For this technique, one individual parameter is
necessary to be measured in preoperative imaging and ap-
plied to this instrument to achieve controlled and accur-
ate puncture of the ventricle via a frontal pre-coronal
approach. The safety and efficacy of this technique was
shown in a single centre study so far [14]. Trial organisation The trial is initiated and designed by the investigators. It
is funded by B. Braun Aesculap. The study is conducted
in 10 German medical centres (see the list at the end of
this paper). These study centres were chosen for their
experience in treatment of hydrocephalic patients and their
willingness to adhere to the study protocol. Discussion The proposed
protocol is designed to prospectively test the method on
th l
b
i
f
lti
t
d
i
d t i l In a secondary step, a per-protocol reporting will be
done, which excludes cross-over patients. In a tertiary analysis as-treated reporting (safety evalu-
ation) is foreseen. Secondary variables The secondary variables (number of cannulation attempts,
the proportion of correct intraventricular positioning of
the perforated part of the catheter and the complication
rate/adverse events) will be reported descriptively based
on an intention-to-treat as well as the per-protocol popu-
lation. The proportion of “early-shunt-failures” as being a
safety-parameter will be reported according to the as-
treated population. Level of significance The one-sided test level for the confirmatory test is 2.5%. The explorative tests will be performed on a two-sided
level of 5%. To adjust for multiple effects and covariates
without running multiple tests, appropriate multivariate
regression models will be used for analysis. Randomisation and blinding H 0: PTG – PFH < = 0 (PTG < =PFH);
H 1: PTG – PFH > 0 (PTG > PFH). Patients’ allocation to both study arms is balanced (1:1). In
order to minimise the size differences between groups,
blocks of random length are used. The randomisation sys-
tem is integrated into a web-based, password protected elec-
tronic case report form (www.studiesportal.com, Aesculap,
Inc., Tuttlingen, Germany). Blinding of the operating sur-
geon is not feasible. The response (catheter position) is rated H 1: PTG – PFH > 0 (PTG > PFH). P is equal to the responder rate (TG – Thomale Guide,
FH – freehand) The analysis will be based on the intention-to-treat
population in order to include catheter guide treatment Schaumann and Thomale Trials 2013, 14:428
http://www.trialsjournal.com/content/14/1/428 Schaumann and Thomale Trials 2013, 14:428
http://www.trialsjournal.com/content/14/1/428 Schaumann and Thomale Trials 2013, 14:428
http://www.trialsjournal.com/content/14/1/428 Page 6 of 8 centre. The trial is conducted in agreement with the
principles of the Declaration of Helsinki and the German
law for medical devices. Patients are enrolled into the trial
only after a signed informed consent is obtained. problems with a change in the intraoperative treat-
ment of the patient (cross-over). In the case of non-
applicability of the catheter guide (for example, uneven
bony surface) in patients randomised into the catheter
guide treatment arm, they will be treated according to the
freehand technique treatment arm. This is also applicable
for patients with unsuccessful cannulation of the ventricle
in the catheter guide treatment arm according to the
decision of the operating surgeon. In the intention-to-
treat analysis, all these cross-over patients will be ana-
lysed in the catheter guide treatment arm. Data management and monitoring The clinical data will be reported on an electronic case
report form (online) by the investigators at the study site. The access to the web-based database is individualised and
password protected. The study shall be conducted accord-
ing to the ISO 14155 guideline for medical device studies. Study monitoring is performed by an authorised and quali-
fied representative of the study administration. Ethical aspects The trial is approved by the ethics committee of the
Charité - Universitätsmedizin Berlin. Secondary approvals
are obtained from the ethics committee of each study Schaumann and Thomale Trials 2013, 14:428
http://www.trialsjournal.com/content/14/1/428 Page 7 of 8 Schaumann and Thomale Trials 2013, 14:428
http://www.trialsjournal.com/content/14/1/428 intensive care units, are not applicable with the current
design of the tool. Our own retrospective analysis of inaccurate placement
of the ventricular catheter reveals that the inaccuracy cor-
relates with smaller ventricles. We found that an FOHR
of smaller than 0.5 resulted in 15% catheter malposi-
tioning at our institution. This cutoff measure represents
a ventricular size in which regularly no technical help is
applied for better accuracy. The catheter guide is thus
meant to be used in a routine manner even if the ven-
tricular size seems not necessarily to be challenging, but
to achieve a higher level of reliability of the lifelong
implanted ventricular catheter. The placement of ventricular catheters is a frequent
procedure in Neurosurgery which is regularly performed
in freehand technique. As recent literature has reported
this is not sufficient for lifelong implants as ventricular
catheters are used in CSF shunts for hydrocephalus pa-
tients. An easy to use surgical instrument used together
with a smartphone software application is tested within
the current study design in a prospective randomised
multicentre fashion in order to evaluate its safety and
efficacy for the routine use in a wide range of patients. The technique itself incorporates innovatively the good
availability of smartphones as a measurement tool for the
surgical instrument. Hereby, an application software was
designed in order to easily apply the measurements in one
coronal brain section image. Virtual visualisation of the
instrument does help to perform the measurements
intuitively and to safely apply the measures to the in-
strument and the patient. Participating hospitals – Charité Universitätsmedizin Berlin, Department of
Neurosurgery, Berlin – Klinikum Kassel, Department of Neurosurgery,
Kassel The image processing can also be done on every ad-
vanced DICOM viewer, which allows in-picture-angle-
measurements. However, we experienced herewith a more
complicated workflow and a higher risk that the measure-
ments are not accurately transferred to the patient. The
images used for the smartphone application must be cor-
onal sections where the frontal horns are depicted [14]. Herein, the trajectory will be projected from oblique to
strictly coronal. This can be transferred as an anonymous
image file, for example, via e-mail to the smartphone and
may then be integrated in the software application. When
higher accuracy is needed, the image section can also be
reconstructed along the virtual trajectory of the catheter
placement using a navigation-software. This will enable
the measurements to be done at the exact localisation of
the entry point. In more difficult cases, such as slit ventri-
cles (FOHWR <0.05), a neuronavigation system for ven-
tricular catheter positioning might still be advantageous. – Unfallkrankenhaus Berlin, Department of
Neurosurgery, Berlin – Unfallkrankenhaus Berlin, Department of Neurosurgery, Berlin – Eberhard Karls University Tübingen, Department of
Neurosurgery, Tübingen – Ruprecht Karls University Heidelberg, Department
of Neurosurgery, Heidelberg – Georg August University Göttingen, Department of
Neurosurgery, Göttingen – Dietrich-Bonhoeffer-Klinikum, Department o
Neurosurgery, Neubrandenburg Neurosurgery, Neubrandenburg – Heinrich Heine University Düsseldorf, Department
of Neurosurgery, Düsseldorf – University Duisburg Essen, Department of
Neurosurgery, Essen – University Duisburg Essen, Department of Neurosurgery, Essen Current status A first investigators’ meeting was held in May 2012, where
the key points of the study design were discussed and agreed
on by all clinical trial centres. In November 2012, the study
protocol was then completed and the trial was approved by
the ethics committee of the Charité - Universitätsmedizin
Berlin on 18 February 2013. On 8 April 2013 the first
patient was enrolled in the GAVCA trial. The estimated
patient recruitment time for 144 patients in 10 study
centres is 24 months. At the time of submission, 24 pa-
tients were recruited in 5 out of 10 participating centres. According to our initial data, the catheter guide used
in combination with the iPhone application revealed a
good reliability in the pilot study. Here, the catheter guide
was used in narrow ventricles with a mean FOHR of
0.38+/−0.05. All catheters could be placed in a func-
tional correct position with primary cannulation. In 8%,
the contralateral ventricle was punctured, which was
observed within the first 10 patients, so a learning curve
in using the catheter guide must be stated [14]. The catheter guide has to be used directly on the
bony surface to ensure the angulation be applied in
relation to the surface of the skull. That may limit
the use of the catheter guide when insufficient bone
area is exposed or the surface may be uneven due to
previous surgeries at the site of the entry point. Transcuta-
neous cannulation of the ventricles, for example, in the
case of an external ventricle drainage as performed on Abbreviations CSF: Cerebrospinal fluid; GAVCA: Guided Application of Ventricular Catheters;
FOHR: Frontal occipital horn ratio; FOHWR: Frontal occipital width ratio;
VC: Ventricular catheters. p
g
AS has no competeing interests. UWT holds the patent for the Guide
instrument. UWT receives compensation for speaker activities from
Aesculap Inc. References 1. Lind CR, Correia JA, Law AJ, Kejriwal R: A survey of surgical techniques for
catheterising the cerebral lateral ventricles. J Clin Neurosci 2008, 15:886–890. 1. Lind CR, Correia JA, Law AJ, Kejriwal R: A survey of surgical techniques for
catheterising the cerebral lateral ventricles. J Clin Neurosci 2008, 15:886–890. 2. Abdoh MG, Bekaert O, Hodel J, Diarra SM, Le Guerinel C, Nseir R,
Bastuji-Garin S, Decq P: Accuracy of external ventricular drainage catheter
placement. Acta Neurochir (Wien) 2012, 154:153–159. 3. Ghajar JB: A guide for ventricular catheter placement. Technical note. J Neurosurg 1985, 63:985–986. 4. Toma AK, Camp S, Watkins LD, Grieve J, Kitchen ND: External ventricular
drain insertion accuracy: is there a need for change in practice? Neurosurgery 2009, 65:1197–1200. discussion 1200–1201. 5. Yamada SM, Yamada S, Goto Y, Nakaguchi H, Murakami M, Hoya K, Matsuno A:
A simple and consistent technique for ventricular catheter insertion using a
tripod. Clin Neurol Neurosurg 2012, 114:622–626. 6. Yamada SM, Kitagawa R, Teramoto A: Relationship of the location of the
ventricular catheter tip and function of the ventriculoperitoneal shunt. J Clin Neurosci 2013, 20:99–101. 7. Stein SC, Guo W: A mathematical model of survival in a newly inserted
ventricular shunt. J Neurosurg 2007, 107(6 Suppl):448–454. 7. Stein SC, Guo W: A mathematical model of survival in a newly inserted
ventricular shunt. J Neurosurg 2007, 107(6 Suppl):448–454. 7. Stein SC, Guo W: A mathematical model of survival in a ne 8. Hsieh CT, Chen GJ, Ma HI, Chang CF, Cheng CM, Su YH, Ju DT, Hsia CC,
Chen YH, Wu HY, Liu MY: The misplacement of external ventricular drain
by freehand method in emergent neurosurgery. Acta Neurol Belg 2011,
111:22–28. 9. Huyette DR, Turnbow BJ, Kaufman C, Vaslow DF, Whiting BB, Oh MY:
Accuracy of the freehand pass technique for ventriculostomy catheter
placement: retrospective assessment using computed tomography
scans. J Neurosurg 2008, 108:88–91. 10. O’Neill BR, Velez DA, Braxton EE, Whiting D, Oh MY: A survey of
ventriculostomy and intracranial pressure monitor placement practices. Surg Neurol 2008, 70:268–273. discussion 273. 11. Reig AS, Stevenson CB, Tulipan NB: CT-based, fiducial-free frameless
stereotaxy for difficult ventriculoperitoneal shunt insertion: experience in
26 consecutive patients. Stereotact Funct Neurosurg 2010, 88:75–80. 12. Sampath R, Wadhwa R, Tawfik T, Nanda A, Guthikonda B: Stereotactic
placement of ventricular catheters: does it affect proximal malfunction
rates? Stereotact Funct Neurosurg 2012, 90:97–103. 13. Competing interests
h AS has no competeing interests. UWT holds the patent for the Guide
instrument. UWT receives compensation for speaker activities from
Aesculap Inc. Page 8 of 8 Schaumann and Thomale Trials 2013, 14:428
http://www.trialsjournal.com/content/14/1/428 16. Kulkarni AV, Drake JM, Armstrong DC, Dirks PB: Measurement of ventricular
size: reliability of the frontal and occipital horn ratio compared to
subjective assessment. Pediatr Neurosurg 1999, 31:65–70. 17. O’Hayon BB, Drake JM, Ossip MG, Tuli S, Clarke M: Frontal and occipital
horn ratio: a linear estimate of ventricular size for multiple imaging
modalities in pediatric hydrocephalus. Pediatr Neurosurg 1998, 29:245–249. doi:10.1186/1745-6215-14-428
Cite this article as: Schaumann and Thomale: Guided Application of
Ventricular Catheters (GAVCA) - multicentre study to compare the ven-
tricular catheter position after use of a catheter guide versus freehand appli-
cation: study protocol for a randomised trail. Trials 2013 14:428. Authors’ contribution
d
d
d AS and UWT designed the study conception and protocol as well as drafted
the manuscript. Both authors have read and approved the manuscript. 17. O’Hayon BB, Drake JM, Ossip MG, Tuli S, Clarke M: Frontal and occipital
horn ratio: a linear estimate of ventricular size for multiple imaging
modalities in pediatric hydrocephalus. Pediatr Neurosurg 1998, 29:245–249. Acknowledgements
f The authors are grateful to Valentin Hermanutz for critical evaluation and
correction of the study protocol design and the manuscript, as well as Viktor
Breul for designing the statistical analysis and calculating the sample size. We
also thank all members of the GAVCA study group for supporting this
project. This study is funded by B. Braun Aesculap. doi:10.1186/1745-6215-14-428
Cite this article as: Schaumann and Thomale: Guided Application of
Ventricular Catheters (GAVCA) - multicentre study to compare the ven-
tricular catheter position after use of a catheter guide versus freehand appli-
cation: study protocol for a randomised trail. Trials 2013 14:428. Received: 16 September 2013 Accepted: 26 November 2013
Published: 12 December 2013 References Villavicencio AT, Leveque JC, McGirt MJ, Hopkins JS, Fuchs HE, George TM:
Comparison of revision rates following endoscopically versus
nonendoscopically placed ventricular shunt catheters. Surg Neurol 2003,
59:375–379. discussion 379–380. Submit your next manuscript to BioMed Central
and take full advantage of:
• Convenient online submission
• Thorough peer review
• No space constraints or color figure charges
• Immediate publication on acceptance
• Inclusion in PubMed, CAS, Scopus and Google Scholar
• Research which is freely available for redistribution
Submit your manuscript at
www.biomedcentral.com/submit Submit your next manuscript to BioMed Central
and take full advantage of:
• Convenient online submission
• Thorough peer review
• No space constraints or color figure charges
• Immediate publication on acceptance
• Inclusion in PubMed, CAS, Scopus and Google Scholar
• Research which is freely available for redistribution
Submit your manuscript at
www.biomedcentral.com/submit 14. Thomale UW, Knitter T, Schaumann A, Ahmadi SA, Ziegler P, Schulz M,
Miethke C: Smartphone-assisted guide for the placement of ventricular
catheters. Childs Nerv Syst 2013, 29:131–139. Submit your next manuscript to BioMed Central
and take full advantage of:
• Convenient online submission
• Thorough peer review
• No space constraints or color figure charges
• Immediate publication on acceptance
• Inclusion in PubMed, CAS, Scopus and Google Scholar
• Research which is freely available for redistribution
Submit your manuscript at
www.biomedcentral.com/submit Submit your next manuscript to BioMed Central
and take full advantage of: Submit your next manuscript to BioMed Central
and take full advantage of: 15. Hayhurst C, Beems T, Jenkinson MD, Byrne P, Clark S, Kandasamy J,
Goodden J, Nandoe Tewarie RD, Mallucci CL: Effect of electromagnetic-
navigated shunt placement on failure rates: a prospective multicenter
study. J Neurosurg 2010, 113:1273–1278. • Convenient online submission
|
https://openalex.org/W1974039452
|
https://ped-rheum.biomedcentral.com/counter/pdf/10.1186/1546-0096-11-S2-P250
|
English
| null |
PReS-FINAL-2260: Provisional findings of an on-going study of musculoskeletal anomalies in a national cohort of patients with trisomy 21
|
Pediatric rheumatology online journal
| 2,013
|
cc-by
| 1,014
|
PReS-FINAL-2260: Provisional findings of an
on-going study of musculoskeletal anomalies
in a national cohort of patients with trisomy 21
C Foley1*, OG Killeen2
From 20th Pediatric Rheumatology European Society (PReS) Congress
Ljubljana, Slovenia. 25-29 September 2013 From 20th Pediatric Rheumatology European Society (PReS) Congress
Ljubljana, Slovenia. 25-29 September 2013 Screening involves completion of a health questionnaire
& a comprehensive musculoskeletal exam. Objectives 1.Take a musculoskeletal history from, and perform a mus-
culoskeletal examination on children with Down syndrome
between the ages of 0.5-18 years. 2.Score hypermobility
using the Beighton & Brighton screening tools. 3.Examine
joints for evidence of past and/or present arthritis. Results Our results support a feature consistently reported in the
limited literature available on DA (table 1). There is
delayed diagnosis, leading to less favourable outcomes. The average time to diagnosis in our cohort was 1.9
years, with the longest delay reported nearly 5 years. This
child developed loss of joint space, generalized osteope-
nia, erosions & subluxations of affected joints. When
compared with a cohort of our newly diagnosed JIA
patients (time to diagnosis 0.6 years), we demonstrate a
significant difference in time to diagnosis (p = 0.025). To
date 59 children have enrolled in the musculoskeletal
screening process, 69% of whom have pes planus. No
atlanto-axial instability has been reported, however one
child had an absent C2 vertebra. There has been one
case of patella instability and 4 new cases of DA diag-
nosed. Anecdotally we have not found the Beighton &
Brighton criteria comprehensive. The majority of children
were found to have hypermobile hips. Neither scoring
systems incorporate hips in their screening criteria. Introduction Musculoskeletal complications of Down syndrome are
common. Joint laxity is almost universal. This, in combi-
nation with low muscle tone, contribute to increased
risk of a number of musculoskeletal disorders e.g. atlanto-axial instability, patella instability & pes planus. Arthritis in children with Down syndrome is also
reported. Down’s Arthropathy (DA) is thought to be 3-6
times more common than JIA in the general paediatric
population. Despite this fact, DA is rarely recognised at
onset & remains under-diagnosed. This contributes to
unnecessary disability & functional impairment. Methods From April 2013 to April 2014, children with Down
syndrome will be invited to attend a screening clinic. 1OLCHC/NCRC, Dublin, Ireland
Full list of author information is available at the end of the article
Table 1
Average
Std. Dev. Minimum
Maximum
Current Down’s Arthropathy Cohort
n = 12
Age (years)
8.5
4.5
0.3
14.9
Active Joint Count*
6
3.7
1
11
Restricted Joint Count*
3
3.3
0
12
Delay in Diagnosis (years)
1.9
1.5
0.2
4.9
*At Presentation
© 2013 Foley and Killeen; licensee BioMed Central Ltd. This is an Open Access article distributed under the terms of the Creative
Commons Attribution License (http://creativecommons.org/licenses/by/2.0), which permits unrestricted use, distribution, and
reproduction in any medium, provided the original work is properly cited. The Creative Commons Public Domain Dedication waiver
(http://creativecommons.org/publicdomain/zero/1.0/) applies to the data made available in this article, unless otherwise stated. 1OLCHC/NCRC, Dublin, Ireland
Full list of author information is available at the end of the article
Table 1
Average
Std. Dev. Minimum
Maximum
Current Down’s Arthropathy Cohort
n = 12
Age (years)
8.5
4.5
0.3
14.9
Active Joint Count*
6
3.7
1
11
Restricted Joint Count*
3
3.3
0
12
Delay in Diagnosis (years)
1.9
1.5
0.2
4.9
*At Presentation
© 2013 Foley and Killeen; licensee BioMed Central Ltd. This is an Open Access article distributed under the terms of the Creative
Commons Attribution License (http://creativecommons.org/licenses/by/2.0), which permits unrestricted use, distribution, and
reproduction in any medium, provided the original work is properly cited. The Creative Commons Public Domain Dedication waiver
(http://creativecommons.org/publicdomain/zero/1.0/) applies to the data made available in this article, unless otherwise stated. Foley and Killeen Pediatric Rheumatology 2013, 11(Suppl 2):P250
http://www.ped-rheum.com/content/11/S2/P250 Open Access Table 1 Table 1
Average
Std. Dev. Minimum
Maximum
Current Down’s Arthropathy Cohort
n = 12
Age (years)
8.5
4.5
0.3
14.9
Active Joint Count*
6
3.7
1
11
Restricted Joint Count*
3
3.3
0
12
Delay in Diagnosis (years)
1.9
1.5
0.2
4.9
*At Presentation 1OLCHC/NCRC, Dublin, Ireland
Full list of author information is available at the end of the article © 2013 Foley and Killeen; licensee BioMed Central Ltd. This is an Open Access article distributed under the terms of the Creative
Commons Attribution License (http://creativecommons.org/licenses/by/2.0), which permits unrestricted use, distribution, and
reproduction in any medium, provided the original work is properly cited. The Creative Commons Public Domain Dedication waiver
(http://creativecommons.org/publicdomain/zero/1.0/) applies to the data made available in this article, unless otherwise stated. Page 2 of 2 Page 2 of 2 Foley and Killeen Pediatric Rheumatology 2013, 11(Suppl 2):P250
http://www.ped-rheum.com/content/11/S2/P250 To date 100% of children screened would not have BHS
using Brighton criteria. Conclusion Pes planus is commonly seen in children with T21,
therefore orthotics & advice regarding correct footwear
is important. Children with T21 often have hypermobile
hips, not accounted for by the current scoring criteria
for BHS. DA is common but often missed, with delayed
diagnosis. Early diagnosis & treatment of DA is impor-
tant to prevent unwanted joint destruction & functional
disability. Children with T21 should have a musculoske-
letal exam as part of their annual screening programme. Disclosure of interest None declared. Authors’ details
1OLCHC/NCRC, Dublin, Ireland. 2OLCHC, Dublin, Ireland. Authors’ details
1OLCHC/NCRC, Dublin, Ireland. 2OLCHC, Dublin, Ireland. Authors’ details
1 Published: 5 December 2013 Published: 5 December 2013 doi:10.1186/1546-0096-11-S2-P250
Cite this article as: Foley and Killeen: PReS-FINAL-2260: Provisional
findings of an on-going study of musculoskeletal anomalies
in a national cohort of patients with trisomy 21. Pediatric Rheumatology
2013 11(Suppl 2):P250. doi:10.1186/1546-0096-11-S2-P250
Cite this article as: Foley and Killeen: PReS-FINAL-2260: Provisional
findings of an on-going study of musculoskeletal anomalies
in a national cohort of patients with trisomy 21. Pediatric Rheumatology
2013 11(Suppl 2):P250. Submit your next manuscript to BioMed Central
and take full advantage of:
• Convenient online submission
• Thorough peer review
• No space constraints or color figure charges
• Immediate publication on acceptance
• Inclusion in PubMed, CAS, Scopus and Google Scholar
• Research which is freely available for redistribution
Submit your manuscript at
www.biomedcentral.com/submit
|
https://openalex.org/W2964724963
|
http://hb.diva-portal.org/smash/get/diva2:1342027/FULLTEXT01
|
English
| null |
A qualitative study on the breastfeeding experience of mothers of preterm infants in the first 12 months after birth
|
International breastfeeding journal - Electronic Edition -
| 2,019
|
cc-by
| 7,784
|
RESEARCH Open Access © The Author(s). 2019 Open Access This article is distributed under the terms of the Creative Commons Attribution 4.0
International License (http://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and
reproduction in any medium, provided you give appropriate credit to the original author(s) and the source, provide a link to
the Creative Commons license, and indicate if changes were made. The Creative Commons Public Domain Dedication waiver
(http://creativecommons.org/publicdomain/zero/1.0/) applies to the data made available in this article, unless otherwise stated. Abstract Keywords: Breastfeeding, Experiences, First year, Mothers, Preterm infant, Qualitative with the infant [4–6]. Due to this vulnerable situation,
breastfeeding support is essential. However, professional
breastfeeding support has been shown to vary widely for
mothers of preterm infants due to the individual style of
various healthcare professionals; support was shown to
be either constructive or destructive [7]. This inconsist-
ency puts the mother in a situation wherein she is ex-
posed to different support styles that are not always
sensitive to her unique situation [7]. This phenomenon
is combined with the fact that preterm infants are imma-
ture in their breastfeeding behaviour and may require a
long time to mature, which may be challenging and A qualitative study on the breastfeeding
experience of mothers of preterm infants in
the first 12 months after birth Lina Palmér1 and Jenny Ericson2,3,4* http://www.diva-portal.org
This is the published version of a paper published in .
Citation for the original published paper (version of record):
Palmér, L., Eriksson, J. (2019)
A qualitative study on the breastfeeding experience of mothers of preterm infants in the
first 12 months after birth
International Breastfeeding Journal, 14(35): 1-8
https://doi.org/10.1186/s13006-019-0229-6
Access to the published version may require subscription.
N.B. When citing this work, cite the original published paper.
Permanent link to this version:
http://urn.kb.se/resolve?urn=urn:nbn:se:hb:diva-21589 http://www.diva-portal.org
This is the published version of a paper published in . Citation for the original published paper (version of record):
Palmér, L., Eriksson, J. (2019)
A qualitative study on the breastfeeding experience of mothers of preterm infants in the
first 12 months after birth
International Breastfeeding Journal, 14(35): 1-8
https://doi.org/10.1186/s13006-019-0229-6
Access to the published version may require subscription. N.B. When citing this work, cite the original published paper. Permanent link to this version:
http://urn.kb.se/resolve?urn=urn:nbn:se:hb:diva-21589 http://www.diva-portal.org This is the published version of a paper published in . Citation for the original published paper (version of record): Access to the published version may require subscription. N.B. When citing this work, cite the original published paper. http://urn.kb.se/resolve?urn=urn:nbn:se:hb:diva-21589 Palmér and Ericson International Breastfeeding Journal (2019) 14:35
https://doi.org/10.1186/s13006-019-0229-6 Abstract Background: Being a mother of a preterm infant (< 37 gestational weeks) puts the mother in a vulnerable and
fragile situation wherein breastfeeding is an important part of becoming a mother and bonding with the infant. Nevertheless, the breastfeeding experience of mothers during the first year after a preterm birth has not been well
studied. To develop professional caring and supporting relationships, it is important to address this knowledge gap. The aim of this study was to describe the breastfeeding experience of mothers of preterm infants from birth up
until 12 months after birth. Methods: The data in this qualitative study are derived from a multicentre randomized controlled trial where 270
mothers of preterm infants provided 496 written comments through questionnaires containing open-ended
questions. The questionnaires were sent to the mother three times during the first 12 months after birth. A
thematic network analysis based on hermeneutical philosophy was used to analyse and interpret the resulting data
to describe the mothers’ experiences of breastfeeding. Results: Three organizing themes, namely, “navigating smoothly,” “navigating with a struggle” and “navigating in
ambiguity” were revealed in the mothers’ narratives regarding their breastfeeding experiences during the first 12
months after birth. These organizing themes were further interpreted as one global theme that was deemed “A
journey to finding one’s unique way in breastfeeding.” Results: Three organizing themes, namely, “navigating smoothly,” “navigating with a struggle” and “navigating in
ambiguity” were revealed in the mothers’ narratives regarding their breastfeeding experiences during the first 12
months after birth. These organizing themes were further interpreted as one global theme that was deemed “A
journey to finding one’s unique way in breastfeeding.” Conclusion: Mothers of preterm infants are in an exposed and vulnerable situation when initiating breastfeeding
during the first year. This situation leads to a unique journey wherein each mother navigates through breastfeeding
depending on her individual situation. An awareness of the diversity of breastfeeding experiences may contribute
to the provision of professional caring and supportive relationships. Conclusion: Mothers of preterm infants are in an exposed and vulnerable situation when initiating breastfeeding
during the first year. This situation leads to a unique journey wherein each mother navigates through breastfeeding
depending on her individual situation. An awareness of the diversity of breastfeeding experiences may contribute
to the provision of professional caring and supportive relationships. Trial registration: www.clinicaltrial.gov NCT01806480 registered 7 March 2013. Trial registration: www.clinicaltrial.gov NCT01806480 registered 7 March 2013. © The Author(s) 2019 Open Access Th
* Correspondence: jenny.ericson@ltdalarna.se
2School of Education, Health and Social Studies, Dalarna University, Falun,
Sweden
3Center for Clinical Research Dalarna, Uppsala University, Falun, Sweden
Full list of author information is available at the end of the article Background Mothers of preterm infants breastfeed to a lesser extent
than mothers of full-term infants [1], which may be due
to infant immaturity and a lack of support [2, 3]. Previ-
ous studies described that mothers of preterm infants
are in a vulnerable situation, and that breastfeeding is an
important part of becoming a new mother and bonding Methods During a randomized controlled trial (RCT), breastfeed-
ing mothers of preterm infants (gestational age < 37
weeks) provided qualitative data in written form about
their experiences of breastfeeding their preterm infants
during the first 12 months after birth. The RCT aimed to
evaluate a proactive breastfeeding support intervention
and included 493 mothers of preterm infants. The study
was conducted after discharge from six neonatal units in
Sweden. The results from this RCT are presented else-
where [7, 10–12]. The follow-up
questionnaire consisted of the following open-ended
question regarding breastfeeding and the feeding experi-
ences of the mothers: “If you want, feel free to write
about what you have experienced while breastfeeding/
bottle-feeding your baby.” The written narratives were
combined in a Microsoft Word document. discharge from the neonatal unit, and 6 (6 m) and 12
months (12 m) after birth between March 2013 and
December 2015, as a part of the RCT. The follow-up
questionnaire consisted of the following open-ended
question regarding breastfeeding and the feeding experi-
ences of the mothers: “If you want, feel free to write
about what you have experienced while breastfeeding/
bottle-feeding your baby.” The written narratives were
combined in a Microsoft Word document. Palmér and Ericson International Breastfeeding Journal (2019) 14:35 Palmér and Ericson International Breastfeeding Journal (2019) 14:35 Page 2 of 8 Table 1 Characteristics of participants (n = 270)
Demographic variables
n (%) median [IQR] mean ± SD
Maternal variables
Age, years
30.5 ± 4.8
Maternal educational level
Higher education
154 (51)
Upper secondary
school or less
116 (49)
Primipara
159 (59)
Mothers not born
in Sweden
16 (6)
Vaginal birth
152 (56)
Multiple birth
32 (12)
Gestational age at
birth, weeks
34 [2]
Exclusive breastfeeding
at discharge
230 (85)
8 weeks after
discharge
171 (63)
6 months after birth
75 (28)
Partial breastfeeding
12 months after birth
39 (15)
SD Standard deviation, IQR Interquartile range Table 1 Characteristics of participants (n = 270)
Demographic variables
n (%) median [IQR] mean ± SD stressful for the mother. The challenges that may occur
include, for example, infant sleepiness, vague feeding
cues, latching difficulties and weak sucking [2, 3, 6]. Ma-
ternal challenges involved with breastfeeding a preterm
infant may include feelings of guilt and failure, an insuf-
ficient milk supply and milk expression. Nevertheless,
breastfeeding a preterm infant may also rebuild a con-
nection and confidence with motherhood; therefore,
breastfeeding can be a healing and a bonding experience
[3, 6, 8]. Previous studies have described maternal experiences
of breastfeeding a preterm infant during a stay at the
neonatal unit and/or a few weeks after discharge. How-
ever, very few studies have described mothers’ experi-
ences of breastfeeding during the first 12 months after
the birth of a preterm infant [9]. Human experience is
complex and cannot be understood by analysing parts or
measuring aspects of breastfeeding, as we are also af-
fected by social context. It is important to examine each
mother’s experience of breastfeeding their preterm infant
because each individual mother has the most knowledge
about her own experience. Illuminating breastfeeding ex-
periences may help health professionals to provide caring
and supportive relationships through learning about the
mothers’ individual experiences. Thus, the aim of this
study was to describe mothers’ experiences of breastfeed-
ing their preterm infants from birth up until 12 months
after birth. SD Standard deviation, IQR Interquartile range discharge from the neonatal unit, and 6 (6 m) and 12
months (12 m) after birth between March 2013 and
December 2015, as a part of the RCT. Analysis y
The epistemological foundation for the analyses in this
study was based on hermeneutical philosophy using a
reflective lifeworld approach [13, 14]. A thematic net-
work analysis, which aims to explore the understanding
of an issue, was used to organize and interpret the data
[15]. First, basic themes were derived from the text;
basic themes support a statement or belief related to the
diversity of the mothers’ experience, but they say very
little by themselves. Then, these basic themes were inter-
preted into organizing themes in relation to the under-
lying narrative and meaning that they put forth. These
organizing themes were more abstract and more reveal-
ing of the meaning within the texts. Finally, a global
theme was deducted, and a comparative analysis among
the organizing themes gave rise to the primary interpret-
ation that linked all of the previous interpretations to-
gether into an overarching theme – the global theme,
suggesting how to understand the phenomenon. During In this study, exclusive breastfeeding was defined as
feeding with breast milk only regardless of feeding
method, but could include medications, fortification and
vitamins. Partial breastfeeding was defined as feeding
with breast milk in combination with formula and/or
solid food. No breastfeeding was defined as full formula
feeding and/or solid food with no breast milk intake. All
the infants in this study were breastfeeding directly at
the breast with the exception of one infant who was
exclusively fed breast milk from a bottle. The character-
istics of the participating mothers are presented in
Table 1. The data consisted of 496 written comments
received from 270 mothers from follow-up question-
naires that were sent to the mothers 8 weeks (8w) after Palmér and Ericson International Breastfeeding Journal (2019) 14:35 Palmér and Ericson International Breastfeeding Journal (2019) 14:35 Page 3 of 8 Page 3 of 8 Page 3 of 8 the analysis, rigor was maintained by trying to be both
open and pliable as well as keeping a bridled attitude. More precisely, the researchers read the written text
with an open mind until it felt familiar, after which the
actual analysis was initiated. During the analysis, there
was a movement between the whole (written text), the
parts (basic themes) and the new whole (the global
theme and the organizing themes). Movement during
analysis and the methodological principles have guided
the process of maintaining rigor. Results The following three organizing themes were revealed in
the mothers’ narratives regarding their breastfeeding
experiences during the first 12 months after birth: “navi-
gating smoothly”, “navigating with a struggle” and “navi-
gating in ambiguity.” These organizing themes were
summarized in one global theme: “a journey to find one’s
unique way in breastfeeding” (Fig. 1). Analysis Quotes were used to
support the interpretations of the text; the quotes that
are presented in the results are labelled with each
mother’s randomization code. After the analysis was fin-
ished and the quotes selected we added the infant’s ges-
tational age (GA) at birth and infant feeding status in
the previous 24 h at the last data collection point to each
quotation to give a sense of each feeding path. the infant and thus strengthen the relationship between
the mother and infant. The mothers experienced a feel-
ing of togetherness with their infants. They described
this feeling of togetherness as a mutual interaction and
an intimate relationship with the infant. Such mutuality
and intimacy was reported to provide a unique closeness
and strong bond with the infant. It feels very good; breastfeeding provides great closeness
and attachment, which gives me a great inner
satisfaction as a parent. Now, after 6 months, I am so
used to breastfeeding. In addition, my son eats more
efficiently now, so it feels more practical to quickly and
easily feed him compared with before, when you
breastfeed all the time. SK5 (GA 36+2, breastfed
exclusively at 6m) When breastfeeding was experienced as smooth, mothers
felt a harmony in life when breastfeeding. Breastfeeding
was described as a way to rest and calm down as well as a
period of relaxation for both the mother and infant. In
such harmonious breastfeeding situations, the mothers
also felt a sense of wellbeing, tranquillity and security in
breastfeeding.For me, breastfeeding is usually a quiet and
harmonious moment - an opportunity to sit down and
relax a little extra. SK59 (GA 34+3, breastfed
exclusively at 6m and partially at 12m) Navigating with a struggle g
g
gg
Navigating with a struggle means that the breastfeeding
journey is experienced more or less as a bodily perform-
ance instead of a smooth relationship with the infant. The most prominent problem or difficulty faced by the
mothers was that the infants’ prematurity complicated
breastfeeding. The mothers reported that their infants
could not or did not want to breastfeed. The infants
could have a weak suck or did not suck. The interaction
and relationship between the mother and infant were, or
could be, complicated. Sometimes, the mother wanted
to breastfeed but the infant did not. Some mothers ac-
cepted this situation, but other mothers wished that they
had tried harder or longer, or they wished for more sup-
port. This failure of acceptance sometimes led to sadness
and regret over the loss of breastfeeding, which was
emotionally difficult to handle. Mothers also felt grateful and were strengthened by the
positive bodily response of their breastfeeding. Pride in
succeeding with breastfeeding was awakened, and breast-
feeding was a privilege to experience. Happiness emerged
because breastfeeding went smoothly and was pain free. Mothers also reported amazement over the body’s ability
to produce enough breast milk for the infant to thrive, as
there was often an inherent apprehension that this would
not happen.Breastfeeding my child makes me proud;
closeness and love grow between us. When I breastfeed
her, we make eye contact and small talk, so I never get
tired of breastfeeding her. I love breastfeeding her; I
couldn’t replace this breastfeeding time with anything
else. K45 (GA 31+0, breastfed exclusively at discharge
and partially at 8w, 6m and 12m) It has been tough when my daughter just fell asleep at
the breast. She has not received the whole meal at the
breast. My milk dried up when she did not suck hard
enough. (I) express milk and feed by bottle, and the
feedings have taken a lot of time and effort. Ö19 (GA
32+2, ceased breastfeeding at 11 months) It has been tough when my daughter just fell asleep at
the breast. She has not received the whole meal at the
breast. My milk dried up when she did not suck hard
enough. (I) express milk and feed by bottle, and the
feedings have taken a lot of time and effort. Navigating with a struggle Ö19 (GA
32+2, ceased breastfeeding at 11 months) Mothers were proud of their ability to manage breast-
feeding and grateful for the closeness and growing love
that breastfeeding provided. One mother of twins was
proud that she managed to breastfeed her infants for 6
months, and she often breastfed the infants together. The positive breastfeeding experiences was etched in the
mother’s memory after she ceased breastfeeding. An-
other mother wrote; I felt happy and satisfied every time
I nursed my son; it was a good feeling that I will not for-
get. K67 (GA 34 + 5, ceased breastfeeding at 5 months). Yet another mother wrote: The struggle could also be over one’s own bodily per-
formance or impairment. Mastitis, breast pain and/or a
low milk supply (both perceived and actual) or maternal
illnesses can disrupt breastfeeding, which can lead to
stress, anxiety and frustration.I had pain from one breast
during breastfeeding ever since he was small. I had
candida in the milk ducts and sore nipples in the first
3 months, so it has been difficult, but I do not want to
stop breastfeeding. Ö60 (GA 36+2, ceased
breastfeeding at 9 months) To breastfeed my child is pure love. I get the chills
in my body when I think about ceasing
breastfeeding (in some years or so) because I like it. It is our moment. SK67 (GA 31+5, breastfed
exclusively at 6m) To breastfeed my child is pure love. I get the chills
in my body when I think about ceasing
breastfeeding (in some years or so) because I like it. It is our moment. SK67 (GA 31+5, breastfed
exclusively at 6m) Struggling with breastfeeding may also be associated
with having too much milk and thus being confronted
with extreme bodily changes in the breasts. Such bodily
changes in the breast may in some cases be experienced
as very trying and, in some cases, these changes of the
breast can be perceived as disgusting. The mothers also described breastfeeding as a gift to the
infant. Breastfeeding gave the mothers a lovely feeling,
and it was satisfying to provide the infant with nutrition
and protection, which was perceived as being the best
possible start. It was described as “awesome” and “beau-
tiful” when the infant became satisfied by being breast-
fed.I give my child the best she can get nutritionally. Navigating smoothly Despite
concerns
about
breastfeeding
that
some
mothers experienced during pregnancy, it was a relief
for many that everything worked well regarding breast-
feeding. Breastfeeding seemed to be a smooth way of Navigating smoothly through one’s breastfeeding journey
means that breastfeeding was experienced in a positive
way without any major problems or difficulties. In this
theme, breastfeeding was a way to become close with Grateful and were strengthened by
positive bodily response
A journey to find one’s unique way in breastfeeding
Navigating in ambiguity
Reward that was fought for
Changing track
Aids that can either help or hinder
Navigating with a struggle
Togetherness
Harmony in life
Navigating smoothly
Breastfeeding as a gift
Exhausting situation
The infant’s prematurity complicated
Bodily performance or impairment
Fig. 1 Basic themes, organizing themes and the global theme that illustrate the findings and the analytic process. The tracks are intertwined with
each other and can be understood as tracks on a map where the goal is to find one’s own unique way in breastfeeding Navigating with a struggle Navigating in ambiguity The infant’s prematurity complicated Exhausting situation Fig. 1 Basic themes, organizing themes and the global theme that illustrate the findings and the analytic process. The tr
each other and can be understood as tracks on a map where the goal is to find one’s own unique way in breastfeeding Fig. 1 Basic themes, organizing themes and the global theme that illustrate the findings and the analytic process. The tracks are intertwined with
each other and can be understood as tracks on a map where the goal is to find one’s own unique way in breastfeeding Palmér and Ericson International Breastfeeding Journal (2019) 14:35 Palmér and Ericson International Breastfeeding Journal (2019) 14:35 Page 4 of 8 It is a great gift to give to your child. It is convenient to
always have the food with you. F23 (GA 34+0, ceased
breastfeeding at 8 months) It is a great gift to give to your child. It is convenient to
always have the food with you. F23 (GA 34+0, ceased
breastfeeding at 8 months) being
with
the
infant
and
thus
strengthened
the
mother’s confidence in parenting. This situation was
characterized by simply being a breastfeeding mother
without
experiencing
any
problems
or
difficulties. Navigating smoothly Breastfeeding was integrated into the mother’s life in an
embodied way.It feels very good to be able to breastfeed,
and it works very well most of the time. SU38 (GA
35+5, breastfed exclusively at 8w and partially at 6m
and 12m) Navigating with a struggle SU61 (GA 32+6, fully formula
fed at 8w) bad due to all the stress and pressure about
breastfeeding and the idea that it (breastfeeding)
would be “the best” option. Bottle-feeding is a pure
dream in comparison. Breastfeeding is not free; it is
the most costly thing I have experienced. I lost myself
completely. SU68 (GA 33+3, partially breastfed 8w) Navigating with a struggle She
gets security and comfort when she lies at the breast. We have a nice emotional contact. Ö71 (GA 35+6,
breastfed exclusively at 6m and partially at 12m) Breastfeeding is sometimes seen by health profes-
sionals as only food, and this mind set can be transferred
to the mothers. This attitude places a huge focus on
breastfeeding as food, infant weight gain, and breast milk
production, which can create feelings of breastfeeding as
an accomplishment or, as one mother wrote: maternal Page 5 of 8 Palmér and Ericson International Breastfeeding Journal (2019) 14:35 Page 5 of 8 Palmér and Ericson International Breastfeeding Journal and infant eating disorders. Ö20 (GA 31 + 4, ceased
breastfeeding at 5 months). and infant eating disorders. Ö20 (GA 31 + 4, ceased
breastfeeding at 5 months). Another mother described a similar experience: with their breast milk when the infants could not breast-
feed directly at the breast.“It has been hard, sometimes
painful, and it was necessary to express a lot of milk. When it (breastfeeding) worked well (most of the time),
it has provided moments of closeness with my baby”. T40 (GA 33+0, ceased breastfeeding at 9 months) Another mother described a similar experience: Now, afterwards, I see that initially my child was “just”
weight gain. Everything was about gaining weight, and
the only thing I saw was the child's needs (nutritional
needs) - not the CHILD, not the interaction through
movements and smiles, just weight. It (breastfeeding)
became “mechanical” - food - weight. No one saw my
little child. F99 (GA 33+0, partially breastfed 8w) Nipple shields were described as either a saving solu-
tion or a hindrance. The nipple shield was experienced
by some mothers as an aid that helped the infants to get
a good latch. However, it was a relief when the infants
started to breastfeed without the nipple shield. On the
other hand, the nipple shield was also experienced by
some mothers as bothersome, a failure and/or difficult
to handle. The mothers who still used a nipple shield be-
cause the infants did not latch without it wished for the
infant to manage breastfeeding without the nipple shield
in order to be free and not bound by the shield. Breastfeeding was sometimes seen as an exhausting situ-
ation that caused the mothers to experience stress. Navigating with a struggle Furthermore, feelings of failure, of being solely respon-
sible, being insufficient or incapable were also expressed. These feelings sometimes led to disappointment and
frustration that breastfeeding did not work out as
expected. Breastfeeding was reported by some mothers
to be mentally tough and unpredictable; moreover, the
need to always be close to the infant and the inability to
leave was tiring for some mothers. Some mothers
wanted to share infant feeding with the father, and some
did so.The worst thing I have experienced. I felt really
bad due to all the stress and pressure about
breastfeeding and the idea that it (breastfeeding)
would be “the best” option. Bottle-feeding is a pure
dream in comparison. Breastfeeding is not free; it is
the most costly thing I have experienced. I lost myself
completely. SU68 (GA 33+3, partially breastfed 8w) Breastfeeding was sometimes seen as an exhausting situ-
ation that caused the mothers to experience stress. Furthermore, feelings of failure, of being solely respon-
sible, being insufficient or incapable were also expressed. These feelings sometimes led to disappointment and
frustration that breastfeeding did not work out as
expected. Breastfeeding was reported by some mothers
to be mentally tough and unpredictable; moreover, the
need to always be close to the infant and the inability to
leave was tiring for some mothers. Some mothers
wanted to share infant feeding with the father, and some
did so.The worst thing I have experienced. I felt really
b d d
t
ll th
t
d
b
t It (breastfeeding) is not as smooth as I thought from
the beginning when I have used a nipple-shield be-
cause my child is unable to suck otherwise. T20 (GA
33+1, ceased breastfeeding at 7 months) Bottles were reported by some mothers to interfere
with breastfeeding. These mothers reported that after
introducing bottles, the infant preferred the bottle over
the breast, which led to the cessation of breastfeeding. Bottles and formula feeding were, in some situations,
also described as a rescue when breastfeeding had been
tough or did not work. The switch to bottle-feeding
facilitated maternal well-being in some cases.Breast-
feeding didn't suit me, and it was the best decision for
me and my family to start giving formula. I felt bad
when breastfeeding. Navigating in ambiguity The breastfeeding journey consists
of navigating one’s own unique needs as a mother as
well as the unique needs of the infant. Such naviga-
tions of multiple, sometimes competing needs, are
complex, which may challenge mothers. The mother
thus navigates with both her own inner wishes and
needs, as well as the perceived wishes and needs of
the
infants
while,
at
the
same
time,
navigating
through
prevailing
norms
and
opinions,
especially
those given by health professionals. Three tracks of
navigation were evident based on the data, namely,
navigating smoothly, navigating with a struggle or
navigating in ambiguity. These tracks were, to some
extent, intertwined with each other and could be
understood as tracks on a map where the goal was to
find one’s own unique way (Fig. 1). Mothers who were not able to fulfil their breastfeeding
wishes or who discovered that they did not enjoy breast-
feeding were an especially vulnerable group, both due to
having a preterm infant and not being able to breastfeed
as they had expected. This extra vulnerability must be
considered when caring for the mother and infant;
otherwise, the suffering of both the mother and infant
can be overwhelming for the mother to handle. Recent
research regarding breastfeeding difficulties showed that
some mothers may feel lost in motherhood, feeling as
though she has an insufficient body, as well as having
difficulties establishing a relationship with the infant
[18]. van Wijlen [19] suggests that breastfeeding needs
to be seen as a relationship, and healthcare attitudes
about breastfeeding must shift towards a relational ap-
proach instead of the disembodied and often mechanis-
tic approach that often exists in the dominant Western
medical model. Furthermore, according to a holistic ap-
proach to care, it is important to discuss the concept of
the lived body (i.e. the human being cannot be separated
from the world or into a separate body and mind) [20]. Human beings must be seen as whole entities wherein
the body-mind-world are integrated into the lived body. According to this, it is not always helpful to simply tell a
mother who wants to breastfeed but cannot due to diffi-
culties that “you are a good mother even if you cannot
breastfeed”. This is because her breastfeeding experi-
ences are intertwined with her view of being a mother. Navigating in ambiguity Navigating in ambiguity means that the breastfeeding
journey could be seen as a reward that was fought for. The initiation of breastfeeding, and occasionally later
during the first year, could be tough and difficult. How-
ever, after a short or long (days to months) struggle,
mothers overcame the difficulties, and breastfeeding be-
came well-functioning and often pleasurable. When navigating breastfeeding with ambiguity, changing
track was one way to find one’s own way in the breastfeed-
ing situation. Changing track means that breastfeeding
was initiated but, after discovering an unwillingness, not
feeling well or feeling uncomfortable, the mothers ceased
breastfeeding. Feeling “disgusting” or “like a cow” was also
reported. For these mothers, breastfeeding felt wrong, they
ceased breastfeeding and even regretted that they had
started to breastfeed.I thought I would be a mom who
would like to breastfeed, but the opposite turned out to
be true. F11 (GA 35+6, breastfed exclusively at 8w and
had ceased before 6m) I had problems with breastfeeding in the beginning
and felt that it (breastfeeding) started to be associated
with anxiety. After about 3 months, things turned
around, and now I love to breastfeed. SU23 (GA 36+4,
breastfed partially at 8w and exclusively at 6m) Expressing breast milk or using nipple shields and bot-
tles are aids that can either help or hinder breastfeeding. Mothers described milk expression as frustrating and
difficult. However, they also reported pride in having
expressed milk for their infants when they later managed
to breastfeed their infants directly at the breast. Some
mothers continued to express milk to provide the infant A journey to find one’s unique way in breastfeeding
The three organizing themes were used to construe
an overarching interpretation, a global theme. The Palmér and Ericson International Breastfeeding Journal (2019) 14:35 Palmér and Ericson International Breastfeeding Journal Page 6 of 8 Page 6 of 8 love, caring, and relationships, which could be both posi-
tive and negative. interpretations suggested that the mothers’ breastfeed-
ing experiences could be described as a journey to
find one’s unique way in breastfeeding when in the
vulnerable situation as a new mother with a preterm
infant. The journey to become a breastfeeding mother
began during pregnancy and was physically stopped
when breastfeeding was ceased. However, breastfeed-
ing did not entirely stop when breastfeeding ended
physically, because the experience continued in the
mother’s memory. Navigating in ambiguity A caring attitude requires openness and a willingness to
encounter each mother’s life situation, to create an
environment in which mothers feel their perspectives
are genuinely heard, valued and respected [21]. This
approach requires that healthcare practitioners genuinely
listen to what the patient expresses, both verbally and
with
body
language
[22,
23]. Hence,
breastfeeding
support should be given according to the individual
mother’s unique wishes and needs regarding mothering
and breastfeeding, as well as to establish a close and pro-
tective relationship with the infant. Furthermore, re-
search [24–27] also suggests that breastfeeding should
be promoted as a woman’s right and a feminist issue and
not only as a women’s duty or responsibility due to the
biological body and the physical health benefits of
breastfeeding. However, social and structural inequities
may influence a woman’s ability to choose to breastfeed
[28], which is important to be aware of when supporting
breastfeeding. Altogether, it is important to approach
breastfeeding as a relation with the mother’s experiences
and life situation in mind. As previously mentioned, the breastfeeding journey can
vary greatly among mothers, and breastfeeding a preterm Discussion Our findings in the mothers’ narratives regarding breast-
feeding experiences during the first 12 months after the
birth of a preterm infant revealed three organizing
themes, namely, “navigating smoothly”, “navigating with
a struggle” and “navigating in ambiguity”, which were
interpreted as one global theme: “A journey to find one’s
unique way in breastfeeding.” q
y
f
g
Breastfeeding as a journey has been described in the
context of full-term infants, as an “engrossing, personal
journey” which is physical and requires maternal commit-
ment, adaptation, and support [16]. Breastfeeding has also
been described as a personal choice, harder than expected
and as being exposed to public debates [17]. In the con-
text of preterm infants, the present study highlights
breastfeeding as a journey in a slightly different way. One
positive finding was that breastfeeding worked out well for
many mothers; they enjoyed breastfeeding and the feelings
and advantages that it entailed for both themselves and
their infant. This finding is not usually described in re-
search but is important to highlight, as it is a further rea-
son to support breastfeeding. The present study also
highlighted breastfeeding a preterm infant as a relational
activity that the mother performs as part of mothering the
infant. The mothers described many aspects of their expe-
riences of breastfeeding their preterm infants; breastfeed-
ing was not solely about nutrition and protection but also
about combinations of many elements, such as emotions, As previously mentioned, the breastfeeding journey can
vary greatly among mothers, and breastfeeding a preterm
infant may be a multifaceted and complex experience. In a Palmér and Ericson International Breastfeeding Journal (2019) 14:35 Palmér and Ericson International Breastfeeding Journal (2019) 14:35 Page 7 of 8 study by Niela Vilén et al., breastfeeding mothers of pre-
term infants were divided into different typologies [9]. This means that the mothers were categorized into typ-
ologies because the mothers in that study maintained their
feelings, either positive or negative, during the follow-ups. There is a risk of categorizing mothers with preterm in-
fants and considering them as a homogeneous group
based on the infant’s prematurity. Our results show that
breastfeeding experiences are more complex than that. Breastfeeding a preterm infant during the first year was
not shown to be straightforward; rather, it was like follow-
ing a winding path, going forward but sometimes back-
ward or sideways. Discussion There may therefore exist a danger in
classifying and categorizing mothers because doing so
takes for granted that the present situation will also persist
in the future. Our results show that there needs to be a
recognition that the journey in breastfeeding takes differ-
ent paths during the entire breastfeeding period. limitation of the study is that it only included mothers
who were breastfeeding at discharge and did not include
experiences of mothers who ceased breastfeeding while in
the neonatal unit potentially impacting generalizability. Acknowledgements Acknowledgements
The authors would like to express their sincere gratitude to the mothers in
the study who shared their experiences with breastfeeding. Funding g
The Centre for Clinical Research Dalarna, Dalarna University, University of
Borås, and the Gillbergska Foundation supported this study. Author details
1 Author details
1Faculty of Caring Science, Work Life and Social Welfare, University of Borås,
Borås, Sweden. 2School of Education, Health and Social Studies, Dalarna
University, Falun, Sweden. 3Center for Clinical Research Dalarna, Uppsala
University, Falun, Sweden. 4Department of Pediatrics, Falu Hospital, Falun,
Sweden. Received: 28 January 2019 Accepted: 22 July 2019 Received: 28 January 2019 Accepted: 22 July 2019 Conclusion Mothers experienced breastfeeding their preterm infants
in different ways, and each mother found her own way in
breastfeeding. An awareness of the diversity of breastfeed-
ing experiences may help provide better professional
caring and supporting relationships. The whole care chain
for preterm infants and their mothers (e.g., maternity,
neonatal and child health care) needs to have a caring
approach and holistically meet the uniqueness in every
mother’s breastfeeding situation. Such an approach en-
ables health care professionals to prepare, care for and
support the breastfeeding mothers of preterm infants, to
encounter each mother where she is and to give each
mother the right to her own breastfeeding journey. Furthermore, when comparing other research about
mothers’ experiences of breastfeeding their preterm infants,
it seems that breastfeeding mothers of preterm infants in
different Western countries experience certain similarities. A recurrent issue in the studies, including ours, was that
the infant’s prematurity complicated breastfeeding [2, 9]. This issue could be addressed in neonatal care to identify
care routines that support the preterm infant’s development
as well as to educate parents about preterm infant feeding
development and how to support the infant during
breastfeeding to enable exclusive breastfeeding. In this
study, with the exception of one mother, all mothers breast-
fed their preterm infants directly at the breast, which
should be taken into account when interpreting the results. In some other countries, it is more common to feed infants
expressed breast milk in a bottle for various reasons,
including care routines, society norms and/or personal
reasons [29, 30]. Competing interests The authors declare that they have no competing interests. Consent for publication
N
l
bl Consent for publication
Not applicable. Authors’ contributions JE is a paediatric nurse and has a PhD, LP is a midwife and has a PhD. JE
was involved in the design of the study and performed the data collection. The authors (JE and LP) performed the analyses of the data reviewed, revised
the manuscript, and approved the final manuscript as submitted. Availability of data and materials The dataset used and analysed during the current study are available from
the corresponding author on reasonable request. The dataset used and analysed during the current study are available from
the corresponding author on reasonable request. Ethics approval and consent to participate The Regional Ethical Review Board, Uppsala, approved the study (Dnr: 2012/
292 and 2012/292/2). The mothers’ received oral and written information of
the study and all participating mothers signed a written consent to
participate in the RCT-study. A strength of this study is the variety of descriptions
during a long follow-up and the relatively representative
sample of breastfeeding mothers of preterm infants in a
Swedish context. One limitation may be that the com-
ments provided by the mothers are, to some extent, short;
however, these comments provide a rich variation in the
descriptions of lived experiences of breastfeeding during
the first year as a mother of a preterm infant. To gain an
even deeper understanding of the possible changes and
their influences during the breastfeeding journey, regular
individual interviews during the breastfeeding period are
warranted. The breastfeeding support given in the RCT in
the first 2 weeks after discharge may have influenced the
mothers’ experiences of breastfeeding. However, the sup-
port was provided for only a short period and not all
mothers took advantage of the support offered. Hence, we
do not believe it to have affected the experiences of breast-
feeding over the 12 m period to any great extent. Another 1.
Flacking R, Nyqvist KH, Ewald U. Effects of socioeconomic status on
breastfeeding duration in mothers of preterm and term infants. Eur J Pub
Health. 2007;17(6):579–84. References
l
k 1. Flacking R, Nyqvist KH, Ewald U. Effects of socioeconomic status on
breastfeeding duration in mothers of preterm and term infants. Eur J Pub
Health. 2007;17(6):579–84. Page 8 of 8 Palmér and Ericson International Breastfeeding Journal (2019) 14:35 Page 8 of 8 Page 8 of 8 2. Dosani A, Hemraj J, Premji SS, Currie G, Reilly SM, Lodha AK, et al. Breastfeeding
the late preterm infant: experiences of mothers and perceptions of public
health nurses. Int Breastfeed J. 2016;12:23. 2. Dosani A, Hemraj J, Premji SS, Currie G, Reilly SM, Lodha AK, et al. Breastfeeding
the late preterm infant: experiences of mothers and perceptions of public
health nurses. Int Breastfeed J. 2016;12:23. 3. Kair LR, Flaherman VJ, Newby KA, Colaizy TT. The experience of breastfeeding
the late preterm infant: a qualitative study. Breastfeed Med. 2015;10(2):102–6. 29. Collins CT, Gillis J, McPhee AJ, Suganuma H, Makrides M. Avoidance of
bottles during the establishment of breast feeds in preterm infants. Cochrane Database Syst Rev. 2016;10:Cd005252. 4. Niela-Vilen H, Axelin A, Melender HL, Salantera S. Aiming to be a
breastfeeding mother in a neonatal intensive care unit and at home: a
thematic analysis of peer-support group discussion in social media. Matern
Child Nutr. 2015;11(4):712–26. 30. Briere CE, McGrath JM, Cong X, Brownell E, Cusson R. Direct-breastfeeding
in the neonatal intensive care unit and breastfeeding duration for
premature infants. Appl Nurs Res. 2016;32:47–51. 5. Flacking R, Ewald U, Nyqvist KH, Starrin B. Trustful bonds: a key to
“becoming a mother” and to reciprocal breastfeeding. Stories of mothers of
very preterm infants at a neonatal unit. Soc Sci Med. 2006;62(1):70–80. Publisher’s Note 6. Ikonen R, Paavilainen E, Kaunonen M. Preterm infants’ mothers’ experiences
with milk expression and breastfeeding: an integrative review. Adv Neonatal
Care. 2015;15(6):394–406. Springer Nature remains neutral with regard to jurisdictional claims in
published maps and institutional affiliations. Springer Nature remains neutral with regard to jurisdictional claims in
published maps and institutional affiliations. 7. Ericson J, Palmer L. Mothers of preterm infants’ experiences of
breastfeeding support in the first 12 months after birth: a qualitative study. Birth. 2019;46:129–36. 7. Ericson J, Palmer L. Mothers of preterm infants’ experiences of
breastfeeding support in the first 12 months after birth: a qualitative study. Birth. 2019;46:129–36. 8. Flacking R, Ewald U, Starrin B. “I wanted to do a good job”: experiences of
‘becoming a mother’ and breastfeeding in mothers of very preterm infants
after discharge from a neonatal unit. Soc Sci Med. 2007;64(12):2405–16. 8. Flacking R, Ewald U, Starrin B. “I wanted to do a good job”: experiences of
‘becoming a mother’ and breastfeeding in mothers of very preterm infants
after discharge from a neonatal unit. Soc Sci Med. 2007;64(12):2405–16. 9. Niela-Vilen H, Axelin A, Salantera S, Melender HL. A typology of
breastfeeding mothers of preterm infants: a qualitative analysis. Adv
Neonatal Care. 2019;19(1):42–50. 10. Ericson J, Eriksson M, Hellstrom-Westas L, Hoddinott P, Flacking R. Proactive
telephone support provided to breastfeeding mothers of preterm infants
after discharge: a randomised controlled trial. Acta Paediatr. 2018;107:791–8. 10. Ericson J, Eriksson M, Hellstrom-Westas L, Hoddinott P, Flacking R. Proactive
telephone support provided to breastfeeding mothers of preterm infants
after discharge: a randomised controlled trial. Acta Paediatr. 2018;107:791–8. 11. Ericson J, Flacking R, Udo C. Mothers’ experiences of a telephone based
breastfeeding support intervention after discharge from neonatal intensive
care units: a mixed-method study. Int Breastfeed J. 2017;12:50. 11. Ericson J, Flacking R, Udo C. Mothers’ experiences of a telephone based
breastfeeding support intervention after discharge from neonatal intensive
care units: a mixed-method study. Int Breastfeed J. 2017;12:50. 12. Ericson J, Eriksson M, Hellstrom-Westas L, Hagberg L, Hoddinott P, Flacking
R. The effectiveness of proactive telephone support provided to
breastfeeding mothers of preterm infants: study protocol for a randomized
controlled trial. BMC Pediatr. 2013;13:73. 12. Ericson J, Eriksson M, Hellstrom-Westas L, Hagberg L, Hoddinott P, Flacking
R. The effectiveness of proactive telephone support provided to
breastfeeding mothers of preterm infants: study protocol for a randomized
controlled trial. BMC Pediatr. 2013;13:73. 13. Gadamer H-G. Publisher’s Note In: Weinsheimer J, Marshall DG, editors. Truth and method. First paperbacktranslation revised by. London: Bloomsbury; 2013. 13. Gadamer H-G. In: Weinsheimer J, Marshall DG, editors. Truth and method. First paperbacktranslation revised by. London: Bloomsbury; 2013. 14. Dahlberg K, Dahlberg H, Nyström M. Reflective lifeworld research. 2. Ed. Lund: Studentlitteratur; 2008. p. 23–94. 14. Dahlberg K, Dahlberg H, Nyström M. Reflective lifeworld research. 2. Ed. Lund: Studentlitteratur; 2008. p. 23–94. 15. Attride-Stirling J. Thematic networks: an analytic tool for qualitative research. Qual Res. 2001;1(3):385–405. 15. Attride-Stirling J. Thematic networks: an analytic tool for qualitative research. Qual Res. 2001;1(3):385–405. 16. Nelson AM. A metasynthesis of qualitative breastfeeding studies. J
Midwifery Womens Health. 2006;51(2):e13–20. 16. Nelson AM. A metasynthesis of qualitative breastfeeding studies. J
Midwifery Womens Health. 2006;51(2):e13–20. 16. Nelson AM. A metasynthesis of qualitative breas
Midwifery Womens Health. 2006;51(2):e13–20. 17. Charlick SJ, McKellar L, Gordon AL, Pincombe J. The private journey: an
interpretative phenomenological analysis of exclusive breastfeeding. Women Birth. 2019;32(1):e34–42. 17. Charlick SJ, McKellar L, Gordon AL, Pincombe J. The private journey: an
interpretative phenomenological analysis of exclusive breastfeeding. Women Birth. 2019;32(1):e34–42. 18. Palmer L, Carlsson G, Mollberg M, Nystrom M. Severe breastfeeding
difficulties: existential lostness as a mother-women's lived experiences
of initiating breastfeeding under severe difficulties. Int J Qual Stud
Health Well-being. 2012;7:10846. 18. Palmer L, Carlsson G, Mollberg M, Nystrom M. Severe breastfeeding
difficulties: existential lostness as a mother-women's lived experiences
of initiating breastfeeding under severe difficulties. Int J Qual Stud
Health Well-being. 2012;7:10846. 19. van Wijlen JE. Breastfeeding woman or lactating object? A critical philosophical
discussion on the influence of Cartesian dualism on breastfeeding in the
neonatal intensive care unit. J Clin Nurs. 2019;28(5–6):1022–31. 19. van Wijlen JE. Breastfeeding woman or lactating object? A critical philosophical
discussion on the influence of Cartesian dualism on breastfeeding in the
neonatal intensive care unit. J Clin Nurs. 2019;28(5–6):1022–31. 20. Merleau-Ponty M. Phenomenology of perception. In: The phenomenal
fields, p. 61–74 and the synthesis of One’s own body, 171–177. London:
Routledge; 1962/2002. 20. Merleau-Ponty M. Phenomenology of perception. In: The phenomenal
fields, p. 61–74 and the synthesis of One’s own body, 171–177. London:
Routledge; 1962/2002. 21. Todre L, Galvin KT, Dahlberg K. ‘Caring for insiderness’: phenomenologically informed
insights that can guide practice. Int J Qual Stud Health Well-Being. 2014;9:21421. 21. Todre L, Galvin KT, Dahlberg K. ‘Caring for insiderness’: phenomenologically informed
insights that can guide practice. Int J Qual Stud Health Well-Being. vulnerability and independency when initiating breastfeeding]. Växjö:
Linnaeus University Press; 2015. Linnaeus University Dissertations; 220/2015
28.
Smith PH. Social justice at the core of breastfeeding protection, promotion
and support: a conceptualization. J Hum Lact. 2018;34(2):220–5.
29.
Collins CT, Gillis J, McPhee AJ, Suganuma H, Makrides M. Avoidance of
bottles during the establishment of breast feeds in preterm infants.
Cochrane Database Syst Rev. 2016;10:Cd005252.
30.
Briere CE, McGrath JM, Cong X, Brownell E, Cusson R. Direct-breastfeeding
in the neonatal intensive care unit and breastfeeding duration for
premature infants. Appl Nurs Res. 2016;32:47–51. Publisher’s Note 2014;9:21421. 22. Dahlberg K. The art of understanding. In: Evaluation for a caring society. Abma MVT, editor. Charlotte: Information Age Publishers; 2018. p. 51–76. 22. Dahlberg K. The art of understanding. In: Evaluation for a caring society. Abma MVT, editor. Charlotte: Information Age Publishers; 2018. p. 51–76. 23. Dahlberg K, Segesten K. Hälsa och vårdande: i teori och praxis [health and
caring: in theory and praxis]. 1st ed. Stockholm: Natur & kultur; 2010. p. 103–56. 23. Dahlberg K, Segesten K. Hälsa och vårdande: i teori och praxis [health and
caring: in theory and praxis]. 1st ed. Stockholm: Natur & kultur; 2010. p. 103–56. 24. Hall Smith P, Hausman LB, Labbok M. Beyond health, beyond choice:
breastfeeding constraints and realities. In: Hausman LB, editor. Feminism
and breastfeeding: rhetoric, ideology, and the material realities of women’s
lives. New Brunswick: Rutgers University Press; 2012. p. 15–24. 25. Labbok MH, Smith PH, Taylor EC. Breastfeeding and feminism: a focus on
reproductive health, rights and justice. Int Breastfeed J. 2008;3:8. 26. Van Esterik P. Breastfeeding and feminism. Int J Gynaecol Obstet. 1994;
47(Suppl):S41–50. 26. Van Esterik P. Breastfeeding and feminism. Int J Gynaecol Obstet. 1994;
47(Suppl):S41–50. 27. Palmer L. Amning och existens: Moderskap, sårbarhet och ömsesidigt
beroende vid inledande amning [breastfeeding and existence. Motherhood,
|
https://openalex.org/W3127735013
|
https://europepmc.org/articles/pmc7851912?pdf=render
|
English
| null |
Overexpression of NELFE contributes to gastric cancer progression via Wnt/β-catenin signaling-mediated activation of CSNK2B expression
|
Journal of experimental & clinical cancer research
| 2,021
|
cc-by
| 9,158
|
RESEARCH Open Access Yu et al. Journal of Experimental & Clinical Cancer Research (2021) 40:54
https://doi.org/10.1186/s13046-021-01848-3 Yu et al. Journal of Experimental & Clinical Cancer Research (2021) 40:54
https://doi.org/10.1186/s13046-021-01848-3 Yu et al. Journal of Experimental & Clinical Cancer Research (2021) 40:54
https://doi.org/10.1186/s13046-021-01848-3 Overexpression of NELFE contributes to
gastric cancer progression via Wnt/β-
catenin signaling-mediated activation of
CSNK2B expression Shijun Yu1†, Li Li1†, Hui Cai2, Bin He1, Yong Gao1* and Yandong Li1* Shijun Yu1†, Li Li1†, Hui Cai2, Bin He1, Yong Gao1* and Yandong Li1* Abstract Background: Accumulating evidence has highlighted the importance of negative elongation factor complex
member E (NELFE) in tumorigenesis. However, the relationship between NELFE and gastric cancer (GC) remains
unclear. This study aimed to explore the expression pattern and specific function of NELFE in GC. Methods: NELFE expression was evaluated by immunohistochemistry and qRT-PCR in GC tissues, respectively. Cell
proliferation, migration and invasion were measured by CCK-8, colony formation, transwell assays, and nude mice
model. Bioinformatics analysis was performed to search potential target genes of NELFE, and a Cignal Finder 10-
Pathway Reporter Array was used to explore potential signaling pathways regulated by NELFE. Dual-luciferase
reporter assays, qRT-PCR and western blotting were conducted to verify their regulatory relationship. The expression
correlations among NELFE, β-catenin and CSNK2B were further explored by immunohistochemistry on consecutive
resections. Results: NELFE was significantly overexpressed in GC tissues both in protein and mRNA level and negatively
correlated with the prognosis of GC patients. Gain- and loss-of-function experiments showed that NELFE
potentiated GC cell proliferation and metastasis in vitro and in vivo. CSNK2B was identified as a downstream
effector of NELFE. Wnt/β-catenin signaling may mediate the regulation of CSNK2B by NELFE. In addition, NELFE, β-
catenin and CSNK2B were all remarkably upregulated in tumor tissues compared with adjacent normal tissues, and
their expression levels in GC were positively correlated with each other. Conclusion: Our findings reveal a new NELFE-Wnt/β-catenin-CSNK2B axis to promote GC progression and provide
new candidate targets against this disease. ds: NELFE, Wnt/β-catenin, CSNK2B, Gastric cancer, Tumorigenesis Keywords: NELFE, Wnt/β-catenin, CSNK2B, Gastric cancer, Tumorigenesis Tissue samples and immunohistochemistry methods p
y
To analyze the protein expression of NELFE in GC, 2 in-
dependent commercial tissue microarrays (TMA) con-
taining 68 primary tumors, 53 adjacent normal tissues
and 35 metastatic tumors from patients with GC were
purchased from Shanghai Outdo Biotech Co., Ltd.,
Shanghai, China (#HStm-Ade076Met-01 and #HStm-
Ade120lym-01). For further IHC analysis of NELFE, β-
catenin and CSNK2B expression and their correlation in
GC, consecutive sections of a TMA containing 75 paired
GC tumor and adjacent normal tissues were purchased
from Shanghai Outdo Biotech Co., Ltd., Shanghai, China
(#HStmA150CS02). The clinical characteristics of the
patients can be checked at the website: http://www. superchip.com.cn/biology/tissue.html. Standard immu-
nohistochemical staining procedures were performed
using a specific antibody against NELFE (#sc-32,912,
Santa Cruz Biotechnology, USA) at a dilution of 1:200,
and the immunostained sections were scanned using an
Aperio ScanScope CS scanner (Aperio, Vista, USA) and
visualized via Aperio ImageScope software (Aperio,
Vista, USA). The results were analyzed by two patholo-
gists blinded to the clinical information independently
based on the percentage of stained cells and staining in-
tensity. Briefly, the percentage of NELFE/β-catenin/
CSNK2B-positive cells was classified into 5 groups: <
10% (0), 10–25% (1), 25–50% (2), 50–75 (3), and > 75%
(4). The staining intensity was divided into 4 groups: no
staining (0), light brown (1), brown (2), and dark brown
(3). The
overall
IHC
scores
of
NELFE/β-catenin/
CSNK2B were calculated using the following formula:
overall score = percentage score×intensity score. The
samples with an overall score ≤6 were defined as weak
staining, and > 6 were defined as strong staining. In
addition, 28 paired of GC samples and adjacent non-
cancer tissues were collected from patients who under-
went surgical resection and signed an informed consent
in Shanghai East Hospital. After resection, these GC tis-
sue samples were immediately frozen in liquid nitrogen
and stored at −80 °C for RNA extraction. The use of hu-
man samples in this study was approved by the ethics
committee of Shanghai East Hospital, Tongji University
School of Medicine. g
NELF has been shown the essential roles in the regula-
tion of promoter-proximal pausing of RNAPII [8, 9]. It
includes four subunits, NELFA, B, C/D, and E, and only
NELFE (also known as RDBP) has a RNA binding ability. © The Author(s). 2021 Open Access This article is licensed under a Creative Commons Attribution 4.0 International License,
which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give
appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if
changes were made. The images or other third party material in this article are included in the article's Creative Commons
licence, unless indicated otherwise in a credit line to the material. If material is not included in the article's Creative Commons
licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain
permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by/4.0/. The Creative Commons Public Domain Dedication waiver (http://creativecommons.org/publicdomain/zero/1.0/) applies to the
data made available in this article, unless otherwise stated in a credit line to the data. Yu et al. Journal of Experimental & Clinical Cancer Research (2021) 40:54 Page 2 of 16 Page 2 of 16 Tissue samples and immunohistochemistry methods The structure of NELFE comprises of an N-terminal leu-
cine zipper motif, a central domain abundant in Arg-
Asp dipeptide repeats (the RD motif) and a C-terminal
RNA recognition motif (RNM) containing two highly
conserved elements dubbed RNP1 and RNP2, respect-
ively [10]. It has been implicated that NELFE binds to a
set of RNA sequences and contributes to the activities of
NELF [10, 11]. More importantly, NELFE has been
found to be involved in cancer progression in several
studies. For instance, oncogenic activation of NELFE
promotes hepatocellular carcinoma (HCC) cell prolifera-
tion and predicts the strong metastatic potential of HCC
cells [12], whereas knockdown of NELFE results in en-
hanced cell colony growth of breast cancer cells when
treated with estrogen-like reagents [13]. While these
findings implicate a critical role of NELFE in cancers, its
specific functions and underlying molecular mechanisms
are not fully understood. Gastric cancer (GC) is one of the most lethal human
malignancies worldwide [14]. Although advances of
modern medical science have improved the survival
rates of patients with GC, there is still a dearth of effect-
ive biomarkers and effective treatments, which repre-
sents a major challenge in the management of GC. In
the present study, we focused on NELFE and investi-
gated its role in the development of GC and explore the
underlying molecular mechanisms. We showed that
NELFE was overexpressed in GC tissues and could pre-
dict poor prognosis. The upregulation of NELFE facili-
tated GC cell proliferation and metastasis in vitro and
in vivo. CSNK2B was identified as a downstream effector
of NELFE. Further mechanistic study revealed that Wnt/
β-catenin signaling was involved in the regulation of
CSNK2B by NELFE. Materials and methods g
Dynamic regulation of transcription elongation shortly
after initiation by RNA polymerase II (RNAPII) plays
key roles in the implementation of gene expression
widespread in metazoans [1, 2]. This step is modulated
by positive and negative transcription elongation factors
which are known as P-TEF and N-TEF, respectively [3,
4]. Acting as N-TEFs, the proteins negative elongation
factor (NELF) and 5,6-dichloro-1β-D-ribofuranosylben-
zimidazole (DRB)-sensitivity-inducing factor (DSIF) co-
operatively inhibit elongation by associating with the
RNAPII complex, and this effect can be alleviated by P-
TEFb-mediated phosphorylation of RNAPII, DSIF and
NELF, thus continuing gene transcription [4, 5]. Dysreg-
ulation of the elongation step has been found to contrib-
ute to human diseases including cancer [6, 7]. Pathway reporter array Cignal™Finder 10-Pathway Reporter Array (Qiagen, Hil-
den, Germany) was performed to concurrently assess
whether NELFE influenced the activity of 10 different
tumor-associated signaling pathways in accordance with
the manufacturer’s specifications, and luciferase activities
of different pathways were determined using a dual lucif-
erase kit (Promega, Wisconsin, USA) as described above. Dual luciferase reporter assay Dual luciferase reporter assay
TCF/LEF1-Luc reporter plasmid was obtained from
Genomeditech
(Shanghai,
China). For
CSNK2B
promoter-reporter plasmid construction, 1200 bp up-
stream promoter region of CSNK2B gene was cloned
into pGL3 luciferase reporter vector. Dual luciferase re-
porter assays were carried out with a dual luciferase kit
(Promega, Wisconsin, USA) as follows: Cells were
seeded into a 24-well plate with a confluence of 70–90%
prior to transfection. The indicated cells were co-
transfected with the TCF/LEF1-Luc reporter plasmids or
CSNK2B promoter-reporter plasmids and Renilla lucif-
erase constructs using Lipofectamine 3000 as described
above. After incubation for 24 h, cells were lysed using
the dual luciferase kit and the luciferase activities were
measured on a Glomax-multi Luminometer (Promega,
Wisconsin, USA) according to the manufacturer’s pro-
tocols. Relative luciferase activity was calculated ac-
cording to the following formula: luciferase activity =
firefly luciferase bioluminescence/ Renilla luciferase
bioluminescence. A full-length cDNA encoding NELFE gene was PCR-
amplified cloned into pcDNA 3.1 mammalian expression
vector (pNELFE), and empty pcDNA 3.1 vector (pVEC)
was used as control group. For transient expression of
NELFE, cells were seeded into a 6-well plate at 80% con-
fluence and were transfected with 2.5 μg of pNELFE or
pVEC using Lipofectamine 3000 (Invitrogen, Thermo
Fisher Scientific, US) according to the manufacturer’s
protocol. NELFE or CSNK2B-specific small inference
RNAs (siRNAs) and negative control siRNA (siNC) were
synthesized by GenePharma (Shanghai, China). The se-
quences are as follow: siNELFE-1: 5′-GGAAAGGGAA
UACUCUCUAdTdT-3′;
siNELFE-2:
5′-GAUUCC
UUGUGCCUCAUAUdTdT-3′;
siCSNK2B-1:
5′-
CUCCGUGGCAAUGAAUUCUdTdT-3′;
siCSNK2B-2:
5′-GUCAAGACGAUUCGCUGAUdTdT-3′; siNC: 5′-
UUCUCCGAACGUGUCACGUdTdT-3′. For
siRNA
transfection, the indicated cells were seeded into a 6-
well plate with a confluence of 20–40%, and the siRNAs
were transfected into these cells using Lipofectamine
3000 (Invitrogen, Thermo Fisher Scientific, US) as per
the
manufacturer’s
instructions. NELFE-knockdown
lentivirus
(shNELFE)
was
packaged
(siNELFE-1
se-
quence was used) and supplied by Genepharma (Shang-
hai, China), the NELFE-overexpression lentivirus (LV-
NELFE) was constructed and packaged by Tuzhu Bio-
tech (Shanghai, China). For lentivirus transduction, the
host cells at a confluence of 50% were transduced with
the lentivirus in the presence of polybrene (8 μg/ml) and
stably infected cells were selected with puromycin (2 μg/
ml). Cell culture and reagents The human GC cell lines AGS, MGC-803, HGC-27,
SGC-7901 and BGC-823 were obtained from the Shang-
hai Cell Bank of the Chinese Academy of Sciences,
Shanghai, China. Cells were maintained in Dulbecco’s
modified Eagle’s medium (DMEM; Corning, Inc., Corn-
ing, NY, USA) supplemented with 10% fetal bovine
serum (FBS) and 1% penicillin/streptomycin (M&C Gene
Technology Ltd., Beijing, China). The cells were cultured Yu et al. Journal of Experimental & Clinical Cancer Research (2021) 40:54 Page 3 of 16 and their KEGG pathway enrichments were analyzed by
STRING database (Version 11.0, https://string-db.org/)
[19]. The KEGG pathway map was generated using
ggplot2 and Cario packages under R v3.5.2. and their KEGG pathway enrichments were analyzed by
STRING database (Version 11.0, https://string-db.org/)
[19]. The KEGG pathway map was generated using
ggplot2 and Cario packages under R v3.5.2. under standard conditions in a humidified incubator
with 5%CO2. To activate Wnt signaling, the indicated
cells were treated with 5 μM Wnt agonist 1 [15]
(#S8178, Selleck Chemicals, Shanghai, China) for 24 h. Plasmid construction, RNA inference and lentivirus
transduction Plasmid construction, RNA inference and lentivirus
transduction Cell proliferation assay The viability of GC cell lines was assessed using the Cell
Counting Kit-8 (CCK-8) assay kit (Dojindo, Kumamoto,
Japan). Briefly, cells were seeded into a 96-well plate at a
density of 3000 cells per well suspended in 100 μl of
complete culture medium. After incubation overnight,
cells were treated with CCK-8 solution (10 μl per well)
at the scheduled time intervals (0, 24, 48, 72, 96 h) and
incubated for 1 h. Subsequently, the optical density (OD)
values were measured at a wavelength of 450 nm on an
automated microplate reader (SpectraMax M5, Molecu-
lar Devices, USA). For colony formation assay, cells were
seeded into a 6-well plate with a density of 1000 cells/
well in triplicate and cultured under normal conditions
for 2–3 weeks. Colonies were fixed with 4% paraformal-
dehyde and stained with 0.5% crystal violet for 20 min,
after which the plates were rinsed, dried, and colonies Microarray data collection and bioinformatic analysis
The microarray data involved in this study were ob-
tained from GEO datasets of NCBI (GEO accession
number: GSE19826 and GSE13911), and the expression
levels of CSNK2B were analyzed by GEO2R tool online. Kaplan-Meier
survival
curves
were
generated
using
Kaplan–Meier plotter (http://kmplot.com/analysis/) [16],
a comprehensive online platform which is capable to as-
sess the effect of over 54,000 genes (mRNA, miRNA,
protein) on survival in
21 cancer types. The co-
expressing genes of NELFE in GC were identified using
the
Coexpedia
database
(http://www.coexpedia.org/)
[17], and their correlation was further verified by Gene
Expression Profiling Interactive Analysis (GEPIA, http://
gepia.cancer-pku.cn/) [18]. CSNK2B-interacting proteins Yu et al. Journal of Experimental & Clinical Cancer Research (2021) 40:54 Page 4 of 16 Fig. 1 (See legend on next page.) Fig 1 (See legend on next page ) Fi
1 (S
l
d
) Fig. 1 (See legend on next page.) Yu et al. Journal of Experimental & Clinical Cancer Research (2021) 40:54 Page 5 of 16 (See figure on previous page.)
Fig. 1 High NELFE expression predicts worse prognosis of patients with GC. a Representative IHC images of NELFE in adjacent normal tissues,
primary tumor tissues and metastatic tumor tissues were shown. Magnification: 40× (upper) and 200× (bottom). b Statistical difference of NELFE
protein expression among different sample types were analyzed using the Chi-square test. The samples with an IHC score ≤6 were defined as
weak, and > 6 were defined as strong. Cell proliferation assay c mRNA levels of NELFE in 28 paired tumor tissues and adjacent normal tissues were determined by qRT-
PCR analysis, relative expression levels were represented as -ΔCT values (left), and fold changes in tumor tissues relative to adjacent normal tissues
were calculated to generate the pie chart (right). d Published NELFE gene expression data were obtained from the online GEO database at
https://www.ncbi.nlm.nih.gov/geo/ (GEO accession number: GSE19826 and GSE13911). e Kaplan–Meier curves for overall survival and first
progression survival analysis of NELFE in GC were generated using the Kaplan–Meier plotter database (http://kmplot.com/analysis). f
Representative images of NELFE immunofluorescence staining (red) in MGC-803 cells. Magnification: 40×. g NELFE protein expression in different
human GC cell lines by western blotting h, the coverslips were mounted with 4,6-Diamidino-2-
Phenylindole
(DAPI)-containing
mounting
medium
(Sigma, USA). Immunofluorescence staining was visual-
ized under a Leica SP8 confocal laser scanning micro-
scope (Leica Biosystems, Germany). were photographed and counted using Image J software
(v1.52, National Institutes of Health Freeware, USA). Cell invasion assay Cell invasion was measured with BD BioCoat Matrigel
invasion chambers (BD Biosciences, #354480, USA). Cells suspended in 400 μl of serum-free DMEM were
seeded into the upper chamber at a density of 5× 104
cells/well, while 800 μl of DMEM containing 10% FBS
was added to the lower chamber as chemoattractant. After incubation for 24 h at 37 °C, cells on the upper side
were wiped off with a cotton swab, and invaded cells on
the lower side of the membrane were fixed with 4%
paraformaldehyde and stained with 0.5% crystal violet
for 30 min. The cells were photographed and counted
under an inverted microscope at × 100 magnification. Western blot analysis y
Total proteins were extracted from GC cell lines using
RIPA lysis buffer (#20–188, Merck Millipore, MA, USA)
with 1× protease and phosphatase inhibitor cocktails
(Sigma-Aldrich, MO, USA). Total protein lysates were ob-
tained by centrifuged at 4 °C and 12,000 rpm for 15 min,
and protein concentration was determined by using a
BCA Protein Assay Kit (#T9300A, Takara Bio, Kyoto,
Japan). 4× SDS loading buffer (#KGP101X, Keygentec
Inc., Shanghai, China) was added into the protein lysates
and the samples were boiled for 5 min prior to electro-
phoresis. Aliquots of the proteins were separated by 10%
SDS-PAGE and transferred onto a nitrocellulose mem-
brane (0.45 μm, Bio-Rad, CA, USA), followed by block of
non-specific binding with 5% non-fat milk in PBS contain-
ing 0.05% Tween-20 (PBST) for 1 h at room temperature. Then membranes were subjected to incubation with spe-
cific primary antibodies at 4 °C overnight and secondary
antibodies at room temperature for 1 h (avoid light). Im-
munoblots were visualized by the Odyssey Digital Infrared
Imaging System (LiCoR Biosciences, NE, USA), and β-
actin was used for internal reference. Primary antibodies
used in the present study are as follow: anti-NELFE (1:
200, # sc-32,912, Santa Cruz Biotechnology, CA, USA),
anti-CSNK2B (1:1000, # 20234–1-AP, Proteintech Group,
Wuhan, China), anti-β-catenin (1:1500, #8480, Cell Signal-
ing Technology, MA, USA) and anti-β-actin (1:5000,
#60008–1-Ig, Proteintech Group, Wuhan, China). Goat
anti-rabbit-DyLight 800 (1:1000, #SA5–35571, Thermo
Fisher Scientific, MA, USA) and anti-mouse-DyLight 800
(1:1000, #SA5–35521, Thermo Fisher Scientific, MA,
USA) served as secondary antibodies. Cell migration assay Migration assays were performed using a 24-well trans-
well chamber system (8 μm pore diameter, Costar #3422,
Corning, USA) according to the protocol of manufac-
turers. The indicated cells suspended with 300 μl FBS-
free medium were seeded into the upper chamber with a
density of 30,000/well, while 600 μl medium containing
10% FBS was added into the lower chamber. After incu-
bating for 24 h at 37 °C, the cells left in the upper surface
were removed with a cotton swab, and the migrated cells
on the lower surface were fixed with 4% paraformalde-
hyde and stained with 0.5% crystal violet for 30 min. Cells that migrated through the membranes were photo-
graphed and counted under an inverted microscope. Immunofluorescence assay Expression and subcellular location of NELFE and
CSNK2B were detected by immunofluorescence assay. The indicated cells were planted onto glass coverslips
and cultured under normal conditions. After fixation
with 4% paraformaldehyde and permeabilized with 0.1%
NP40 in PBS, the coverslips were blocked with 50%
horse serum in PBS for 1 h at 4 °C and subjected to incu-
bation of anti-NELFE (1:50) and anti-CSNK2B (1:50) at
4 °C
overnight. After
incubated
with
fluorescein-
conjugated secondary antibody (Invitrogen, USA) for 1 Quantitative Real Time PCR (qRT-PCR) analysis
Total RNA was isolated using TRI Reagent (Sigma-Al-
drich, MO, USA) and then reverse transcribed into
cDNA by PrimeScript RT reagent Kit with gDNA Eraser Quantitative Real Time PCR (qRT-PCR) analysis Yu et al. Journal of Experimental & Clinical Cancer Research (2021) 40:54 Page 6 of 16 Fig. 2 (See legend on next page.) Fig. 2 (See legend on next page.) Fig. 2 (See legend on next page.) Yu et al. Journal of Experimental & Clinical Cancer Research (2021) 40:54 Page 7 of 16 (See figure on previous page.)
Fig. 2 NELFE potentiates GC cell proliferation in vitro and in vivo. a NELFE-expressing plasmids or siRNAs against NELFE were transfected into the
indicated GC cell lines, respectively. After 48 h, cells were lysed and western blot analyses were performed to verify the efficacies of
overexpression and knockdown. b-c Cell viabilities were determined using CCK-8 cell proliferation assays at every 24 h after transfection of the
above plasmids (b) or siRNAs (c). d-e Colony formation assays were conducted using the indicated GC cell lines with lentivirus-mediated stable
overexpression (d) or knockdown (e) of NELFE. Representative colony images (left) and statistical analysis of the colony number in different
groups (right) were shown. f-g Subcutaneous xenograf model of nude mice was established using MGC-803 or BGC-823 cells with stable
overexpression (f) or knockdown (g) of NELFE (n=5 per group). Images of tumors (left), final tumor weight (middle) and tumor volume curve
(right) were shown. *P < 0.05, **P < 0.01 (Takara Bio, Kyoto, Japan) according to the manufac-
turers’ protocols. Synthesized cDNA was subjected to
qPCR analysis using TB Green Premix Ex Taq II (Takara
Bio, Kyoto, Japan) on an ABI QuantStudio 6 Flex Real-
time PCR system (Applied Biosystems, CA, USA) in ac-
cordance with the manufacturer’s instructions. Animal procedures Four to six weeks male athymic BALB/c nude mice were
obtained
from
SLAC
Laboratory
Animal
Company
(Shanghai, China) and were raised under pathogen-free
conditions with access to distilled food and water. To es-
tablish a subcutaneous xenograft model, 2× 106 of MGC-
803/BGC-823 cells suspended in 100 μl of PBS were inoc-
ulated subcutaneously into left (LV-VEC/shNC) or right
(LV-NELFE/shNELFE) flank axillary region of nude mice
(5 mice per group). The size of tumors was measured
every 7 days, and tumor volume was calculated using the
following formula: Tumor volume (mm3) = length ×
width2/2. Tumor bearing mice were euthanatized on day
28 after cell injection, histologically intact tumors were re-
moved, weighted and photographed. For the lung metasta-
sis model, 3× 106 the above-mentioned GC cells in 100 μl
of PBS were injected intravenously into the nude mice (6
mice per group) through tail vein. On day 42, the nude
mice were sacrificed, lung tissues were collected and the
number of metastatic lung nodules was counted. The lung
tissues were immersed in 4% paraformaldehyde for subse-
quent histological examination, and potential lung meta-
static lesions were finally confirmed by hematoxylin and
eosin (H&E) staining method. All animal experiments
were approved by the Medical Ethics Committee of
Shanghai East Hospital. Results NELFE is significantly upregulated in GC tissues
To investigate the expression pattern of NELFE in
GC, we performed immunohistochemistry (IHC) ana-
lysis
using
commercial
tissue
microarrays
(TMAs)
containing 68 primary gastric tumor tissues, 35 meta-
static tumor tissues and 53 adjacent normal tissues
derived from 102 GC patients, and representative IHC
images of NELFE were shown in Fig. 1a. Compared
with adjacent normal tissues, elevated NELFE expres-
sion was observed in both primary and metastatic GC
tissues (Fig. 1b). Intriguingly, metastatic tumor tissues
exhibited significantly higher NELFE expression than
primary tumor tissues (P < 0.05, Fig. 1b), suggesting
that NELFE might be a potential biomarker to distin-
guish metastatic gastric malignancies from primary
gastric malignancies. Besides, additional tumor tissues
and matched adjacent normal tissues from 28 GC pa-
tients were collected and subjected to quantitative
real-time PCR (qRT-PCR) analysis. The result showed
that the mRNA level of NELFE was also remarkably
elevated in GC tissues (P < 0.0001, Fig. 1d), among
which 18 cases (64.29%) exhibited an increase of
more than 1.5-fold change (Fig. 1c). In order to con-
firm the reliability of above data, we analyzed the ex-
pression profiles of NELFE in human gastric cancer
samples
from
public
Gene
Expression
Omnibus
(GEO) datasets (GEO accession number: GSE19826
and GSE13911). Consistently, NELFE expression in
GC tumor samples was higher than that in normal
tissues (Fig. 1d). To evaluate the prognostic signifi-
cance of elevated NELFE expression in GC, survival
analysis was conducted on an online Kaplan-Meier
Plotter database using published microarray data from Immunofluorescence assay β-actin
was used as the internal reference gene and relative gene
expression level was calculated using the comparative
cycle threshold (ddCt) method. Primer sequences for
qPCR
analysis
were
synthesized
HuaGene
Biotech
(Shanghai, China) and listed in Supplementary Table 1. GraphPad Prism software 7.0. The student’s t-test or one-
way analysis of variance (ANOVA) was used to compare
the mean difference between binary or multiple variables,
the χ2 test was used to explore the relationships between
NELFE/β-catenin/CSNK2B expression and clinicopatho-
logical parameters of GC patients. Pearson correlation co-
efficient analysis was used for comparing the correlation
among the expression of three genes. A P value of < 0.05
was considered statistically significant. (Takara Bio, Kyoto, Japan) according to the manufac-
turers’ protocols. Synthesized cDNA was subjected to
qPCR analysis using TB Green Premix Ex Taq II (Takara
Bio, Kyoto, Japan) on an ABI QuantStudio 6 Flex Real-
time PCR system (Applied Biosystems, CA, USA) in ac-
cordance with the manufacturer’s instructions. β-actin
was used as the internal reference gene and relative gene
expression level was calculated using the comparative
cycle threshold (ddCt) method. Primer sequences for
qPCR
analysis
were
synthesized
HuaGene
Biotech
(Shanghai, China) and listed in Supplementary Table 1. Statistical analysis All in vitro and in vivo experiments were performed inde-
pendently at least three times in triplicate, and continuous
variables were presented as mean ± standard deviation
(SD). Statistical
analysis
was
performed
by
using Yu et al. Journal of Experimental & Clinical Cancer Research (2021) 40:54 Page 8 of 16 Fig. 3 Positive effects of NELFE on GC cell migration in vitro and metastatis in vivo. a-b Transwell migration assays were carried out at 24 h after
transfection of NELFE-expressing plasmids (a) or siRNAs (b), and representative images (upper) and corresponding statistical analysis of migratory
cell number (bottom) were shown. Magnification: 100×. c-d A pulmonary metastasis model of nude mice was generated using cells with stable
overexpression or knockdown of NELFE (n=6 per group). Representative lung tissues (left), HE staining images of the lung tissues (middle) and
statistical analysis of the number of metastatic nodules on lung surface per mouse (right). *P < 0.05, **P < 0.01 Fig. 3 Positive effects of NELFE on GC cell migration in vitro and metastatis in vivo. a-b Transwell migration assays were carried out at 24 h after
transfection of NELFE-expressing plasmids (a) or siRNAs (b), and representative images (upper) and corresponding statistical analysis of migratory
cell number (bottom) were shown. Magnification: 100×. c-d A pulmonary metastasis model of nude mice was generated using cells with stable
overexpression or knockdown of NELFE (n=6 per group). Representative lung tissues (left), HE staining images of the lung tissues (middle) and
statistical analysis of the number of metastatic nodules on lung surface per mouse (right). *P < 0.05, **P < 0.01 Fig. 3 Positive effects of NELFE on GC cell migration in vitro and metastatis in vivo. a-b Transwell migration assays were carried out at 24 h after
transfection of NELFE-expressing plasmids (a) or siRNAs (b), and representative images (upper) and corresponding statistical analysis of migratory
cell number (bottom) were shown. Magnification: 100×. c-d A pulmonary metastasis model of nude mice was generated using cells with stable
overexpression or knockdown of NELFE (n=6 per group). Representative lung tissues (left), HE staining images of the lung tissues (middle) and
statistical analysis of the number of metastatic nodules on lung surface per mouse (right). *P < 0.05, **P < 0.01 Fig. 3 Positive effects of NELFE on GC cell migration in vitro and metastatis in vivo. Abnormal upregulation of NELFE is essential for GC cell
migration, invasion and metastasis g
Considering the IHC results that NELFE expression was
significantly upregulated in metastatic tumor tissues in
comparison with primary tumor tissues, we assumed
that NELFE plays a role in tumor metastasis of GC. To
verify this hypothesis, transwell migration and invasion
assays were carried out in GC cells following transfec-
tion with pNELFE or siNELFE-1/2. The results showed
that NELFE overexpression drastically increased the
number of migratory cells (Fig. 3a), while siRNAs-
mediated silencing of NELFE resulted in opposite results
(Fig. 3b). As the same time, similar phenotypes were also
observed in cell invasion assays (Supplementary Fig. 1a
and b). Moreover, a lung metastasis model of athymic
nude mice was generated using GC cells with stable
overexpression or knockdown of NELFE. As expected,
NELFE overexpression increased pulmonary metastatic
nodules of MGC-803 cells (Fig. 3c), whereas NELFE
knockdown in BGC-823 cells reduced pulmonary meta-
static nodules compared with the control group (Fig. 3d). Collectively, the above findings suggested that
NELFE acts as a facilitating factor for metastasis of GC. NELFE promotes GC cell proliferation in vitro and in vivo
We next assessed the biological functions of NELFE
in GC. To achieve NELFE overexpression or knock-
down in GC cell lines, NELFE-overexpression plasmid
(pNELFE) was transfected into MGC-803 and HGC-
27 cells with a relatively low NELFE expression, and
specific
siRNAs
against
NELFE
(siNELFE-1
and
siNELFE-2)
were
constructed
and
transfected
into
BGC-823 and AGS cells with high levels of NELFE
expression, after which western blot analyses con-
firmed the transfection efficiencies (Fig. 2a). Subse-
quently,
cell
viability
was
measured
using
CCK-8
assays. The growth curves indicated that NELFE over-
expression facilitated the proliferation of MGC-803
and HGC-27 cells (Fig. 2b), whereas BGC-823 and
AGS cells with NELFE silencing exhibited reduced
proliferation
rates
as
compared
to
their
control
groups (Fig. 2c). The effects of NELFE on long-term
proliferation abilities of GC cell lines were investi-
gated using colony formation assays after the indi-
cated cells were transduced with lentivirus for stable
overexpression or silencing of NELFE. As expected,
exogenous NELFE augmented colony formation of
MGC-803 and HGC-27 cells (Fig. 2d), while reduced
NELFE expression dramatically inhibited the colony
growth of BGC-823 and AGS cells (Fig. 2e). To verify
whether these in vitro findings were relevant to GC
tumor growth in vivo, a xenograft model of nude
mice was generated by subcutaneous injection of GC
cells with stable expression or knockdown of NELFE. Statistical analysis a-b Transwell migration assays were carried out at 24 h after
transfection of NELFE-expressing plasmids (a) or siRNAs (b), and representative images (upper) and corresponding statistical analysis of migratory
cell number (bottom) were shown. Magnification: 100×. c-d A pulmonary metastasis model of nude mice was generated using cells with stable
overexpression or knockdown of NELFE (n=6 per group). Representative lung tissues (left), HE staining images of the lung tissues (middle) and
statistical analysis of the number of metastatic nodules on lung surface per mouse (right). *P < 0.05, **P < 0.01 Yu et al. Journal of Experimental & Clinical Cancer Research (2021) 40:54 Page 9 of 16 Fig. 4 (See legend on next page.) Fig. 4 (See legend on next page.) Yu et al. Journal of Experimental & Clinical Cancer Research (2021) 40:54 Page 10 of 16 (See figure on previous page.)
Fig. 4 CSNK2B mediates the cancer-promoting functions of NELFE in GC. a Data from Coexpedia database (http://www.coexpedia.org/) was
analyzed by informatics method to identify co-expressing genes of NELFE in GC, which were ranked by sum of their edges’ log-likelyhood scores. b RNA-sequencing data from TCGA database were analyzed using the online GEPIA tool (http://gepia.cancer-pku.cn/) to validate the results
obtained from Coexpedia database, and CSNK2B and BAG6 were considered as candidate genes using the following criteria: Pearson correlation
coefficient ≥0.7 and P value < 0.05. c CSNK2B and BAG6 mRNA levels in AGS cells with stable knockdown of NELFE were analyzed using qRT-PCR
analysis. d Western blot analyses confirming the regulation of CSNK2B protein expression by NELFE in GC cells were shown. e After transfection
with siRNAs against CSNK2B, cells were lysed and western blot analysis was performed to validate the knockdown efficacies. f-h CCK-8 assays (f),
colony formation assays (g) and transwell migration assays (h) were carried out at 24 h after transfection of siNC or siCSNK2B-2 into MGC-803 cells
with stable NELFE overexpression or the control cells, respectively. **P < 0.01, n.s. not statistically significant GC patients. As shown in Fig. 1e, high NELFE ex-
pression was negatively correlated with overall sur-
vival (P < 0.05) and first progression survival (P <
0.05) rates of GC patients. By immunofluorescence
assay, we observed that the subcellular location of
NELFE in MGC-803 cells was mostly in nucleus and
partly in cytoplasm (Fig. 1f). Statistical analysis In addition, the expres-
sion pattern of NELFE in various GC cell lines was
shown in Fig. 1g. Taken together, these data sug-
gested that NELFE is overexpressed in GC and may
serve as a potential biomarker for the progression of
GC. expression of NELFE substantially increased tumor
size and tumor weight (Fig. 2f), while knockdown of
NELFE led to opposite results (Fig. 2g). These find-
ings revealed that NELFE plays important roles in GC
cell growth in vitro and in vivo. Abnormal upregulation of NELFE is essential for GC cell
migration, invasion and metastasis Consistent
with
the
in
vitro
observations,
forced Oncogenic functions of NELFE in GC is mediated by
CSNK2B Notably, a concomitant decrease in CSNK2B
mRNA was observed, while BAG6 expression showed no
significant change in AGS-shNELFE cells as compared
to the control cells (Fig. 4c), indicating that CSNK2B
was regulated by NELFE. Further western blot results
confirmed the positive effects of NELFE on CSNK2B
protein expression (Fig. 4d). CSNK2B is a regulatory
subunit of casin kinase 2 (CK2) and plays a key role in
regulating its kinase activity [20]. Previous studies have
shown that CSNK2B is an oncogene in different human
malignancies by directly trans-activating NF-κB signaling
pathway [21, 22]. Cignal Finder 10-pathway reporter array to screen
NELFE-associated signaling pathways in GC cells. Among the 10 signaling pathways directly related
with tumor progression, Wnt/β-catenin pathway was
strongly activated by NELFE overexpression (Fig. 5a). Next, a TCF/LEF1 luciferase reporter assay was per-
formed to validate the effect of NELFE on the activ-
ity of canonical Wnt/β-catenin signaling pathway. Consistent with the data described above, higher
levels of Wnt-dependent activities were observed in
cells with NELFE overexpression, whereas silenced
NELFE led to contrary results (Fig. 5b). Further-
more, upon NELFE overexpression or knockdown in
GC cell lines, we examined the protein and mRNA
level of β-catenin, a major component of Wnt sig-
naling pathway [25], respectively. Interestingly, the
protein level of β-catenin was enhanced by NELFE
(Fig. 5c), while no significant change in mRNA level
was observed after NELFE overexpression or knock-
down (Fig. 5d), implying the regulation of β-catenin
by NELFE is not at the transcriptional level. To ex-
plore whether CSNK2B is associated with Wnt/β-ca-
tenin
signaling
pathway,
CSNK2B-interacting
proteins were identified using the STRING database
and further KEGG pathway analysis was performed. Notably, CSNK2B-interacting proteins were highly
enriched in Wnt/β-catenin signaling pathway (Fig. 5e), suggesting a relationship between CSNK2B and
Wnt
signaling. Thereby,
we
aimed
to
explore
whether
Wnt/β-catenin
signaling
pathway
could
regulate CSNK2B expression in GC cells. By com-
paring sequences, 2 potential TCF/LEF response ele-
ments with the consensus sequence CANNTG (E-
box motifs) were recognized in the promoter region
of CSNK2B (Fig. 5f), indicating that CSNK2B might
be a target gene of TCF/LEF transcription factors. To
verify
this
possibility,
CSNK2B
promoter-
luciferase reporter plasmid was constructed and a
Wnt/β-catenin pathway activator (Wnt agonist 1)
was
used
for
dual-luciferase
reporter
assays. Oncogenic functions of NELFE in GC is mediated by
CSNK2B To
explore
potential
mechanisms
possible
for
the
cancer-promoting functions of NELFE, we performed
gene co-expression analysis using data from Coexpedia
database, and 12 co-expressed genes of NELFE were
identified in GC (Fig. 4a). For further validation, Pearson
correlation analysis was conducted to compare the rela-
tionships between NEFLE and these genes using an on-
line Cancer Genome Atlas (TCGA) data visualization
web tool GEPIA. Among the 12 genes, casein kinase 2
beta (CSNK2B, r = 0.77, P < 0.01) and BAG cochaperone
6 (BAG6, r = 0.70, P < 0.01) showed the highest correl-
ation with NELFE in GC (Fig. 4b). Subsequently, we Yu et al. Journal of Experimental & Clinical Cancer Research (2021) 40:54 Page 11 of 16 Fig. 5 (See legend on next page.) Fig. 5 (See legend on next page.) Yu et al. Journal of Experimental & Clinical Cancer Research (2021) 40:54 Page 12 of 16 (See figure on previous page.)
Fig. 5 NELFE promotes CSNK2B expression via activating Wnt/β-catenin signaling pathway. a At 24 h after transfection of pNELFE into MGC-803
cells, Cignal Finder 10-Pathway Reporter Array was employed to uncover potential downstream signaling pathways regulated by NELFE. b The
indicated cells were transiently co-transfected with the TCF/LEF1 firefly luciferase reporter construct and the Renilla luciferase vector for 24 h, and
dual-luciferase reporter assays were conducted to detect relative TCF/LFE1 luciferase activity. c-d β-catenin (c) protein and mRNA (d) expression
levels in the indicated GC cell lines were determined by western blot and qPCR analyses, respectively. e CSNK2B-interacting proteins were
obtained from the STRING database at http://string-db.org, and Kyoto Encyclopedia of Genes and Genomes (KEGG) pathway enrichment analysis
was performed by R programming language v3.5.2. f Schematic representation of two predicted TCF/LEF-binding sites in the promoter region of
CSNK2B. g MGC-803 and AGS cells were transfected with CSNK2B promoter-driven luciferase constructs, 5 μM of Wnt agonist 1 was added into
the cells concurrently, and CSNK2B promoter activity was measured using dual-luciferase assays after 24 h. h AGS cells with stable knockdown of
NELFE and the control cells were treated with 5 μM of Wnt agonist 1, and DMSO were used as the control group. Then dual-luciferase reporter
assays were performed to compare the difference of CSNK2B promoter activity between the two groups. **P < 0.01, n.s. not statistically significant performed qRT-PCR analysis to determine their mRNA
expression levels after knockdown of NELFE in AGS
cells. Oncogenic functions of NELFE in GC is mediated by
CSNK2B In
agreement
with
our expectations,
treatment
with
Wnt agonist 1 dramatically increased the activity of On the basis of these findings, we next synthesized
two
siRNAs
against
CSNK2B
(siCSNK2B-1
and
siCSNK2B-2) to test whether NELFE exerts its func-
tion dependent on CSNK2B expression. Western blot
results showed that both siCSNK2B-1 and siCSNK2B-
2 successfully silenced CSNK2B expression in MGC-
803 cells, while siCSNK2B-2 exhibited a better knock-
down
efficiency
than
siCSNK2B-1
(Fig. 4e),
thus
siCSNK2B-2 was selected for further cell biology ex-
periments. CCK-8 assays were carried out after trans-
fection
of
siCSNK2B-2
into
MGC-803
cells
with
NELFE overexpression. In line with previous reports
[22–24], knockdown of CSNK2B significantly arrested
cell
proliferation,
and
more
importantly,
it
also
strongly weakened the positive effects of NELFE on
proliferation ability of MGC-803 cells (Fig. 4f). Similar
results were also observed in colony formation and
transwell migration assays (Fig. 4g, h), hinting us that
NELFE enhanced GC cell proliferation and migration
through activating CSNK2B expression. Accordingly,
it can be deduced that CSNK2B is a downstream fac-
tor of NELFE and mediates positive effects of NELFE
on GC progression. NELFE enhances CSNK2B gene expression via activation
of Wnt/β-catenin pathway Discussion It is well established that tumorigenesis is a complex
process commonly accompanied with aberrant acti-
vation of oncogenic genes [28]. NELFE, a member of
RNA-binding proteins (RBP) which is involved in
regulating transcription elongation and RNA homeo-
stasis, has received increasing attention in recent
years in the development of human malignancies for
its key roles in dysregulated transcriptional events
[29–31]. Nevertheless up to date, the relationship be-
tween NELFE and GC progression remains largely
unknown. In the present study, we for the first time
investigated the expression features of NELFE in GC. We found NELFE was dramatically overexpressed in
GC tumor tissues than in adjacent normal tissues by
IHC and qPCR analysis, and metastatic tumors ex-
hibited an increased NELFE expression than primary
tumors, indicating the participation of NELFE in GC
progression. Gene expression data from GEO data-
base further confirmed these results, and a negative
correlation of NELFE expression with the overall and
first progression survival rates of GC patients were
found based on the online survival database Kaplan-
Meier plotter. Through gain- or loss-of-function ex-
periments, we showed that NELFE facilitated GC cell
proliferation, migration and invasion in vitro, and
further animal experiments confirmed this conclu-
sion in vivo. These observations unambiguously sup-
ported the cancer-promoting role of NELFE in GC,
consistent with previous studies on NELFE in HCC
or pancreatic cancer [12, 30]. GC tissues are positively correlated with each other
To investigate possible clinical relevance of these 3
genes in GC, we performed an IHC analysis using
consecutive sections on a TMA containing 75 paired
tumor/adjacent normal tissues from GC patients. Consistent with the IHC data described above, there
was a significant increase in NELFE expression in
tumor tissues than in adjacent normal tissues (P <
0.01, Fig. 6a), and the same trend was observed in
β-catenin expression (P < 0.001, Fig. 6b), which has
been well understood in previous studies [26, 27]. Similarly, high CSNK2B expression was also found
in tumor tissues when compared with adjacent nor-
mal tissues (P < 0.05, Fig. 6c). Further analysis of
the same locations on consecutive tumor tissues
sections indicated a consistent trend among the ex-
pression levels of all three genes in GC (Fig. 6d). Moreover, Pearson correlation analyses were applied
based on the IHC scores of NELFE, β-catenin and
CSNK2B (Fig. 6e). NELFE enhances CSNK2B gene expression via activation
of Wnt/β-catenin pathway In an attempt to better understand how NELFE reg-
ulates
CSNK2B
expression
in
GC,
we
used
the Yu et al. Journal of Experimental & Clinical Cancer Research (2021) 40:54 Page 13 of 16 Fig. 6 (See legend on next page.) Fig. 6 (See legend on next page.) Fig. 6 (See legend on next page.) Yu et al. Journal of Experimental & Clinical Cancer Research (2021) 40:54 Page 14 of 16 (See figure on previous page.)
Fig. 6 Overexpressed NELFE, β-catenin and CSNK2B in GC are positively correlated with each other. a-c Representative IHC images of NELFE (a),
β-catenin (b) and CSNK2B (c) in GC tumor tissues and adjacent normal tissues. Magnification: 40× (upper) and 200× (bottom). Bar graphs showed
the statistical analysis of three proteins expression between tumor tissues and adjacent normal tissues. d Representative IHC images of NELFE, β-
catenin and CSNK2B at the same location on serial tumor sections. e The correlation among these three genes in GC tumor tissues were
analyzed using Pearson’s correlation coefficient analysis (n=75) according to their respective IHC scores, and some of the dots represented more
than one specimen CSNK2B promoter in MGC-803 and AGS cells (Fig. 5g), indicating that Wnt/β-catenin pathway mediated
transcriptional activation of CSNK2B gene. Further
immunofluorescence analysis showed a significant
enhancement
of
CSNK2B
staining
intensity
after
Wnt agonist 1 treatment (Supplementary Fig. 1c). Combined with the above analyses, we questioned
whether NELFE regulated CSNK2B through activa-
tion of Wnt/β-catenin pathway. Dual luciferase as-
says
demonstrated
that
NELFE
inhibition
significantly reduced CSNK2B promoter activity (P
< 0.01), which could be rescued by treatment of
Wnt agonist 1 (Fig. 5h). Collectively, these data
strongly suggested that Wnt/β-catenin pathway me-
diate the activation of CSNK2B expression induced
by NELFE. Discussion As expected, both β-catenin (R =
0.5072, P < 0.0001) and CSNK2B (R = 0.5292, P <
0.0001) expression showed a strong correlation with
NELFE expression in GC, and both of them were
positively associated with each other (R = 0.6143, P
< 0.0001). Thus, these findings further supported the
regulatory relationship among NELFE, β-catenin and
CSNK2B in GC. In mechanistic studies, we performed comprehensive
analysis combining the bioinformatics data and our ex-
perimental evidence to elucidate the exact mechanisms
by which NELFE exerts its effects during GC tumorigen-
esis, and CSNK2B, a regulatory subunit of casein kinase
II (CK2), was identified as a downstream effector of
NELFE. Studies have demonstrated that CSNK2B is es-
sential for the kinase activity of CK2, and overexpression
of CSNK2B has been reported to promote HCC tumori-
genesis via trans-activation of NF-κB [20, 21]. Although
there is currently no available evidence regarding the
roles of CSNK2B in GC, our cell biology experiments
confirmed that CSNK2B facilitated GC cell proliferation
and migration, and it also mediated the pro-oncogenic
functions of NELFE in GC. Yu et al. Journal of Experimental & Clinical Cancer Research (2021) 40:54 Page 15 of 16 (2021) 40:54 Fig. 7 Schematic illustration of the NELFE/β-catenin/CSNK2B signaling axis in GC progression. Increased NELFE expression activates the Wnt/β-
catenin pathway via upregulation of β-catenin. In the nucleus, interaction of β-catenin with TCF/LEF transcription factors results in activation of
CSNK2B expression, which finally promotes GC cell proliferation and metastasis Fig. 7 Schematic illustration of the NELFE/β-catenin/CSNK2B signaling axis in GC progression. Increased NELFE expression activates the Wnt/β-
catenin pathway via upregulation of β-catenin. In the nucleus, interaction of β-catenin with TCF/LEF transcription factors results in activation of
CSNK2B expression, which finally promotes GC cell proliferation and metastasis Another important finding of our work is that Wnt/β-
catenin signaling pathway was uncovered to mediate the
regulation of NELFE on CSNK2B expression. As we
known, activation of β-catenin is then able to translocate
into the nucleus and form a complex with T-cell factor/
lymphoid enhancer factor (TCF/LEF) proteins to induce
the transcription of genes involved in cell growth and
proliferation, such as c-myc and cyclin D1 [32]. Consist-
ently, dual-luciferase assays showed that CSNK2B gene
was
transcriptionally
activated
by
TCF/LEF,
which
means Wnt/β-catenin signaling pathway is a positive
regulator of CSNK2B expression. Acknowledgements
Not applicable. Acknowledgements
Not applicable. Competing interests
The authors declare there is no conflict of interest. Supplementary Information
The online version contains supplementary material available at https://doi. org/10.1186/s13046-021-01848-3. Additional file 1: Supplementary Table 1. Primer sequences for qRT-
PCR analysis. pp
y
The online version contains supplementary material available at https://doi. org/10.1186/s13046-021-01848-3. Authors’ contributions Yandong Li and Yong Gao conceived and designed the study; Shijun Yu and
Li Li performed most of experiments; Li Li and Hui Cai collected and
analyzed the data; Bin He performed animal experiments; Shijun Yu wrote
the paper, and Yandong Li revised the paper. All authors have read and
approved the final manuscript. Supplementary Information Competing interests
The authors declare there is no conflict of interest. Availability of data and materials All data generated or analyzed during this study are included in this
published article and its supplementary information files. Discussion Therefore, these find-
ings identified a NELFE/β-catenin/CSNK2B axis in pro-
moting the initiation and development of GC. However,
the exact mechanism for how NELFE enhanced the pro-
tein expression of β-catenin, still needs more in-depth
studies. In addition, it is worth noting that a recent study
reported that NELFE is able to decrease the stabilization
of
N-Myc
downstream-regulated
gene
2
(NDRG2),
which is involved in the inhibition of the Wnt/β-catenin
signaling pathway [30], thus providing a possible explan-
ation for the effects of NELFE on β-catenin expression. Certainly, more experimental evidence is needed to ver-
ify this hypothesis. Finally, our IHC analysis on consecu-
tive sections of a TMA confirmed that NELFE, β-catenin
and CSNK2B expression were all highly expressed in
tumor tissues as compared to adjacent normal tissues,
and all these three genes expression in GC were posi-
tively correlated with each other, which provides further
support to the above experimental findings. Additional file 2: Supplementary Figure 1. a-b Cell invasion assays
were carried out at 24 h after transfection of NELFE-expressing plasmids
(a) or siRNAs (b), and representative images (upper) and quantification
analysis (bottom) were shown. Magnification: 100×. *P < 0.05, **P < 0.01. c Representative images of CSNK2B immunofluorescence staining (red) in
MGC-803 cells after Wnt agonist 1 treatment (5 μM). Magnification: 40×. Funding This study was supported by grants from the National Key Research and
Development Program of China (Code: 2017YFC1308900), the National
Natural Science Foundation of China (Code: 81772568 and 81772567), Key
Disciplines Group Construction Project of Pudong Health Bureau of Shanghai
(Code: PWZxq2017–13), and the Outstanding Clinical Discipline Project of
Shanghai Pudong (Code: PWYgy2018–02). Conclusion This study revealed that NELFE functions as an onco-
gene in GC and a new oncogenic NELFE/β-catenin/
CSNK2B signaling axis was identified, which may help
to
develop
new
diagnosis
and
treatment
strategies
against GC (Fig. 7). Abbreviations NELFE: Negative elongation factor complex member E; GC: Gastric cancer;
IHC: Immunohistochemistry; qPCR: Quantitative real-time PCR; TMA: Tissue
microarray; RNAPII: RNA polymerase II; NELF: Negative elongation factor;
DRB: 5,6-dichloro-1β-D-ribofuranosylbenzimidazole; DSIF: DRB-sensitivity-
inducing factor; RNM: RNA recognition motif; HCC: Hepatocellular carcinoma;
CCK-8: Cell Counting Kit-8; OD: Optical density; PBST: PBS containing 0.05%
Tween-20; H&E: Hematoxylin and eosin; ANOVA: One-way analysis of
variance; SD: Standard deviation; GEO: Gene Expression Omnibus; BAG6: BAG
cochaperone 6; CSNK2B: Casein kinase 2 beta; TCGA: Cancer Genome Atlas;
RBP: RNA-binding proteins; CK2: Casein kinase II; TCF/LEF: T-cell factor/
lymphoid enhancer factor; NDRG2: N-Myc downstream-regulated gene 2 Ethics approval and consent to participate This study has been approved by the ethics committee of Shanghai East
Hospital, Tongji University School of Medicine, China. Consent for publication
Not applicable. Author details
1 Author details
1Department of Oncology, Shanghai East Hospital, Tongji University School
of Medicine, Shanghai 200120, China. 2Department of Geriatrics, The Second
Affiliated Hospital of Zhejiang University, Hangzhou 310009, China. Additional file 1: Supplementary Table 1. Primer sequences for qRT-
PCR analysis. Page 16 of 16 Page 16 of 16 Page 16 of 16 Yu et al. Journal of Experimental & Clinical Cancer Research (2021) 40:54 Yu et al. Journal of Experimental & Clinical Cancer Research (2021) 40:54 Received: 13 October 2020 Accepted: 17 January 2021 References Liang K, Smith ER, Aoi Y, Stoltz KL, Katagi H, Woodfin AR, et al. Targeting
Processive transcription elongation via SEC disruption for MYC-induced
Cancer therapy. Cell. 2018;175(3):766–79 e17. 30. Han L, Zan Y, Huang C, Zhang S. NELFE promoted pancreatic cancer
metastasis and the epithelialtomesenchymal transition by decreasing the
stabilization of NDRG2 mRNA. Int J Oncol. 2019;55(6):1313–23. 8. Vos SM, Farnung L, Urlaub H, Cramer P. Structure of paused transcription
complex pol II-DSIF-NELF. Nature. 2018;560(7720):601–6. 8. Vos SM, Farnung L, Urlaub H, Cramer P. Structure of paused transcription
complex pol II-DSIF-NELF. Nature. 2018;560(7720):601–6. 31. Dang H, Pomyen Y, Martin SP, Dominguez DA, Yim SY, Lee JS, et al. NELFE-
dependent MYC signature identifies a unique Cancer subtype in
hepatocellular carcinoma. Sci Rep. 2019;9(1):3369. 9. Narita T, Yamaguchi Y, Yano K, Sugimoto S, Chanarat S, Wada T, et al. Human transcription elongation factor NELF: identification of novel subunits
and reconstitution of the functionally active complex. Mol Cell Biol. 2003;
23(6):1863–73. 32. Ashihara E, Takada T, Maekawa T. Targeting the canonical Wnt/beta-catenin
pathway in hematological malignancies. Cancer Sci. 2015;106(6):665–71. 32. Ashihara E, Takada T, Maekawa T. Targeting the canonical Wnt/beta-catenin
pathway in hematological malignancies. Cancer Sci. 2015;106(6):665–71. 10. Yamaguchi Y, Inukai N, Narita T, Wada T, Handa H. Evidence that negative
elongation factor represses transcription elongation through binding to a
DRB sensitivity-inducing factor/RNA polymerase II complex and RNA. Mol
Cell Biol. 2002;22(9):2918–27. Received: 13 October 2020 Accepted: 17 January 2021 downregulation of CSNK2B blocks nuclear factor-kappaB activation in
hepatocellular carcinoma. EBioMedicine. 2020;51:102603. downregulation of CSNK2B blocks nuclear factor-kappaB activation in
hepatocellular carcinoma. EBioMedicine. 2020;51:102603. downregulation of CSNK2B blocks nuclear factor-kappaB activation in
hepatocellular carcinoma. EBioMedicine. 2020;51:102603. 23. Chua MMJ, Lee M, Dominguez I. Cancer-type dependent expression of CK2
transcripts. PLoS One. 2017;12(12):e0188854. transcripts. PLoS One. 2017;12(12):e0188854. 24. Wang W, Wang X, Li C, Chen T, Zhang N, Liang Y, et al. Huaier suppresses
breast Cancer progression via linc00339/miR-4656/CSNK2B signaling
pathway. Front Oncol. 2019;9:1195. References 1. Chen FX, Smith ER, Shilatifard A. Born to run: control of transcription
elongation by RNA polymerase II. Nat Rev Mol Cell Biol. 2018;19(7):464–78 25. Nusse R. Wnt signaling in disease and in development. Cell Res. 2005;15(1):
28–32. 2. Uptain SM, Kane CM, Chamberlin MJ. Basic mechanisms of transcript
elongation and its regulation. Annu Rev Biochem. 1997;66:117–72. 26. Wu C, Zhuang Y, Jiang S, Liu S, Zhou J, Wu J, et al. Interaction between
Wnt/beta-catenin pathway and microRNAs regulates epithelial-
mesenchymal transition in gastric cancer (review). Int J Oncol. 2016;48(6):
2236–46. 3. Peterlin BM, Price DH. Controlling the elongation phase of transcription
with P-TEFb. Mol Cell. 2006;23(3):297–305. 4. Yamaguchi Y, Takagi T, Wada T, Yano K, Furuya A, Sugimoto S, et al. NELF, a
multisubunit complex containing RD, cooperates with DSIF to repress RNA
polymerase II elongation. Cell. 1999;97(1):41–51. 4. Yamaguchi Y, Takagi T, Wada T, Yano K, Furuya A, Sugimoto S, et al. NELF, a
multisubunit complex containing RD, cooperates with DSIF to repress RNA
polymerase II elongation. Cell. 1999;97(1):41–51. 27. Song Y, Li ZX, Liu X, Wang R, Li LW, Zhang Q. The Wnt/beta-catenin and
PI3K/Akt signaling pathways promote EMT in gastric cancer by epigenetic
regulation via H3 lysine 27 acetylation. Tumour Biol. 2017;39(7):
1010428317712617. 5. Yamaguchi Y, Wada T, Watanabe D, Takagi T, Hasegawa J, Handa H. Structure and function of the human transcription elongation factor DSIF. J
Biol Chem. 1999;274(12):8085–92. 28. Hanahan D, Weinberg RA. Hallmarks of cancer: the next generation. Cell. 2011;144(5):646–74. 6. Lin C, Smith ER, Takahashi H, Lai KC, Martin-Brown S, Florens L, et al. AFF4, a
component of the ELL/P-TEFb elongation complex and a shared subunit of
MLL chimeras, can link transcription elongation to leukemia. Mol Cell. 2010;
37(3):429–37. 6. Lin C, Smith ER, Takahashi H, Lai KC, Martin-Brown S, Florens L, et al. AFF4, a
component of the ELL/P-TEFb elongation complex and a shared subunit of
MLL chimeras, can link transcription elongation to leukemia. Mol Cell. 2010;
37(3):429–37. 29. Dang H, Takai A, Forgues M, Pomyen Y, Mou H, Xue W, et al. Oncogenic
activation of the RNA binding protein NELFE and MYC signaling in
hepatocellular carcinoma. Cancer Cell. 2017;32(1):101–14 e8. 7. Liang K, Smith ER, Aoi Y, Stoltz KL, Katagi H, Woodfin AR, et al. Targeting
Processive transcription elongation via SEC disruption for MYC-induced
Cancer therapy. Cell. 2018;175(3):766–79 e17. 7. Publisher’s Note Springer Nature remains neutral with regard to jurisdictional claims in
published maps and institutional affiliations. Springer Nature remains neutral with regard to jurisdictional claims in
published maps and institutional affiliations. 11. Rao JN, Neumann L, Wenzel S, Schweimer K, Rosch P, Wohrl BM. Structural
studies on the RNA-recognition motif of NELF E, a cellular negative
transcription elongation factor involved in the regulation of HIV
transcription. Biochem J. 2006;400(3):449–56. 12. Iida M, Iizuka N, Tsunedomi R, Tsutsui M, Yoshida S, Maeda Y, et al. Overexpression of the RD RNA binding protein in hepatitis C virus-related
hepatocellular carcinoma. Oncol Rep. 2012;28(2):728–34. 13. Aiyar SE, Sun JL, Blair AL, Moskaluk CA, Lu YZ, Ye QN, et al. Attenuation of
estrogen receptor alpha-mediated transcription through estrogen-
stimulated recruitment of a negative elongation factor. Genes Dev. 2004;
18(17):2134–46. 14. Bray F, Ferlay J, Soerjomataram I, Siegel RL, Torre LA, Jemal A. Global cancer
statistics 2018: GLOBOCAN estimates of incidence and mortality worldwide
for 36 cancers in 185 countries. CA Cancer J Clin. 2018;68(6):394–424. 15. Liu J, Wu X, Mitchell B, Kintner C, Ding S, Schultz PG. A small-molecule
agonist of the Wnt signaling pathway. Angew Chem. 2005;44(13):1987–90. 16. Szasz AM, Lanczky A, Nagy A, Forster S, Hark K, Green JE, et al. Cross-
validation of survival associated biomarkers in gastric cancer using
transcriptomic data of 1,065 patients. Oncotarget. 2016;7(31):49322–33. 17. Yang S, Kim CY, Hwang S, Kim E, Kim H, Shim H, et al. COEXPEDIA:
exploring biomedical hypotheses via co-expressions associated with
medical subject headings (MeSH). Nucleic Acids Res. 2017;45(D1):D389–D96. 18. Tang Z, Li C, Kang B, Gao G, Li C, Zhang Z. GEPIA: a web server for cancer
and normal gene expression profiling and interactive analyses. Nucleic
Acids Res. 2017;45(W1):W98–W102. 19. Szklarczyk D, Gable AL, Lyon D, Junge A, Wyder S, Huerta-Cepas J, et al. STRI
NG v11: protein-protein association networks with increased coverage,
supporting functional discovery in genome-wide experimental datasets. Nucleic Acids Res. 2019;47(D1):D607–D13. 20. Litchfield DW. Protein kinase CK2: structure, regulation and role in cellular
decisions of life and death. Biochem J. 2003;369(Pt 1):1–15. 21. Yu M, Yeh J, Van Waes C. Protein kinase casein kinase 2 mediates inhibitor-
kappaB kinase and aberrant nuclear factor-kappaB activation by serum
factor(s) in head and neck squamous carcinoma cells. Cancer Res. 2006;
66(13):6722–31. 22. Xiao Y, Huang S, Qiu F, Ding X, Sun Y, Wei C, et al. Publisher’s Note Tumor necrosis factor
alpha-induced protein 1 as a novel tumor suppressor through selective
|
https://openalex.org/W3132985183
|
https://www.medecinesciences.org/articles/medsci/pdf/2021/02/msc200239.pdf
|
French
| null |
[Latest insights on X-chromosome inactivation: When general principles should be revisited].
| null | 2,021
|
cc-by
| 6,230
|
Dernières nouvelles
du chromosome X
Des principes généraux
nuancés
Madeleine Moscatelli, Claire Rougeulle
l’ICX. Ces nouvelles données nous conduisent aujourd’hui à revisiter et
nuancer certains des principes généraux initialement établis.
Université de Paris,
Épigénétique et Destin Cellulaire,
CNRS, F-75006 Paris, France.
claire.rougeulle@u-paris.fr Dernières nouvelles
du chromosome X
Des principes généraux
nuancés
Madeleine Moscatelli, Claire Rougeulle
l’ICX. Ces nouvelles données nous conduisent aujourd’hui à revisiter et
nuancer certains des principes généraux initialement établis. Université de Paris,
Épigénétique et Destin Cellulaire,
CNRS, F-75006 Paris, France. claire.rougeulle@u-paris.fr médecine/sciences
médecine/sciences médecine/sciences 2021 ; 37 : 152-8 médecine/sciences 2021 ; 37 : 152-8 Dernières nouvelles
du chromosome X
Des principes généraux
nuancés > L’inactivation d’un des deux chromosomes X des
femelles mammifères est un processus vital et
emblématique des régulations épigénétiques. Elle
est déclenchée par l’accumulation d’un ARN non
codant, XIST, qui isole le chromosome concerné
de la machinerie transcriptionnelle ; l’état inactif
persiste ensuite de manière stable au cours des
divisions cellulaires successives. Cependant, des
découvertes récentes conduisent à revisiter cer-
tains principes généraux de l’inactivation du chro-
mosome X initialement établis. Ainsi le chercheur,
tout comme le poète1, est-il invité à « vingt fois
sur le métier remettre son ouvrage ». < > L’inactivation d’un des deux chromosomes X des
femelles mammifères est un processus vital et
emblématique des régulations épigénétiques. Elle
est déclenchée par l’accumulation d’un ARN non
codant, XIST, qui isole le chromosome concerné
de la machinerie transcriptionnelle ; l’état inactif
persiste ensuite de manière stable au cours des
divisions cellulaires successives. Cependant, des
découvertes récentes conduisent à revisiter cer-
tains principes généraux de l’inactivation du chro-
mosome X initialement établis. Ainsi le chercheur,
tout comme le poète1, est-il invité à « vingt fois
sur le métier remettre son ouvrage ». < Madeleine Moscatelli, Claire Rougeulle l’ICX. Ces nouvelles données nous conduisent aujourd’hui à revisiter et
nuancer certains des principes généraux initialement établis. Université de Paris,
Épigénétique et Destin Cellulaire,
CNRS, F-75006 Paris, France. claire.rougeulle@u-paris.fr l’ICX. Ces nouvelles données nous conduisent aujourd’hui à revisiter et
nuancer certains des principes généraux initialement établis. Chez les mammifères, l’unique différence entre les
mâles et les femelles d’un point de vue génétique se
situe au niveau de la 23e paire de chromosomes, les
chromosomes sexuels X et Y. Chez les femelles, la pré-
sence de deux copies du chromosome X (XX) par rapport
au mâle (XY) est compensée, très tôt au cours du déve-
loppement de l’embryon, par l’inactivation d’un des
deux X – c’est-à-dire que la quasi-totalité des gènes du
chromosome ainsi inactivé ne s’exprime plus. Dernières nouvelles
du chromosome X
Des principes généraux
nuancés
Madeleine Moscatelli, Claire Rougeulle
l’ICX. Ces nouvelles données nous conduisent aujourd’hui à revisiter et
nuancer certains des principes généraux initialement établis.
Université de Paris,
Épigénétique et Destin Cellulaire,
CNRS, F-75006 Paris, France.
claire.rougeulle@u-paris.fr Cet état
silencieux est ensuite maintenu, de manière stable et
pourtant réversible, dans toutes les cellules de l’orga-
nisme (à l’exception des cellules sexuelles qui ne seront
pas discutées ici). Ce processus d’inactivation du chro-
mosome X (ICX) constitue un paradigme de régulation
épigénétique qui implique un traitement différent de
deux chromosomes homologues au sein d’un même
noyau cellulaire. L’hypothèse de l’inactivation du X a
été proposée au début des années
1960 par Mary Lyon, une éminente
généticienne britannique [1] (➜). L’étude de ce processus, principa-
lement chez la souris, a, depuis, permis d’en comprendre
les principes généraux mais aussi les mécanismes molé-
culaires. Plus récemment, les progrès techniques et
l’analyse d’autres espèces, l’homme en particulier, ont
contribué à compléter et préciser nos connaissances sur
(➜) Voir le Forum de
J. L. Guenet et al.,
m/s n° 6-7, juin-
juillet 2015, page 687 Vignette (Photo © Inserm/Heard, Edith/Institut Curie).
1 Nicolas Boileau (1636-1711), L’art poétique, Chant 1. Principe n° 1 : en présence de XIST, l’inactivation
se déclenchera XIST, l’ARN aux multiples facettes, chef d’orchestre
de l’inactivation du X XIST, l’ARN aux multiples facettes, chef d’orchestre
de l’inactivation du X de l’inactivation du X
L’inactivation du chromosome X est sous le contrôle d’une région du
chromosome X, le centre d’inactivation (X inactivation center, XIC),
dont l’une des particularités est d’inclure de nombreux gènes ne
codant pas des protéines (Figure 1A). Parmi ceux-ci, XIST (X-inactive
specific transcript) est l’acteur principal de ce processus ; sa trans-
cription conduit à la synthèse d’un long ARN non-codant (lARNnc)
qui s’accumule dans le noyau autour du chromosome X dont il est
issu. XIST est présent uniquement chez les euthériens (ou mammi-
fères placentaires) et sa séquence est peu conservée d’une espèce à
l’autre, à l’exception des régions A à F, riches en répétitions en tandem
(Figure 1A), et qui jouent chacune un rôle dans la localisation de
l’ARN XIST ou dans l’extinction des gènes. Via de multiples interactions
ribonucléoprotéiques, notamment au niveau de ces régions [2-4]
(Figure 1A), XIST est à l’origine de la série de bouleversements, dont
la cinétique et les mécanismes sont progressivement élucidés, qui
mènent à la formation d’un compartiment chromosomique réprimé. Parmi les premiers changements, l’interaction entre la région A de XIST
et le co-répresseur SPEN (split-ends), un effecteur majeur de l’inac-
tivation, conduit, successivement, au recrutement de SPEN sur les
promoteurs actifs et les séquences enhancers du chromosome X ciblé,
puis à l’exclusion de la machinerie de transcription (ARN polymérase
II et facteurs généraux) et donc à l’extinction de la transcription des L’inactivation du chromosome X est sous le contrôle d’une région du
chromosome X, le centre d’inactivation (X inactivation center, XIC),
dont l’une des particularités est d’inclure de nombreux gènes ne
codant pas des protéines (Figure 1A). Parmi ceux-ci, XIST (X-inactive
specific transcript) est l’acteur principal de ce processus ; sa trans-
cription conduit à la synthèse d’un long ARN non-codant (lARNnc)
qui s’accumule dans le noyau autour du chromosome X dont il est
issu. XIST est présent uniquement chez les euthériens (ou mammi-
fères placentaires) et sa séquence est peu conservée d’une espèce à
l’autre, à l’exception des régions A à F, riches en répétitions en tandem
(Figure 1A), et qui jouent chacune un rôle dans la localisation de
l’ARN XIST ou dans l’extinction des gènes. Vignette (Photo © Inserm/Heard, Edith/Institut Curie).
1 Nicolas Boileau (1636-1711), L’art poétique, Chant 1. Principe n° 1 : en présence de XIST, l’inactivation
se déclenchera Via de multiples interactions
ribonucléoprotéiques, notamment au niveau de ces régions [2-4]
(Figure 1A), XIST est à l’origine de la série de bouleversements, dont
la cinétique et les mécanismes sont progressivement élucidés, qui
mènent à la formation d’un compartiment chromosomique réprimé. (➜) Voir le Forum de
J. L. Guenet et al., Parmi les premiers changements, l’interaction entre la région A de XIST
et le co-répresseur SPEN (split-ends), un effecteur majeur de l’inac-
tivation, conduit, successivement, au recrutement de SPEN sur les
promoteurs actifs et les séquences enhancers du chromosome X ciblé,
puis à l’exclusion de la machinerie de transcription (ARN polymérase
II et facteurs généraux) et donc à l’extinction de la transcription des m/s n° 2, vol. 37, février 2021
https://doi.org/10.1051/medsci/2020278
m/s n° 2, vol. 37, février 2021
https://doi.org/10.1051/medsci/ 152 m/s n° 2, vol. 37, février 2021
https://doi.org/10.1051/medsci/2020278
m/s n° 2, vol. 37, février 2021
https://doi.org/10.1051/medsci/ K9
K27
H3 H4
H2A
H2B
K4
H4Ac
XIC
Ch
rom
osome
X
Ppnx Linx
Xist
Ftx
Slc16a2
Rlim
Cdx4 Chic1
Tsix Jpx
Tsx
Zcchc13
Acétylation
Tri-méthylation
Xist
Ubiquitination
Méthylation
A
B
C
F
D
E
ARN XIST
Gène non codant
Gène codant
YY1
SAF-A
LBR
K9
K27
K4
1
2
3
K27
K119
4
5
6
SPEN
HDAC3
Pol II
PRC2
PRC1
DNMT3B
SMCHD1
A
F
B
C
D
E
Locus de XIST
Exon 1
Exon 7
2
3
4
5
6
8
A
B A REVUES Zcchc13 K9
K27
H3 H4
H2A
H2B
K4
H4Ac
Acétylation
Tri-méthylation
Xist
Ubiquitination
Méthylation
A
B
C
F
D
E
ARN XIST
YY1
SAF-A
LBR
K9
K27
K4
1
2
3
K27
K119
4
5
6
SPEN
HDAC3
Pol II
PRC2
PRC1
DNMT3B
SMCHD1
B B Figure 1. XIST et la mise en place de l’inactivation d’un des deux chromosomes X chez la souris. A. Panneau du haut : zoom sur le centre d’inactiva-
tion du chromosome X de la souris, où sont représentés les gènes codants (traits pleins) et les gènes non codants (pointillés) dont XIST (en vert). Panel du milieu : zoom sur la structure du gène de XIST chez la souris, avec les exons (de 1 à 8) représentés par les rectangles gris et les introns
par une ligne noire. Les éléments répétés annotés de A à F sont représentés en couleurs. Panneau du bas : représentation schématique de l’ARN
XIST avec ses partenaires protéiques. B. Principe n° 1 : en présence de XIST, l’inactivation
se déclenchera Par ailleurs, SPEN, formant un com-
plexe avec HDAC3 (histone déacétylase 3), entraîne la déacétylation
des histones 3 (H3) et 4 (H4), puis la perte d’autres marques de
chromatine active, telles que la di- ou triméthylation de la lysine 4
de l’histone H3 (H3K4me2/3) [6, 7]. Le recrutement successif de deux
autres complexes chromatiniens répresseurs de type Polycomb, PRC1
(polycomb repressive complex) par les régions B et C de XIST [8, 9] puis
PRC2, conduit respectivement à l’apposition des marques répressives
H2AK119ub (monoubiquitination de l’histone H2A sur la lysine 119) et
H3K27me3 (triméthylation de la lysine 27 de l’histone H3) [7, 10]. Plus
tardivement, les îlots CpG des promoteurs des gènes du X inactif (Xi)
sont méthylés par l’ADN méthyltransférase DNMT3B (DNA méthyltrans-
ferase 3 B) [11] (Figure 1B). f
g
L’inactivation du X s’accompagne également d’une réorganisation globale
de l’architecture du chromosome dans l’espace nucléaire qui fait intervenir
la protéine structurale SMCHD1 (structural maintenance of chromosomes
flexible hinge domain containing 1) [12], aussi impliquée dans la méthy-
lation de certaines régions du Xi [11]. De plus, l’interaction de la protéine
LBR (lamina B receptor) avec l’ARN XIST pourrait permettre d’ancrer le
chromosome Xi à la périphérie du noyau et ainsi contribuer à sa répression
[3, 13], mais cela est encore débattu [14]. Par ailleurs, la localisation
de l’ARN XIST sur le chromosome impliquerait à la fois des interactions
entre les régions C et F de XIST, d’une part, et E, d’autre part, avec, res-
pectivement, le facteur de transcription YY1 [15] et la protéine de matrice
nucléaire SAF-A (scaffold attachment factor A) [16] (Figure 1B). Par les remaniements qu’il engendre, XIST est l’initiateur principal
de l’ICX lors de l’embryogénèse précoce. Des éléments régulateurs,
en particulier au sein du XIC, assurent une expression correcte de
XIST, c’est-à-dire chez les femelles exclusivement, à partir d’un seul
des deux chromosomes X et à une période précise du développement
embryonnaire. Cette régulation fine de l’expression et de la localisa-
tion de XIST est fondamentale puisque l’absence d’expression de XIST
au cours du développement entraîne la létalité des embryons femelles
[17], et dans la mesure où XIST est potentiellement capable d’inac-
tiver, avec plus ou moins d’efficacité, n’importe quel chromosome
dans lequel il a été introduit artificiellement. Principe n° 1 : en présence de XIST, l’inactivation
se déclenchera Représentation schématique, au niveau d’un nucléosome, de la cinétique d’inactivation du chromosome
X. Avant l’inactivation, les marques activatrices H3K27ac, H3K9ac, H3K4me3, H4ac ainsi que l’ARN polymérase II (Pol II) sont présentes sur le
chromosome X. L’expression de XIST et son interaction avec SPEN conduit à (1) l’expulsion de la Pol II, (2) au recrutement de HDAC3 et à la déacé-
tylation des histones 3 et 4, (3) la perte de la marque H3K4me3. (4) L’interaction de XIST avec PRC1 permet l’apposition de la marque répressive
H2AK119ub, (5) puis de la marque H3K27me3 par PRC2 ; (6) enfin, DNMT3B catalyse la méthylation de l’ADN. La localisation de XIST au niveau du
X inactif se fait par le biais des protéines SAF-A et YY1. Enfin, les protéines SMCHD1 et LBR entraînent, respectivement, une réorganisation topolo-
gique du X inactif et sa localisation en périphérie du noyau cellulaire. m/s n° 2, vol. 37, février 2021 153 m/s n° 2, vol. 37, février 2021 fécondation, dès l’activation du génome zygotique, cette
expression est détectable chez les embryons mâles [20] et
l’ARN XIST s’accumule au niveau des deux chromosomes X
chez les femelles [21]. Plus surprenant encore, ces chro-
mosomes couverts par XIST restent actifs [21-23]. Ainsi,
durant le développement pré-implantatoire humain, l’ac-
cumulation de XIST n’entraîne ni l’apposition de marques
répressives (H3K27me3), ni l’inactivation massive des
chromosomes (Figure 2). L’hypothèse d’un mécanisme
alternatif et transitoire de compensation de dose, faisant
intervenir une atténuation globale du niveau d’expression
des gènes des deux chromosomes X, a été proposée [22]. Ce modèle controversé s’oppose à celui d’une initiation
partielle de l’inactivation de certains gènes entre le jour
5 et le jour 6 du développement embryonnaire [23, 24]. La
limitation de l’accès aux stades péri- et post-implanta-
toires empêche, pour l’instant, de répondre à l’ensemble
des questions concernant la mise en place de l’inactivation
du X chez l’homme. Le développement de procédures pour
cultiver des embryons jusqu’au jour 14 du développement
devrait permettre de lever ces interrogations, au moins
partiellement. Les premières analyses suggèrent une mise
en place tardive et graduelle [25], confirmant ainsi l’im-
portant décalage entre la présence de XIST et la répression
du chromosome X lors de l’embryogenèse humaine. gènes du chromosome [3, 5, 6]. 1 Primate supérieur dépourvu de queue. m/s n° 2, vol. 37, février 2021 Principe n° 1 : en présence de XIST, l’inactivation
se déclenchera L’expression de XACT, qui débute de manière concomitante à celle de
XIST après la fécondation, est régulée par les facteurs de pluripotence
(SOX2 [sex-determining region Y-box2], NANOG [Nanog homebox], OCT4
[octamer-binding protein]) [28]. Propriété la plus marquante, l’ARN
XACT s’accumule simultanément à XIST au niveau des deux chromosomes
X actifs (Xa) durant les premiers stades du développement [26, 27]. Les
molécules XIST et XACT occupent cependant des territoires bien distincts
qui ne se chevauchent pas [23] (Figure 2). Dans une première étude
fonctionnelle de XACT utilisant des cellules souches embryonnaires de
souris qui modélisent la mise en place de l’inactivation lors de la diffé-
renciation, le locus XACT a été artificiellement intégré sur l’un des deux
chromosomes X murins. Alors que l’inactivation du X est normalement
aléatoire (touchant l’un ou l’autre des chromosomes X avec une égale
probabilité), dans ces cellules transgéniques, la présence de XACT tend à
empêcher l’accumulation de XIST sur le chromosome modifié et donc son
inactivation [23]. Cette expérience renforce l’hypothèse selon laquelle
XACT pourrait avoir un rôle antagoniste à XIST en affectant son accu-
mulation et/ou sa fonction. Cette hypothèse nécessite désormais d’être
testée dans un contexte plus physiologique. rine et post-natale. C’est donc un état extrêmement
stable qui, en théorie, est maintenu indépendamment
du signal de départ, donc de la présence de l’ARN XIST
[19, 29]. Cependant, des données récentes conduisent
à remettre en question ce principe et à s’interroger
sur le rôle de XIST dans le maintien de l’inactivation,
notamment dans le système hématopoïétique et dans
certains cancers. Par ailleurs, environ 20 % des gènes du X (5 % chez
la souris) échappent à l’inactivation, soit de façon
constitutive, dans l’ensemble des tissus, soit, de
façon spécifique, dans certains types cellulaires [30]. L’impact de ce processus d’échappement n’est pas
encore établi, mais il pourrait contribuer au dimor-
phisme sexuel [30]. Le système immunitaire : l’exception qui confirme
la règle ? Le lignage hématopoïétique est un système bien
caractérisé dans lequel des cellules souches adultes
assurent, par différentes voies de différenciation, la
genèse de l’ensemble des cellules sanguines. Il joue
un rôle essentiel, aussi bien dans le transport de
l’oxygène que dans l’immunité et dans la distinction
entre le soi et le non-soi. Mais une observation est
intrigante : les femmes ont une réponse immunitaire Principe n° 1 : en présence de XIST, l’inactivation
se déclenchera 37, février 2021 154 2 cellules
4 cellules
8 cellules
Morula
Blastocyste
Fœtus
Xa Xa
Xi Xa
EGA
Xa Xa
Développement
Expression de XIST
Inactivation de l’X
Expression de XACT
H3K27me3
XACT
XIST (forme décondensée)
XIST (forme condensée)
Implantation 2 cellules
4 cellules
8 cellules
Morula
Blastocyste
Fœtus
Xa Xa
Xi Xa
EGA
Xa Xa
Développement
Expression de XIST
Inactivation de l’X
Expression de XACT
H3K27me3
XACT
XIST (forme décondensée)
XIST (forme condensée)
Implantation Xa Xa Xa Xa Xa Xa REVUES SYNTHÈSE Figure 2. Inactivation du chromosome X au cours du développement humain. Aux stades 4-8 cellules, XIST (en vert) et XACT (en rouge) sont expri-
més simultanément et recouvrent les deux chromosomes X sur des territoires bien distincts ; cependant, les deux chromosomes X sont toujours
actifs (Xa). L’inactivation du chromosome X se met en place plus tardivement lors du développement embryonnaire (en orange), concomitamment
à l’enrichissement de la marque épigénétique répressive H3K27me3 (en bleu). Notons que les cinétiques exactes qui aboutissent, dans le fœtus, à
l’accumulation de XIST et à l’inactivation aléatoire d’un des deux chromosome X restent à préciser. L’expression de XACT, qui débute de manière concomitante à celle de
XIST après la fécondation, est régulée par les facteurs de pluripotence
(SOX2 [sex-determining region Y-box2], NANOG [Nanog homebox], OCT4
[octamer-binding protein]) [28]. Propriété la plus marquante, l’ARN
XACT s’accumule simultanément à XIST au niveau des deux chromosomes
X actifs (Xa) durant les premiers stades du développement [26, 27]. Les
molécules XIST et XACT occupent cependant des territoires bien distincts
qui ne se chevauchent pas [23] (Figure 2). Dans une première étude
fonctionnelle de XACT utilisant des cellules souches embryonnaires de
souris qui modélisent la mise en place de l’inactivation lors de la diffé-
renciation, le locus XACT a été artificiellement intégré sur l’un des deux
chromosomes X murins. Alors que l’inactivation du X est normalement
aléatoire (touchant l’un ou l’autre des chromosomes X avec une égale
probabilité), dans ces cellules transgéniques, la présence de XACT tend à
empêcher l’accumulation de XIST sur le chromosome modifié et donc son
inactivation [23]. Cette expérience renforce l’hypothèse selon laquelle
XACT pourrait avoir un rôle antagoniste à XIST en affectant son accu-
mulation et/ou sa fonction. Cette hypothèse nécessite désormais d’être
testée dans un contexte plus physiologique. Principe n° 1 : en présence de XIST, l’inactivation
se déclenchera Cette « propriété » a été
élégamment exploitée pour rendre non-fonctionnel un des trois chro-
mosomes 21 dans des cellules issues de patients atteints de trisomie,
atténuant ainsi, in vitro, certains phénotypes moléculaires associés au
syndrome de Down [18]. Cependant, le contexte est important, l’inac-
tivation ne pouvant se mettre en place que dans certaines situations
développementales ou de différenciation cellulaire [19]. Qui pour contrecarrer l’action de XIST ? Qui pour contrecarrer l action de XIST ? Plusieurs hypothèses peuvent être proposées pour
expliquer l’absence d’inactivation des chromosomes X
en présence de XIST durant le développement précoce
humain. L’une implique les effecteurs de XIST, dont on
connait peu de choses chez l’homme, mais qui pour-
raient être absents, non fonctionnels ou incapables
d’interagir avec XIST. Selon une hypothèse alternative,
un mécanisme antagoniste empêcherait XIST d’inac-
tiver les chromosomes X. En accord avec ce modèle,
le nuage que forme XIST à ces stades paraît plus
dispersé que dans les cellules somatiques présentant
un chromosome Xi [23] (Figure 2). Ceci suggère que
l’ARN XIST ne peut, à ce stade, interagir avec le chro-
mosome X et/ou que celui-ci est réfractaire à l’action
de XIST. XACT (X active coating transcript), un lARNnc
également codé par le chromosome X, pourrait entrer
en compétition avec XIST lors des premiers stades du
développement. D’évolution récente puisque présent
uniquement chez les hominoïdes1, XACT est exprimé
principalement dans les embryons préimplanta-
toires et dans les cellules souches
embryonnaires humaines (CSEh)
qui en dérivent [23, 26, 27] (➜). (➜) Voir la Dernière
heure de C. Vallot et
C. Rougeulle, m/s n° 2,
février 2013, page 253 Découplage entre accumulation de XIST et inactivation
chromosomique : le cas du développement précoce humain
L’analyse de la cinétique d’inactivation du X et du rôle de XIST chez la souris
montrent que l’accumulation de XIST lors du développement embryonnaire
précoce entraîne l’inactivation du chromosome concerné. Ce couplage ne
semble cependant pas systématiquement transposable aux autres espèces
de mammifères chez lesquels différentes stratégies semblent à l’œuvre,
dans l’espèce humaine notamment (Figure 2). Si, comme chez la souris,
l’expression de XIST chez l’homme est initiée très rapidement après la m/s n° 2, vol. Principe n° 2 : l’inactivation du chromosome X, chez l’adulte,
stable, elle sera Il en est
de même dans les macrophages et les cellules dendri-
tiques qui l’expriment [32-34]. TLR7 est impliqué dans la
détection d’ARN viral, et constitue l’un des facteurs clés
du lupus érythémateux, une maladie auto-immune qui
touche préférentiellement les femmes (80 % des cas). Ainsi, cette plasticité atypique de l’état inactif pourrait
avoir des conséquences physiologiques à l’origine de
dimorphismes sexuels. TLR7 (Toll-like receptor 7), par exemple, échappe à l’ICX
durant la différenciation des lymphocytes B. Il en est
de même dans les macrophages et les cellules dendri-
tiques qui l’expriment [32-34]. TLR7 est impliqué dans la
détection d’ARN viral, et constitue l’un des facteurs clés
du lupus érythémateux, une maladie auto-immune qui
touche préférentiellement les femmes (80 % des cas). Ainsi, cette plasticité atypique de l’état inactif pourrait
avoir des conséquences physiologiques à l’origine de
dimorphismes sexuels. accrue à certaines infections2 par rapport aux hommes. Elles sont
aussi plus susceptibles de développer des maladies auto-immunes,
suggérant un système immunitaire globalement plus réactif [31]. Le
chromosome X est porteur de nombreux gènes aux fonctions cruciales
dans la régulation des mécanismes immunitaires [31]. Le processus
d’inactivation du X pourrait donc participer à ce dimorphisme sexuel
immunitaire. accrue à certaines infections2 par rapport aux hommes. Elles sont
aussi plus susceptibles de développer des maladies auto-immunes,
suggérant un système immunitaire globalement plus réactif [31]. Le
chromosome X est porteur de nombreux gènes aux fonctions cruciales
dans la régulation des mécanismes immunitaires [31]. Le processus
d’inactivation du X pourrait donc participer à ce dimorphisme sexuel
immunitaire. Le chromosome Xi présente en effet des caractéristiques particulières
dans certaines cellules du lignage hématopoïétique qui suggèrent une
dynamique de régulation de l’état inactif spécifique de ce lignage adulte
(Figure 3) [32]. Lors de la différenciation des lymphocytes B et T, l’ARN
XIST se délocalise progressivement et transitoirement du Xi, tout en res-
tant présent dans le noyau. Le paysage chromatinien du Xi est également
modifié : la perte de XIST est associée à la disparition des marques chro-
matiniennes répressives H3K27me3 et H2AK119ub. Cependant, lorsque
les lymphocytes sont activés, XIST recouvre à nouveau le Xi (Figure 3)
[32]. Ces changements n’entraînent pas de réactivation globale du
Xi, mais certains gènes, notamment des gènes jouant un rôle dans
l’immunité, sont ré-exprimés [32]. Le gène codant le récepteur Toll-like Inactivation du X et cancers
Des anomalies du Xi ont été mises en évidence dans
certains contextes cancéreux. 2 Telles que par : les virus MERS-CoV, de l’Hépatite B, Ebola ou les bactéries Mycobacterium tuberculosis,
Legionella pneumophila ou Lesptospira spp. Principe n° 2 : l’inactivation du chromosome X, chez l’adulte,
stable, elle sera Stabilité et maintien de l’état inactif L’état inactif du chromosome X, qui s’installe avant que l’embryon
ne soit totalement formé, persiste tout au long de la vie intra-uté- m/s n° 2, vol. 37, février 2021 m/s n° 2, vol. 37, février 2021 155 m/s n° 2, vol. 37, février 2021 CSH
PLC
pro-T
pro-B
pré-B
B imm
B naïf
LB circulant
LB activé
pré-T
CD4+/CD8+
LT4
LT8
LT activé
Expression de XIST
Accumulation de XIST au Xi
H3K27me3
H2AK119Ub
Nuage de XIST
Figure 3. Dynamique de l’inactivation du chromosome X dans le lignage lymphoïde. Au cours de la différenciation des progéniteurs lymphoïdes
communs (PLC) en lymphocytes B et T, le nuage que forme XIST au Xi (en vert clair), ainsi que les marques répressives H3K27me3 (en bleu) et
H2AK119ub (en violet) associées sont progressivement perdus, cependant l’expression de XIST reste inchangée (vert foncé). Lorsque les lympho-
cytes B et T sont activés, le nuage de XIST et les marques épigénétiques répressives sont à nouveau localisés au niveau du Xi. CSH : Cellule souche
hématopoïétique ; pro-B/T : lymphocyte pro-B/T ; pré-B/T : lymphocyte pré-B/T ; B imm : lymphocyte B immature ; B naïf : lymphocyte B naïf ; LB
circulant : lymphocyte B circulant ; CD4+/CD8+ : précurseur lymphocytaire double positif ; LT4/8 : lymphocytes T CD4+/CD8+. CSH
PLC
pro-T
pro-B
pré-B
B imm
B naïf
LB circulant
LB activé
pré-T
CD4+/CD8+
LT4
LT8
LT activé
Nuage de XIST Figure 3. Dynamique de l’inactivation du chromosome X dans le lignage lymphoïde. Au cours de la différenciation des progéniteurs lymphoïdes
communs (PLC) en lymphocytes B et T, le nuage que forme XIST au Xi (en vert clair), ainsi que les marques répressives H3K27me3 (en bleu) et
H2AK119ub (en violet) associées sont progressivement perdus, cependant l’expression de XIST reste inchangée (vert foncé). Lorsque les lympho-
cytes B et T sont activés, le nuage de XIST et les marques épigénétiques répressives sont à nouveau localisés au niveau du Xi. CSH : Cellule souche
hématopoïétique ; pro-B/T : lymphocyte pro-B/T ; pré-B/T : lymphocyte pré-B/T ; B imm : lymphocyte B immature ; B naïf : lymphocyte B naïf ; LB
circulant : lymphocyte B circulant ; CD4+/CD8+ : précurseur lymphocytaire double positif ; LT4/8 : lymphocytes T CD4+/CD8+. TLR7 (Toll-like receptor 7), par exemple, échappe à l’ICX
durant la différenciation des lymphocytes B. Conclusion 6. Dossin F, Pinheiro I, Z.ylicz JJ, et al. SPEN integrates transcriptional and
epigenetic control of X-inactivation. Nature 2020 ; 578 : 455-60. 6. Dossin F, Pinheiro I, Z.ylicz JJ, et al. SPEN integrates transcriptional and
epigenetic control of X-inactivation. Nature 2020 ; 578 : 455-60. Des avancées majeures ont récemment révélé l’extraordinaire plasti-
cité du processus d’inactivation du X et les particularités d’espèce et
de contexte qui y sont associées. Les divers mécanismes d’action de
XIST sont encore à explorer en détail, tant dans la mise en place du
processus d’inactivation du X que dans son maintien, et plus générale-
ment dans l’homéostasie cellulaire. 7. Z.ylicz JJ, Bousard A, Žumer K, et al. The implication of early chromatin
changes in X chromosome inactivation. Cell 2019 ; 176 : 182-97.e23. 7. Z.ylicz JJ, Bousard A, Žumer K, et al. The implication of early chromatin
changes in X chromosome inactivation. Cell 2019 ; 176 : 182-97.e23. 8. Bousard A, Raposo AC, Z.ylicz JJ, et al. The role of Xist-mediated Polycomb
recruitment in the initiation of X-chromosome inactivation. EMBO Rep
2019 ; 20 : e48019. 8. Bousard A, Raposo AC, Z.ylicz JJ, et al. The role of Xist-mediated Polycomb
recruitment in the initiation of X-chromosome inactivation. EMBO Rep
2019 ; 20 : e48019. 9. Pintacuda G, Wei G, Roustan C, et al. hnRNPK recruits PCGF3/5-PRC1 to the
Xist RNA B-repeat to establish polycomb-mediated chromosomal silencing. Mol Cell 2017 ; 68 : 955-969.e10. La découverte que, loin d’être figé, l’état du X inactif reste modulable,
ouvre des perspectives d’étude du dimorphisme sexuel et des maladies
qui y sont associées. Si la stabilité de l’inactivation a jusqu’alors été
principalement étudiée dans le système immunitaire, il est intéressant
de noter que le chromosome X porte un nombre important de gènes
jouant un rôle dans des fonctions cérébrales et que des gènes échap-
pant à l’inactivation sont impliqués dans des maladies, telles que
l’autisme et l’épilepsie [39]. L’approfondissement de notre connais-
sance de la régulation et de la stabilité de l’inactivation du X chez
l’adulte pourrait donc révéler de nouveaux mécanismes à l’origine
de ces diverses maladies (cancer, autisme, épilepsie et maladies du
système immunitaire) et, idéalement, nous permettre d’identifier des
biomarqueurs et de nouvelles cibles thérapeutiques spécifiques. ‡ 10. Almeida M, Pintacuda G, Masui O, et al. PCGF3/5-PRC1 initiates Polycomb
recruitment in X chromosome inactivation. Science 2017 ; 356 : 1081-4. 11. RÉFÉRENCES 1. Guénet JL, Panthier JJ, Avner P, et al. L’héritage de Mary F. Lyon (1925-2014). Med Sci (Paris) 2015 ; 31 : 687-9. 1. Guénet JL, Panthier JJ, Avner P, et al. L’héritage de Mary F. Lyon (1925-2014). Med Sci (Paris) 2015 ; 31 : 687-9. REVUES 2. Chu C, Zhang QC, da Rocha ST, et al. Systematic discovery of Xist RNA
binding proteins. Cell 2015 ; 161 : 404-16. 2. Chu C, Zhang QC, da Rocha ST, et al. Systematic discovery of Xist RNA
binding proteins. Cell 2015 ; 161 : 404-16. 3. McHugh CA, Chen CK, Chow A, et al. The Xist lncRNA interacts directly with
SHARP to silence transcription through HDAC3. Nature 2015 ; 521 : 232-6. 4. Minajigi A, Froberg J, Wei C, et al. Chromosomes. A comprehensive Xist
interactome reveals cohesin repulsion and an RNA-directed chromosome
conformation. Science 2015 ; 349 : 10.1126/science. aab2276 aab2276. 4. Minajigi A, Froberg J, Wei C, et al. Chromosomes. A comprehensive Xist
interactome reveals cohesin repulsion and an RNA-directed chromosome
conformation. Science 2015 ; 349 : 10.1126/science. aab2276 aab2276. 5. Monfort A, Di Minin G, Postlmayr A, et al. Identification of spen as a crucial
factor for Xist function through forward genetic screening in haploid
embryonic stem cells. Cell Rep 2015 ; 12 : 554-61. 5. Monfort A, Di Minin G, Postlmayr A, et al. Identification of spen as a crucial
factor for Xist function through forward genetic screening in haploid
embryonic stem cells. Cell Rep 2015 ; 12 : 554-61. SUMMARY 17. Marahrens Y, Panning B, Dausman J, et al. Xist-deficient mice are defective in
dosage compensation but not spermatogenesis. Genes Dev 1997 ; 11 : 156-66. LIENS D’INTÉRÊT LIENS D’INTÉRÊT surexpression pourrait mener à une croissance cellulaire anormale. Cependant, il n’existe actuellement aucune preuve d’un lien causal
entre anomalie d’inactivation du X et tumorigenèse, notamment dans
l’espèce humaine. Chez la souris, la délétion de XIST dans les cellules
souches hématopoïétiques entraîne, chez les femelles uniquement, le
développement d’hémopathies malignes très agressives et associées à
une surexpression de certains gènes du chromosome X [38]. Dans ce
contexte, XIST a donc un rôle de suppresseur de tumeur. Cependant,
le statut d’inactivation du X n’ayant pas été étudié en détail dans
ces souris mutantes, la question de l’action de XIST sur le maintien de
l’état inactif du chromosome X reste ouverte. Les auteures déclarent n’avoir aucun lien d’intérêt concernant les don-
nées publiées dans cet article. Latest insights on X-chromosome inactivation: When general
principles should be revisited 18. Czermin΄ski JT, Lawrence JB. Silencing trisomy 21 with XIST in neural stem cells
promotes neuronal differentiation. Dev Cell 2020 ; 52 : 294-308.e3. 19. Wutz A, Jaenisch R. A shift from reversible to irreversible X inactivation is
triggered during ES cell differentiation. Mol Cell 2000 ; 5 : 695-705. The inactivation of one of the two X chromosomes of female mam-
mals is a vital process and a paradigm for epigenetic regulations. X-inactivation is triggered, early during embryo development, by the
accumulation of a peculiar noncoding RNA, XIST, which interacts with
a plethora of molecular complexes and ultimately protects the coa-
ted chromosome from the expression machinery. Once installed, the
inactive state is locked by multiple layers of chromatin modifications,
ensuring its stable perpetuation across cell divisions. However, recent
discoveries made in various model organisms urge us to revisit some of
the general principles of the X-inactivation process. ‡ 20. Briggs SF, Dominguez AA, Chavez SL, et al. Single-cell XIST expression in
human preimplantation embryos and newly reprogrammed female induced
pluripotent stem cells: XIST in human preimplantation embryos and iPSCs. Stem Cells 2015 ; 33 : 1771-81. 21. Okamoto I, Patrat C, Thépot D, et al. Eutherian mammals use diverse
strategies to initiate X-chromosome inactivation during development. Nature 2011 ; 472 : 370-4. 22. Petropoulos S, Edsgärd D, Reinius B, et al. Single-cell RNA-Seq reveals
lineage and X chromosome dynamics in human preimplantation embryos. Cell 2016 ; 165 : 1012-26. 23. Vallot C, Patrat C, Collier AJ, et al. XACT noncoding RNA competes with XIST
in the control of X chromosome activity during human early development. Cell Stem Cell 2017 ; 20 : 102-11. 24. Moreira de Mello JC, Fernandes GR, Vibranovski MD, et al. Early X
chromosome inactivation during human preimplantation development
revealed by single-cell RNA-sequencing. Sci Rep 2017 ; 7 : 1-12. 24. Moreira de Mello JC, Fernandes GR, Vibranovski MD, et al. Early X
chromosome inactivation during human preimplantation development
revealed by single-cell RNA-sequencing. Sci Rep 2017 ; 7 : 1-12. Conclusion Gendrel AV, Apedaile A, Coker H, et al. Smchd1-dependent and independent
pathways determine developmental dynamics of Cpg island methylation on
the inactive X chromosome. Dev Cell 2012 ; 23 : 265-79. 12. Gdula MR, Nesterova TB, Pintacuda G, et al. The non-canonical SMC protein
SmcHD1 antagonises TAD formation and compartmentalisation on the
inactive X chromosome. Nat Commun 2019 ; 10 : 30. 13. Chen CK, Blanco M, Jackson C, et al. Xist recruits the X chromosome to the
nuclear lamina to enable chromosome-wide silencing. Science 2016 ; 354 :
468-72. 14. Nesterova TB, Wei G, Coker H, et al. Systematic allelic analysis defines the
interplay of key pathways in X chromosome inactivation. Nat Commun
2019 ; 10 : 1-15. 15. Jeon Y, Lee JT. YY1 Tethers Xist RNA to the inactive X nucleation center. Cell
2011 ; 146 : 119-33. 16. Yamada N, Hasegawa Y, Yue M, et al. Xist exon 7 contributes to the stable
localization of Xist RNA on the inactive X-chromosome. PLoS Genet 2015 ;
11 : e1005430. Principe n° 2 : l’inactivation du chromosome X, chez l’adulte,
stable, elle sera Par exemple, la perte
de l’expression de XIST et de marques d’hétérochro-
matine associées au Xi est observée dans le cancer du
sein [35]. Une perte totale du Xi, souvent accompa-
gnée d’une duplication du Xa, est également décrite
dans de nombreux types de cancers (cancer du sein,
des ovaires) [36, 37]. De ces processus résulte une
expression accrue de gènes liés à l’X [35], en particulier
des facteurs de croissance ou des oncogènes, dont la m/s n° 2, vol. 37, février 2021 156 REMERCIEMENTS Nous remercions Pauline Andreu, Project Manager du Labex Who Am I?, ainsi que Céline
Morey pour leurs relectures avisées du manuscrit. Nous remercions également la Ligue
contre le Cancer qui soutient ce travail. 25. Zhou F, Wang R, Yuan P, et al. Reconstituting the transcriptome and DNA
methylome landscapes of human implantation. Nature 2019 ; 572 : 660-4. RÉFÉRENCES The X chromosome in immune functions: when a chromosome makes
the difference. Nat Rev Immunol 2010 ; 10 : 594-604. 39. Fang H, Disteche CM, Berletch JB. X Inactivation and escape: epigenetic and
structural features. Front Cell Dev Biol 2019 ; 7 : 219. 32. Syrett CM, Anguera MC. When the balance is broken: X-linked gene dosage from two X
chromosomes and female biased a toimm nit J Le koc Biol 2019 106 919 32 32. Syrett CM, Anguera MC. When the balance is broken: X-linked gene dosage from two X
chromosomes and female-biased autoimmunity. J Leukoc Biol 2019 ; 106 : 919-32. chromosomes and female-biased autoimmunity. J Leukoc Biol 2019 ; 106 : 919-32. TIRÉS À PART
C. Rougeulle 33. Souyris M, Cenac C, Azar P, et al. TLR7 escapes X chromosome inactivation in immune cells. Sci
Immunol 2018 ; 3 : eaap8855. m/s n° 2, vol. 37, février 2021 157 RÉFÉRENCES 34. Souyris M, Mejía JE, Chaumeil J, et al. Female predisposition to TLR7-
driven autoimmunity: gene dosage and the escape from X chromosome
inactivation. Semin Immunopathol 2019 ; 41 : 153-64. 26. Vallot C, Rougeulle C. Inactivation du chromosome X chez l’humain: XACT et XIST, à chacun son
chromosome. Med Sci (Paris) 2013 ; 29 : 223-5. 35. Chaligné R, Popova T, Mendoza-Parra M-A, et al. The inactive X chromosome
is epigenetically unstable and transcriptionally labile in breast cancer. Genome Res 2015 ; 25 : 488-503. 27. Vallot C, Huret C, Lesecque Y, et al. XACT, a long noncoding transcript coating the
chromosome in human pluripotent cells. Nat Genet 2013 ; 45 : 239-41. 7. Vallot C, Huret C, Lesecque Y, et al. XACT, a long noncoding transcript coa chromosome in human pluripotent cells. Nat Genet 2013 ; 45 : 239-41. 36. Sirchia SM, Tabano S, Monti L, et al. Misbehaviour of XIST RNA in breast
cancer cells. PLoS One 2009 ; 4 : e5559. 28. Casanova M, Moscatelli M, Chauvière LÉ, et al. A primate-specific retroviral enhancer wires the
XACT lncRNA into the core pluripotency network in humans. Nat Commun 2019 ; 10 : 5652. 37. Kawakami T, Zhang C, Taniguchi T, et al. Characterization of loss-of-inactive
X in Klinefelter syndrome and female-derived cancer cells. Oncogene 2004 ;
23 : 6163-9. 29. Csankovszki G, Panning B, Bates B, et al. Conditional deletion of Xist disrupts histone macroH2A
localization but not maintenance of X inactivation. Nat Genet 1999 ; 22 : 323-4. 30. Carrel L, Brown CJ. When the Lyon(ized chromosome) roars: ongoing expression from an inactive X
chromosome. Philos Trans R Soc Lond B Biol Sci 2017 ; 372. 30. Carrel L, Brown CJ. When the Lyon(ized chromosome) roars: ongoing expression from an inactive X
chromosome. Philos Trans R Soc Lond B Biol Sci 2017 ; 372. 38. Yildirim E, Kirby JE, Brown DE, et al. Xist RNA is a potent suppressor of
hematologic cancer in mice. Cell 2013 ; 152 : 727-42. 38. Yildirim E, Kirby JE, Brown DE, et al. Xist RNA is a potent suppressor of
hematologic cancer in mice. Cell 2013 ; 152 : 727-42. 31. Libert C, Dejager L, Pinheiro I. The X chromosome in immune functions: when a chromosome makes
the difference. Nat Rev Immunol 2010 ; 10 : 594-604. 31. Libert C, Dejager L, Pinheiro I. m/s n° 2, vol. 37, février 2021 158
|
https://openalex.org/W4242893380
|
https://jneuroengrehab.biomedcentral.com/track/pdf/10.1186/s12984-021-00837-z
|
English
| null |
A usability study in patients with stroke using MERLIN, a robotic system based on serious games for upper limb rehabilitation in the home setting.
|
Research Square (Research Square)
| 2,021
|
cc-by
| 10,902
|
Guillén‑Climent et al.
J NeuroEngineering Rehabil (2021) 18:41
https://doi.org/10.1186/s12984-021-00837-z Guillén‑Climent et al. J NeuroEngineering Rehabil (2021) 18:41
https://doi.org/10.1186/s12984-021-00837-z Open Access A usability study in patients with stroke
using MERLIN, a robotic system based
on serious games for upper limb rehabilitation
in the home setting Silvia Guillén‑Climent1* , Ainara Garzo2, María Nieves Muñoz‑Alcaraz1,3,4, Pablo Casado‑Adam3,
Javier Arcas‑Ruiz‑Ruano2, Manuela Mejías‑Ruiz3 and Fernando Jesús Mayordomo‑Riera3,5 Abstract Background: Neuroscience and neurotechnology are transforming stroke rehabilitation. Robotic devices, in addition
to telerehabilitation, are increasingly being used to train the upper limbs after stroke, and their use at home allows
us to extend institutional rehabilitation by increasing and prolonging therapy. The aim of this study is to assess the
usability of the MERLIN robotic system based on serious games for upper limb rehabilitation in people with stroke in
the home environment. Methods: 9 participants with a stroke in three different stages of recovery (subacute, short-term chronic and long-
term chronic) with impaired arm/hand function, were recruited to use the MERLIN system for 3 weeks: 1 week training
at the Maimonides Biomedical Research Institute of Cordoba (IMIBIC), and 2 weeks at the patients’ homes. To evaluate
usability, the System Usability Scale (SUS), Adapted Intrinsic Motivation Inventory (IMI), Quebec User Evaluation of
Satisfaction with assistive Technology (QUEST), and the ArmAssist Usability Assessment Questionnaire were used in
the post-intervention. Clinical outcomes for upper limb motor function were assessed pre- and post-intervention. Results: 9 patients participated in and completed the study. The usability assessment reported a high level of satis‑
faction: mean SUS score 71.94 % (SD = 16.38), mean QUEST scale 3.81 (SD = 0.38), and mean Adapted IMI score 6.12
(SD = 1.36). The results of the ArmAssist Questionnaire showed an average of 6 out of 7, which indicates that MERLIN
is extremely intuitive, easy to learn and easy to use. Regarding clinical assessment, the Fugl-Meyer scores showed
moderate improvements from pre- to post-intervention in the total score of motor function (p = 0.002). There were
no significant changes in the Modified Ashworth scale outcomes (p = 0.169). Conclusions: This usability study indicates that home-based rehabilitation for upper limbs with the MERLIN system is
safe, useful, feasible and motivating. Telerehabilitation constitutes a major step forward in the use of intensive rehabili‑
tation at home. Trial registration ClinicalTrials.gov, NCT04405609. Registered 06 January 2020—Retrospectively registered, https://clini
caltrials.gov/ct2/show/NCT04405609 *Correspondence: silvia.guillen@imibic.org
1 Maimonides Biomedical Research Institute of Cordoba (IMIBIC), Reina
Sofia University Hospital, University of Cordoba, Córdoba, Spain
Full list of author information is available at the end of the article *Correspondence: silvia.guillen@imibic.org
1 Maimonides Biomedical Research Institute of Cordoba (IMIBIC), Reina
Sofia University Hospital, University of Cordoba, Córdoba, Spain
Full list of author information is available at the end of the article Participants In order to detect most problems of usability which can
affect a product, Jakob Nielsen’s theory [19] regarding the
sufficient number of users to evaluate a system is widely
accepted. According to Nielsen, between three and five
users can identify 85% of the most relevant usability
problems. In this case, due to the heterogeneity of the
study population, it was decided to recruit 12 patients
at different stages of post-stroke upper limb recovery, in
order to test as many system features as possible. fi
Continuous advances in neuroscience and neurotech-
nology are transforming stroke rehabilitation [10]. At a
time when the rehabilitation services resources are una-
ble to meet the demand, robot-assisted rehabilitation and
home-based telerehabilitation are gaining greater accept-
ance [11]. Robot-based neurorehabilitation systems pro-
vide a solution to increase the number of movements,
involve safe, intensive rehabilitation exercises [12, 13] and
have the advantage that the precise movements of the
robot are able to measure the patients’ movements objec-
tively [14, 15]. On the other hand, home-based telereha-
bilitation allows us to extend institutional rehabilitation
by increasing and prolonging the therapy [16]. What is
more, the combination of game-based telerehabilitation
and robotic systems creates a motivating, engaging envi-
ronment for patients [17]. The enjoyment patients derive
from playing these so called ‘serious games’, designed spe-
cifically for the rehabilitation tasks, can greatly increase
the quality and quantity of the therapy delivered [18]. Participants were recruited at the Reina Sofía Univer-
sity Hospital in Cordoba, Spain. The participants were
divided in three different groups, depending on their
stage of recovery: subacute (2–6 months of recovery of
stroke), short-term chronic (6–12 months) and long-term
chronic (over 12 months). Four patients were recruited
from each stage. The inclusion criteria to participate in
the study were: subjects over 18 with upper limb hemi-
paresis after stroke, unilateral paresis and cognitive abil-
ity to understand, accept and actively participate in the
usability study. Having Wi-Fi at home and a table meas-
uring 110 × 68 cm on which the MERLIN system can be
set up was considered also a requirement to participate
in the study. Patients who presented bilateral motor defi-
cit, severe spasticity, psychiatric illness, and/or cognitive
impairment were excluded. All the subjects were duly informed about the study
and all of them gave their written consent before the first
session. © The Author(s) 2021. Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which
permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the
original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or
other third party material in this article are included in the article’s Creative Commons licence, unless indicated otherwise in a credit line
to the material. If material is not included in the article’s Creative Commons licence and your intended use is not permitted by statutory
regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this
licence, visit http://creativecommons.org/licenses/by/4.0/. The Creative Commons Public Domain Dedication waiver (http://creativeco
mmons.org/publicdomain/zero/1.0/) applies to the data made available in this article, unless otherwise stated in a credit line to the data. Guillén‑Climent et al. J NeuroEngineering Rehabil (2021) 18:41 Page 2 of 16 Keywords: Stroke, Neurological rehabilitation, Upper extremity, Telerehabilitation, Serious games, Home training,
Robot Participants MERLIN is a robotic system based on serious games
for the upper limb tele rehabilitation in patients with a
stroke. It is presented as an affordable and easy to use
solution to allow the patient to carry out an intensive
rehabilitation at home, with a continuous remote moni-
toring and communication with the therapist. The sys-
tem is composed of an upper-limb rehabilitation robot
and a software platform which guides and measures the
patient’s movements and allows physicians to custom-
ize the therapeutic plan and to monitor the patients’
evolution. Background evaluate the ease to use, consistency and acceptance of
the system have been evaluated. The research carried
out also aims to demonstrate the feasibility of including
the robotic therapy as a complement to a regular daily
rehabilitation program. g
Strokes are among the leading causes of death, physical
disability and economic burden worldwide [1, 2]. The
prevalence of people living with the effects of stroke has
increased over the last few years, thus creating a higher
demand for rehabilitation services [3]. The paralysis of
the upper limbs is a common impairment after strokes,
and only 10–20% of patients recover completely [4, 5]:
for these patients, the main aim of arm rehabilitation is
to recover lost functions [6]. Nowadays, the key aspects
to make rehabilitation effective for people with stroke are
considered to be intensity, repetition and using suitably
challenging and function-oriented activities [7–9]. How-
ever, the increase in the number of people affected and
the current limitation of health resources make it very
difficult to provide services using a traditional approach. MERLIN unactuated robotic telerehabilitation systemh The MERLIN system has been developed to bring neu-
rorehabilitation to the post-stroke patients’ homes with
the aim of providing daily, intensive, motivating and
patient-tailored rehabilitation, with the indirect super-
vision of the therapist [20]. The system is composed of
ArmAssist (AA), a cost-effective robotic system based on
serious games developed by TECNALIA, and the Antari
Home Care platform [21] to supervise, organize and
customize the patients’ daily training remotely, which
has been developed by GMV [22]. The AA system is a
modular solution which includes an affordable, portable
robotic device for a complete upper limb rehabilitation,
and a software platform based on serious games to moti-
vate the patients and assess their training [18]. In the MERLIN system, the non-actuated version of
the AA robotic device has been used to ensure a safe
use in the home environment when continuous supervi-
sion is not feasible. The AA device includes several sen-
sors to measure the patient’s active self-directed active
movements during the games, which are performed
on a normal table to control the games (see Fig. 1). The
device can be easily fastened on the forearm, and allows
natural movements with low resistance. The posi-
tion and orientation of the robot are calculated using
the information from the camera, which reads the QR
codes on the mat below, and the encoders included on
the wheels. Wrist angle, hand grasping force and verti-
cal arm force are calculated by a potentiometer, and two With the MERLIN system, the patients can access the
daily therapy previously organized by the therapist, as
well as viewing a summary of the results obtained during
therapy (see Fig. 2). It also features a messaging tool to
communicate with the therapist, similar to mailing.h The AA system has been previously tested in a clinical
setting by therapists and patients with positive results of
acceptance [11] and effectiveness, with improvements in
the patients’ motor function after use [24]. Previous stud-
ies have also demonstrated that the therapy using serious
games and the AA system is enjoyable and motivating
because the patients feel more engaged [27]. In the pre-
sent study, the system was adapted for home use using
the non-actuated version of the robot for greater safety. Study design
Thi i t
t The games have been
adapted to for the target group taking into account any
possible cognitive or visual problems [26]. Arm and hand functions were evaluated at baseline (on
day 1 before starting the training), and on the last day. The usability of the system and the participants’ motiva-
tion was evaluated on the last day using different vali-
dated scales, as explained below. Study design
Thi i t
t This interventional study is an open label trial with a sin-
gle group and a longitudinal design. Each patient used
the MERLIN system for 3 weeks: 1 week training at the
IMIBIC (Maimonides Biomedical Research Institute in
Córdoba, Spain) with the supervision of a physiothera-
pist, 1 week at the patient’s home with similar supervi-
sion and 1 week at patient’s home on their own with
remote support and supervision of a physiotherapist to
organize the rehabilitation sessions. The purpose of this manuscript is to present the usa-
bility validation of MERLIN system. In this study, we Page 3 of 16 Page 3 of 16 Page 3 of 16 Guillén‑Climent et al. J NeuroEngineering Rehabil (2021) 18:41 Guillén‑Climent et al. J NeuroEngineering Rehabil (2021) 18:41 Guillén‑Climent et al. J NeuroEngineering Rehabil (2021) 18:41 Force Sensing Resistors (FSR) and a load cell are included
on the hand module and arm support, respectively. The
key movements that can be measured are: three types of
movements over the table, horizontal shoulder abduc-
tion-adduction, flexion–extension in the elbow (verti-
cal force), wrist prono-supination movements and hand
opening and closing [23]. This version of the system is
aimed at patients who can actively carry out the move-
ments and is thus more appropriate for patients who
have mild or moderate motor impairment according to
the Fugl-Meyer scale, who are more suited to continu-
ing the therapy at home. The movements are used to
interact with the implemented serious games on the
software, which are divided into different levels depend-
ing on the patients’ stage of recovery and cognitive capa-
bilities. The games include assessment and training [24]
and they were co-designed by patients and physiothera-
pists [25]. 7 training games are available, such as choos-
ing letters to make a word, discovering pairs, solitaire or
doing a puzzle, for example. Additionally, the option of
using some online games is also available. This option is
recommended for patients with good movement control
and cognitive capabilities. The games can be configured
for only some movements or a combination of different
ones. The exercises involve extending the user’s range
of exercises beyond their normal threshold, which has
been previously set by the assessment games, and can be
modified when necessary, i.e. when motor improvement
is detected by the physiotherapist. MERLIN unactuated robotic telerehabilitation systemh With this aim, the software was programmed to work on
a tablet to make the system more compact and adaptable
to the home-setting and package was designed to trans-
port the system, and the mat was adapted to make the
system easy to transport and store (see Fig. 3). Fig. 1 AA system As previously explained, the Antari HomeCare plat-
form has also been integrated into the MERLIN system. This telecare platform designed for managing patients’ Fig. 1 AA system Fig. 1 AA system Fig. 1 Guillén‑Climent et al. J NeuroEngineering Rehabil (2021) 18:41 Page 4 of 16 Fig. 2 Adaptations made to the system for home use. a Adaptation of mat. b Package Fig. 2 Adaptations made to the system for home use. a Adaptation of mat. b Package Fig. 2 Special emphasis was placed on the correct positioning
of the arm and shoulder for rehabilitation. In addition,
each participant received a copy of the user’s manual,
which also included a telephone number to contact in
case of any technical or clinical problems. The thera-
pist installed the system at the participant’s home at the
beginning of the second week, adjusted the chair height
and explained to the patient the correct back and shoul-
der positions for doing the training as well as the most
comfortable arm position for rehabilitation when super-
vision was not available (see Fig. 4b). 3 sessions of 30 min
were held on alternate days. During this week, the patient
carried out the therapy assigned for each day with the
supervision of the therapist. The third week followed the
same timetable, except that the participants were trained
to use the system at home completely autonomously. The
therapist used the tele-care platform daily to follow up
remotely the participants’ progress and their use of the
system, as well as for organizing the following sessions. treatments and online follow-up, was adapted for remote
customizing of rehabilitation therapies. The therapist
used the online platform to plan each patient’s therapy
and selected the games to be used, the movements to
train, the number of days to be repeated and the length
of each game. Monitoring the patient’s evolution and
therapy (duration of training, frequency, points obtained,
etc.) was performed online using this software (see
Fig. 2). The messaging tool was accessible via the Antari
HomeCare system. MERLIN unactuated robotic telerehabilitation systemh The aim of this usability study was to evaluate the users’
acceptance of the new features and remote monitoring
carried out by the therapist, instead of the face-to-face
monitoring usually performed in previous evaluations
of AA [11, 24, 27]. The system safety when patients work
independently at home was also evaluated, and this feed-
back on usability, acceptance, motivation and safety is an
important input for demonstrating how easy the system
is to learn and use. The training movements and games used for this pur-
pose were selected by the therapist, who decided on the
intensity level and movements to train in the therapy
according to the patient’s evolution or cognitive condi-
tions. The therapist organized the rehabilitation ses-
sions beforehand, using the tele-care platform designed
for this purpose. Prior to commencing therapy, patients
were requested to perform a calibration process to set
up the threshold according to their range of motion. This Intervention sessions Rehabilitation therapy included 11 sessions using the
MERLIN system performed over a period of 3 weeks. The
first week was used as training to teach users and caregiv-
ers how to use the system correctly, as well as for getting
used to the rehabilitation system, robot movements and
protocol times. The training sessions were organized
every day for 1 h at the IMIBIC facilities (see Fig. 4a). Guillén‑Climent et al. J NeuroEngineering Rehabil (2021) 18:41 Page 5 of 16 g
Fig. 3 MERLIN system. a Results of the therapy shown on the therapist’s panel. b Evolution of the therapy shown on the patient’s panel Fig. 3 MERLIN system. a Results of the therapy shown on the therapist’s panel. b Evolution of the therapy shown on the patient’s panel Fig 3 MERLIN system a Results of the tem. a Results of the therapy shown on the therapist’s panel. b Evolution of the therapy shown on the patient’s panel Guillén‑Climent et al. J NeuroEngineering Rehabil (2021) 18:41 Page 6 of 16 Fig. 4 Usability of study pictures. a Session at IMIBIC facilities. b Session in a home environment Fig. 4 Usability of study pictures. a Session at IMIBIC facilities. b Session in a home environment Fig. 4 allowed the participants to set the level of challenge in
the exercises at their maximum capacity. After that, while
using the system, the range of motion for each game and
patient was controlled by the system itself. (IMI adapted) [36], the Quebec User Evaluation of Satis-
faction with assistive Technology (QUEST) [37], and the
AA Usability Assessment Questionnaire [18].h The SUS scale was used to evaluate the usability of the
system. The scores ranged from 0 to 100%, where a higher
score means better usability, with a threshold of 68%. Assessmenth The IMI scale is a multidimensional measurement
method designed to assess participants’ subjective expe-
rience related to a given activity. The full version of the
questionnaire includes 45 items and 7 subscales. Each
item is used to rate the statement on a scale ranging
from 1 ‘strongly agree’ to 7 ‘strongly disagree’ [36, 37]. In accordance with the self-determination theory, this
scale allows the researcher to decide which items to use
in order to create a shorter version of the questionnaire
[36]. The version used in the current study consisted of
twenty items divided into six subscales: interest/enjoy-
ment, perceived competence, effort, pressure/tension,
perceived choice and value/usefulness. The version used
can be found in the complementary documentation (see
Additional file 1) and is very similar to the IMI question-
naires used in other usability studies with robotic devices
[38, 39]. The study data were collected and managed using RED-
Cap [28] electronic tool hosted at FIBICO (Foundation
for Biomedical Research in Córdoba, Spain) [29]. RED-
Cap (Research Electronic Data Capture) is a secure,
web-based software platform designed to support data
capture for research studies, providing (1) an intuitive
interface for validated data capture; (2) audit trails for
tracking data manipulation and export procedures; (3)
automated export procedures for seamless data down-
loads to common statistical packages; and (4) procedures
for data integration and interoperability with external
sources [30, 31]. REDCap is HIPAA (Health Insurance
Portability and Accountability Act) [32] and 21 CFR Part
11 [33] compliant, which means that it meets the mini-
mum level of security for data in clinical investigations. However, no personal data were recorded on REDCap, in
compliance with the European General Data Protection
Regulation (GDPR) [34], as the participants were Euro-
pean citizens. The purpose of QUEST is to evaluate the patient’s sat-
isfaction with the device and with the services they have
used. It consists of 12 questions: 8 related to the device
and 4 related to the services, which must be rated on a
Likert scale from 1 ‘Not satisfied at all’ to 5 ‘Very satisfied’. The AA Usability Assessment Questionnaire consists of
a 17-item survey and was specifically designed for the
AA device used in MERLIN system. The questions are
rated by patients and therapists from 1 ‘strongly agree’ to
7 ‘strongly disagree’ to evaluate the satisfaction with the
system and the therapy. It also includes three open-ended Secondary outcome measurements. Clinical information y
With the aim of quantifying general arm function and
any effects the system has on it, clinical standardized
scales were used, before the patients started with the
therapy using MERLIN and after finishing the clinical
trial. The Fugl Meyer Upper Extremity Assessment Scale
(Fugl-Meyer) [40] and the Modified Ashworth Scale
(MAS) [41] were used to evaluate the patients’ clinical
condition before their enrolment in the study to confirm
their participation according to the inclusion criteria. The
same scales were repeated at the end of the therapy using
MERLIN to confirm the safety of the system and that no
negative effects had been caused in the patients such as
reduction of arm function. Fugl-Meyer and MAS could
also be used to measure the effectiveness of the system,
although only small improvements were expected due to
the short duration of the intervention, in which only lim-
ited clinical evidence could be obtained. 66.7% of the participants were men and 33.3% women. Their age range was between 41 and 89, with an average
of 63.9. 66.7% had left hemiparesis and 33.3% right. In addition to the therapy received with the MERLIN
system during the usability study, 88.9% of the patients
also received other Occupational Therapy or Physiother-
apy sessions (public or private). The characteristics of the
nine participants are shown in Table 1.h The participants were also asked at the end of each ses-
sion about any adverse effects that may have happened
during the clinical trial: no relevant adverse effects were
reported. The participants reported only two drawbacks
that have been previously foreseen in the user’s manual:
(1) chafing on the skin due to contact with the robot’s
protruding elements; and (2) mild shoulder fatigue at the
end of the session. To avoid any chafing, strips of foam
were provided to attach to any parts of the robot that
could cause the problem. This solution should be cus-
tomized to the participant’s arm/hand size. As regards
shoulder fatigue, the timing of the sessions was adjusted
to 20–30 min for all the participants with the aim of
avoiding this adverse side effect. The Fugl-Meyer is an index to assess the sensorimo-
tor impairment in individuals. The MAS measures mus-
cle tone during passive soft tissue stretching by rotating
a joint and estimating the resistance, and it is used as a
simple measure of spasticity. Statistical analysish The statistical outcomes were analysed using IBM SPSS
Statistics© [42] software for Windows© Operating Sys-
tem. Descriptive summary statistics (mean with standard
deviation, SD) was used to process the quantitative data
provided by the Likert-scale items in SUS, QUEST and
Adapted IMI. The qualitative data obtained in the open-
ended questions were analysed using thematic analysis. Primary outcomes measurements. Usability and acceptance
data The feasibility of use of the system and motivation were
evaluated by the patients using semi-structured inter-
views and different usability questionnaires with Likert
scales during the clinical trials, which lasted for 3 weeks. The validated scales used were: the System Usability Scale
(SUS) [35], the Adapted Intrinsic Motivation Inventory Guillén‑Climent et al. J NeuroEngineering Rehabil (2021) 18:41 Page 7 of 16 the recruitment period due to the pandemic. A special
effort was made to recruit additional participants to com-
plete the study, but no patients were willing to partici-
pate in the study during the research period. The patients
with a post-stroke upper limb impairment included in
this research usually had a comorbidity such as hyper-
tension, mellitus diabetes, atherosclerosis, heart disease,
etc. which includes this population in the risk group for
COVID-19. We were fully aware of this situation, as well
as the mobility restrictions imposed by the government,
and because of fear of infection, it was not possible to
include additional participants at the end of the study. However, it was considered that the results are also reli-
able with 9 patients, as was previously explained. questions about the participant’s subjective opinion, such
as the aspects liked most, any negative aspects identified,
and any proposals for improving the system. In addition, two short questions were added to ask the
participants about their willingness to pay for the MER-
LIN system as therapy. Usability and acceptance of resultsh The quantitative data obtained using SUS, QUEST and
IMI questionnaires is summarized in Figures below,
which show the results of user acceptance and experi-
ence. The usability perception has been rated with a
mean score of 71.94 % (SD = 16.38) on the SUS scale (see
Fig. 5), which means that the system usability is consid-
ered “Good”, according to the research by Bangor et al. [43] For the clinical assessment, a one-tailed paired t-test
with a significance level of p < 0.05 was used to compare
pre- and post-intervention Fugl Meyer and MAS out-
come measures. Participantsh According to the IMI (see Fig. 6b), the participants
considered the intervention useful (6.06 ± 1.93), con-
sidered themselves competent (5.67 ± 1.79), did not feel
pressurized (6.11 ± 0.73) and reported high levels of
interest and enjoyment (6.00 ± 1.82). They also agreed
that they had participated voluntarily (6.78 ± 0.55) and
evaluated their own effort positively (6.11 ± 1.34). The This clinical study was planned to start in Septem-
ber 2019 and finish in June 2020. During the study, the
COVID-19 global pandemic broke out, which had a
significant impact in the study. It was foreseen that 12
patients with a post-stroke hemiparesis would participate
in the research, but only 9 of these completed the inter-
vention study. 3 patients dropped out of the study after Guillén‑Climent et al. J NeuroEngineering Rehabil (2021) 18:41 Page 8 of 16 Table 1 Demographic information of patients
Gender: M = Male, F = Female; Hemiparesis/Dominant Hand: L = Left, R = Right; Other therapies: Y = Yes, N = No
Age
Gender
Hemiparesis
Dominant Hand
Months
since stroke
Employment situation
Other
therapies
P1
59
M
L
R
2
Medical leave
Y
P2
60
M
R
R
5
Medical leave
Y
P3
70
F
L
R
6
Retired
Y
P4
74
M
R
R
6
Retired
Y
P5
41
M
L
R
37
Medical leave
Y
P6
42
F
L
R
17
Unemployed
Y
P7
89
M
L
R
4
Retired
N
P8
69
F
L
R
7
Retired
Y
P9
71
M
R
R
35
Retired
Y Table 1 Demographic information of patients Some participants also expressed negative aspects. The
height and size of the robot did not allow all the patients
to adopt a completely relaxed posture. Regarding the
serious games, some participants complained about the
limited range of games and how simple they were. The
participants, both patients and clinicians, proposed the
following future improvements: motivation and satisfaction levels were positive, as
reflected in a mean score on the QUEST scale of 3.81
out of 5 (SD = 0.38) (see Fig. 6a). On both those scales,
a higher score means that the participant is more moti-
vated or satisfied (the result of the pressure/tension
subscale has been normalized). The results of the AA Questionnaire test are pre-
sented in Table 2. Participantsh The participants rated the system
with an average of 6 on a scale of 7, and it can therefore
be concluded that they considered it easy to learn, easy
to use and intuitive. They also considered that the sys-
tem affects their treatment positively because it allows
them to train longer, and they reported that this ther-
apy could be more entertaining compared to the regu-
lar therapy (6 out of 7). All the participants agreed that
they would recommend the system to other patients,
but some improvements were proposed. Some exam-
ples of the participants’ proposals can be found in the
next section (see Open-ended questions). • Improve robot design to allow a relaxed arm posi-
tion. • Adjust the dimensions and reduce the size and the
area to better suit home use. • Include more games and make them more attractive. • Develop more complex games with greater cognitive
involvement. Finally, the participants were asked about their willing-
ness to pay for a therapy using the MERLIN system and
what price they would be willing to pay. 88.9% of the par-
ticipants said they were in favour of paying for a system
like MERLIN for more than 6 weeks and they would be
willing to rent it for 40–60 € per month. Open‑ended questionh The participants were motivated to participate both dur-
ing the clinical trials and the rehabilitation sessions. At
the end of the trials, all of them gave positive answers to
the questions regarding the use of the MERLIN system
at home. The patients stated that they had enjoyed their
participation and valued positively this different style of
therapy for the rehabilitation of their affected arm. Some
of them stressed the aspects they liked most: Clinical results Although the aim of these trials was not to measure the
effectiveness of the system, some measurements of the
mobility status of the participants were also taken before
and after the use of MERLIN, with the aim of detecting
any unforeseen negative effects in the patients. As pre-
viously explained, the Fugl-Meyer and MAS scales were
used with this aim. The outcome measurements are
shown in Tables 3 and 4. From the data gathered, it can
be stated that there was a visible improvement on the
Fugl-Meyer scale after rehabilitation using the system
(T1), with significant changes in upper limb and coordi-
nation (p = 0.008 and p = 0.004, respectively), and in the
total score for motor function (p = 0.002). • “It is a new attractive and motivating therapy”. • “The MERLIN system is an entertaining and easy to
use therapy, which allows you to repeat an exercise
many times”. • “We can decide on the most convenient time for
doing the therapy”. Guillén‑Climent et al. J NeuroEngineering Rehabil (2021) 18:41 Page 9 of 16 Table 2 Patients’ results from the AA questionnaire
Questions
P1
P2
P3
P4
P5
P6
P7
P8
P9
MEAN
1. It has been easy to learn how
to use the system, both the
hardware and the software
7
7
7
5
7
7
3
7
7
6.33
2. I think I will often need the
support of a technical person
to be able to use this system
2
4
5
5
1
1
6
2
6
3.56
3. Using this system, I need to
spend a lot of time in non-
training activities
2
2
4
2
1
4
2
1
1
2.11
4. I can remember without
problem how to use the
system effectively
7
7
4
3
7
7
3
7
7
5.78
5. It took a long time to be able
to use the system without
problems
2
7
2
1
1
1
1
1
1
1.89
6. I think that I will benefit from
using this system
7
7
7
1
7
7
7
7
7
6.33
7. Using this system I am moti‑
vated to train longer
7
6
7
1
5
7
4
7
7
5.67
8. I think that this system is
uncomfortable to use
5
3
5
7
4
5
4
7
2
4.67
9. Clinical results I enjoyed training with this
system
7
6
7
1
5
7
5
7
7
5.78
10. I would recommend other
people to use this system
7
7
6
4
7
7
6
7
7
6.44
11. I think that this system
needs to be improved
7
6
6
7
7
5
4
7
6
6.11
12. I had internet connection
problems while using the
system
4
2
5
6
1
3
2
6
1
3.33
13. I feel uncomfortable using
a system like this, because I
have no experience in using
a pc
1
1
4
1
1
1
6
1
2
2.00
14. I don’t think using this
system will make any change
to my condition
2
1
2
7
1
1
5
1
2
2.44 Guillén‑Climent et al. J NeuroEngineering Rehabil (2021) 18:41 Page 10 of 16 On the scale 1 indicates‘Strongly disagree’and 7‘Strongly agree’
Table 2 (continued)
Questions
P1
P2
P3
P4
P5
P6
P7
P8
P9
MEAN
15. I feel that the games are
inadequate for the training
2
1
1
4
6
1
1
1
2
2.11
16. I am familiar with this kind
of technology
6
4
2
2
4
4
1
7
2
3.56
17. I feel myself safe using this
system
6
7
6
2
7
7
7
7
6
6.11 Guillén‑Climent et al. Clinical results J NeuroEngineering Rehabil (2021) 18:41 Page 11 of 16 Table 3 Fugl-Meyer Motor function outcomes
T0 = baseline and T1 = post-training sessions
SD standard deviation, P Participant
a Maximum score for each motor component of the assessment
Upper limb (36a)
Wrist (10a)
Hand (14a)
Coordination (6a)
Motor function (66a)
T0
T1
T0
T1
T0
T1
T0
T1
T0
T1
P1
31
36
9
9
14
14
5
6
58
63
P2
23
28
7
10
10
11
4
4
44
52
P3
32
36
10
10
13
14
5
6
60
66
P4
30
31
9
9
13
13
4
4
56
57
P5
24
26
3
4
8
8
3
4
38
42
P6
26
30
7
9
14
14
5
6
52
59
P7
32
32
10
10
12
12
4
5
58
59
P8
9
10
3
3
6
7
3
3
21
23
P9
32
32
9
10
12
12
3
4
56
58
MEAN
26.56
29.00
7.44
8.22
11.33
11.67
4.00
4.67
49.22
53.22
SD
7.49
7.84
2.74
2.73
2.78
2.60
0.87
1.12
12.84
13.25
T-test
t = − 3.55; p= 0.008
t = − 2.14; p=0.065
t = − 2.0; p= 0.081
t = − 4.0; p= 0.004
t = − 4.54; p= 0.002 Table 3 Fugl-Meyer Motor function outcomes T0 = baseline and T1 = post-training sessions SD standard deviation, P Participant a Maximum score for each motor component of the assessment However, since there were no significant changes in
MAS outcomes (p = 0.169), this kind of robot-based ther-
apy does not seem to influence the spasticity of the upper
limb. post-stroke patients’ home. This is the first time that
this system has been tested in a home environment, as
it was previously tested in a clinical setting [24, 18, 27].h The positive results for usability obtained in this
study agree with other home-based studies published
in recent years [5, 52, 53]. 9 patients completed the
study and their overall ratings on the different scales
were positive. The mean obtained for the SUS scale was
71.94, which is considered as ‘Good’ on the usability
scale [44] and is higher than other similar studies that
investigated system usability for similar technology for
rehabilitation, such as Nijenhuis et al., which was rated
at 69.0 [54], or Radder et al. with 70.1 [55]. Discussionh J NeuroEngineering Rehabil (2021) 18:41 Page 12 of 16 Table 4 MAS outcomes at T0–T1
P participant
a Maximum score in assessment = 5
T0
T1
P1
0
0
P2
1
1
P3
1
0
P4
0
0
P5
1
1
P6
0
0
P7
0
0
P8
2
1
P9
0
0
MEAN
0.56
0.33
SD
0.73
0.50
T-test
t = 1.512; p = 0.169 Table 4 MAS outcomes at T0–T1
P participant
a Maximum score in assessment = 5
T0
T1
P1
0
0
P2
1
1
P3
1
0
P4
0
0
P5
1
1
P6
0
0
P7
0
0
P8
2
1
P9
0
0
MEAN
0.56
0.33
SD
0.73
0.50
T-test
t = 1.512; p = 0.169 Table 4 MAS outcomes at T0–T1 this feedback option, which helped them engage better
with the therapy. It should be highlighted that the 100% of the partici-
pants would recommend using the MERLIN system to
other patients, and 88.9% would use it for more than 6
weeks. All the participants reported low or no levels
of stress when using MERLIN at home on their own. Despite the general positive perception, participants also
considered that MERLIN would need some improve-
ments, which will be considered when the system is
developed further. The size and height of the device, as
well as the inclusion of new games and improvements
in the design are some examples of proposals for usabil-
ity made by the participants. Although the participants
recruited in the current evaluation were not cognitively
affected due to the inclusion criteria, these games were
designed for all types of post-stroke patients, and so not
all the software was properly adapted for the participants. However, the MERLIN system also offers the option of
selecting online games, but only if the therapist considers
that the patient has enough mobility control to play them
and not get frustrated. The participants agreed that the
number of games was limited and some of them became
too easy after a few sessions of training at home. Despite
the fact that during the project, the number of the objec-
tives for each game was increased (number of images,
words, etc. Discussionh The aim of this study was to analyse the user accept-
ance and usability of the MERLIN system in the home
environment in patients with upper limb motor impair-
ment after a stroke. In this study, since we also aimed
to observe the possible changes in the patients’ clinical
condition after participating in the trials, some additional
clinical measurements were also added. When new tech-
nology for rehabilitation is developed, the final users’ spe-
cial needs and their mobility limitations must be taken
into consideration in order to guarantee that the system
design meets their real needs and requirements [44]. In
addition, the usability of the system must be evaluated to
build a system which is comfortable and motivating, and
which patients are willing to use, in order to ensure bet-
ter adherence to treatment in the future. In addition, the mean of 6.06 out of 7 points for IMI,
with a minimum of 5.67 on each subscale, indicates that
participants perceived the system as an interesting and
motivational system to use. The abovementioned stud-
ies also evaluated the system using this scale, obtain-
ing a mean of 5.2 and 5.1 respectively [54, 55], which
is lower than the results obtained with the MERLIN
system. Regarding the results of the QUEST and AA Ques-
tionnaires, the participants perceived this therapy as
interesting and motivating, as well as simple, intui-
tive and easy to use. However, P4 and P8 expressed a
slightly more negative view compared to the rest. P4
was not used to using new technology and felt frus-
trated when interacting with the computer and games. P8 has severe motor impairment, which meant much
more effort was required to move the arm compared
with the other participants. Despite these drawbacks,
both patients decided to complete the intervention. Numerous studies over the last few decades have
shown the effectiveness and the advantages of using
robotic systems for neurorehabilitation [45–47] and
tele-rehabilitation in last decades [48–50]. Other stud-
ies have also demonstrated that the use of an exo-
skeleton with a patient-driven control strategy for
rehabilitation, in which the patient plays an active role
during the therapy sessions, helps to make the treat-
ment more attractive and therefore more effective
[51]. In this study, a robotic tele-rehabilitation system
based on serious games was set up for testing in the Guillén‑Climent et al. Discussionh to be achieved), we feel that, considering the
feedback given and other studies using virtual reality [56,
57] or collaboration-based games [58, 59] the number of The open-ended questions about the participants’
subjective opinion showed that they enjoyed the ther-
apy using the MERLIN system, which indicates the
potential for robotic systems based on serious games
to be used for making people with stroke more actively
involved in their rehabilitation sessions. In addition,
the researchers observed and perceived that the use of
feedback like games’ scores or positive messages and
the possibility of following up their own evolution,
motivated them: the patients were very positive about 0
10
20
30
40
50
60
70
80
90
100
1
2
3
4
5
6
7
8
9
SUS score
SCORE
MEAN
Fig. 5 Individual results on SUS Scale (0–100 %) Fig. 5 Individual results on SUS Scale (0–100 %) Guillén‑Climent et al. J NeuroEngineering Rehabil (2021) 18:41 Page 13 of 16 0.00
1.00
2.00
3.00
4.00
5.00
6.00
ID 1
ID 2
ID 3
ID 4
ID 5
ID 6
ID 7
ID 8
ID 9
Device subscale score
Services subscale score
Final quest score
Mean Device score
Mean service subscale score
Mean Final quest Score
0
1
2
3
4
5
6
7
Interest/enjoyment
Perceived
competence
Effort
Pressure/tension
Perceived choice
Value/usefulness
a QUEST Score
b IMI Adapted Score
ID 1
ID 2
ID 3
ID 4
ID 5
ID 6
ID 7
ID 8
ID 9
MEAN
Fig. 6 Individual results. a QUEST Scale. b IMI Adapted Scale 0.00
1.00
2.00
3.00
4.00
5.00
6.00
ID 1
ID 2
ID 3
ID 4
ID 5
ID 6
ID 7
ID 8
ID 9
a QUEST Score 0
1
2
3
4
5
6
7
Interest/enjoyment
Perceived
competence
Effort
Pressure/tension
Perceived choice
Value/usefulness
b IMI Adapted Score
ID 1
ID 2
ID 3
ID 4
ID 5
ID 6
ID 7
ID 8
ID 9
MEAN b Fig. games should be extended with the help of health profes-
sionals to engage patients in longer therapies. On the other hand, since there were no significative
changes to MAS (p = 0.169), it can be concluded that
the use of this system does not have any influence on
the upper limb spasticity. Discussionh This result can be considered
positive because, in addition to the intrinsic factors that
cause spasticity, it has been suggested that extrinsic fac-
tors (noxious triggers) may increase the spasticity. These
factors could include mental stress, physical contact,
anxiety, pain, muscle fatigue, muscle contractures, cer-
tain body postures, jerky movements or changes of posi-
tion, among others [64], which could induce or aggravate
high muscle tone and induce pain. None of the patients
reported an increase in spasticity or a reduction in mobil-
ity after using the system, which demonstrates that MER-
LIN does not cause any noxious triggers. As stated above, the aim of this study was to evalu-
ate the usability of the system, and, in that respect, the
intervention timeline and the number of patients used
were too small to fully demonstrate the effectiveness of
the therapy. However, the clinical assessment showed
moderate improvements in motor function in the upper
limbs which agrees with results found in other stud-
ies carried out with similar robotic systems in the home
environment [60, 61]. On an individual level, the nine
participants were classified for motor function as mildly
(6 patients), moderately (2) and severely (1) impaired on
the baseline measurements using Fugl-Meyer scale. The
participants achieved a notable improvement, which
was evaluated using the Fugl-Meyer motor function
scale after training. The intervention involved patients
in the subacute, short-term chronic and long-term
chronic phases, according to the inclusion criteria. More
changes were expected in patients considered subacute
after training, as theoretically most recovery from spe-
cific impairments occurs during the 6 first months after
stroke [62, 63]. However no great differences were found
in their progress. In fact, P6, a patient in the chronic
phase, experienced the greatest motor change. The most
significant changes were experienced by P6, P2 and P3:
these patients were in different phases of evolution and
had a different baseline, as can be seen in Tables 1 and
3, respectively. Therefore, in this study, no relationship
was found between motor function improvements and
progress, and as most of the participants were also taking
part in additional therapies, the improvements cannot be
ascribed exclusively to the use of the MERLIN system. According to the clinical results, none of the patients
reported a reduction of the mobility or increased the
spasticity after using of the system, which demonstrates
that its use at least safe. Consent for publication Informed consent for publication of images was included in the patients’
participation informed consent form. The patients were explicitly informed
about the use of images and consent was explicitly requested for this action
and ensuring their privacy after their publication. Availability of data and materials The datasets used and analysed during the current study are available from
the uicec@imibic.org on reasonable request. Conclusionsh This study demonstrates the usability of the home-based
MERLIN system in patients with upper limb motor func-
tion impairment after stroke in different stages. The usa-
bility analysis showed that almost 100% of the patients
who participated found the system useful, safe and
motivating, and all of them achieved moderate clinical
improvement in motor function, according to the average
Fugl-Meyer score. Discussionh Additionally, there were no serious adverse side effects
during the study. Mild shoulder pain, mild fatigue and
chafing of the skin were the only adverse effects noted. Therefore, this study provides evidence that the MERLIN
system may be used safely in a home setting. Future studies might consider greater using a number
of participants including a control group, and a longer
duration of training. Identifying the different patients’
factors like age, sex, severity, or evolution of stroke
could be also helpful with the aim of identifying the tar-
get group. In addition, incorporating additional or more
complex games and wider range of movements in the
MERLIN system, would probably enhance motivation,
which might lead to a more effective therapy. Guillén‑Climent et al. J NeuroEngineering Rehabil (2021) 18:41 Page 14 of 16 Page 14 of 16 the project. The authors would also like to thank QJ Rehab for providing an
enhanced industrialized version of AA. Funding g
This research is part of a MERLIN project, which has received funding from EIT
Health (Grant no. 20649). EIT Health is supported by the European Institute
of Innovation and Technology (EIT), a body of the European Union which
receives support from the European Union’s Horizon 2020 Research and
innovation programme. Competing interests
Th
h
d
l
h The authors declare that they have no competing interests. The authors declare that they have no competing interests. The online version contains supplementary material available at https://doi. org/10.1186/s12984-021-00837-z. References 1. Johnson CO, Nguyen M, Roth GA, Nichols E, Alam T, Abate D, et al. Global, regional, and national burden of stroke, 1990–2016: a systematic
analysis for the Global Burden of Disease Study 2016. Lancet Neurol. 2019;18(5):439–58. Ethics approval and consent to participate Further research with a larger sample of participants,
including a control group and a longer intervention,
would help to explore the efficacy of the system and to
identify the factors, in order to gather better results on
neurorehabilitation. The patients who participated in the research were informed about the
implications of their participation and agreed to participate voluntarily, giv‑
ing their informed consent, as approved by the ethical committee, prior to
starting the trials.The experimental protocol was approved on 13/09/2019
(18/09/2019-Acta 20) by the Ethical Committee for Research with Drugs in
Cordoba ref. num. 4237, and by the Spanish Agency of Drugs and Medical
Devices (Agencia Española de Medicamentos y Productos Sanitarios—AEMPS)
on 15/10/2019, according to Directive 93/42 on using medical devices in
research. The feasibility of using this low-cost, easy to learn,
easy to use and easily transportable rehabilitation sys-
tem might constitute a major step forward in transferring
intensive rehabilitation to the home setting. Nowadays,
this factor is of special importance due to impact of the
COVID-19 pandemic. Authors’ contributions MNMA, FJMR, and MMR defined the study protocol. FJMR, PCA and MMR
were responsible for participants’ recruitment. SGC carried out the study and
collected the patients’ data. SGC y AG performed the data analysis and wrote
the manuscript. FJMR participated in the study coordination and design. AG
and JARR participated in the trials, providing the technology that was used
during the evaluation, and gave technological support during the trials. They
also provided information regarding new technology in the manuscript. The
therapists who participated in the trials also received training in the technol‑
ogy which was organized by TECNALIA. All authors read and approved the
final manuscript. In this study, the participants carried out 8 additional
hours of upper limb rehabilitation at home with an inno-
vative approach in neurological rehabilitation, based on
serious games and using intensive, repetitive, interac-
tive, and individualized practice. The results of this study
reflect that home training with the MERLIN system with
the indirect supervision of a therapist could be an inter-
esting approach that includes the most important specific
features of neurorehabilitation: it is high-intensity, task-
specific, goal-setting, repetitive, functional, meaningful
and challenging for the patient. Abbreviations AA: ArmAssist; ADL: Activities of Daily Living; ARAT: Action Research Arm Test;
CFR: Code of Federal Regulations; FIBICO: Foundation for Biomedical Research
in Cordoba; Fundación para la Investigación Biomédica de Cordoba in Span‑
ish; FSR: Force Sensing Resistors; Fugl-Meyer: Upper Limb Fugl Meyer Scale;
GDPR: General Data Protection Regulation; HIPPA: Health Insurance Portability
and Accountability Act; IMI: Intrinsic Motivation Inventory; IMIBIC: Maimonides
Biomedical Research Institute (Instituto Maimónides de Investigación Bio‑
médica, in Spanish); MAS: Modified Ashworth Scale; P: Participant; QR: Quick
response; QUEST: Quebec User Evaluation of Satisfaction with assistive Tech‑
nology; REDCap: Research Electronic Data Capture; SD: Standard deviation;
SUS: System Usability Scale; T0: Evaluation measurement pre-intervention; T1:
Evaluation measurement at post-training sessions. Received: 24 September 2020 Accepted: 10 February 2021 Additional file 1. Intrinsic Motivation Inventory (IMI). Additional file 1. Intrinsic Motivation Inventory (IMI). Author details 1 Maimonides Biomedical Research Institute of Cordoba (IMIBIC), Reina Sofia
University Hospital, University of Cordoba, Córdoba, Spain. 2 Neurorehabilita‑
tion area, Health Division of TECNALIA, Basque Research and Technology
Alliance (BRTA), San Sebastián, Spain. 3 Interlevel Clinical Management Unit
of Physical Medicine and Rehabilitation, Reina Sofía University Hospital of Cór‑
doba, Córdoba, Spain. 4 Córdoba and Guadalquivir Health District, Andalusia
Health Service, Córdoba, Spain. 5 Department of Applied Physics, Radiology
and Physical Medicine, University of Córdoba, Córdoba, Spain. 2.
Lindsay MP, Norrving B, Sacco RL, Brainin M, Hacke W, Martins S, et al.
World Stroke Organization (WSO): Global Stroke Fact Sheet 2019. Interna‑
tional Journal of Stroke. 2019;14(8):806–17. 1.
Johnson CO, Nguyen M, Roth GA, Nichols E, Alam T, Abate D, et al.
Global, regional, and national burden of stroke, 1990–2016: a systematic
analysis for the Global Burden of Disease Study 2016. Lancet Neurol.
2019;18(5):439–58. 3.
Stinear CM, Lang CE, Zeiler S, Byblow WD. Advances and challenges in
stroke rehabilitation. Lancet Neurol. 2020;19(4):348–60. Acknowledgements Authors would like to thank the patients who gave up their free time to
participate in the study. Special thanks to C. Rodriguez-de-Pablo, I. Dorronsoro,
G. Gaminde, B. Garcia-Mendizabal, E. Hernandez-Jimenez, J.H. Jung, T. Keller, A. Belloso, J. Perry, D. Valencia and H. Zabaleta for their contributions to develop‑
ing the system and to our MERLIN project partners for their participation in 2. Lindsay MP, Norrving B, Sacco RL, Brainin M, Hacke W, Martins S, et al. World Stroke Organization (WSO): Global Stroke Fact Sheet 2019. Interna‑
tional Journal of Stroke. 2019;14(8):806–17. 3. Stinear CM, Lang CE, Zeiler S, Byblow WD. Advances and challenges in
stroke rehabilitation. Lancet Neurol. 2020;19(4):348–60. 3. Stinear CM, Lang CE, Zeiler S, Byblow WD. Advances and challenges in
stroke rehabilitation. Lancet Neurol. 2020;19(4):348–60. Guillén‑Climent et al. J NeuroEngineering Rehabil (2021) 18:41 Page 15 of 16 Page 15 of 16 4. Coupar F, Pollock A, Rowe P, Weir C, Langhorne P. Predictors of upper limb
recovery after stroke: a systematic review and meta-analysis. Clin Rehabil. 2012;26(4):291–313. 23. Perry JC, Zabaleta H, Belloso A, Rodriguez-De-Pablo C, Cavallaro FI, Keller
T. ArmAssist. Development of a functional prototype for at-home teler‑
ehabilitation of post-stroke arm impairment. In: Proc IEEE RAS EMBS int
conf biomed robot biomechatronics. 2012;1561–6. 5. Sivan M, Gallagher J, Makower S, Keeling D, Bhakta B, O’Connor RJ, et al. Home-based Computer Assisted Arm Rehabilitation (hCAAR) robotic
device for upper limb exercise after stroke: results of a feasibility study in
home setting. J Neuroeng Rehabil. 2014;11:16. 24. Tomic TJD, Savic AM, Vidakovic AS, Rodic SZ, Isakovic MS, Rodriguez-de-
Pablo C, et al. ArmAssist robotic system versus matched conventional
therapy for poststroke upper limb rehabilitation: a randomized clinical
trial. Biomed Res Int. 2017:7. 6. Masiero S, Armani M, Ferlini G, Rosati G, Rossi A. Randomized trial of a
robotic assistive device for the upper extremity during early inpatient
stroke rehabilitation. Neurorehabil Neural Repair. 2014;28(4):377–86. 25. Rodríguez-De-Pablo C, Perry JC, Balasubramanian S, Belloso A, Saviç A,
Tomiç TD, et al. Serious games for assessment and training in post-stroke
robotic upper-limb telerehabilitation. In: NEUROTECHNIX 2014-Proc 2nd
int congr neurotechnology, electron informatics. 2014;126–34. 7. Wattanapan P, Lukkanapichonchut P, Massakulpan P, Suethanapornkul
S, Kuptniratsaikul V. Effectiveness of stroke rehabilitation compared
between intensive and nonintensive rehabilitation protocol: a multi‑
center study. J Stroke Cerebrovasc Dis. 2020;29:104809. 26. Rodriguez-De-Pablo C, Perry JC, Cavallaro FI, Zabaleta H, Keller T. Acknowledgements Devel‑
opment of computer games for assessment and training in post-stroke
arm telerehabilitation. In: Proc Annu Int Conf IEEE Eng Med Biol Soc
EMBS. 2012;4571–4. 8. Veerbeek JM, Langbroek-Amersfoort AC, van Wegen EEH, Meskers CGM,
Kwakkel G. Effects of robot-assisted therapy for the upper limb after
stroke: a systematic review and meta-analysis. Neurorehabil Neural
Repair. 2017;31(2):107–21. 27. Rodríguez-de-Pablo C, Savić A, Keller T. Game-based assessment in
upper-limb post-stroke telerehabilitation. In: Ibáñez J, González-Vargas J,
Azorín JM, Akay M, Pons JL, editors. Converging clinical and engineering
research on neurorehabilitation II. Cham: Springer International Publish‑
ing; 2017. p. 413–7. 9. Veerbeek JM, van Wegen E, van Peppen R, van der Wees PJ, Hen‑
driks E, Rietberg M, Kwakkel G. What is the evidence for physical
therapy poststroke? A systematic review and meta-analysis. PLoS ONE. 2014;9(2):e87987. g
28. Vanderbilt. REDCap. [Updated]. https://projectredcap.org/resources/citat
ions/. Accessed June 2020. 10. Morone G, Paolucci S, Mattia D, Pichiorri F, Tramontano M, Iosa M. The 3Ts
of the new millennium neurorehabilitation gym: therapy, technology,
translationality. Expert Rev Med Devices. 2016;13(9):785–7. 29. IMIBIC. FIBICO-Fundación para la Investigación Biomédica de Córdoba. https://www.imibic.org/fibico. 11. Perry JC, Rodriguez-de-Pablo C, Cavallaro FI, Belloso A, Keller T. Assess‑
ment and training in home-based telerehabilitation of arm mobility
impairment. J Access Des All (CC) JACCES. 2013;3(2):117–35. 30. Harris PA, Taylor R, Thielke R, Payne J, Gonzalez N, Conde JG. Research
electronic data capture (REDCap)—a metadata-driven methodology and
workflow process for providing translational research informatics sup‑
port. J Biomed Inform. 2009;42(2):377–81. 12. Winstein CJ, Stein J, Arena R, Bates B, Cherney LR, Cramer SC, et al. Guidelines for adult stroke rehabilitation and recovery: a guideline for
healthcare professionals from the American Heart Association/American
Stroke Association. Stroke. 2016;47:98–169. 31. Harris PA, Taylor R, Minor BL, Elliott V, Fernandez M, O’Neal L, et al. The
REDCap consortium: building an international community of software
platform partners. J Biomed Inform. 2019;95:103208. 13. Brewer BR, McDowell SK, Worthen-Chaudhari LC. Poststroke upper
extremity rehabilitation: a review of robotic systems and clinical results. Top Stroke Rehabil. 2007;14(6):22–44. 32. HHS.gov. Health Information Privacy. https://www.hhs.gov/hipaa/index
.html. 33. U.S. Food. & Drug Administration. Part11, Electronic records; electronic
signatures- scope and application. https://www.fda.gov/regulatory-infor
mation/search-fda-guidance-documents/part-11-electronic-records-
electronic-signatures-scope-and-application. 14. O’Neill C, Proietti T, Nuckols K, Clarke ME, Hohimer C, Cloutier A, et al. Inflatable soft wearable robot for reducing therapist fatigue during
upper extremity rehabilitation in severe stroke. IEEE Robot Autom Lett. 2020;5(3):3899–906. 34. Acknowledgements The European Parliament and The Council of the European Union. Regulation (EU) 2016/679 of the European Parliament and of the Council
of 27 April 2016 on the protection of natural persons with regard to the
processing of personal data and on the free movement of such data, and
repealing Directive 95/46/EC. Brussels; 2016. https://eur-lex.europa.eu/eli/
reg/2016/679/oj. 15. Burgar CG, Lum PS, Shor PC, Van Der Loos HFM. Development of robots
for rehabilitation therapy: the Palo Alto VA/Stanford experience. J Rehabil
Res Dev. 2000;37(6):663–73. 16. Nijenhuis SM, Prange GB, Amirabdollahian F, Sale P, Infarinato F, Nasr N,
et al. Feasibility study into self-administered training at home using an
arm and hand device with motivational gaming environment in chronic
stroke. J Neuroeng Rehabil. 2015;12(1):1–13. https://doi.org/10.1186/
s12984-015-0080-y. 35. Brooke J. SUS: a quick and dirty usability scale. In: Jordan PW, Thomas B,
Weerdmeester BA, McClelland IL, editors. Usability evaluation in industry. London: Taylor & Francis; 1996. 17. Popović MD, Kostić MD, Rodić SZ, Konstantinović LM. Feedback-mediated
upper extremities exercise: increasing patient motivation in poststroke
rehabilitation. Biomed Res Int. 2014;2014:11. 36. Self-Determination Theory. An approach to human motivation and
personality-Questionnaires, Intrinsic Motivation Inventory. https://selfd
eterminationtheory.org/intrinsic-motivation-inventory/. 37. Colombo R, Pisano F, Mazzone A, Delconte C, Micera S, Carrozza MC,
et al. Design strategies to improve patient motivation during robot-aided
rehabilitation. J Neuroeng Rehabil. 2007;4:12. 18. Rodríguez-de-Pablo C, Popović M, Savić A, Perry JC, Belloso A, Tomić
TD, et al Post-stroke robotic upper-limb telerehabilitation using serious
games to increase patient motivation: first results from ArmAssist System
Clinical Trial. In: Londral AR, Encarnação P, editors. Advances in neurotech‑
nology, electronics and informatics: revised selected papers from the 2nd
International Congress on Neurotechnology, Electronics and Informat‑
ics (NEUROTECHNIX 2014), October 25–26, Rome, Italy. Cham: Springer
International Publishing; 2016. p. 63–78. https://doi.org/10.1007/978-3-
319-26242-0_5. 38. Novak D, Nagle A, Keller U, Riener R. Increasing motivation in robot-aided
arm rehabilitation with competitive and cooperative gameplay. J Neuro‑
eng Rehabil. 2014;11:1–15. 39. Demers L, Weiss-lambrou R, Ska B. The Quebec user evaluation of satisfac‑
tion with assistive technology (QUEST 2.0): an overview and recent
progress. Technol Disabil. 2002;14:101–5. 40. Sanford J, Moreland J, Swanson LR, Stratford PW, Gowland C. Reliability
of the Fugl-Meyer assessment for testing motor performance in patients
following stroke. Phys Ther. 1993;73(7):447–54. 19. Nielsen J, Landauer TK, Garzo A, Ruiz-ruano JA. A mathematical model
of the finding of usability problems. Acknowledgements In: Proceedings of ACM INTER‑
CHI’93 conference (Amsterdam, The Netherlands, 24–29 April 1993),
pp. 206–213. g
y
41. Luo ZC, Lo WLA, Bian RH, Wong S, Li L. Advanced quantitative estimation
methods for spasticity: a literature review. J Int Med Res. 2019(13). 20. Dorronsoro I, Gaminde G, Jung JH, Tellez J, Keller T. MERLIN: upper-limb
rehabilitation robot system for home environment. In: Converging clinical
and engineering research on neurorehabilitation IV. ICNR 2020 proceed‑
ings. 2020. (In Press). 42. IBM Corp. IBM SPSS Statistics for Windows [v 25.0.]. Armonk, NY: IBM Corp;
2017. https://www.ibm.com/support/home/. 42. IBM Corp. IBM SPSS Statistics for Windows [v 25.0.]
2017. https://www.ibm.com/support/home/. 43. Bangor A, Kortum P, Miller J. Determining what individual SUS scores
mean: adding an adjective rating scale. J Usability Stud. 2009;4(3):114–23. 21. GMV. Antari Home Care. http://www.gmv.com. 21. GMV. Antari Home Care. http://www.gmv.com. 22. Rodriguez-De-Pablo C, Rozevink SG, Mayordomo F, Téllez J, Kocić S,
Keller T. MERLIN: homecare arm rehabilitation. In: INRS2019 international
neurorehabilitation symposium. Toronto; 2019. 44. Nasr N, Leon B, Mountain G, Nijenhuis SM, Prange G, Sale P, et al. The
experience of living with stroke and using technology: opportunities to 22. Rodriguez-De-Pablo C, Rozevink SG, Mayordomo F, Téllez J, Kocić S,
Keller T. MERLIN: homecare arm rehabilitation. In: INRS2019 international
neurorehabilitation symposium. Toronto; 2019. Guillén‑Climent et al. J NeuroEngineering Rehabil (2021) 18:41 Page 16 of 16 Page 16 of 16 engage and co-design with end users. Disabil Rehabil-Assist Technol. 2016;11(8):653–60. rehabilitation: a feasibility study on the assistive functionality. J Rehabil
Assist Technol Eng. 2016;3:2055668316670553. engage and co-design with end users. Disabil Rehabil-Assist Technol. 2016;11(8):653–60. 56. Díaz de Greñu L, Lozano H, Garzo A, Santisteban L, Bengoetxea A,
Arambarri J, et al. Diseño de un Sistema de Rehabilitación del Miembro
Superior para Supervivientes de ACV. In: XXXV Congreso Anual de la
Sociedad Española de Ingeniería Biomédica (CASEIB2017). Bilbao; 2017. 45. Lum PS, Burgar CG, Shor PC, Majmundar M, Van der Loos M. Robot-
assisted movement training compared with conventional therapy tech
niques for the rehabilitation of upper-limb motor function after stroke. Arch Phys Med Rehabil. 2002;83(7):952–9. 57. Shahmoradi L, Almasi S, Ahmadi H, Bashiri A, Azadi T, Mirbagherie A,
et al. Virtual reality games for rehabilitation of upper extremities in stroke
patients. J Bodywork Mov Ther. 2020;27(12):3473–8. 46. Chang WH, Kim YH. Robot-assisted therapy in stroke rehabilitation. J
Stroke. 2013;15(3):174–81. 47. Acknowledgements Straudi S, Severini G, Da Roit M, Pizzongolo LD, Martinuzzi C, Basaglia
N. The dose of robot-assisted gait therapy may influence functional
recovery in a multidisciplinary rehabilitation program: an exploratory
retrospective study. Int J Rehabil Res. 2020;43(2):175–82. 58. Pereira F, Bermudez i Badia S, Ornelas R, Cameirao MS. Impact of game
mode in multi-user serious games for upper limb rehabilitation: a
within-person randomized trial on engagement and social involvement. J Neuroeng Rehabil. 2019;16(1):1–13. 48. Peretti A, Amenta F, Tayebati SK, Nittari G, Mahdi SS, Telerehabilitation. Review of the state-of-the-art and areas of application. JMIR Rehabil
Assist Technol. 2017;4(2):e7. 59. Gorsic M, Cikajlo I, Goljar N, Novak D. A multisession evaluation of an
adaptive competitive arm rehabilitation game. J Neuroeng Rehabil. 2017;14:1–15. 49. Rogante M, Grigioni M, Cordella D, Giacomozzi C. Ten years of telereha‑
bilitation: a literature overview of technologies and clinical applications. Neurorehabilitation. 2010;27(4):287–304. 60. Ferreira FMR, Chaves MEA, Oliveira VC, Van Petten A, Vimieiro CBS. Effec‑
tiveness of robot therapy on body function and structure in people with
limited upper limb function: a systematic review and meta-analysis. PLoS
ONE. 2018;13(7):21. 50. Medina J, Acosta-Vargas P, Rybarczyk Y. A systematic review of usability
and accessibility in tele-rehabilitation systems. In: Assistive and rehabilita‑
tion engineering. IntechOpen. 2019. 61. Prange GB, Jannink MJ, Groothuis-Oudshoorn CG, Hermens HJ, Ijzerman
MJ. Systematic review of the effect of robot-aided therapy on recovery of
the hemiparetic arm after stroke. J Rehabil Res Dev. 2006;43(2):171–84. g
g
p
51. Rehmat N, Zuo J, Meng W, Liu Q, Xie SQ, Liang H. Upper limb rehabilita‑
tion using robotic exoskeleton systems: a systematic review. Int J Intell
Robot Appl. 2018;2(3):283–95. 62. Roach ES, Bettermann K, Biller J. Toole’s cerebrovascular disorders. 6th ed. Cambridge: Cambridge University Press; 2010. 52. Wilk R, Johnson MJ, Ieee, editors. Usability feedback of patients and thera‑
pists on a conceptual mobile service robot for inpatient and home-based
stroke rehabilitation. In: 5th IEEE RAS/EMBS international conference on
biomedical robotics and Biomechatronics (BioRob); 2014 Aug 12–15; Sao
Paulo: IEEE; 2014. 63. Lee KB, Lim SH, Kim KH, Kim KJ, Kim YR, Chang WN, et al. Six-month func‑
tional recovery of stroke patients: a multi-time-point study. Int J Rehabil
Res. 2015;38(2):173–80. 64. Phadke CP, Balasubramanian CK, Ismail F, Boulias C. Revisiting physiologic
and psychologic triggers that increase spasticity. Am J Phys Med Rehabil. 2013;92(4):357–69. 53. Acknowledgements Held JP, Ferrer B, Mainetti R, Steblin A, Hertler B, Moreno-Conde A, et al. Autonomous rehabilitation at stroke patients home for balance and gait:
safety, usability and compliance of a virtual reality system. Eur J Phys
Rehabil Med. 2018;54(4):545–53. Publisher’s note 54. Nijenhuis SM, Prange GB, Amirabdollahian F, Sale P, Infarinato F, Nasr N,
et al. Feasibility study into self-administered training at home using an
arm and hand device with motivational gaming environment in chronic
stroke. J Neuroeng Rehabil. 2015;12:12. Springer Nature remains neutral with regard to jurisdictional claims in pub‑
lished maps and institutional affiliations. Springer Nature remains neutral with regard to jurisdictional claims in pub‑
lished maps and institutional affiliations. 55. Radder B, Prange-Lasonder GB, Kottink AI, Gaasbeek L, Holmberg J, Meyer
T, et al. A wearable soft-robotic glove enables hand support in ADL and •
fast, convenient online submission
•
thorough peer review by experienced researchers in your field
•
rapid publication on acceptance
•
support for research data, including large and complex data types
•
gold Open Access which fosters wider collaboration and increased citations
maximum visibility for your research: over 100M website views per year
•
At BMC, research is always in progress. Learn more biomedcentral.com/submissions
Ready to submit your research
Ready to submit your research ? Choose BMC and benefit from:
? Choose BMC and benefit from: •
fast, convenient online submission
•
thorough peer review by experienced researchers in your field
•
rapid publication on acceptance
•
support for research data, including large and complex data types
•
gold Open Access which fosters wider collaboration and increased citations
maximum visibility for your research: over 100M website views per year
•
At BMC, research is always in progress. Learn more biomedcentral.com/submissions
Ready to submit your research
Ready to submit your research ? Choose BMC and benefit from:
? Choose BMC and benefit from: •
fast, convenient online submission
•
thorough peer review by experienced researchers in your field
•
rapid publication on acceptance
•
support for research data, including large and complex data types
•
gold Open Access which fosters wider collaboration and increased citations
maximum visibility for your research: over 100M website views per year
•
At BMC, research is always in progress. Learn more biomedcentral.com/submissions
Ready to submit your research
Ready to submit your research ? Choose BMC and benefit from:
? Publisher’s note Choose BMC and benefit from: •
fast, convenient online submission
•
thorough peer review by experienced researchers in your field
•
rapid publication on acceptance
•
support for research data, including large and complex data types
•
gold Open Access which fosters wider collaboration and increased citations
maximum visibility for your research: over 100M website views per year
•
At BMC, research is always in progress. Learn more biomedcentral.com/submissions
Ready to submit your research
Ready to submit your research ? Choose BMC and benefit from:
? Choose BMC and benefit from:
|
https://openalex.org/W2112362317
|
https://europepmc.org/articles/pmc3961023?pdf=render
|
English
| null |
Is Fasting Duration Time Important in Post Adenotonsillectomy Feeding Time?
|
Anesthesiology and pain medicine
| 2,013
|
cc-by
| 4,392
|
Yalda Jabbari Moghaddam
1, Mahin Seyedhejazi
2,*, Masoud NaderPour
1, Yoosef Yaghooblua
1, Samad Golzari
2 Yalda Jabbari Moghaddam
1, Mahin Seyedhejazi
2,*, Masoud NaderPour
1, Yoosef Yaghooblua
1, Samad Golzari
2 1Department of Otolaryngology, Head and Neck Surgery, Tabriz University of Medical Sciences, Tabriz, Iran
2Department of Anesthesiology, Tabriz University of Medical Sciences, Tabriz, Iran Corresponding author: Mahin Seyedhejazi, Department of Anesthesiology, Tabriz University of Medical Sciences, Tabriz, Iran. Tel.: +98-9141150981, Fax: +98-4
dhejazie@yahoo.com. : Mahin Seyedhejazi, Department of Anesthesiology, Tabriz University of Medical Sciences, Tabriz, Iran. Tel.: +98-9141150981, Fax: +98-4115262280, E-mail: se
m. Received: January 15, 2013; Revised: July 7, 2013; Accepted: July 30, 2013 Received: January 15, 2013; Revised: July 7, 2013; Accepted: July 30, 2013 Background: Adenotonsillectomy is a common otolaryngology surgery. Nausea and vomiting are the most common complications of
this procedure with a prevalence ranging from 49% to 73 %.f p
p
g
g
Objectives: Our aim was to evaluate the effects of short time fasting protocol on decreasing postoperative pain, nausea and vomiting, and
initiation of oral feeding after adenotonsillectomy. Patients and Methods: 120 children aged 4 to 14 years candidates for adenotonsillectomy were randomly divided into intervention and
control groups (n = 120, 60 in each group). Each patient of the intervention group was given oral dextrose 10% as much volume as he could
consume at 3 and 6 hours prior to the operation. All the data including pain severity, nausea and vomiting of the patients, the time of oral
feeding initiation etc. were gathered in checklists after the operation. Statistical analyses were then performed using Statistical Package
for the Social Sciences (SPSS) software version 16. Descriptive statistical methods and mean difference test for independent groups and
chi square test or Fisher exact test, and if regression needed model test were applied. A P value of 0.05 or less was considered statistically
significant.ii i
Results: The amount of Acetaminophen administered for the intervention group was significantly lower than the control group, and also
the time of oral feeding initiation was significantly shorter in the intervention group than the control group (P < 0.005). Pain severity at
all occasions following surgery was significantly lower in the intervention group than the control group (P < 0.001). Although frequency
of nausea at recovery time was significantly lower in the intervention group than the control group (P < 0.002), there were no significant
differences in frequency of nausea between the two groups at other postoperative occasions. Implication for health policy/practice/research/medical education:
Adenotonsillectomy is considered a common otolaryngology surgery, with nausea and vomiting as the most prevalent associated complications. Chil
dren are prone to post-adenotonsillectomy pain, nausea and vomiting, and more than a half of children who undergo surgery have vomiting. Unfortu
nately, postoperative pain, nausea and vomiting delay oral feeding initiation, which in turns leads to dehydration, delay on operation site healing, and
on the other hand prolongation of convalescence period.
Copyright © 2014, Iranian Society of Regional Anesthesia and Pain Medicine (ISRAPM); Published by Kowsar Corp. This is an open-access article distributed under the
terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work
is properly cited. Yalda Jabbari Moghaddam
1, Mahin Seyedhejazi
2,*, Masoud NaderPour
1, Yoosef Yaghooblua
1, Samad Golzari
2 Postoperative vomiting frequency was not
significant between the two groups at any occasions.if i
Conclusions: The findings of this survey showed that shortening the duration of pre-adenotonsillectomy fasting period and hydration
of patients several hours prior to the operation might be effective in decreasing postoperative pain and facilitating postoperative oral
feeding initiation. Nevertheless this method does not seem to prevent postoperative nausea and vomiting. Keywords: Pain; Nausea; Vomiting; Children; Fasting; Preoperative Period Anesth Pain Med. 2014 February; 4(1): e10256. Anesth Pain Med. 2014 February; 4(1): e10256. OI: 10.5812/aapm.10256 Research Article Published online 2014 February 26. 1. Background new surgery techniques (13), and pain relieving tech
niques have been studied in pain control, nausea and
vomiting postoperatively, but there is yet no satisfactory
cure for postoperative pain, nausea and vomiting. Adenotonsillectomy is a common otolaryngology sur
gery, with nausea and vomiting as the most prevalent
complications of this procedure with a prevalence rang
ing from 49% to 73% (1-4). This complication affects more
than 41% of patients (5-7). Thus pain, nausea and vomit
ing are common problems in children after adenotonsil
lectomy, and more than a half of children who undergo
surgery have vomiting (8, 9).f Unfortunately postoperative pain, nausea and vomiting
delay oral feeding initiation which consequently induce
dehydration and delay in operation site healing, and on
the other hand cause prolongation of convalescence pe
riod (14). It seems that other nonpharmacologic factors
play role in inducing postoperative nausea and vomit
ing. To reduce the risk of gastric content aspiration dur
ing general anesthesia, each child should become fasting Although effects of applying some systemic medica
tions prior to, during or after adenotonsillectomy (10),
local anesthesia (11), administering some antibiotics (12), Jabbari Moghaddam Y et al. prior to transportation to the operating room (15). prior to transportation to the operating room (15). 0 1 2 3 4 5
No Hurt
Hurts
Litle Bit
Hurts
Litle More
Hurts
Even More
Hurts
Whole Lot
Hurts
Worst
Figure 1. Wong-Baker FACES Pain Rating Scale 0 1 2 3 4 5
No Hurt
Hurts
Litle Bit
Hurts
Litle More
Hurts
Even More
Hurts
Whole Lot
Hurts
Worst
Figure 1. Wong-Baker FACES Pain Rating Scale 5
Hurts
Worst p
p
p
g
( 5)
However studies have shown that postoperative nau
sea associate with prolonged fasting (16) and some other
studies have suggested that short period of fasting is safe
for children and even allows children intake clear fluids
and solid food two and four to six hours prior to opera
tion respectively (17, 18). Egeli et al. (19) in their study sug
gested that 24 hours hydration can reduce postoperative
morbidity following tonsillectomy in children. Even
recently Elqueta et al. (20) studied the effect of large in
traoperative crystalloid administration as prophylaxis of
postoperative vomiting in children undergoing tonsil
lectomy in a randomized controlled trial, and concluded
that super-hydration during tonsillectomy is an alterna
tive way to decrease the risk of postoperative vomiting in
children (20). 3. Patients and Methods After getting approval from the Medical Ethics Com
mittee, and obtaining written informed consent from
the parents of the participants, 120 children aged 4 to 14
candidates for adenotonsillectomy were randomly divid
ed into intervention and control groups (n = 120, 60 in
each group). To distribute the trial blemishing variables
such as age, gender, educational status and cultural class,
patients were randomly allocated into intervention and
control groups. Exclusion criteria were diabetes, gastro
intestinal tract disorders, upper respiratory tract infec
tion, and weighting more than 50 kg. According to our study aims, a checklist was designed
according to the former studies. This checklist included
questions concerning demographic data of the patients
and other needed information. All data regarding the
pain, nausea and vomiting complaints, oral feeding ini
tiation instant, and oral analgesic drugs administered by
an invariable accessory, were then gathered in the men
tioned checklists postoperatively. In the intervened group, all of the patients were given
oral dextrose 10% solution for a maximum dose of 10 mg/
kg depending on their appetite. There was no obligation
to intake an amount of maximal dose. Case group mem
bers were kept fasting prior to adenotonsillectomy as
otolaryngology department routine (all children must
be NPO from 12:00 am preoperative day, till the opera
tion, and during this time only receiving 1/3-2/3 serum as
much to keep vein open (KVO)). 2. Objectives The aim of this single blinded clinical trial was to score
short time fasting protocol effects on decreasing pain,
nausea and vomiting, and establishing oral feeding after
adenotonsillectomy in hospitalized children of Tabriz
Children Hospital. 3.1. Four Point Severity Scale The four-point severity scale included no symptom,
mild symptom, severe nausea or up to two vomits, and
more than two vomits (23). Adenotonsillectomy was
performed by a same surgeon, and invariable anesthetic
drugs were used in the both groups. In all patients, pre
medication was performed using Midazolam (0.03 mg/
kg/IV) and Fentanyl 1 μg/kg/IV few minutes before separa
tion from parents, and anesthesia induction was initiated
using Lidocaine 1 mg/kg and Propofol with a dose of 3.5
mg/kg, Atracurium 0.5 mg/kg and 0.1 mg/kg Dexametha
sone single dose, and a stat 40 mg/kg of Acetaminophen
suppository just after intubation (with an appropriate
size ETT), and maintenance was performed with Isoflu
rane 1-1.5%, and a mixture of N2O 50% and O2 50%. Fifteen
mg/kg of oral Acetaminophen was administered postop
eratively to the patients with a pain score more than 5 in
the both groups. 1. Background Despite this change in fasting period, most
of the children candidates for surgery have yet to with
stand long periods of hunger. Hurts
Even More Figure 1. Wong-Baker FACES Pain Rating Scale Wong-Baker FACES pain rating scale, a 0 to 10 scale, was
employed to score the pain severity of patients (Figure
1) (21, 22). Another scale which its criteria are among the
accredited ones in assessing postoperative nausea and
vomiting was applied to assess the nausea and vomiting. Sounds like if we could simply take some non-pharma
cological steps to keep child vivacious postoperatively e.g. by administering fasting protocol exquisitely, it would be
very helpful in decreasing postoperative complications. Anesth Pain Med. 2014;4(1):e10256 4. Results One hundred and twenty patients aged 4 to 14 years
were allocated into two groups of control and interven
tion each containing 60 patients. Amount of administered Acetaminophen in 30 patients
was a single dose of 15 mg/kg which comprised most pa
tients, and their fasting period was shorter than routine
in our department.i Table 1. Postoperative Oral Feeding Initiation Hours
Hours
Intervention Group
Control Group
1
40
8
2
14
2
3
4
5
4
2
5
5
-
20
6
-
20
Total
60
60
Table 2. Acetaminophen Prescription Times
Acetaminophen
Perception Times
Intervention Group,
Patients No. Control
Group, Pa
tients No. Once
30
10
Twice
10
24
Three times
-
20
Four times
1
3
Five times
1
-
Not use
18
3
Total
60
60
Allocating patients into two groups was performed by
randomly permuted blocks method. There was no signif
icantly meaningful difference in age distribution of the
two groups (P = 0.247). Among the intervention group, oral feeding was estab
lished in the first, second, third, and the fourth hours
postoperatively for 40, 14, 4, and 2 patients respectively. Therefore oral feeding was initiated in all patients during
four hours postoperatively. Table 3. Comparison of Nausea Frequency in Various Postoperative
R
H
P
t
H Table 1. Postoperative Oral Feeding Initiation Hours There was significantly higher amounts of Acetamino
phen applied in the control group compared to the in
tervention group, i.e. in 10, 24, and 20 patients the Acet
aminophen was administered one, two, and three times
respectively at a dose of 15 mg/kg (P < 0.005) (Table 2). f Pain severity in the patients of both groups was assessed
in 6 different occasions (at recovery time, 2, 4, 6, 8, and 24
hours postoperatively) according to the guidelines de
scribed in methods section of this article. At all occasions,
pain severity in the intervention group was meaningfully
less than the control group (P < 0.001).f Table 2. Acetaminophen Prescription Times Nausea prevalence was different in various occasions
postoperatively, the most of which was seen in recovery
time and 2 hours postoperatively, affecting a total num
ber of 21 patients (17.5%).if Acetaminophen
Perception Times
Intervention Group,
Patients No. Control
Group, Pa
tients No. 3.2. Statistical Analysis Data was then imported in SPSS analysis software ver
sion 16 after gathering and later were analyzed. Acquired
data was statistically analyzed using descriptive statisti
cal methods (frequency, age (%), mean and standard de
viation), independent groups T-test for means, chi square
test, and Fisher’s exact test or regression models. A P val
ue of 0.05 or less was considered statistically significant. As a medical ethics rule, study method was completely
explained to the parents and their questions were an
swered. All of the parents provided their written consent
form to enroll their children in the study. Anesth Pain Med. 2014;4(1):e10256 2 2 Jabbari Moghaddam Y et al.i But among controlled group, oral feeding initiation in
the first hour was significantly less than the intervention
group, i.e. 8 patients (Table 1). 4. Results Once
30
10
Twice
10
24
Three times
-
20
Four times
1
3
Five times
1
-
Not use
18
3
Total
60
60 As shown in Table 2 and Table 4, although there was a
significant difference in pain severity between the two
groups, no significant difference was noted regarding
nausea and vomiting. Comparing the frequency of nausea between the con
trol and intervention groups revealed that there was
a statistically significant lower nausea frequency (no
symptom) only at recovery time, as shown in Table 3. In
recovery time, only 4 of 60 patients in the intervention
group complained of nausea (mild or severe symptoms),
while 17 of 60 patients in the control group complained
of nausea (mild or severe symptoms) (P = 0.002). In other
5 occasions, i.e. 2, 4, 6, 8, and 24 hours postoperatively,
there was a lower nausea frequency reported in interven
tion group compared to control group, but this differ
ence was not statistically significant. Allocating patients into two groups was performed by
randomly permuted blocks method. There was no signif
icantly meaningful difference in age distribution of the
two groups (P = 0.247).i Among the intervention group, oral feeding was estab
lished in the first, second, third, and the fourth hours
postoperatively for 40, 14, 4, and 2 patients respectively. Therefore oral feeding was initiated in all patients during
four hours postoperatively. i
In this study, among 120 patients, the highest vomiting
(moderate to severe symptoms) prevalence, i.e. 10 patients 3
Anesth Pain Med. 2014;4(1):e10256
Table 3. Comparison of Nausea Frequency in Various Postoperative Occasions Between the Intervention and Control Groups
Recovery
2 Hours Post
Operation
4 Hours Post
Operation
6 Hours Post
Operation
8 Hours Post
Operation
24 Hours Post
Operation
Intervention group
4
9
3
4
1
0
Control group
17
11
4
10
3
2
P value
0.002
0.387
0.491
0.7
0.303
0.248
Table 4. Comparison of Vomiting Frequency in Various Postoperative Occasions Between the Intervention and Control Groups
Recovery
2 Hours Post
Operation
4 Hours Post
Operation
6 Hours Post
Operation
8 Hours Post
Operation
24 Hours Post
Operation
Intervention group
3
3
2
2
0
0
Control group
4
7
1
5
0
1
P value
0.509
0.154
0.506
0.219
-
0.5 Anesth Pain Med. 2014;4(1):e10256 3 3 Jabbari Moghaddam Y et al. 5. Discussion Our aim in this study was to assess the effects of short
time fasting protocol on post adenotonsillectomy pain,
nausea, vomiting, and instant oral feeding initiation in
children hospitalized in Tabriz Children Hospital. Adenotonsillectomy complications might increase the
risk of aspiration which may complicate the patients
(24). Numerous authors have searched for the ideal pre
anesthetic medication, and also for the best medication
route. The premedication must be acceptable to patients,
and an atraumatic route of administration should be
available, in addition to the other characteristics re
quired for such a drug (25, 26).f Findings showed that in all postoperative occasions (re
covery time, 2, 4, 6, 8, and 24 hours postoperatively) pain
severity in the intervention group was less than the con
trol group. Similar finding were reported by Klemetti et
al. (34). In the mentioned study the intervention group
that experienced shorter fasting time period than con
trol group members experienced more severe pain (P
= 0.0002) (34). None of the two groups differed in nau
sea degree in the post anesthesia care unit, but as time
passed, nausea and vomiting level increased in the both
groups. Although there was no statistically significant dif
ference in nausea and vomiting between the two groups,
but there was a higher frequency and severity in the con
trol group. An effective pain therapy to block or modify the physi
ological responses to stress has become an essential
component of modern pediatric anesthesia and surgical
practice (27). Comparing oral feeding initiation between the two
groups revealed that there was an earlier oral feeding
initiation among intervention group compared to the
control group, i.e. patients who had received oral dex
trose prior to the operation were able to start postopera
tive oral feeding earlier than those who were kept fasting
as department routine preoperatively. Among control
group, the number of patients who had started oral feed
ing in the first hour postoperatively was significantly
lower than the intervention group. Also the number of
patients starting oral feeding in the following occasions
was higher in the control group. This study stated that shorter preoperative fasting time
plus suitable and controlled feeding in patients has a sig
nificant decreasing effect on postoperative pain, and this
intervention could increase patients’, particularly chil
dren’s resistance to postoperative nausea and vomiting
(34).ifif Carithers et al. (28) and Guida et al. 4. Results than the control group, which could indicate the fact that
among intervention group, patients’ pain degrees were
milder than the control group, and they experienced a
better recovery than the control group. As like as our find
ings, Nygren stated in his study that despite some former
surveys which believed that preoperative fasting time
limitation and hydration of patient do not affect his or
her status postoperatively, but it seems to be effective in
patient condition and recovery. This survey claimed that
preoperative hydration of the patients is significantly ef
fective in recovery, lowering nausea and vomiting, and
earlier discharge of patient from hospital. This study also
showed that a carbohydrate serum preparation prior to
the operation is significantly effective in recovery after
the operation (32). In another study Dr. Seyedhejazi and
his colleagues (33) reported that, infiltration of bupiva
caine and clonidine in children undergoing tonsillecto
my is more efficacious than single IV fentanyl to decrease
postoperative pain. This approach is also safer regarding
the intraoperative complications. (8.3%), was reported to be at the second hour postopera
tively. As delineated in Table 4, vomiting frequency compari
son between the intervention and control groups in vari
ous occasions (recovery time, 2, 4, 6, 8, and 24 hours post
operatively) showed that there was a lower frequency of
vomiting in intervention group, but this difference was
not statistically significant. Anesth Pain Med. 2014;4(1):e10256 5. Discussion (29) believed that
postoperative nausea is the most common postopera
tive complication in children despite advancements in
surgery and anesthesia. Different surveys have reported
prevalence rates ranging from 62% to 73% (30, 31). As in Klemetti’s survey (34), in our study there were sig
nificant differences in pain, but no significant difference
in postoperative nausea and vomiting. Considering nau
sea severity in various postoperative occasions, there was
a significantly lower nausea in intervention group, only
in recovery time. In our study, the prevalence of these two complications,
nausea and vomiting, in various postoperative hours was
different. Highest nausea prevalence was reported to be
at recovery time and two hours after the operation, which
generally affected 21 patients (17.5%), a lower rate in com
parison to the former studies, and the highest vomiting
prevalence after adenotonsillectomy was at the second
postoperative hour with a lower rate in comparison to
the former studies. Hamid et al. (15) reported a vomiting
prevalence rate in 80% for post adenotonsillectomy chil
dren who did not receive any prevention for vomiting.i In recovery occasion, only four patients of the interven
tion group complained of nausea, compared to 17 in the
control group. In other postoperative occasions (2, 4, 6, 8,
and 24 hours postoperatively) nausea severity was lower
in the intervention group than the control group, but
this difference was not statistically significant. We think
that our study limitation is its small study population,
and also usage of cold knife techniques for adenoton
sillectomy, so we suggest further multi central studies
in large groups and with other scales for comparison of
study groups and usage of different surgical instruments In this survey, the frequency of Acetaminophen admin
istration for intervention group was significantly lower 4 Jabbari Moghaddam Y et al. for adenotonsillectomy. for adenotonsillectomy. 13. Shapiro NL, Bhattacharyy N. Cold dissection versus cobla
tion-assisted adenotonsillectomy in children. Laryngoscope. 2007;117(3):406–10. This study showed that shortening of preoperative fast
ing period and hydration of the patient few hours prior
to the adenotonsillectomy may lower postoperative pain
and accelerate oral feeding initiation time. Nevertheless
this method lonely does not seem to be effective in pre
vention of postoperative nausea and vomiting. ( )
14. White MC, Nolan JA. An evaluation of pain and postoperative
nausea and vomiting following the introduction of guidelines
for tonsillectomy. Paediatr Anaesth. 2005;15(8):683–8. 15. Hamid SK, Selby IR, Sikich N, Lerman J. References 23. Dahan A, Miller JM, Hoffman A, Amidon GE, Amidon GL. The sol
ubility-permeability interplay in using cyclodextrins as pharma
ceutical solubilizers: mechanistic modeling and application to
progesterone. J Pharm Sci. 2010;99(6):2739–49. 1. Splinter W, Roberts DJ. Prophylaxis for vomiting by children af
ter tonsillectomy: dexamethasone versus perphenazine. Anesth
Analg. 1997;85(3):534–7.f 2. Pappas AL, Sukhani R, Hotaling AJ, Mikat-Stevens M, Javorski JJ,
Donzelli J, et al. The effect of preoperative dexamethasone on
the immediate and delayed postoperative morbidity in children
undergoing adenotonsillectomy. Anesth Analg. 1998;87(1):57–61.f 24. Watcha MF, White PF. Postoperative nausea and vomiting. Its eti
ology, treatment, and prevention. Anesthesiology. 1992;77(1):162–
84.fi 25. Hosseini Jahromi SA, Hosseini Valami SM, Javadi A. Effects of
suppository Acetaminophenbupivacaine wound infiltration,
and caudal block with bupivacaine on postoperative pain in
Pediatric inguinal herniorrhaphy. Anesth Pain Med. 2012;1(4,
Spring):243–7. 3. Ved SA, Walden TL, Montana J, Lea DE, Tefft MC, Kataria BK, et al. Vomiting and recovery after outpatient tonsillectomy and ad
enoidectomy in children. Comparison of four anesthetic tech
niques using nitrous oxide with halothane or propofol. Anesthe
siology. 1996;85(1):4–10. 26. KVerma R, Paswan A, De A, Gupta S. Premedication with Midazol
am nasal spray: an alternative to oral midazolam in children. Anesth Pain. 2011;1(4, Spring):248–51.f 4. Kanerva M, Tarkkila P, Pitkaranta A. Day-case tonsillectomy in
children: parental attitudes and consultation rates. Int J Pediatr
Otorhi. 2003;67(7):777–84. 27. Hosseini Jahromi SA, Hosseini Valami SM, Hatamian S. Com
parison between effect of lidocaine, morphine and ketamine
spray on post-tonsillectomy pain in children. Anesth Pain Med. 2012;2(1):17–21. 5. Kotiniemi LH, Ryhanen PT, Valanne J, Jokela R, Mustonen A, Pouk
kula E. Postoperative symptoms at home following day-case sur
gery in children: a multicentre survey of 551 children. Anaesthe
sia. 1997;52(10):963–9.f 28. Carithers JS, Gebhart DE, Williams JA. Postoperative risks of pe
diatric tonsilloadenoidectomy. Laryngoscope. 1987;97(4):422–9. 6. Smith BL, Manford ML. Postoperative vomiting after paediatric
adenotonsillectomy. A survey of incidence following differing
pre- and postoperative drugs. Br J Anaesth. 1974;46(5):373–8. 29. Guida RA, Mattucci KF. Tonsillectomy and adenoidectomy: an in
patient or outpatient procedure? Laryngoscope. 1990;100(5):491–
3. 7. Kermode J, Walker S, Webb I. Postoperative vomiting in children. Anaesth Intensive Care. 1995;23(2):196–9. 30. Furst SR, Rodarte A. Prophylactic antiemetic treatment with on
dansetron in children undergoing tonsillectomy. Anesthesiology. 1994;81(4):799–803. 8. Mukherjee K, Esuvaranathan V, Streets C, Johnson A, Carr AS. Ade
notonsillectomy in children: a comparison of morphine and fen
tanyl for peri-operative analgesia. Anaesthesia. Authors’ Contribution All of the contributors had role in data collection, writ
ing paper. 19. Egeli E, Harputluoglu U, Ozturk O, Oghan F, Kocak S. Can post-
adenotonsillectomy morbidity be reduced by intravenous 24 h
hydration in pediatric patients following adenotonsillectomy? Int J Pediatr Otorhi. 2004;68(8):1047–51.fl Financial Disclosurei 20. Elgueta MF, Echevarria GC, De la Fuente N, Cabrera F, Valderrama
A, Cabezon R, et al. Effect of intravenous fluid therapy on post
operative vomiting in children undergoing tonsillectomy. Br J
Anaesth. 2013;110(4):607–14. The authors declared no financial interest. The authors declared no financial interest. Acknowledgements 17. Emerson BM, Wrigley SR, Newton M. Pre-operative fasting for
paediatric anaesthesia. A survey of current practice. Anaesthesia. 1998;53(4):326–30. The authors declared nothing for acknowledgment. The authors declared nothing for acknowledgment. The authors declared nothing for acknowledgment. 18. Ewah BN, Robb PJ, Raw M. Postoperative pain, nausea and vom
iting following paediatric day-case tonsillectomy. Anaesthesia. 2006;61(2):116–22. 5. Discussion Vomiting after adenoton
sillectomy in children: a comparison of ondansetron, dimenhy
drinate, and placebo. Anesth Analg. 1998;86(3):496–500.l 16. Smith AF, Vallance H, Slater RM. Shorter preoperative fluid fasts
reduce postoperative emesis. BMJ. 1997;314(7092):1486. Tabriz University of Medical Sciences. 22. Constance LM, Sabine KB, Myron Y. Smith's pediatric anesthesia for
infants and children. 8th ed. USA: Elsevier Mosby; 2011.f Funding/Support 21. Wong D, Baker C. Reference manual for the Wong-Baker FACES pain
rating scale. Tulsa, USA: Wong & Baker; 1995. Tabriz University of Medical Sciences. References 2001;56(12):1193–7.fi 31. Litman RS, Wu CL, Catanzaro FA. Ondansetron decreases emesis
after tonsillectomy in children. Anesth Analg. 1994;78(3):478–81.il 9. Jones JE, Tabaee A, Glasgold R, Gomillion MC. Efficacy of gastric
aspiration in reducing posttonsillectomy vomiting. Arch Otolar
yngol Head Neck Surg. 2001;127(8):980–4. 32. Nygren J, Thorell A, Ljungqvist O. Are there any benefits from
minimizing fasting and optimization of nutrition and fluid
management for patients undergoing day surgery? Curr Opin
Anaesth. 2007;20(6):540–4. 10. Bergendahl HT, Lonnqvist PA, Eksborg S, Ruthstrom E, Norden
berg L, Zetterqvist H, et al. Clonidine vs. midazolam as premedi
cation in children undergoing adeno-tonsillectomy: a prospec
tive, randomized, controlled clinical trial. Acta Anaesth Scand. 2004;48(10):1292–300.i 33. Seyedhejazi M, Jabbari Moghaddam Y, Rezazade Jodi M, Rahimi
Panahi J, Bilajani E, Ghojazade M, et al. Comparison of intrave
nous fentanyl and infiltration of bupivacaine and clonidine in
decreasing post tonsillectomy pain and complications in chil
dren. J Pharm Sci. 2012;18(2):141–9.f 11. Segal N, Puterman M, Rotem E, Niv A, Kaplan D, Kraus M, et al. A
prospective randomized double-blind trial of fibrin glue for re
ducing pain and bleeding after tonsillectomy. Int J Pediatr Otorhi. 2008;72(4):469–73. 34. Klemetti S, Kinnunen I, Suominen T, Antila H, Vahlberg T, Gren
man R, et al. The effect of preoperative fasting on postoperative
pain, nausea and vomiting in pediatric ambulatory tonsillec
tomy. Int J Pediatr Otorhi. 2009;73(2):263–73. 12. Colreavy MP, Nanan D, Benamer M, Donnelly M, Blaney AW,
O’Dwyer TP, et al. Antibiotic prophylaxis post-tonsillectomy: is it
of benefit? Int J pediatr otorhi. 1999;50(1):15–22. Anesth Pain Med. 2014;4(1):e10256 5
|
https://openalex.org/W2234292234
|
https://publications.aston.ac.uk/id/eprint/40848/1/Atkins2016_Article_SimulatedConsultationsASocioli.pdf
|
English
| null |
Simulated consultations: a sociolinguistic perspective
|
BMC medical education
| 2,016
|
cc-by
| 9,437
|
* Correspondence: Sarah.Atkins@nottingham.ac.uk
1Centre for Research in Applied Linguistics, Trent Building, University of
Nottingham, Nottingham NG7 2RD, UK
Full list of author information is available at the end of the article Sarah Atkins1*, Celia Roberts2, Kamila Hawthorne3 and Trisha Greenhalgh4 Sarah Atkins1*, Celia Roberts2, Kamila Hawthorne3 and Trisha Greenhalgh4 © 2016 Atkins et al. Open Access This article is distributed under the terms of the Creative Commons Attribution 4.0
International License (http://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and
reproduction in any medium, provided you give appropriate credit to the original author(s) and the source, provide a link to
the Creative Commons license, and indicate if changes were made. The Creative Commons Public Domain Dedication waiver
(http://creativecommons.org/publicdomain/zero/1.0/) applies to the data made available in this article, unless otherwise stated. Atkins et al. BMC Medical Education (2016) 16:16
DOI 10.1186/s12909-016-0535-2 Atkins et al. BMC Medical Education (2016) 16:16
DOI 10.1186/s12909-016-0535-2 Open Access Abstract Background: Assessment of consulting skills using simulated patients is widespread in medical education. Most
research into such assessment is sited in a statistical paradigm that focuses on psychometric properties or
replicability of such tests. Equally important, but less researched, is the question of how far consultations with
simulated patients reflect real clinical encounters – for which sociolinguistics, defined as the study of language in its
socio-cultural context, provides a helpful analytic lens. Discussion: In this debate article, we draw on a detailed empirical study of assessed role-plays, involving
sociolinguistic analysis of talk in OSCE interactions. We consider critically the evidence for the simulated
consultation (a) as a proxy for the real; (b) as performance; (c) as a context for assessing talk; and (d) as potentially
disadvantaging candidates trained overseas. Talk is always a performance in context, especially in professional
situations (such as the consultation) and institutional ones (the assessment of professional skills and competence). Candidates who can handle the social and linguistic complexities of the artificial context of assessed role-plays
score highly – yet what is being assessed is not real professional communication, but the ability to voice a credible
appearance of such communication. Summary: Fidelity may not be the primary objective of simulation for medical training, where it enables the
practising of skills. However the linguistic problems and differences that arise from interacting in artificial settings
are of considerable importance in assessment, where we must be sure that the exam construct adequately
embodies the skills expected for real-life practice. The reproducibility of assessed simulations should not be
confused with their validity. Sociolinguistic analysis of simulations in various professional contexts has identified
evidence for the gap between real interactions and assessed role-plays. The contextual conditions of the simulated
consultation both expect and reward a particular interactional style. Whilst simulation undoubtedly has a place in
formative learning for professional communication, the simulated consultation may distort assessment of
professional communication These sociolinguistic findings contribute to the on-going critique of simulations in
high-stakes assessments and indicate that further research, which steps outside psychometric approaches, is
necessary. Keywords: Simulated consultations, OSCE, Communication skills, Interpersonal skills, Assessment, Diversity,
Sociolinguistics © 2016 Atkins et al. * Correspondence: Sarah.Atkins@nottingham.ac.uk
1Centre for Research in Applied Linguistics, Trent Building, University of
Nottingham, Nottingham NG7 2RD, UK
Full list of author information is available at the end of the article
© 2016 Atkins et al. Open Access This article is distributed under the terms of the Creative Commons Attribution 4.0
International License (http://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and
reproduction in any medium, provided you give appropriate credit to the original author(s) and the source, provide a link to
the Creative Commons license, and indicate if changes were made. The Creative Commons Public Domain Dedication waiver
(http://creativecommons.org/publicdomain/zero/1.0/) applies to the data made available in this article, unless otherwise stated. Background However, given that such exams can
demonstrate significant differences in pass rates between
demographic groups, such as, the Membership of the Royal
College of Physicians ‘Practical Assessment of Clinical
Examination Skills’ (PACES) exams and Membership of the
Royal College of General Practitioners’ Clinical Skills
Assessment (CSA) [9–11], the construct validity of simula-
tions is an important research question. Recreating a lin-
guistically authentic medical interaction may not be the
primary objective of simulation when used for medical
training, where it enables practice and focussing in on par-
ticular skills, particularly technical skills. Recent research
evidence has also shown the important uses of simulation
in communication skills training for medical teams, not
least because of the facility it provides practitioners to rec-
ord and reflect on how an interaction has unfolded [12]. However, the linguistic differences and difficulties of simu-
lation are of considerable importance in assessment. In
many high-stakes summative exams, the simulated consult-
ation is not used to enable the medical practitioner to re-
flect on and develop their own communication skills, but
rather for an external party to measure a candidate’s com-
petence and assign a grade. When using the simulated con-
sultation for assessment, particularly where interpersonal
and communication skills are being marked, we must be
sure that the exam construct and the linguistic require-
ments placed on candidates adequately embody the skills
expected for real-life practice. Sociolinguistic analysis can unpick the subtly different
interactional discrepancies that a simulation produces. It looks at direct evidence from the interactions pro-
duced in situ, rather than relying on asking participants
to reflect back on an interaction. Sociolinguistics often
employs techniques of ‘discourse analysis’ to identify
important features and fine-grained characteristics of
talk, such as grammatical structure, turn-taking be-
tween speakers, intonation and the integration of non-
verbal communication [22]. These can systematically
evidence the characteristic linguistic features that occur
in a particular setting - and how the talk is created by,
as well as constitutive of, the social relationships in that
context. Sociolinguistics is therefore interested in the
choices that speakers make when they use language and
what those choices and variations might mean for the
evaluation of speakers (page 16) [23]. The authenticity of the interaction is particularly per-
tinent when assessing communication skills. The clinical
consultation is not a technical procedure, but an emo-
tionally charged interpersonal interaction of high social
significance and linguistic complexity [13]. Background language and society inter-relate. It looks at how people
use language in their everyday lives and how language-
in-context creates the complex social world. The tools of
sociolinguistic research are real recordings of spoken
language, to examine evidence of how different contexts
and social backgrounds affect the talk we produce and
how it is evaluated by others. This focus on evidence
from actual interactions is an important one here. Prior
research on simulated consultations has largely addressed
psychometric properties of particular tests and scenarios
such as internal consistency (e.g. using Cronbach’s alpha),
generalisability, inter-rater reliability, predictability (e.g. of
subsequent examination success), discriminatory power
(ability to distinguish consistently between ‘good’ and
‘poor’ examinees), as well as protocols and procedures for
quality control [3, 14–18]. A crucial question remains as
to how far such scenarios reflect real consulting abilities. Both simulated patients and those being assessed, when
asked after the event, tend to rate their experience as ‘real-
istic’ [19, 20]. Various studies have also found that un-
announced simulated patients, trained to present with a
particular scenario in a kind of ‘mystery shopper’ ap-
proach, went undetected by medical practitioners [6, 21]. Yet there remain crucial differences between real and sim-
ulated consultations, particularly as they are used in as-
sessment, that cannot easily be evidenced by the mystery
shopper method or participants’ retrospective accounts. This paper addresses issues arising from the use of simu-
lated patients in assessments of clinical consulting, in
particular the linguistic difficulties of interacting in such
settings and how far they reflect a practitioner’s real con-
sulting abilities. Simulated patients are lay people or pro-
fessional actors trained to portray a patient with a
particular condition in a standardised way. As well as
their use in practice and training for medical practi-
tioners, they play an important role in formal assess-
ment,
such
as
the
objective
structured
clinical
examination (OSCE) for undergraduates [1, 2] and licen-
sure examinations for postgraduates [3–5]. One advan-
tage of this assessment format is that it helps ensure
everyone has a standardised, equitable and repeatable
experience [6–8]. Abstract Open Access This article is distributed under the terms of the Creative Commons Attribution 4.0
International License (http://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and
reproduction in any medium, provided you give appropriate credit to the original author(s) and the source, provide a link to
the Creative Commons license, and indicate if changes were made. The Creative Commons Public Domain Dedication waiver
(http://creativecommons.org/publicdomain/zero/1.0/) applies to the data made available in this article, unless otherwise stated. Atkins et al. BMC Medical Education (2016) 16:16 Page 2 of 9 Atkins et al. BMC Medical Education (2016) 16:16 Page 2 of 9 Background When con-
sidering the appropriateness of simulation in assessing
this complexity, we need a different kind of research -
sited in a humanistic rather than psychometric para-
digm. Sociolinguistic research provides a useful means
of interrogating and debating these issues. Sociolinguis-
tics is a field which systematically studies the way Integrated with this theoretical approach in sociolinguis-
tics, is a fundamental interest in how language produces
power relations in society. Professional discourses in par-
ticular, which largely consist of goal-oriented encounters,
often demonstrate a degree of power-imbalance [24]. We
can evidence these power relationships in the ‘microphys-
ics’ and fine-grained detail of our everyday practices [25],
including our professional talk. There are asymmetrical
relationships in terms of who is expected to speak at cer-
tain points, who should show politeness and which Atkins et al. BMC Medical Education (2016) 16:16 Page 3 of 9 Page 3 of 9 speakers are meant to demonstrate their domain-specific,
professional knowledge [26]. Power relations in the setting
of an exam role-play differ from real-life clinical encoun-
ter, since the role-player has a very different position to
that of a patient and there is also a more powerful third
participant in the examiner, observing the interaction. Given the different relationships of participants, inter-
actional differences in simulated consultations compared
to real-life are perhaps to be expected. The environment
in assessed simulations is (intentionally) decontextualized
and scenarios involving invasive physical examinations [5]
and a range of patient groups such as those with multi-
morbidity, limited English or a very different communica-
tive style [27] are often (though not invariably) excluded. Yet candidates are asked to behave as if these scenarios
were real [28, 29]. degree of standardisation of any scenario (hence, its replic-
ability) and its reflection of the real (its authenticity), since
100 % standardisation would require the simulated patient
to reproduce a script robotically. In reality, while the sim-
ulated patient plays a character and helps depict a context-
ual hinterland in answer to candidates’ questions, he or
she must draw on their own interactional resources to
manage the interaction itself [28]. To understand how simulations are experienced differ-
ently from real consultations, we must ask, “what ‘main-
tenance work’ needs to be done by both parties to
maintain the semblance of reality?”. Background To do this, we draw
on the work of sociologist Erving Goffman, whose sem-
inal essay ‘Frame Analysis’ addressed the question,
“under what circumstances do we think things are real?”
[39]. Goffman argued that the sense of feeling an activity
is real depends upon our sense of self as we relate to
others. Each interaction creates and reinforces a shared
reality to keep the relationship going [40]. We attend to
others, become involved in the to and fro of talk, how-
ever momentarily, since we have what Goffman calls a
moral requirement to display ourselves in ways that
others expect of us. This
sociolinguistic
approach
to
language
and
professional communication was recently used in a 3-year
study of the Royal College of General Practitioners’
Clinical Skills Assessment (CSA) [27, 30], which we draw
upon in this paper. As well as this we draw on analytic
work around final-year undergraduate medical OSCEs by
Roberts et al. [31]. De La Croix and Skelton on under-
graduate OSCEs [32–34], Seale et al.’s linguistic study of
OSCE examinations [28], Niements on the use of role-
play in training [35], and O’Grady and Candlin on the
Royal Australian College of General Practitioners' licens-
ing exam [36]. Some linguistic evidence on the use of
simulation for assessing professional communication out-
side the medical field is also drawn upon, particularly Sto-
koe’s [37, 38] account of the linguistic patterns in
simulated police interviews. Evaluating this evidence col-
lectively helps us flesh out this debate paper with a fuller
picture of the complexities of using simulation. Although
simulated consultations and OSCE exams do vary in their
setup, some of the essential commonalities, like the semi-
scripted standardised part of role-players, the timed cases
and marking descriptors, render our sociolinguistic dis-
cussion relevant to this wide genre of assessment. The
paper addresses the following four themes, identified from
the literature review and our own research, in considering
the simulated consultation: (a) as a proxy for the real; (b)
as performance; (c) as a context for assessing talk and (d)
as potentially disadvantaging candidates trained overseas. p
His concept of ‘frame’ describes this socially defined
reality. In any given stage of an encounter, speakers and
listeners establish or negotiate what is going on: we are
in the frame of a passing conversation, a preliminary
chat about the weather before the consultation proper
begins, an examination and so on. Background The frame constitutes
what is happening and also works as a filtering process
through which general principles of conduct apply. For
example, when a doctor tells a patient that chances of
recovery are high, both sides can understand that they
are in a ‘reassurance’ frame within this shared moment
of reality. Different frames can be invoked, and indeed
evidenced, through changes in linguistic behaviour by
the participants. For example, in a case from our CSA
research [27], a simulated patient presents with menor-
rhagia. The candidate indicates that he wants to do a
"quick abdominal examination", in the frame of a rou-
tine element of the diagnosis. But when he responds to
the simulated patient’s query by saying “we look for any
abnormal
growth”,
the
simulated
patient
becomes
alarmed. The candidate then shifts the frame from infor-
mation giving to one of reassurance and self-correction. Discussion At any time in an encounter, Goffman argues (page
156–200) [39], we can experience multiple frames. For
example, in an OSCE-style exam, the frame of showing
empathy to a role-playing patient is nested in a frame of
displaying competence to an examiner, which in turn is
nested in the institutional frame of the overall assessment
process. For this reason, the values associated with em-
pathy are not seriously committed to or felt as real be-
cause they are anchored in a more fundamental frame, The simulated consultation as a proxy for the real
Social interaction cannot ultimately be standardised. While there are some relatively stable, overarching fea-
tures of a consultation, such as the phases to be per-
formed and the general history to be conveyed by the
patient, at the minute level of turn-by-turn talk stand-
ardisation becomes difficult and small differences in de-
livery are inevitable. There is a tension, then, between the Atkins et al. BMC Medical Education (2016) 16:16 Page 4 of 9 Page 4 of 9 related to simulated performance in the exam. While role-
player and trainee/candidate can put on a surface per-
formance that is realistic, the assessor must decide
whether the candidate is demonstrating ‘real caring’. This
makes any simulated consultation a hybrid activity in
which real qualities (subjectively experienced) are assessed
through the unreal, requiring a considerable amount of
interactional work to sustain the talk and illusion of a real
consultation [28]. roles and identities at play in simulations and these can
be evidenced in the communication. In answering the
question on ‘authenticity’, Seale et al. ultimately do sug-
gest that experience of participants in a role-play is fun-
damentally different from that of a real-life interaction
and that the candidate must do much more inter-
actional ‘work’ to keep the illusion up (page 181). Of course, real consultations also require some level of
performance, but to properly understand the differences
we must unpack what ‘acting’ and ‘performing’ mean in
these interactional situations. To do so, we can draw on
Goffman’s depiction of life as drama – i.e. we present our-
selves on the world as a stage, ‘performing’ in different
ways to different ‘audiences’ in different settings (everyday,
professional, institutional and so on) [43]. We perform all
the time in the everyday, managing impressions of our-
selves in what Goffman called ‘facework’ [44]. Discussion Goffman calls an activity that does not fit within the
frame of the moment a ‘frame-break’. For example,
candidates in simulated consultations often do not
know whether they are expected to carry out a physical
examination ‘for real’. They may commence a physical
examination frame, only to be interrupted by the
examiner either verbally or by handing them a card
with key physical findings. Candidates must then rap-
idly shift frame to the preliminaries of diagnosis. We
found such shifts were typically marked by disfluencies
and/or hesitations, even with highly successful candi-
dates, as the candidate worked to maintain the simu-
lated
case
and
ignore
any
interaction
with
the
examiner (page 53) [27]. To justify a simulated con-
sultation as a proxy for the real obscures its limitations
and complexities, many of which only become appar-
ent when analysing their interactional detail. It is in
this linguistic detail of simulations that we can really
identify the different communicative competences that
come to the fore in simulated consultations, which
may not be the competences required for real-life
practice. Goffman distinguished the banal and intimate per-
formances of the everyday that occur ‘backstage’ from
professional behaviour, which is largely ‘front-stage’
[43] – a term he used to refer to activities like the
waiter at table, the doctor in the surgery or the teacher
in class. Here there are constraints on behaviour in
terms of manner, quality of attention and emotions,
and the performance has an ‘audience’ that evaluate
the competence displayed [45, 46]. Importantly, under-
standing professional behaviour as a performance does
not undercut its values. For example, to care for a pa-
tient may involve masking frustration or fatigue in
order to care better. When institutions require this
professional behaviour to be monitored and assessed,
however, it becomes an institutional performance. Evaluation of professional performance becomes insti-
tutionalised as observers rate and record performance
and implement rewards and sanctions. There is a
heightened awareness of the need, on the part of the
professional, to perform expressively, a “heightened
mimicry” [28] and, on the part of the assessor, “a license
… to regard the act of expression and the performer
with special intensity” (page 11) [46]. However, it is im-
portant to make the distinction between a heightened
performance for institutional purposes (e.g. Discussion someone
pointedly looking in the mirror when taking a driving
test) and a simulated performance (someone pretending
to look in the mirror). The simulated consultation as performance p
One of the concerns voiced about OSCE examina-
tions is that they test acting skills as much as they do
professional communication [30, 41]. Niements de-
scribes how role-played interactions "cannot reproduce
the orientations of real interactions...[W]hat is authentic
to those users when they “live” a specific situation can-
not be authentic to trainers/trainees when they play it"
[42] (p. 317). A number of studies have addressed the
types of ‘acted’ behaviour such settings consequently
produce, what de la Croix and Skelton have called "the
language game of role-play" [32]. Seale et al. explore
how different ‘frames’, real and fictitious, are invoked
through talk in simulations [28]. They find subtle mo-
ments in which attention is drawn to the fictitious na-
ture of role-play, citing an example of humorous
comments made about an entirely invented paediatric
patient, that both the candidate and the role-player are
pretending is present (page 183). In their analysis, using
the fictitious nature of role-play to create humour is a
means for the candidate to achieve rapport with the
actor, not rapport with a ‘patient’. So there are multiple In simulation, the environment is mutually con-
structed as an unreal activity. In her analysis of emotions
in theatre acting, Konijin discusses the way actors must
monitor how far the emotions they are acting out accord
with the inner model of what the play should convey
[47]. The actor’s task is not to convey sincere emotions
but to play out words and actions that convince the
audience of the authenticity of their character within the
terms of the drama. At the same time they monitor their Page 5 of 9 Atkins et al. BMC Medical Education (2016) 16:16 simultaneous amplification and reduction is most appar-
ent in the interpersonal domains of assessment, as we
discuss below. own experience of acting and so experience a ‘dual con-
sciousness’. In a simulation, likewise, the trainee or candi-
date has to work hard to create a synthetic reality – one
that convinces the audience/observer, but not one that is
real to candidates in terms of consequences for patients:
an institutionalised display rather than a professional in-
vestment, all the while monitoring their conduct vis-à-vis
the examiner. In sum, simulation is a multi-layered per-
formance for both role-player and candidate requiring
some of the skills of an actor. The design and timing of cases The design of cases for simulated consultations moves the
focus from the how of patient care to the why of the par-
ticular selected case. Both students and candidates are
primed to fear the trip-wire that comes with the case:
“learners sometimes think there are hidden aspects…they
are being asked to discover, akin to peeling away the skins
of an onion until the flesh is found” (page 67) [52]. The simulated consultation as a context for assessing talk
The design of OSCE-style exams bring five other com-
plexities, relating to the quality of talk to the candidate’s
task, adding burdens and reducing the ‘real’. We con-
sider: (i) the talk-heavy nature of the consultations; (ii)
the design and timing of cases; (iii) the shift of power to
the role-player; (iv) standardised scenarios but individual
emotional responses and (v) who fails such assessments
– and why? We establish these themes from the authors’
study of the CSA [27] and from an overview of the lin-
guistic research on simulations [28, 32–37], but draw on
these findings to debate the particular implications for
assessment. g
In a high-stakes examination, this ‘Sherlock Holmes’
factor can mar or make success [27]. It turns the candi-
date into a timekeeper, dealing with concerns superfi-
cially so that the putative puzzle of the case can be
resolved. They may stop in mid-sentence when the whis-
tle blows or pack in questions or information as the last
minute ticks by. The strictly timed structure for simu-
lated consultations produces very different openings and
closings from that identified in real consultations. For
example, in real clinical encounters, doctors raising new
topics at the likely end of the encounter is rare but, con-
versely,
closings
are
often
extended
conversational
exchanges which build the doctor-patient relationship
more generally [53]. The talk-heavy nature of the consultations In simulated consultations, it is primarily talk-in-interaction
that is assessed. To succeed in simulated scenarios, candi-
dates must work harder or ‘over perform’, holding a higher
proportion of the conversational floor (between 67–77 %)
than in everyday consultations [34]. Research by Seale et al. identifies the complex, additional linguistic work required
from candidates in simulations [28] and research on the
Royal Australian College of General Practitioners’ licensing
examination identified how role-played scenarios require a
complex, hybrid discourse from the GP candidate [36]. Col-
lectively, these findings suggest that simulated consultations
require actions and skills to be verbalised by the candidate
to a much greater degree than in everyday clinical work. Who fails such assessments – and why? We have noted the heavy focus on talk in simulated con-
sultations. Communication or interpersonal skills are
often explicitly assessed with their own marking criteria
in medical OSCEs, but can also become implicitly
judged across all other domains, since professional ac-
tions like data-gathering and clinical management must
also be performed through effective communication
[27]. The metric of reliability also tends to reinforce the
unspoken assumption that there is an implicit ‘best way’
of scoring highly in the interpersonal domain. Candidates
in simulated consultations routinely produce formulaic
phrases such as "Can you tell me a bit more about…" "I
understand how you feel" or "I’m sorry to hear about that",
with a greater frequency and often in different sequential
positions, than is found in real life practice [27]. Some of
these mimic the phrases recommended in communication
skills textbooks and their extensive use in simulations may
be inevitable in an environment in which talk is being ob-
served and assessed. This is a finding corroborated by
Roberts et al. [31] in a study of undergraduate medical
OSCEs, where the use of elicitation phrases such as "How
do you feel about that?" could be interpreted as sounding
overly trained if used in the wrong location (page 8–9). In
an essay on the experience of being a role-player evalu-
ating candidates in US medical exams, Jamison points
out that to gain marks, empathy and compassion must
be ‘voiced’ and that (perhaps as a consequence) candi-
dates seemed either aggressively formulaic in their in-
sistence, "that must be really hard", or saturated with
humility "Would you mind if I – listened to your heart?"
(page 4–5) [56]. There have been similar findings on
simulations in professional settings outside medicine,
such as Stokoe’s research on police interview role-plays,
in which communication directives from training man-
uals are overtly used in the openings, in a way which
they are not in real-life police interviews, potentially for
the benefit of a marker [38]. Standardised scenarios but individual emotional responses
The role-player’s power is made more complex by the
shift from the institutional persona of the actor/patient
to the instinctive resources of the private person. In
other words, the role-player works with a hybrid of act-
ing behaviour and their own, individual interactional re-
sources. While careful training is used to standardise
‘patients’, the role-player is usually not working to a
tightly scripted part. Who fails such assessments – and why? He or she is given guidance to react
to the candidate in a natural way, to fulfil interactional
criteria (Table 1). If the candidate’s performance is un-
clear or irritating to the role-player, then the role-player
can respond in accord with their inner emotions (the ir-
ritation feels real, even though the setting is simulated). g
g
All parties in fact, must draw on their own inter-
actional resources to make sense of the encounter. Even
where a middle class actor acts a convincingly troubled
and inarticulate teenager, they cannot gainsay their own
interpretive processes (e.g. they can mumble or remain
silent but they cannot not understand). Examiners not
only have to judge this hybrid of simulation and instinct-
ive resources, they also have to manage their own mix of
instinctive reactions to how others interact, their own
professional expectations and the formal categories of
the examination. As one said, “A lot of the time, I am
comparing them to me and what I’m used to” [27]. This
mix of habits of talk, interpretation and evaluation (on
the one hand) and standardised judgements (on the
other) are most problematic in the domain of interper-
sonal skills, where subjective interpretation is necessary
to interpret what counts as ‘rapport’ or ‘sensitivity’. This It seems to be a consequence of the assessed, simulated
setting then, that participants use these formulaic, trained
professional phrases and interactional moves with a much
higher frequency than real-life. In exams such as the CSA,
high scoring candidates also produce 32 % more of these
exam-modelled utterances than weaker candidates. Yet
these phrases appear much less frequently in real consul-
tations [27]. Interestingly, weaker CSA candidates who
also produce these types of phrases, albeit slightly less fre-
quently, were assessed as formulaic in examiner feedback: Table 1 Examples of instructions to role-player from the CSA:
Behaviour/Demeanour/Body language
‘Case 1’
Reticent, trying to appear unconcerned. A bit resentful if the doctor appears to be telling you off. ‘Case 2’
Try to build a good rapport with doctor and don’t keep information back. You are familiar with GPs and hospitals, so you are comfortable with
the doctor. Table 1 Examples of instructions to role-player from the CSA:
Behaviour/Demeanour/Body language Try to build a good rapport with doctor and don’t keep information back. You are familiar with GPs and hospitals, so you are comfortable with
the doctor. The shift of power to the role-player account of the conversational inauthenticities of role-play
for police interviews [37], particularly the more elaborate
and sometimes humorous way in which conversational ac-
tions are performed in these false settings, where the
stakes for participants are entirely different from those
where a real defendant is being interviewed. There is lin-
guistic evidence, then, for how participants must orient
themselves in acted, simulated settings, monitoring their
performance and conducting extra linguistic work to
maintain the illusion of a real interaction. account of the conversational inauthenticities of role-play
for police interviews [37], particularly the more elaborate
and sometimes humorous way in which conversational ac-
tions are performed in these false settings, where the
stakes for participants are entirely different from those
where a real defendant is being interviewed. There is lin-
guistic evidence, then, for how participants must orient
themselves in acted, simulated settings, monitoring their
performance and conducting extra linguistic work to
maintain the illusion of a real interaction. Who fails such assessments – and why? The shift of power to the role-player The shift of power to the role player
Sociolinguistic research has identified how asymmetrical
interactions, where one speaker has more power than
another, show small-scale differences in talk. Medical
consultations are necessarily asymmetrical. The move-
ment in recent years towards patient-centredness and
shared decision making has not fundamentally altered
this, since asymmetry stems at least partly from the
doctor’s knowledge [54]. But in simulated consultations,
candidates must manage the fact that “the power rela-
tion is inverted, because knowledge and judgment rest
with the simulated patient rather than with the physician
student” (page 266) [55]. De la Croix and Skelton iden-
tify a higher number of interruptions from role-players
across 100 third-year OSCE exams, suggesting a position
of greater interactional power compared to findings on
the linguistic behaviour of real-life patients [32–34]. Not
only do the simulated patients know the case and how it
should play out [34], but in examiner feedback sessions
for research on simulations [27], examiners noted that
the simulated patients positioned themselves in an
actorly manner. They put demands on candidates that
patients usually would not and showed familiarity with
the exigencies of the case through their language [27]. Talk in simulated assessments is also relatively decon-
textualized, without the shaping role of the computer
[27, 48] or any of the other props and interruptions of
real consultations. Decontextualised environments incur
more talk [49] and, in an environment such as this, lead
to talk becoming intensely focussed on. In addition,
there is no continuity of care, so shared, unspoken
knowledge between doctor and patient can play no part. This potentially diminishes the types of relationships
and interactions that can be experienced by doctor and
patient in the simulated consultation. Relationship build-
ing over time and the deep values inherent in building
professional capability [50, 51] are overshadowed by an
externally timed case where surface skills must be made
explicit (e.g. enacted or voiced) for assessment. This Evidence from outside medical education has also
shown the shift in power relations between speakers when
interacting in simulations. Stokoe conducts a nuanced Atkins et al. BMC Medical Education (2016) 16:16 Page 6 of 9 became particularly evident in feedback sessions with
examiners, as the following section explores. Who fails such assessments – and why? It seems just very formulaic and a lot of it seems
learned. ‘I understand why you would be worried’, Atkins et al. BMC Medical Education (2016) 16:16 Page 7 of 9 Page 7 of 9 Atkins et al. BMC Medical Education (2016) 16:16 As indicated above, simulations lead to more talk,
more formulaic phrases and more work to ensure that
such talk sounds sincere. This focus on talk and how it
sounds in contexts of intense assessment puts particular
pressure on those whose style of communicating is dif-
ferent from the majority of examiners and also, perhaps,
the patient role-players. Small differences in such subtle
features as intonation, word stress and other small
markers of speech can be amplified and read off as
showing negative characteristics, such as formulaic re-
sponses or not engaging, attracting lower marks in the
interpersonal (page 32–73) [27]. Additionally, since it
can be difficult to make standardised, simulated cases re-
flect the same variation as real-life consulting, perform-
ance will not reflect the ability to interact effectively and
flexibly with a diverse patient population. In many such
exams, while UK graduates are not assessed on consult-
ing in linguistically challenging situations, International
Medical Graduates, many of whom consult regularly in
another expert language within the British multi-cultural
context, have no opportunity to display this skill as they
might use it in their everyday practice. Such competence
in linguistically and culturally challenging situations is
increasingly important for medical practitioners treating
diverse patient populations, both in the UK and globally. It is perhaps the biggest challenge for assessing medical
practitioners’ interpersonal competence in our modern-
day context of globalised, mobile and diverse societies. ‘What kind of thought went through your mind when
you made this appointment’ which kind of is an
attempt to do the right thing but to me it just felt
very crass… [27] ‘What kind of thought went through your mind when
you made this appointment’ which kind of is an
attempt to do the right thing but to me it just felt
very crass… [27] Detailed analysis of stronger candidates’ talk showed
that they knew how to play the game: they customised
formulaic phrases so they sounded more real and sin-
cere, adding in little hesitations, colloquialisms and
changes in intonation (page 59–61) [27]. Summary A review of sociolinguistic approaches to simulations
demonstrates that simulated assessment, even when it is
‘realistic’, shows some crucial differences to the communi-
cative competences found in real-life practice. Talk is al-
ways a performance in context and in simulations, the
role-playing patient, the candidate and the examiner all
have to work hard to maintain the illusion. Candidates
who can handle the social and linguistic complexity of this
somewhat artificial, standardised situation score highly –
yet what is being assessed is not real communication but
the ability to voice a credible appearance of such commu-
nication. It follows that if communication skills are
assessed purely through simulated patients, this may not
reflect the real consulting abilities of candidates. We must
question whether simulations replace the values–led de-
velopment of medical students with ‘playing the game’ of
simulation [50, 51, 59]. The ability of doctors to form en-
during therapeutic relationships with patients may not be
adequately reflected in the “colonisation [of medicine] by
the technologies of the unreal” [60]. Who fails such assessments – and why? In such cir-
cumstances, the ‘empathy telling’, already a simulation of
feeling and perception, has to be further worked on to
invoke a convincing suspension of disbelief: a double
simulation or to extend Konijin’s concept of ‘dual con-
sciousness’ [47], a ‘triple consciousness’, consisting of the
candidate’s own sense of themselves as professionals, the
consciousness that they must simulate a professional en-
counter – and in addition within the institutional frame
– must work on the formulaic phrases of the simulation
so that they sound more sincere to examiners. It is in
these small details of talk, here in the small variations in
delivery of exam-modelled phrases, that we can see how
power and social relationships are constituted in the
micropractices of interaction and its evaluation [25]. In terms of construct validity, does the simulated con-
sultation measure what it purports: the interpersonal
capabilities expected of a doctor? The answer is a com-
plex one. Though the simulation may be good at testing
skills such as giving explanations and structuring the
consultation, there are a number of linguistic features
which do not mimic real-life practice. For example, it
examines competence in using additional communica-
tive resources to make exam-induced ‘voiced’ phrases
sound sincere and to manage the triple consciousness
required to perform to examiners. In terms of assess-
ment theory, there is a "construct-irrelevant variance"
[57] in which certain know-how is assessed which is not
a requirement of real consultations. Received: 18 September 2015 Accepted: 6 January 2016 Received: 18 September 2015 Accepted: 6 January 2016 Received: 18 September 2015 Accepted: 6 January 2016 Acknowledgements
Th
h
f The authors are grateful to the research funders who facilitated the work
with the Royal College of General Practitioners, which is referred to in this
paper. This included a Knowledge Transfer Partnership award (KTP008346,
2011–2013), from the Technology Strategy Board and the Academy of
Medical Royal Colleges in the United Kingdom. SA was additionally funded
by an Economic and Social Research Council ‘Future Research Leaders’ grant
at the University of Nottingham (ES/K00865X/1, 2013–2016). The authors are grateful to the research funders who facilitated the work
with the Royal College of General Practitioners, which is referred to in this
paper. This included a Knowledge Transfer Partnership award (KTP008346,
2011–2013), from the Technology Strategy Board and the Academy of
Medical Royal Colleges in the United Kingdom. SA was additionally funded
by an Economic and Social Research Council ‘Future Research Leaders’ grant
at the University of Nottingham (ES/K00865X/1, 2013–2016). We are also grateful to the Royal College of General Practitioners for the
access and close advice they gave the authors throughout the original
research, on which this debate article is built, and to all the exam candidates
who gave their consent to be part of the study. We are also grateful to the Royal College of General Practitioners for the
access and close advice they gave the authors throughout the original
research, on which this debate article is built, and to all the exam candidates
who gave their consent to be part of the study. Simulated consultations as potentially disadvantaging
candidates trained overseas While all assessments may require ‘exam skills’ to some
degree, when one group of candidates fares much worse
than another, as occurs in many of these assessments [27],
the fairness of these exam-constructed requirements
needs to be carefully considered. There is wide recogni-
tion that for many candidates trained outside their home
country for the assessment, simulations are often a new
phenomenon and that, like any type of assessment, lack of
familiarity affects performance [58]. The simple solution
offered is that this group need more practice with simula-
tions. However, detailed sociolinguistic analysis suggests
that simulations may cause difficulties for this group of
candidates in other ways as well. The discipline of sociolinguistics offers an evidenced
approach to these questions around professional com-
munication. In this paper, we have introduced three core
sociolinguistic concepts relevant to the assessment of Atkins et al. BMC Medical Education (2016) 16:16 Page 8 of 9 Page 8 of 9 Page 8 of 9 communication in medicine: that the particular variety
of talk in simulated consultations separates it out from
the talk in real consultations; that the notion of ‘frame’
is used to understand how we relate to and make our
talk real to each other and that this reality breaks down
in institutionally assessed communication; and that
micro-features of talk feed constantly into our evaluation
of others and, in high-stakes assessments, can have large
consequences on the trajectory of an interaction. While
a single awkward moment is unlikely to lead to failure,
in settings of intense evaluation, perceived infelicities
such as an unfilled pause or formulaic phrase become
amplified. The cumulative effect of such micro-features
may lead to a candidate being judged as "not developing
rapport" or as showing inadequate responsiveness to
"verbal and non-verbal cues" and an overall negative im-
pression of interpersonal abilities. as an adviser on the original project with the Royal College of General
Practitioners. TG has conducted extensive research in the field of medical
education and has drawn on this in substantially rewriting the initial and
subsequent versions of this article. All authors contributed to conceptualizing
and writing the paper, sourcing material. All authors have seen and
approved the final manuscript. Competing interests
KH
i
f KH was an examiner for the Royal College of General Practitioners’ (RCGP) Clinical
Skills Assessment during the data-collection and analytic phases of the research
cited in this article. The authors have no other competing interests to declare. KH was an examiner for the Royal College of General Practitioners’ (RCGP) Clinical
Skills Assessment during the data-collection and analytic phases of the research
cited in this article. The authors have no other competing interests to declare. 11. McManus IC, Wakeford R. PLAB and UK graduates' performance on MRCP(UK)
and MRCGP examinations: data linkage study. Br Med J. 2014;348:g2621. 12. Korkiakangas T, Weldon S-M, Bezemer J, Kneebone R. Video-Supported
Simulation for Interactions in the Operating Theatre (ViSIOT). Clin Simul
Nurs. 2015;11(4):203–7. Abbreviations
CSA
li i
l kill CSA: clinical skills assessment; MRCGP: Examinations for ‘Membership of the
Royal College of General Practitioners’; OSCE: objective structured clinical
examination; RCGP: Royal College of General Practitioners. 9. Dewhurst NG, McManus C, Mollon J, Dacre JE, Vale AJ. Performance in the
MRCP(UK) Examination 2003–4: analysis of pass rates of UK graduates in
relation to self-declared ethnicity and gender. BMC Med. 2007;5:8. 10. McManus IC, Elder AT, Dacre J. Investigating possible ethnicity and sex bias
in clinical examiners: an analysis of data from the MRCP(UK) PACES and
nPACES examinations. BMC Med Educ. 2013;13:103. References 1. Khan KZ, Gaunt K, Ramachandran S, Pushkar P. The Objective Structured
Clinical Examination (OSCE): AMEE Guide No. 81. Part II: Organisation &
Administration. Med Teach. 2013;35(9):e1447–63. 1. Khan KZ, Gaunt K, Ramachandran S, Pushkar P. The Objective Structured
Clinical Examination (OSCE): AMEE Guide No. 81. Part II: Organisation &
Administration. Med Teach. 2013;35(9):e1447–63. 2. Khan KZ, Ramachandran S, Gaunt K, Pushkar P. The Objective Structured
Clinical Examination (OSCE): AMEE Guide No. 81. Part I: An historical and
theoretical perspective. Med Teach. 2013;35(9):e1437–46. 2. Khan KZ, Ramachandran S, Gaunt K, Pushkar P. The Objective Structured
Clinical Examination (OSCE): AMEE Guide No. 81. Part I: An historical and
theoretical perspective. Med Teach. 2013;35(9):e1437–46. 3. Swanson DB, van der Vleuten CP. Assessment of clinical skills with
standardized patients: state of the art revisited. Teaching and Learning in
Medicine. 2013;25(sup1):S17–25. 3. Swanson DB, van der Vleuten CP. Assessment of clinical skills with
standardized patients: state of the art revisited. Teaching and Learning in
Medicine. 2013;25(sup1):S17–25. 4. First LR, Chaudhry HJ, Melnick DE. Quality, cost, and value of clinical skills
assessment. N Engl J Med. 2013;368(10):963–4. 4. First LR, Chaudhry HJ, Melnick DE. Quality, cost, and value of clinical skills
assessment. N Engl J Med. 2013;368(10):963–4. 5. Boulet JR, Smee SM, Dillon GF, Gimpel JR. The use of standardized patient
assessments for certification and licensure decisions. Simul Healthc. 2009;
4(1):35–42. 5. Boulet JR, Smee SM, Dillon GF, Gimpel JR. The use of standardized patient
assessments for certification and licensure decisions. Simul Healthc. 2009;
4(1):35–42. 6. Cleland JA, Abe K, Rethans J-J. The use of simulated patients in medical
education: AMEE Guide No 42 1. Med Teach. 2009;31(6):477–86. 6. Cleland JA, Abe K, Rethans J-J. The use of simulated patients in medical
education: AMEE Guide No 42 1. Med Teach. 2009;31(6):477–86. 7. Holmboe ES, Ward DS, Reznick RK, Katsufrakis PJ, Leslie KM, Patel VL, et al. Faculty development in assessment: the missing link in competency-based
medical education. Acad Med. 2011;86(4):460–7. 7. Holmboe ES, Ward DS, Reznick RK, Katsufrakis PJ, Leslie KM, Patel VL, et al. Faculty development in assessment: the missing link in competency-based
medical education. Acad Med. 2011;86(4):460–7. 8. Gormley G, Sterling M, Menary A, McKeown G. Keeping it real! Enhancing
realism in standardised patient OSCE stations. Clin Teach. 2012;9(6):382–6. 8. Gormley G, Sterling M, Menary A, McKeown G. Keeping it real! Enhancing
realism in standardised patient OSCE stations. Clin Teach. 2012;9(6):382–6. 13.
Sarangi S. Healthcare interaction as an expert communicative system. New
Advent Lang Interac. 2010;196:167. 11.
McManus IC, Wakeford R. PLAB and UK graduates' performance on MRCP(UK)
and MRCGP examinations: data linkage study. Br Med J. 2014;348:g2621. 12.
Korkiakangas T, Weldon S-M, Bezemer J, Kneebone R. Video-Supported
Simulation for Interactions in the Operating Theatre (ViSIOT). Clin Simul
Nurs. 2015;11(4):203–7. Author details
1 1Centre for Research in Applied Linguistics, Trent Building, University of
Nottingham, Nottingham NG7 2RD, UK. 2Department of Education &
Professional Studies, King’s College London, Franklin-Wilkins Building,
Waterloo Road, London SE1 9NH, UK. 3Duke of Kent Building, Faculty of
Health and Medical Sciences, University of Surrey, Surrey GU2 7XH, UK. 4Nuffield Department of Primary Care Health Sciences, University of Oxford Although a number of studies have identified that sim-
ulated interactions show important differences from
real-life professional communication [27, 28, 33–37], we
are not arguing that simulation has no place in teaching
or assessment. Much of medical practice consists of
skills that are more or less technical in nature and which
can be both taught and assessed effectively using simu-
lated patients (the rationale behind the ‘skills lab’) [61]. Formative simulated consultations have great value in
the safety they afford learners to make and learn from
mistakes, as well as to ‘slow down’ the consultation to
study what has happened. Summative simulated assess-
ments, however, must carefully consider the difficulties
of assessing interpersonal skills in this setting. Hence, we
do not seek to bury the OSCE, but in introducing the
sociolinguistic perspective, we do seek to debate its level
of validity for assessing communicative and interactional
aspects of clinical performance. Furthermore, we believe
the evidence identified in a number sociolinguistic studies
of simulated interaction [27, 30, 58] requires us to con-
sider carefully what we mean by ‘fairness’ in assessment
and how we might better assess communication skills in
settings of cultural and linguistic diversity. 4Nuffield Department of Primary Care Health Sciences, University of Oxford,
Oxford OX2 6GG, UK. 10.
McManus IC, Elder AT, Dacre J. Investigating possible ethnicity and sex bias
in clinical examiners: an analysis of data from the MRCP(UK) PACES and
nPACES examinations. BMC Med Educ. 2013;13:103. 6.
Cleland JA, Abe K, Rethans J-J. The use of simulated patients in medical
education: AMEE Guide No 42 1. Med Teach. 2009;31(6):477–86. CSA: clinical skills assessment; MRCGP: Examinations for ‘Membership of the
Royal College of General Practitioners’; OSCE: objective structured clinical
examination; RCGP: Royal College of General Practitioners. 8.
Gormley G, Sterling M, Menary A, McKeown G. Keeping it real! Enhancing
realism in standardised patient OSCE stations. Clin Teach. 2012;9(6):382–6. 7.
Holmboe ES, Ward DS, Reznick RK, Katsufrakis PJ, Leslie KM, Patel VL, et al.
Faculty development in assessment: the missing link in competency-based
medical education. Acad Med. 2011;86(4):460–7. 14.
Brannick MT, Erol-Korkmaz HT, Prewett M. A systematic review of the
reliability of objective structured clinical examination scores. Med Educ.
2011;45(12):1181–9. Authors’ contributions
S
ll
h Bosse HM, Nickel M, Huwendiek S, Jünger J, Schultz JH, Nikendei C. Peer
role-play and standardised patients in communication training: a
comparative study on the student perspective on acceptability, realism, and
perceived effect. BMC Med Educ. 2010;10(1):27. 46. Bauman R. Verbal art as performance. Prospect Heights, Illinois: Waveland
Press; 1984. 47. Konijn E. Acting emotions: shaping emotions on stage. Amsterdam:
Amsterdam University Press; 2000. 47. Konijn E. Acting emotions: shaping
Amsterdam University Press; 2000. 21. Kinnersley P, Ben-Shlomo Y, Hawthorne K, Donovan J, Chaturvedi N. The
acceptability and practicality of simulated patients for studying general
practice consultations in Britain. Educ Prim Care. 2005;16:540–6. 48. Swinglehurst D, Greenhalgh T, Roberts C. Computer templates in chronic
disease management: ethnographic case study in general practice. BMJ
Open. 2012;2(6). 22. Gumperz J. On interactional sociolinguistic method. In: Sarangi S, Roberts C,
editors. Talk, work and institutional order: Discourse in medical, mediation
and management settings. New York: Mouton de Gruyter; 1999. p. 453–71. 49. Levelt W. Speaking: from intention to articulation. Cambridge: MIT Press;
1989. 23. Holmes J. An Introduction to Sociolinguistics. London: Longman; 1992. 50. Bleakley A. ‘Good’ and ‘poor’ communication in an OSCE: education or
training? Med Educ. 2003;37(3):186–7. 24. Holmes J, Marra M. The Routledge Handbook of Language and
Professional Communication. In: Vajay B, Stephen B, editors. The
Routledge Handbook of Language and Professional Communication. Abingdon: Routledge; 2014. p. 112–26. 51. Fraser SW, Greenhalgh T. Coping with complexity: educating for capability. Br Med J. 2001;323(7316):799–803. 52. Kurtz SM, Silverman JD. The Calgary—Cambridge Referenced Observation
Guides: an aid to defining the curriculum and organizing the teaching in
communication training programmes. Med Educ. 1996;30(2):83–9. 25. Foucault M. Discipline and Punish: The Birth of the Prison. Harmondsworth:
Penguin; 1979. 26. Morand DA. Language and power: an empirical analysis of linguistic
strategies used in superior-subordinate communication. J Organ Behav. 2000;21(3):235–48. 53. West C. Co-ordinating closings in primary care visits: producing continuity
of care. In: Communication in medical care: Interaction between primary
care physicians and patients. Volume 20, edn. Edited by Heritage J, Maynard
DW. Cambridge: Cambridge University Press; 2006. 27. Roberts C, Atkins S, Hawthorne K. Performance features in clinical skills
assessment: Linguistic and cultural factors in the Membership exam in the
Royal College of General Practitioners. London: King's College London with
the University of Nottingham; 2014. DW. Cambridge: Cambridge University Press; 2006. 54. Peräkylä A. Communicating and responding to diagnosis. Authors’ contributions
S
ll
h SA was originally the Research Associate who conducted the analytic work
on the Clinical Skills Assessment described in this article and drafted the first
version of this debate article. CR was originally the Principle Investigator on
the research project with the Royal College of General Practitioners and
substantially contributed to the first draft and subsequent versions of this
article. KH contributed a significant amount of analytic work and discussion 14. Brannick MT, Erol-Korkmaz HT, Prewett M. A systematic review of the
reliability of objective structured clinical examination scores. Med Educ. 2011;45(12):1181–9. 14. Brannick MT, Erol-Korkmaz HT, Prewett M. A systematic review of the
reliability of objective structured clinical examination scores. Med Educ. 2011;45(12):1181–9. Page 9 of 9 Atkins et al. BMC Medical Education (2016) 16:16 Page 9 of 9 15. Ilgen JS, Ma IW, Hatala R, Cook DA. A systematic review of validity evidence
for checklists versus global rating scales in simulation‐based assessment. Med Educ. 2015;49(2):161–73. analysis: Changing institutional practices. Basingstoke: Palgrave
Macmillan; 2011. p. 119–39. analysis: Changing institutional practices. Basingstoke: Palgrave
Macmillan; 2011. p. 119–39. 39. Goffman E. Frame analysis: an essay on the organization of experience. New
York: Harper and Row; 1974. 16. Lievens F, Sackett PR. The validity of interpersonal skills assessment via
situational judgment tests for predicting academic success and job
performance. J Appl Psychol. 2012;97(2):460. 40. Goffman E. Interaction ritual: essays in face to face behavior. New York:
Doubleday; 1967. 41. Harrison S. How do you make a medical student feel stupid
latex breasts and silicone ears. In: The Guardian. 2008. 17. Nestel D, Tabak D, Tierney T, Layat-Burn C, Robb A, Clark S, et al. Key
challenges in simulated patient programs: An international comparative
case study. BMC Med Educ. 2011;11(1):69. latex breasts and silicone ears. In: The Guardian. 2008. 42. Niemants NSA. From Role-Playing to Role-Taking: Interpreter’s Role(s) in
Healthcare. In: From Role-Playing to Role-Taking: Interpreter’s Role(s) in
Healthcare. Amsterdam/Philadelphia: John Benjamins; 2013. p. 305–19. 18. Miller GE. The assessment of clinical skills/competence/performance. Acad
Med. 1990;65(9):S63–7. 43. Goffman E. The presentation of self in everyday life. 1959. 19. Allen J, Rashid A. What determines competence within a general practice
consultation? Assessment of consultation skills using simulated surgeries. Br J Gen Pract. 1998;48(430):1259–62. 44. Goffman E. On face-work: An analysis of ritual elements in social interaction. Psychiatry. 1955;18(3):213–31. 45. Drew P, Heritage J. Analyzing talk at work: An introduction. London: Able
Books; 1992. 20. Authors’ contributions
S
ll
h In:
Communication in medical care: Interaction between primary care
physicians and patients. 2006. p. 214–47. 28. Seale C, Butler CC, Hutchby I, Kinnersley P, Rollnick S. Negotiating frame
ambiguity: A study of simulated encounters in medical education. Commun
Med. 2007;4(2):177–87. 55. Hanna M, Fins JJ. Power and communication: why simulation training
ought to be complemented by experiential and humanist learning. Acad
Med. 2006;81(3):265–70. 29. Sanci L, Day N, Coffey C, Patton G, Bowes G. Simulations in evaluation of
training: a medical example using standardised patients. Eval Program
Plann. 2002;25(1):35–46. 56. Jamison L. The empathy exams: essays. Minnesota: Graywolf Press; 2014. 56. Jamison L. The empathy exams: essays. Minnesota: Graywolf Press; 2014. 57. Haladyna TM, Downing SM. Construct‐irrelevant variance in high‐stakes
testing. Educ Meas. 2004;23(1):17–27. 57. Haladyna TM, Downing SM. Construct‐irrelevant variance in high‐stakes
testing. Educ Meas. 2004;23(1):17–27. 30. Mohanna K. Exploring the Royal College of General Practitioners' Clinical
Skills Assessment (unpublished thesis in partial completion of Ed D). London: University College London Institute of Education; 2011. 58. Esmail A, Roberts C. Independent review of the membership of the Royal
College of General Practitioners (MRCGP) examination. Gen Med Counc. 2013:1–44. http://www.gmc-uk.org/MRCGP_Final_Report__18th_
September_2013.pdf_53516840.pdf. 58. Esmail A, Roberts C. Independent review of the membership of the Royal
College of General Practitioners (MRCGP) examination. Gen Med Counc. 2013:1–44. http://www.gmc-uk.org/MRCGP_Final_Report__18th_
September_2013.pdf_53516840.pdf. 31. Roberts C, Wass V, Jones R, Sarangi S, Gillett A. A discourse analysis study of
'good' and 'poor' communication in an OSCE: a proposed new framework
for teaching students. Med Educ. 2003;37(3):192–201. 59. Skelton JR. Everything you were afraid to ask about communication skills. Br J Gen Pract. 2005;55(510):40–6. 59. Skelton JR. Everything you were afraid to ask about communication skills. Br J Gen Pract. 2005;55(510):40–6. 32. de la Croix A. The language game of role-play: an analysis of assessed
consultations between third year medical students and Simulated Patients
(SPs). Birmingham: University of Birmingham; 2010. 60. Greenhalgh T. Future-proofing relationship-based care: a priority for general
practice. Br J Gen Pract. 2014;64(628):580. 61. Ziv A, Ben-David S, Ziv M. Simulation based medical education: an
opportunity to learn from errors. Med Teach. 2005;27(3):193–9. 33. de la Croix A, Skelton J. The simulation game: an analysis of
interactions between students and simulated patients. Med Educ. 2013;
47(1):49–58. 34. de la Croix A, Skelton J. The reality of role‐play: interruptions and amount of
talk in simulated consultations. Med Educ. 2009;43(7):695–703. Submit your next manuscript to BioMed Central
and we will help you at every step: • We accept pre-submission inquiries
• Our selector tool helps you to find the most relevant journal
• We provide round the clock customer support
• Convenient online submission
• Thorough peer review
• Inclusion in PubMed and all major indexing services
• Maximum visibility for your research
Submit your manuscript at
www.biomedcentral.com/submit
Submit your next manuscript to BioMed Central
and we will help you at every step: • We accept pre-submission inquiries
• Our selector tool helps you to find the most relevant journal
• We provide round the clock customer support
• Convenient online submission
• Thorough peer review
• Inclusion in PubMed and all major indexing services
• Maximum visibility for your research
Submit your manuscript at
www.biomedcentral.com/submit
y
p
and we will help you at every step: 35. Niements N. From Role-Playing to Role-Taking: Interpreter’s Role(s) in
Healthcare. In: Schäffner C, Fowler Y, Kredens K, editors. Interpreting in a
changing landscape: selected papers from critical link. Amsterdam/
Philadelphia: John Benjamins; 2013. p. 305–19. • We accept pre-submission inquiries 36. O'Grady C, Candlin CN. Engendering trust in a multiparty consultation
involving an adolescent patient. In: Candlin C, Crichton J, editors. Discourses
of trust. London: Palgrave Macmillan; 2013. p. 52–69. 36. O'Grady C, Candlin CN. Engendering trust in a multiparty consultation
involving an adolescent patient. In: Candlin C, Crichton J, editors. Discourses
of trust. London: Palgrave Macmillan; 2013. p. 52–69. 37. Stokoe E. The (in) authenticity of simulated talk: comparing role-played and
actual interaction and the implications for communication training. Res
Lang Soc Interact. 2013;46(2):165–85. 37. Stokoe E. The (in) authenticity of simulated talk: comparing role-played and
actual interaction and the implications for communication training. Res
Lang Soc Interact. 2013;46(2):165–85. 38. Stokoe E. Simulated interaction and communication skills training: The
“Conversation Analytic Role-play Method. In: Applied conversation 38. Stokoe E. Simulated interaction and communication skills training: The
“Conversation Analytic Role-play Method. In: Applied conversation
|
https://openalex.org/W4362621544
|
https://aacr.figshare.com/articles/journal_contribution/Supplementary_Table_S1_from_Combined_Pan-RAF_and_MEK_Inhibition_Overcomes_Multiple_Resistance_Mechanisms_to_Selective_RAF_Inhibitors/22500829/1/files/39960106.pdf
|
English
| null |
Supplementary Table S1 from Combined Pan-RAF and MEK Inhibition Overcomes Multiple Resistance Mechanisms to Selective RAF Inhibitors
| null | 2,023
|
cc-by
| 2,639
|
Table S1. RNAi gene enrichment (RIGER) analysis of the primary screen. The number of reads per shRNA was determined by Illumina sequencing and
normalized to the total number of reads per sample. The data were then log2
transformed. RIGER
analysis
was
performed
in
Gene-E
(http://www.broadinstitute.org/cancer/software/GENE-E/) as previously described [14]. Table S1. RNAi gene enrichment (RIGER) analysis of the primary screen. Gene
# Hairpins
NES
Gene rank
p-value
MET
30
0.00082
1
0.0001
MLL
5
0.00089
2
0.0001
PTPN11
5
0.00099
3
0.0001
CPA4
5
0.0062
4
0.0001
CSTF3
5
0.0064
5
0.0001
BCL2L12
5
0.0074
6
0.0001
CACNG5
5
0.0076
7
0.0001
SHOC2
10
0.0085
8
0.0004
RAF1
10
0.01
9
0.0004
CPN1
5
0.012
10
0.0003
VIP
5
0.013
11
0.0007
SORD
5
0.014
12
0.0008
ADAMTS12
5
0.014
13
0.0008
PPARA
5
0.015
14
0.001
ZFAND1
5
0.015
15
0.001
DLGAP5
5
0.016
16
0.0011
SIRT5
5
0.016
17
0.0011
CSAG2
3
0.017
18
0.0004
BAP1
5
0.018
19
0.0012
XCL2
5
0.018
20
0.0012
FAM65C
5
0.019
21
0.0012
ITK
9
0.019
22
0.0017
LOC139542
5
0.019
23
0.0012
FBXO11
5
0.02
24
0.0012
SLC39A9
5
0.02
25
0.0013
USP47
5
0.02
26
0.0013
PNMA5
5
0.021
27
0.0013
ZNF141
5
0.021
28
0.0013
THNSL2
10
0.021
29
0.0022
PIK3R5
10
0.021
30
0.0022
PXMP2
4
0.022
31
0.0014
ARID1B
4
0.023
32
0.0014
C13orf15
10
0.023
33
0.0025
SLC38A10
5
0.024
34
0.0018
ZNF641
5
0.024
35
0.0019
CPD
5
0.025
36
0.0019
CCDC58
5
0.025
37
0.0019
FBXL2
5
0.025
38
0.002
KCNH6
5
0.026
39
0.0021
MCHR1
5
0.026
40
0.0021
CD59
5
0.026
41
0.0021
GFRA2
5
0.026
42
0.0022
GTF2H1
5
0.027
43
0.0023
POLK
5
0.027
44
0.0023 Gene
# Hairpins
NES
Gene rank
p-value
EDF1
5
0.027
45
0.0023
TOP1
5
0.028
46
0.0025
CRKL
5
0.028
47
0.0025
ODAM
10
0.029
48
0.0039
AASDHPPT
5
0.029
49
0.0025
SOX17
3
0.029
50
0.001
GPR6
3
0.029
51
0.001
PDZK1IP1
5
0.029
52
0.0025
HDAC3
10
0.029
53
0.004
RNF13
5
0.029
54
0.0025
RAC1
5
0.03
55
0.0026
DDX3X
5
0.031
56
0.0028
NOX1
5
0.031
57
0.0028
LOC402434
5
0.031
58
0.0028
ATP5S
5
0.031
59
0.0028
GPR44
5
0.032
60
0.0029
MPPED1
5
0.032
61
0.0029
IFITM2
4
0.032
62
0.0033
SPA17
4
0.032
63
0.0033
GINS2
4
0.033
64
0.0034
ATG2B
4
0.033
65
0.0034
INPP5A
5
0.034
66
0.0035
COL8A2
5
0.034
67
0.0035
C20orf70
4
0.034
68
0.0037
GFP
24
0.034
69
0.0059
LOC440945
5
0.034
70
0.0035
SYBL1
5
0.034
71
0.0035
TMEM45A
5
0.035
72
0.0037
CD58
5
0.035
73
0.0038
AFMID
5
0.035
74
0.0038
PPFIA1
5
0.035
75
0.0038
ASB6
9
0.035
76
0.0059
PF4V1
5
0.035
77
0.0038
SNAPC3
5
0.036
78
0.0039
WNT4
4
0.036
79
0.0039
SNAP23
10
0.036
80
0.0065
PLAA
4
0.036
81
0.0039
IFT172
5
0.036
82
0.0043
SPANXN1
5
0.037
83
0.0043
WNT11
3
0.037
84
0.0021
C10orf137
5
0.037
85
0.0043
ELL
5
0.037
86
0.0043
PRG2
10
0.037
87
0.0071
GDPD1
5
0.039
88
0.0048 Gene
# Hairpins
NES
Gene rank
p-value
CEACAM7
5
0.039
89
0.0049
OPCML
5
0.039
90
0.0049
GNG7
8
0.04
91
0.0062
PIK3C2G
9
0.04
92
0.0083
MXRA8
5
0.04
93
0.0051
CCDC82
5
0.041
94
0.0051
OPTC
10
0.041
95
0.0082
BTNL8
5
0.041
96
0.0053
GTSE1
5
0.041
97
0.0054
ANTXR1
5
0.042
98
0.0054
FUBP1
5
0.042
99
0.0057
MARK3
9
0.042
100
0.0087
SQSTM1
5
0.042
101
0.0058
EEF1A1
5
0.043
102
0.0059
CST5
5
0.043
103
0.0059
TMEM31
5
0.043
104
0.006
CCNT1
5
0.043
105
0.0061
CD2
5
0.044
106
0.0061
HLA-DPB1
5
0.044
107
0.0061
ITGAV
5
0.044
108
0.0061
TSPAN4
5
0.044
109
0.0062
GMPPA
9
0.044
110
0.0098
C12orf44
5
0.044
111
0.0064
WBSCR23
5
0.045
112
0.0064
CBR3
5
0.045
113
0.0064
SIVA1
4
0.045
114
0.0065
tAKR
5
0.045
115
0.0065
KHK
8
0.045
116
0.0083
RBMS3
5
0.045
117
0.0065
TRA2A
5
0.045
118
0.0065
ZC3H13
5
0.046
119
0.0065
CXYorf3
5
0.046
120
0.0065
WIRE
5
0.046
121
0.0066
SAP130
5
0.046
122
0.0066
STRADA
4
0.046
123
0.0069
MOBKL2C
5
0.046
124
0.0067
SLC17A2
5
0.047
125
0.0067
MAP2K2
10
0.047
126
0.01
RALGPS1
5
0.047
127
0.0067
PLK2
10
0.047
128
0.011
TTPAL
4
0.047
129
0.0072
CYP4B1
5
0.048
130
0.0069
ATG4C
5
0.048
131
0.0072
BRAP
5
0.048
132
0.0072 Gene
# Hairpins
NES
Gene rank
p-value
BTNL2
5
0.048
133
0.0072
EFNB2
5
0.048
134
0.0072
SFTPA2B
7
0.049
135
0.0095
RIOK1
9
0.049
136
0.012
TLX2
5
0.049
137
0.0073
NCKAP1
5
0.049
138
0.0074
DMRT3
4
0.049
139
0.0074
C11orf16
5
0.05
140
0.0077
MCRS1
10
0.05
141
0.011
BMP6
5
0.05
142
0.0079
SLC35D1
5
0.05
143
0.008
OR2L13
5
0.05
144
0.008
C19orf10
5
0.05
145
0.008
RAI14
5
0.051
146
0.008
ADAMTS5
5
0.051
147
0.0081
CSTF2T
10
0.051
148
0.012
DNAJB12
5
0.051
149
0.0082
IGFBP1
7
0.051
150
0.011
PRC1
10
0.051
151
0.012
C10orf116
4
0.052
152
0.0082
LOC163223
5
0.052
153
0.0085
PRIMA1
10
0.052
154
0.012
ARHGAP29
5
0.053
155
0.0086
ENTPD3
5
0.053
156
0.0086
LTA
5
0.053
157
0.0086
CLCN7
5
0.053
158
0.0086
PDK2
9
0.053
159
0.014
PDE2A
5
0.053
160
0.0086
RAB24
5
0.053
161
0.0086
LRRTM4
5
0.053
162
0.0086
ENG
5
0.054
163
0.0086
SEC63D1
10
0.054
164
0.014
PTPN6
5
0.054
165
0.0088
FLJ35834
4
0.054
166
0.0092
CRYBA1
5
0.054
167
0.0089
FAM116A
5
0.055
168
0.0089
LIN7B
4
0.055
169
0.0092
TJP2
5
0.055
170
0.0089
POLRMT
5
0.055
171
0.0089
ZNF431
5
0.055
172
0.0089
TMEM219
9
0.055
173
0.015
SS18
5
0.056
174
0.0093
SSBP3
5
0.056
175
0.0093
DDX21
5
0.056
176
0.0093 Gene
# Hairpins
NES
Gene rank
p-value
TGM1
5
0.056
177
0.0093
SFRS11
5
0.057
178
0.0093
RTDR1
5
0.057
179
0.0094
PIGX
5
0.057
180
0.0094
TBR1
5
0.057
181
0.0094
KBTBD2
4
0.057
182
0.01
GNB2
5
0.058
183
0.0097
H2AFJ
5
0.058
184
0.0097
C21orf81
5
0.058
185
0.0097
KIAA0247
5
0.058
186
0.0099
C6orf227
10
0.058
187
0.016
LOC138932
5
0.059
188
0.01
LUZP1
9
0.059
189
0.016
NDRG4
9
0.059
190
0.016
CLUL1
10
0.06
191
0.016
KIAA0101
5
0.06
192
0.011
RGN
4
0.061
193
0.012
DAZ4
9
0.061
194
0.018
SLC2A11
4
0.061
195
0.012
PPCDC
5
0.062
196
0.011
ZC3H3
5
0.062
197
0.011
NRBP2
9
0.062
198
0.018
NARS
5
0.062
199
0.012
ADAMTS19
5
0.062
200
0.012
IGSF10
5
0.062
201
0.012
BRD1
5
0.062
202
0.012
TIMM10
9
0.063
203
0.019
OTOP2
5
0.063
204
0.012
FNTA
5
0.063
205
0.012
ZFX
5
0.063
206
0.012
PPP3CB
5
0.063
207
0.012
C6orf170
5
0.063
208
0.012
LRRC23
4
0.063
209
0.012
OR8U1
5
0.063
210
0.012
FRMD5
5
0.063
211
0.012
MS4A7
3
0.064
212
0.0068
NDFIP2
5
0.064
213
0.012
SHPK
5
0.064
214
0.012
ECM1
5
0.064
215
0.012
BIRC1
5
0.064
216
0.012
PRKDC
10
0.064
217
0.018
MYO1C
5
0.065
218
0.012
ERBB3
16
0.065
219
0.021
MS4A6E
5
0.065
220
0.012 Gene
# Hairpins
NES
Gene rank
p-value
IL17C
5
0.065
221
0.013
FLJ35220
5
0.065
222
0.013
GSPT1
5
0.065
223
0.013
OSGEP
5
0.065
224
0.013
NPM2
10
0.066
225
0.019
LEPREL2
5
0.066
226
0.013
MAD2L1BP
5
0.066
227
0.013
ABCB1
5
0.066
228
0.013
DBN1
5
0.066
229
0.013
MX2
5
0.067
230
0.013
ZNF100
5
0.067
231
0.013
C2orf85
4
0.067
232
0.013
PATZ1
5
0.067
233
0.014
RPS6KA1
14
0.067
234
0.022
RAB28
5
0.068
235
0.014
SSU72
5
0.068
236
0.014
DSCC1
10
0.068
237
0.02
FXN
5
0.068
238
0.014
GAB1
5
0.068
239
0.014
UGT2B17
5
0.068
240
0.014
TPM3
5
0.068
241
0.014
SLC39A14
5
0.068
242
0.014
MYF5
5
0.068
243
0.014
KLRC1
5
0.068
244
0.014
RDH8
5
0.068
245
0.014
HAPLN1
5
0.069
246
0.014
PPIA
4
0.069
247
0.014
SENP5
5
0.069
248
0.014
HMX3
5
0.069
249
0.014
CXorf58
14
0.069
250
0.023
OR5I1
5
0.069
251
0.014
ZNF507
5
0.069
252
0.014
MDFI
5
0.069
253
0.014
SLC30A5
5
0.069
254
0.014
CSNK2B
10
0.069
255
0.021
PELI1
4
0.069
256
0.014
ACR
5
0.07
257
0.014
AURKA
9
0.07
258
0.022
PAPOLA
5
0.07
259
0.014
ASXL1
5
0.07
260
0.014
DNAJC21
5
0.07
261
0.014
BTBD11
10
0.07
262
0.022
WEE1
10
0.071
263
0.022
EDG5
4
0.071
264
0.015 Gene
# Hairpins
NES
Gene rank
p-value
CPA3
5
0.071
265
0.014
SNX9
5
0.071
266
0.014
CSRP1
5
0.071
267
0.014
LACTB
5
0.071
268
0.014
ALS2
5
0.071
269
0.015
ANGPTL7
5
0.072
270
0.015
C6orf35
5
0.072
271
0.015
MARVELD2
5
0.072
272
0.015
SIRPG
5
0.072
273
0.015
RWDD3
5
0.072
274
0.015
SFTPC
5
0.072
275
0.015
LIG3
5
0.073
276
0.015
ERGIC2
5
0.073
277
0.015
GPR113
5
0.073
278
0.015
MTPN
4
0.073
279
0.015
UGT2B10
5
0.073
280
0.015
GAL3ST2
5
0.073
281
0.015
KIF3B
5
0.074
282
0.015
GBE1
5
0.074
283
0.015
ILKAP
5
0.074
284
0.015
ZFYVE1
9
0.074
285
0.024
TP53I3
5
0.074
286
0.016
NTRK2
10
0.074
287
0.024
TMOD1
5
0.074
288
0.016
HOXA11
5
0.074
289
0.016
FGD4
5
0.075
290
0.016
PLA2G7
5
0.075
291
0.016
PVALB
4
0.075
292
0.016
FBXO6
5
0.076
293
0.017
FGF5
5
0.076
294
0.017
ZNRF2
5
0.076
295
0.017
ZNF587
5
0.076
296
0.017
TSPAN8
5
0.076
297
0.017
SPRN
9
0.076
298
0.026
SMS
5
0.076
299
0.017
CCDC55
4
0.076
300
0.017
NFAM1
5
0.076
301
0.017
NSDHL
5
0.076
302
0.017
CLEC4C
5
0.076
303
0.017
GPX8
5
0.077
304
0.017
ADRBK2
10
0.077
305
0.026
GINS1
9
0.077
306
0.026
CTSZ
4
0.077
307
0.017
HSPA1B
5
0.077
308
0.018 Gene
# Hairpins
NES
Gene rank
p-value
METTL6
5
0.077
309
0.018
COL5A1
5
0.077
310
0.018
LOC402280
5
0.078
311
0.018
FBXO4
5
0.078
312
0.018
FLJ39501
5
0.078
313
0.018
IGSF1
5
0.078
314
0.018
OVGP1
4
0.079
315
0.017
LYSMD4
5
0.079
316
0.018
H2-ALPHA
4
0.079
317
0.017
BAAT
5
0.079
318
0.018
PDE1B
5
0.079
319
0.018
DLK1
5
0.079
320
0.018
MID1IP1
10
0.079
321
0.028
FRMD7
5
0.079
322
0.018
CA5B
10
0.079
323
0.028
TFE3
5
0.079
324
0.018
ZXDC
4
0.079
325
0.018
CLTB
5
0.079
326
0.018
NTSR2
5
0.08
327
0.019
AZI1
5
0.08
328
0.019
MGMT
5
0.08
329
0.019
PLD1
5
0.08
330
0.019
SEC14L2
5
0.08
331
0.019
AQP5
4
0.08
332
0.018
ZNF780B
4
0.08
333
0.018
S100A7A
5
0.081
334
0.019
GPR23
5
0.081
335
0.019
TMEM60
5
0.081
336
0.019
C11orf24
4
0.081
337
0.018
PTTG1
5
0.081
338
0.019
CPEB2
5
0.081
339
0.02
GP1BA
5
0.082
340
0.02
NPY5R
5
0.082
341
0.02
DCTN1
5
0.082
342
0.02
CMTM6
5
0.082
343
0.02
CTCFL
5
0.082
344
0.02
POLR3K
4
0.082
345
0.019
RFXANK
5
0.083
346
0.02
ANKLE1
10
0.083
347
0.031
SFRS4
5
0.083
348
0.02
PRDX4
5
0.083
349
0.02
MURC
9
0.083
350
0.03
DPY19L1P1
5
0.083
351
0.02
UBR4
5
0.083
352
0.02 Gene
# Hairpins
NES
Gene rank
p-value
RRP12
5
0.083
353
0.02
SETD7
5
0.083
354
0.02
ADAM11
3
0.083
355
0.013
RPL39
5
0.084
356
0.02
LOC442428
5
0.084
357
0.02
PDZD7
10
0.084
358
0.032
SNF1LK
5
0.084
359
0.02
PTHLH
5
0.084
360
0.02
DKK3
5
0.084
361
0.02
CLDN10
4
0.084
362
0.019
GPR128
4
0.084
363
0.019
SLC26A1
5
0.084
364
0.021
ACCSL
5
0.084
365
0.021
ETV1
5
0.084
366
0.021
LY86
10
0.085
367
0.032
GPR25
5
0.085
368
0.021
RAG2
5
0.085
369
0.021
BPY2
5
0.085
370
0.021
SLAMF6
5
0.085
371
0.021
PKIB
5
0.085
372
0.021
ZMYND11
5
0.085
373
0.021
LOC442313
5
0.086
374
0.021
ARC
5
0.086
375
0.021
SPIC
5
0.086
376
0.022
SYT13
5
0.086
377
0.022
SLC25A44
5
0.087
378
0.022
GPR85
5
0.087
379
0.022
PHKB
10
0.087
380
0.034
KLKB1
5
0.087
381
0.022
TNFAIP8L1
10
0.087
382
0.034
HPCA
5
0.087
383
0.022
MAT2B
5
0.087
384
0.022
PGCP
5
0.088
385
0.022
TMEM217
5
0.088
386
0.022
TMEM14B
5
0.088
387
0.022
HDGF
5
0.088
388
0.023
PEX1
5
0.088
389
0.023
PRDX2
5
0.089
390
0.023
CCL8
5
0.089
391
0.023
GRB2
5
0.089
392
0.023
TMPRSS11B
5
0.089
393
0.023
C9orf139
10
0.09
394
0.035
RANGRF
10
0.09
395
0.035
TRIM34
5
0.09
396
0.023 Gene
# Hairpins
NES
Gene rank
p-value
ZNF37A
5
0.09
397
0.023
GDF9
5
0.09
398
0.023
NUDT14
4
0.09
399
0.021
WBP1
5
0.091
400
0.024
KRT34
5
0.091
401
0.024
MAPRE3
2
0.091
402
0.0064
SLAMF9
5
0.091
403
0.024
NUP155
5
0.091
404
0.024
MLLT1
5
0.091
405
0.024
ESCO2
5
0.091
406
0.024
ETF1
10
0.091
407
0.036
C1orf87
5
0.091
408
0.024
DRD2
5
0.091
409
0.024
ADAL
5
0.091
410
0.024
DDX25
9
0.092
411
0.035
C5orf4
5
0.092
412
0.024
C6orf127
5
0.092
413
0.024
MAGEA2
3
0.092
414
0.016
RASSF2
4
0.092
415
0.021
PSTK
5
0.092
416
0.024
TLL2
5
0.092
417
0.024
HKDC1
5
0.092
418
0.024
ISOC1
10
0.092
419
0.036
ST14
5
0.093
420
0.024
PSMF1
5
0.093
421
0.025
FCGR2B
5
0.093
422
0.025
IMP4
5
0.093
423
0.025
PLK4
19
0.093
424
0.046
ZFYVE21
5
0.093
425
0.025
KAAG1
5
0.093
426
0.025
SORCS3
5
0.093
427
0.025
CEP290
5
0.093
428
0.025
FLJ20273
5
0.094
429
0.025
CTU2
9
0.094
430
0.036
HLA-DRB4
5
0.094
431
0.025
CDH24
5
0.094
432
0.025
ZNF140
5
0.094
433
0.025
P2RY14
3
0.094
434
0.017
SMARCD3
5
0.094
435
0.025
PHF16
5
0.094
436
0.025
SLC35A5
5
0.094
437
0.025
ZC3H7A
9
0.095
438
0.037
LOC441708
4
0.095
439
0.023
LOC345829
5
0.095
440
0.025 Gene
# Hairpins
NES
Gene rank
p-value
COX5B
5
0.095
441
0.025
ACSL6
5
0.095
442
0.025
MAGEA4
10
0.095
443
0.039
HOXB3
5
0.095
444
0.025
GPR27
4
0.095
445
0.023
MAP3K7IP2
5
0.095
446
0.025
CCDC136
5
0.095
447
0.025
FAM192A
5
0.095
448
0.025
OR5T1
5
0.095
449
0.025
ABCC12
5
0.095
450
0.025
BAT4
10
0.095
451
0.04
PHC2
10
0.096
452
0.04
DHRS4L2
5
0.096
453
0.026
PYGO2
5
0.096
454
0.026
USP48
5
0.096
455
0.026
EPDR1
3
0.096
456
0.017
MPN2
4
0.096
457
0.023
HAGHL
5
0.096
458
0.026
GTPBP4
5
0.096
459
0.026
PPIL6
5
0.097
460
0.026
C4orf34
5
0.097
461
0.026
C5orf35
5
0.097
462
0.026
GLE1
5
0.097
463
0.026
GTF3C6
5
0.097
464
0.026
BRE
5
0.097
465
0.026
SPINK2
4
0.097
466
0.024
RGS18
5
0.097
467
0.026
METTL9
5
0.097
468
0.026
ATP2C1
5
0.098
469
0.026
RNASE8
4
0.098
470
0.024
SMARCD2
5
0.098
471
0.026
C9orf103
5
0.098
472
0.026
SLC31A1
5
0.098
473
0.026
C5orf33
5
0.098
474
0.026
TRIM2
5
0.098
475
0.026
DLST
4
0.098
476
0.024
SLC15A3
5
0.098
477
0.026
MGC4172
5
0.098
478
0.026
CNTFR
5
0.098
479
0.026
UGT8
5
0.099
480
0.026
PLAU
9
0.099
481
0.04
TMED8
5
0.099
482
0.026
CDK4
13
0.099
483
0.048
AREG
4
0.099
484
0.025 Gene
# Hairpins
NES
Gene rank
p-value
EPB41L3
5
0.099
485
0.026
KIAA1639
3
0.099
486
0.019
ELP4
5
0.099
487
0.026
KIAA0556
5
0.099
488
0.026
GPR142
5
0.099
489
0.026
ALDH8A1
4
0.099
490
0.025
TFAM
5
0.1
491
0.027
RECK
5
0.1
492
0.027
PSMD8
5
0.1
493
0.027
REG1A
5
0.1
494
0.027
KCNE3
5
0.1
495
0.027
CDR2L
10
0.1
496
0.043
LOC390377
4
0.1
497
0.026
FCGRT
5
0.1
498
0.027
IP6K2
4
0.1
499
0.026
|
https://openalex.org/W2953360276
|
https://europepmc.org/articles/pmc6580143?pdf=render
|
English
| null |
EEG-Based Prediction of Cognitive Load in Intelligence Tests
|
Frontiers in human neuroscience
| 2,019
|
cc-by
| 7,173
|
EEG-Based Prediction of Cognitive
Load in Intelligence Tests Nir Friedman 1,2, Tomer Fekete 2, Kobi Gal 1,3* and Oren Shriki 2,4,5* 1 Department of Software and Information Systems Engineering, Ben-Gurion University of the Negev, Beersheba, Israel,
2 Department of Cognitive and Brain Sciences, Ben-Gurion University of the Negev, Beersheba, Israel, 3 School of Informatics,
University of Edinburgh, Edinburgh, United Kingdom, 4 Department of Computer Science, Ben-Gurion University of the
Negev, Beersheba, Israel, 5 Zlotowski Center for Neuroscience, Ben-Gurion University of the Negev, Beersheba, Israel Measuring and assessing the cognitive load associated with different tasks is crucial for
many applications, from the design of instructional materials to monitoring the mental
well-being of aircraft pilots. The goal of this paper is to utilize EEG to infer the cognitive
workload of subjects during intelligence tests. We chose the well established advanced
progressive matrices test, an ideal framework because it presents problems at increasing
levels of difficulty and has been rigorously validated in past experiments. We train classic
machine learning models using basic EEG measures as well as measures of network
connectivity and signal complexity. Our findings demonstrate that cognitive load can be
well predicted using these features, even for a low number of channels. We show that by
creating an individually tuned neural network for each subject, we can improve prediction
compared to a generic model and that such models are robust to decreasing the number
of available channels as well. Keywords: brain-computer interface, electroencephalography, cognitive load, machine learning, Raven’s matrices Edited by:
Tamer Demiralp,
Istanbul University, Turkey ORIGINAL RESEARCH
published: 11 June 2019
doi: 10.3389/fnhum.2019.00191 Received: 30 December 2018
Accepted: 22 May 2019
Published: 11 June 2019 Received: 30 December 2018
Accepted: 22 May 2019
Published: 11 June 2019 1. INTRODUCTION Reviewed by:
Ahmet Omurtag,
Nottingham Trent University,
United Kingdom
Pietro Aricò,
Sapienza University of Rome, Italy The performance of complex tasks requires the integration of various mental resources,
such as task-related knowledge, working memory, attention and decision making. However,
our
brains
have
limited
resources
for
processing
and
integrating
information. The
concept of cognitive load generally refers to the relative load on these limited resources
(Sweller et al., 1998; Coyne et al., 2009). Cognitive workload has been explored from different perspectives. Brouwer et al. (2012) refer
to workload as the working memory load in an n-back task. Mills et al. (2017) use simple true-
false questions for eliciting low workload and open-ended questions, which require more precise
memory, for eliciting high workload. Other studies have emphasized the role of skill acquisition
in modeling cognitive load (Sweller et al., 1998). Logan (1985) show that when subjects acquire a
skill and learn how to perform a task in an automatic manner, their cognitive workload decreases
(Borghini et al., 2017) . Thus, the cognitive load depends not only on task complexity but also on the
subject’s skill at the given task. A highly complex task performed by a non-skilled individual would
result in high cognitive load, whereas a simple task performed by a skilled individual would result
in low cognitive load. For example, Stevens et al. (2006) assessed subjects as they were learning to
diagnose disorders of organ systems and Mak et al. (2013) focused on performance improvement
in a visual-motor task. Both studies showed a decrease in cognitive load metrics, with an increase
in task familiarity. In all of these studies, the relative difficulty of the task is seen as a proxy for
its associated cognitive load. The difficulty was assessed using a variety of approaches, such as the Citation: Friedman N, Fekete T, Gal K and Shriki
O (2019) EEG-Based Prediction of
Cognitive Load in Intelligence Tests. Front. Hum. Neurosci. 13:191. doi: 10.3389/fnhum.2019.00191 June 2019 | Volume 13 | Article 191 Frontiers in Human Neuroscience | www.frontiersin.org EEG-Based Prediction of Cognitive Load Friedman et al. focused on Lempel-Ziv (Tononi and Edelman, 1998) complexity,
Multi Scale Entropy (MSE) (Abásolo et al., 2006) and Detrended
Fluctuation Analysis (DFA) (Rubin et al., 2013). type of questions (true-false vs. open ended), subject performance
and even participant subjective ratings. A major limitation of
many studies is that the levels of difficulty were not rigorously
defined. Here, we chose a setting in which problem difficulty was
rigorously validated and is commonly used in the psychological
literature (see below). Another limitation of previous studies is
that cognitive workload was assessed using discrete levels, often
only two or three levels (Aricò et al., 2016a,b). In the present
study, we use a continuous scale for workload. focused on Lempel-Ziv (Tononi and Edelman, 1998) complexity,
Multi Scale Entropy (MSE) (Abásolo et al., 2006) and Detrended
Fluctuation Analysis (DFA) (Rubin et al., 2013). The results of this paper demonstrate the applicability of using
EEG and machine learning for quantifying cognitive load in well-
validated problem-solving tasks. In particular, as EEG and other
measures of brain activity become more pervasive, quantitative
cognitive load measures could be used to facilitate the design of
domains involving real-time problem-solving, such as e-learning,
psychometric exams, military training, and more (Ikehara and
Crosby, 2005; Mills et al., 2017). In addition to behavioral measures, there is a growing interest
in assessing cognitive workload using physiological measures,
such as pupil diameter (Palinko et al., 2010). The focus of
this paper is on quantifying cognitive workload using measures
based on electroencephalography (EEG). Several studies have
previously developed EEG-based measures for cognitive load. In
particular, it was found that the ratio between the theta power (4–
8 Hz) and the alpha power (8–12 Hz), as well as the ratio between
the beta power (12–30 Hz) and the alpha power and several
related combinations, provided informative indices concerning
task engagement and cognitive workload (Pope et al., 1995;
Stevens et al., 2006; Mills et al., 2017). Other researchers came
to similar conclusions, namely that the relation between different
spectral features can help predict cognitive load from EEG (Gerˇe
and Jaušcvec, 1999; McDonald and Soussou, 2011; Conrad and
Bliemel, 2016). Citation: This study aimed to further expand these studies
and develop continuous and more accurate EEG-based measures
of cognitive load. Furthermore, we tried to examine the utility
of additional measures, in particular network connectivity and
signal complexity. 2. METHODS We recorded EEG from subjects while they solved the Advanced
Progressive Matrices set II (Raven test). The 36 problems in the
test were presented in increasing levels of difficulty. The raw
EEG data were then passed through an artifact removal pipeline
(see details below) before extracting EEG-based measures of
spectral activity, neural complexity and network connectivity. These measures served as input to machine learning algorithms,
which were trained to predict problem difficulty. 2.1. Participants p
Fifty-two subjects (26 female and 26 male; age range 21–28,
Mean = 24.55 years, SD = 1.76 years) participated voluntarily in
the experiment, provided written informed consent and received
compensation for participating. The experiment was approved
by the Ben-Gurion University ethics committee. All subjects
reported that they are right-handed, have normal or corrected
vision, and that they have never completed any sort of intelligence
test in the past. Four participants were excluded from the study
because they required 10 min or less to solve the entire test or
answered correctly 16 problems or less. An additional participant
was excluded due to a compromised recording (several electrodes
did not record any signal throughout the entire session). We focused on recording EEG during performance of a well-
known psychological assessment tool, the advanced progressive
matrices test (Raven, 2000), which is commonly used to measure
general intelligence. The test is composed of different problems
that involve the manipulation of shapes. Problems are presented
to subjects at increasing levels of difficulty. The difficulty of each
problem is validated across a large number of subjects (Forbes,
1964; Arthur
et al., 1999), in the sense that more difficult
questions lead to a higher error rate in the population. Frontiers in Human Neuroscience | www.frontiersin.org 2.2. Experimental Paradigm II Example Problem. The subject is asked to choose the missing shape from the 8 possible options. The correct answer here is FIGURE 1 | Illustration of Raven’s set II Example Problem. The subject is asked to choose the missing shape from the 8 pos
option 8. FIGURE 1 | Illustration of Raven’s set II Example Problem. The subject is asked to choose the missing shape from the 8 possible options. The correct answer here is
option 8. [12–30 Hz], and gamma [30–50 Hz]) was calculated for each
channel across the whole trial duration. This resulted in 310
features (62 channels × 5 bands) for each trial. (AgCl electrolyte gel). Electrodes were positioned according
to the standard 10/20 system with linked ears reference. An
impedance test and adjustment were carried out at the beginning
of the session, and impedances of all electrodes were kept below
5 k. The signal was sampled at 256 Hz with a high-pass filter
of 1 Hz. The data were recorded using Matlab Simulink g.Tec
plug-ins. • Neural complexity metrics- We focused on three measures of
complexity, specifically, Lempel-Ziv complexity (LZC) (Zhang
et al., 2001), Multi Scale Entropy (MSE) (Abásolo et al., 2006)
and Detrended Fluctuation Analysis (DFA) (Peng et al., 1995;
Rubin et al., 2013). The LZC measure was computed as the
mean of the measure across all channels, resulting in a single
feature for each trial. In comparison, the MSE and DFA
measures were first computed for each individual channel, and
for DFA we also computed the metric for each frequency band
(as described above), and a broadband [1–50Hz]. We then
computed the mean, variance, maximum, minimum,
mean
variance
and maximum
minimum , resulting in 6 features for the MSE, and 36
features for DFA [6 measures × (5 bands + 1 broadband)],
resulting in 43 complexity features for each trial. Because these
metrics are affected by trial duration, we calculated them for
the last 2,500 samples (≈10 s) of each trial. 2.2. Experimental Paradigm g
Subjects performed the Raven’s APM Set II problems (36
items in increasing difficulty level), and instructions were
delivered before the test started (see Figure 1 for an example
problem). The test was run with no time limit, with all
the key requirements and administration instructions carefully
following the manual (Raven et al., 1998). Subjects sat in
a comfortable chair facing a computer screen 60 cm away. The test was conducted by displaying the problems on the
computer screen (23′′, 1,920 × 1,080 resolution, with a 2.3◦
visual angle between each answer’s corners), where the subjects
were required to press a keyboard key (with their right hand) in
accordance with their chosen answer number. The experiment
was programmed in MATLAB ( www.mathworks.com, version
2015), using the Psychophysics Toolbox extensions (Brainard
and Vision, 1997; Pelli, 1997; Kleiner et al., 2007). Each trial lasted
from the presentation of the corresponding problem until subject
response, and thus trial duration was variable. Here, we adopt problem difficulty as the operational definition
of cognitive load and demonstrate that it can be predicted from
the subject’s EEG readings. Specifically, we employ a variety of
EEG measures as input to machine-learning algorithms and train
them to predict problem difficulty. As mentioned above, previous studies of EEG-based measures
of cognitive load were limited in several ways. In particular,
they relied mostly on spectral features and produced a simple
discrete measure of either low or high load. In contrast, this paper
models cognitive load in a continuous manner. In addition, we
go beyond basic spectral features and examine how measures of
network connectivity and signal complexity affect the prediction
of cognitive load. To measure network connectivity, we used
complex network analysis (CNA), which provides measures to
examine functional connectivity in the brain (Bullmore and
Sporns, 2009; Fekete et al., 2014). Features of neural complexity
are often computed using measures of entropy, reflecting the
proportion of ordered patterns that can be detected in a
signal (Bullmore et al., 2009). To measure neural complexity, we EEG was recorded through the whole session using the g.Tec
HIamp system (g.Tec, Austria) with 64 gel-based electrodes June 2019 | Volume 13 | Article 191 2 Friedman et al. EEG-Based Prediction of Cognitive Load FIGURE 1 | Illustration of Raven’s set II Example Problem. The subject is asked to choose the missing shape from the 8 possible options. The correct answer here is
option 8. 2.3. Feature Extraction Data were analyzed using a combination of the EEGLAB Matlab
toolbox (Delorme and Makeig, 2004) routines and custom
code. Data were first high-pass filtered (cut-off1 Hz), then a
customized adaptive filter was applied to suppress line-noise. This was followed by Artifact Subspace Reconstruction (Mullen
et al., 2015), re-referencing to the mean, and low-pass filtering
(cutoff60 Hz). Next, Infomax ICA was carried out (Bell and
Sejnowski, 1995). The resulting ICs were evaluated automatically
for artifacts by combining spatial, spectral and temporal analysis
of ICs. ICs identified as containing ocular, muscular or cardiac
artifacts were removed from data. • Connectivity metrics - These features are based on a graph
reflecting the connectivity of the underlying network. The
graph comprises 62 vertices (channels); edges in the graph
represent correlations between channels (there are no self • Connectivity metrics - These features are based on a graph
reflecting the connectivity of the underlying network. The
graph comprises 62 vertices (channels); edges in the graph
represent correlations between channels (there are no self Various features were extracted from the EEG data: Various features were extracted from the EEG data: • Power spectrum metrics (PS) - The power in 5 frequency
bands (delta [1–4 Hz], theta [4–8 Hz], alpha [8–12 Hz], beta June 2019 | Volume 13 | Article 191 Frontiers in Human Neuroscience | www.frontiersin.org 3 EEG-Based Prediction of Cognitive Load Friedman et al. edges). There are two approaches regarding the weight of each
edge. One is to take the absolute value of the correlation as the
weight of each edge. Another is to give the same weight to all
edges that were kept after the thresholding process described
below. We kept only the top x% (we tried several thresholds)
of the edges with the highest values, for example 5% (which
was what we ultimately used), meaning that we were left
with 190 edges out of the 622 (minus the 62 self edges). The graph was used to extract graph-theoretical features such
as average shortest distance between nodes, small-worldness,
etc. (Bullmore and Sporns, 2009). We ultimately used the
mean and standard deviation of the small-worldness measure
and its components, across the different thresholds. participants for incorrect answers may not reflect the true level of
the question. After excluding the subjects (180 trials) and specific
problems (94 trials), as stated above, and the incorrect trials (366
trials), we were left with 1,232 trials. 2.3. Feature Extraction For each of the 1232 correct trials, we computed different
features (as detailed in the Methods) and assigned them with
the corresponding difficulty level (a number between 1 and 36)
as the target value. Several types of machine learning algorithms
were tested in order to predict cognitive load - ”Random Forest”
(RF) from the sklearn python package (Buitinck et al., 2013) ,
which is a bagging decision-tree based model (Ho, 1995), and
”XGBoost” (XGB) and its corresponding python package (Chen
and Guestrin, 2016). XGB is also a decision tree-based model,
though it comes from the ”boosting” family (Zhou, 2012). They
were chosen because of the virtues of an ensemble learning
algorithm, along with their usual good fit with temporal data. Additionally, we applied an artificial neural network (ANN),
using the keras python package (Chollet et al., 2015). Lastly, we
used simple Linear Regression (LR), also from the sklearn python
package, as a baseline for comparison. The hyper-parameters
of these models were found using a grid search. The best
performance was exhibited by the XGBoost classifier with a
step size of 0.05. For the optimal feature group, the number of
boosting rounds was 300. All other parameters were run with
the default settings. All results shown were cross-validated by
dividing the data randomly to training and validation sets (80%
of the data were used for training, 20% of the data were used for
validation) and repeating the process 10–20 times (determined
by the time complexity of the analysis). Our main measure
of model performance was r2, which is simply the Pearson
correlation squared. It is commonly interpreted as the proportion
of the variance for a dependent variable that is explained by an
independent variable or variables. • Basic - Simple demographic features of subjects’ age and sex
were used. In addition, the time it took to answer each problem
was used as a feature. These features were added to all the
above feature groups in the prediction phase. 3.2. Effect of Discretizing the Workload 3.2. Effect of Discretizing the Workload
In our analysis, the target variable (difficulty of each problem)
had 34 possible values. We analyzed the influence of reducing the
number of levels of the target variable. We used different sized
bins, to reduce the number of different values to 6, 9, 18, 34. For
example, to obtain 6 levels, values were binned to [1–6], [7–12],
[13–18], [19–24], [25–30], and [31–36]. As evident in Figure 5,
prediction quality generally decreased with the number of levels. This is not surprising, because the prediction task becomes more
complex with the number of levels. In addition, we show that
using only the best 12 electrodes found earlier to compute the
connectivity features (combined with the PS features of those
electrodes), we obtain r2 = 0.713 (p < 0.05) for 6 levels, which is
the best prediction quality we obtained. 3. RESULTS Even though ANN can capture very complex
relationships, they require a large training set. On the other hand,
LR and RF do not require significant amounts of training data,
but their model complexity is significantly more constrained
than XGB. FIGURE 3 | This figure shows the Pearson correlation between the XGBoost
model’s prediction and the true label of each instance. The model shown here
uses the PS, connectivity and basic features, which is the one that produced
the best prediction. The equation of the linear fit is y = 1.19x −2.33. conducted a two step analysis. Firstly, we ran 1,000 simulations,
where in each, ten electrodes were chosen randomly out of the
total of 62. For each electrode combination, only the relevant
PS features were used (five per channel, in addition to the
basic features) to generate a workload prediction using the XGB
algorithm. We then sorted the electrodes based on the percentage
of simulations each electrode was involved in that yielded a score
above a specified threshold, out of all simulations it participated
in. The top thirty electrodes were chosen in descending order and
were taken for the second step, where the effect of the number of
best electrodes on the r2 was examined. As seen in Figure 4, a
relatively high r2 of 0.7 (p < 0.01) can be obtained using only 12
electrodes (and in fact over 95 percent of peak performance for
only 8). Additionally, using the same features of the 12 electrodes,
the model produces a Spearman correlation of 0.82 (p < 0.05). These 12 electrodes were: CP1, CPz, CP4, TP8, TP10, P3, P4,
PO7, O1, O2, AF3, FT8. Next, we compared the utility of each of the three different
feature types. PS and connectivity features obtained the highest
score, and adding the complexity features to either of the two did
not contribute significantly to the prediction. This suggests that
complexity features do not add any further information beyond
spectral features and connectivity features. To test whether this
was not due to high model complexity resulting in over-fitting, we
conducted a feature selection process. We found that even after
reducing the number of features, no combination of complexity,
connectivity and PS features yielded better results than using only
the PS and connectivity features together with the basic features. the PS and connectivity features together with the basic features. 3. RESULTS LR
RF
XGB
ANN
PS
0.007
0.383*
0.655*
0.346*
Complexity
0.323*
0.055
0.508*
0.286*
Connectivity
0.335*
0.186
0.5*
0.267*
PS & Complexity
0
0.322*
0.641*
0.186*
PS & Connectivity
0.07
0.44*
0.67*
0.32*
Complexity & Connectivity
0.339*
0.122
0.519*
0.331*
All Features
0.05
0.358*
0.628*
0.297*
summarizes the r2 results for this analysis (all results marked
with an * were significant). The best results were obtained using
XGB for all feature types as seen in a variance test(F(36, 3) =
16.79, p
<
0.001, Tukey multiple comparisons: p
<
0.05
for all XGB pairs). XGB provides a good trade-offbetween
model complexity and the number of samples required to
reach robustness. Even though ANN can capture very complex
relationships, they require a large training set. On the other hand,
LR and RF do not require significant amounts of training data,
but their model complexity is significantly more constrained
than XGB. Next we compared the utility of each of the three different FIGURE 3 | This figure shows the Pearson correlation between the XGBoost
model’s prediction and the true label of each instance. The model shown here
uses the PS, connectivity and basic features, which is the one that produced
the best prediction. The equation of the linear fit is y = 1.19x −2.33. TABLE 1 | The table shows the Pearson correlation (r2) of each Feature group-
Model Pair. summarizes the r2 results for this analysis (all results marked
with an * were significant). The best results were obtained using
XGB for all feature types as seen in a variance test(F(36, 3) =
16.79, p
<
0.001, Tukey multiple comparisons: p
<
0.05
for all XGB pairs). XGB provides a good trade-offbetween
model complexity and the number of samples required to
reach robustness. Even though ANN can capture very complex
relationships, they require a large training set. On the other hand,
LR and RF do not require significant amounts of training data,
but their model complexity is significantly more constrained
than XGB. summarizes the r2 results for this analysis (all results marked
with an * were significant). The best results were obtained using
XGB for all feature types as seen in a variance test(F(36, 3) =
16.79, p
<
0.001, Tukey multiple comparisons: p
<
0.05
for all XGB pairs). XGB provides a good trade-offbetween
model complexity and the number of samples required to
reach robustness. 3. RESULTS Figure 3 shows a scatter plot of the best model’s prediction
together with the true label of each instance in the test set. The
Pearson correlation of the best model is r2 = 0.67 (p < 0.01). The model was trained on the problem serial number, which
should, in principle, produce a linear relationship. However, as
evident in Figure 2, the relationship between problem number
and error rate is slightly non-linear. This suggests that the
relationship between problem number and the EEG measure
could also be non-linear. We therefore also computed the
Spearman correlation, which relates to a general monotonic
relationship rather than a linear one, and obtained a value of 0.81
(p < 0.01). One of the features used by the algorithm was the
duration of each segment, namely the time it took the subject to
answer. We also examined the performance with only this feature
and found a r2 of 0.23 (p < 0.01) and a Spearman correlation of
0.41 (p < 0.01). 3. RESULTS After removal of subjects who did not meet the inclusion criteria
(see Methods), we were left with 47 subjects for the analysis (24
female and 23 male; age range 21–28, Mean = 24.55 years, SD =
1.79 years). Our goal was to estimate the cognitive workload of
subjects as they were trying to solve each problem during the test. To this end, we assumed that the difficulty level increased
with every problem, as validated in previous studies (Forbes,
1964; Arthur et al., 1999). Figure 2 shows the rate of incorrect
responses over all problems in our data, reflecting the established
relationship between problem number and difficulty level. Interestingly, problems 24 and 29 deviated significantly from the
trend (more than 3 standard deviations). For this reason, both
problems were also excluded from our analysis. In addition, we
only considered trials where subjects answered correctly. This
is based on the premise that the cognitive load exhibited by At first, we compared the different feature types in
the prediction process with the different classifiers. Table 1 FIGURE 2 | Subject Error rate as a function of problem number. The mean error rate across subjects is plotted for each problem (circles) together with a quadratic fit
(dark gray curve). The equation corresponding to the fit is y = 0.0004793x2 −0.000897x −0.08256. The color of each point indicates the number of standard
deviations from the fit, with bright colors indicating a higher value. FIGURE 2 | Subject Error rate as a function of problem number. The mean error rate across subjects is plotted for each problem (circles) together with a quadratic fit
(dark gray curve). The equation corresponding to the fit is y = 0.0004793x2 −0.000897x −0.08256. The color of each point indicates the number of standard
deviations from the fit, with bright colors indicating a higher value. June 2019 | Volume 13 | Article 191 Frontiers in Human Neuroscience | www.frontiersin.org 4 EEG-Based Prediction of Cognitive Load Friedman et al. TABLE 1 | The table shows the Pearson correlation (r2) of each Feature group-
Model Pair. 3.3. Individualized Prediction Using Neural
N
k From an applicative point of view, the number of electrodes
affects both the cost and the complexity of using EEG. We
therefore examined the extent to which reducing the number
of electrodes affects the prediction quality. To this end, we Lastly, because different individuals might experience different
levels of cognitive load for the same problem, we wanted to assess June 2019 | Volume 13 | Article 191 Frontiers in Human Neuroscience | www.frontiersin.org 5 EEG-Based Prediction of Cognitive Load Friedman et al. FIGURE 4 | Performance as a function of the number of best channels. Channels were ordered according to their contribution to the prediction quality (see text for
details). The curve depicts the prediction quality (r2) for the XGBoost algorithm as a function of the number of best channels taken into account. FIGURE 5 | Difficulty level discretization effect on prediction quality (r2). Each line corresponds to different feature types. PS red are the PS features of the 12 best
channels. Error bars reflect standard error of the mean. th
i fl
f i di id
li i
th
di ti
d l T
thi
t
i
(M
0 43 SD
0 06)
hi h
i ld d
i
ifi
t FIGURE 4 | Performance as a function of the number of best channels. Channels were ordered according to their contribution to the prediction quality (see text for
details). The curve depicts the prediction quality (r2) for the XGBoost algorithm as a function of the number of best channels taken into account. FIGURE 4 | Performance as a function of the number of best channels. Channels were ordered according to their contribution to the prediction quality (see text for
details). The curve depicts the prediction quality (r2) for the XGBoost algorithm as a function of the number of best channels taken into account. e as a function of the number of best channels. Channels were ordered according to their contribution to the prediction quality (see
ts the prediction quality (r2) for the XGBoost algorithm as a function of the number of best channels taken into account. nction of the number of best channels. Channels were ordered according to their contribution to the prediction quality (see text for
ediction quality (r2) for the XGBoost algorithm as a function of the number of best channels taken into account. FIGURE 5 | Difficulty level discretization effect on prediction quality (r2). 3.3. Individualized Prediction Using Neural
N
k Each line corresponds to different feature types. PS red are the PS features of the 12 best
channels. Error bars reflect standard error of the mean. FIGURE 5 | Difficulty level discretization effect on prediction quality (r2). Each line corresponds to different feature types. PS red are the PS features of the 12 best
channels. Error bars reflect standard error of the mean. tization effect on prediction quality (r2). Each line corresponds to different feature types. PS red are the PS features of the 12 best
ard error of the mean. tuning (M = 0.43, SD = 0.06), which yielded a significant
difference (t = −4.75, p = 0.001) in favor of the individualized
network models. the influence of individualizing the prediction model. To this
end, we first built a three layer artificial neural network (ANN),
trained with data from all subjects using the PS and connectivity
features of the 12 best electrodes. We then fixed the parameters
of the first and second layers, and for each subject continued
to train the weights of the output layer (Figure 6). This is a
common practice in the field of neural networks (Gruber et al.,
2017). We conducted a paired t-test (Figure 7), by calculating
the mean correlation with the correct answer over several folds
using the general model (M = 0.39, SD = 0.06) and after Frontiers in Human Neuroscience | www.frontiersin.org 4. DISCUSSION We recorded EEG from subjects while solving the advanced
progressive matrices test (Raven’s matrices test) and used EEG
features and machine learning to predict problem difficulty,
our chosen operationalization of cognitive workload. Problem June 2019 | Volume 13 | Article 191 Frontiers in Human Neuroscience | www.frontiersin.org 6 Friedman et al. EEG-Based Prediction of Cognitive Load FIGURE 6 | Diagram explaining the architecture of the ANN that was used. There were 2 hidden layers, and all layers were dense (e.g., all connections were present). The parameters between the input layer and hidden layer 1, and the parameters between hidden layer 1 and hidden layer 2 were held during the individualization
phase. FIGURE 6 | Diagram explaining the architecture of the ANN that was used. There were 2 hidden layers, and all layers were dense (e.g., all connections were present). The parameters between the input layer and hidden layer 1, and the parameters between hidden layer 1 and hidden layer 2 were held during the individualization FIGURE 7 | Difference of r2 score using an ANN before and after
individualizing the last layer for each subject. Error bars reflect standard error
of the mean. Paired t-test results are shown. ***p = 0.001. obtained using EEG measures. This could be very useful for
many applications in which there is a wide range of cognitive
workload levels. These findings extend those of previous studies
which used a small number (2-3) of discrete levels of cognitive
workload (Gerˇe and Jaušcvec, 1999; McDonald and Soussou,
2011; Conrad and Bliemel, 2016). Indeed, we found that reducing
the number of difficulty levels improves the results significantly. We examined several machine learning algorithms and found
that XGBoost outperformed all other algorithms with all three
feature groups. XGBoost was more accurate than the simpler
models of linear regression and Random Forest. The lower scores
of the ANN are probably due to the fact that they typically require
a much larger training dataset than we had at our disposal (Chen
and Guestrin, 2016). Furthermore, even though the ANN scored
lower than XGBoost, we showed that prediction quality can be
improved by tuning the last layer of the ANN to each individual. With a larger dataset, the personalized ANN could potentially
attain better prediction than XGBoost. Additionally, in this study
we did not use individual features such as individual frequency
bands. In general, this could improve the performance of the
algorithm. REFERENCES scikit-learn project,” in ECML PKDD Workshop: Languages for Data Mining
and Machine Learning (Prague), 108–122. Abásolo, D., Hornero, R., Gómez, C., García, M., and López, M. (2006). Analysis
of eeg background activity in Alzheimer’s disease patients with lempel–ziv
complexity and central tendency measure. Med. Eng. Phys. 28, 315–322. doi: 10.1016/j.medengphy.2005.07.004 Bullmore,
E.,
Barnes,
A.,
Bassett,
D. S.,
Fornito,
A.,
Kitzbichler,
M.,
Meunier, D., et al. (2009). Generic aspects of complexity in brain
imaging data and other biological systems. Neuroimage 47, 1125–1134. doi: 10.1016/j.neuroimage.2009.05.032 Bullmore, E., and Sporns, O. (2009). Complex brain networks: graph theoretical
analysis of structural and functional systems. Nat. Rev. Neurosci. 10, 186. doi: 10.1038/nrn2575 Aricò, P., Borghini, G., Di Flumeri, G., Colosimo, A., Bonelli, S., Golfetti, A.,
et al. (2016a). Adaptive automation triggered by eeg-based mental workload
index: a passive brain-computer interface application in realistic air traffic
control environment. Front. Hum. Neurosci. 10:539. doi: 10.3389/fnhum.201
6.00539 Chen, T., and Guestrin, C. (2016). “Xgboost: a scalable tree boosting system,”
in Proceedings of the 22nd ACM Sigkdd International Conference on
Knowledge Discovery and Data Mining (San Francisco, CA: ACM), 785–794. doi: 10.1145/2939672.2939785 Aricò, P., Borghini, G., Di Flumeri, G., Colosimo, A., Pozzi, S., and
Babiloni, F. (2016b). A passive brain–computer interface application for
the mental workload assessment on professional air traffic controllers
during realistic air traffic control tasks. Prog. Brain Res. 228, 295–328. doi: 10.1016/bs.pbr.2016.04.021 Chollet, F. et al. (2015). Keras. Available online at: https://keras.io. (accessed
May, 2019). Conrad, C. D., and Bliemel, M. (2016). “Psychophysiological measures of cognitive
absorption and cognitive load in E-learning applications,” in Proceedings of the
37th International Conference on Information Systems, December 11–14, eds P. Agerfalk, N. Levina, and S. S. Kien (Dublin). Arthur, W. Jr., Tubre, T. C., Paul, D. S., and Sanchez-Ku, M. L. (1999). College-sample psychometric and normative data on a short form of the
raven advanced progressive matrices test. J. Psychoeduc. Assess. 17, 354–361. doi: 10.1177/073428299901700405 Agerfalk, N. Levina, and S. S. Kien (Dublin). Coyne, J. T., Baldwin, C., Cole, A., Sibley, C., and Roberts, D. M. (2009). “Applying real time physiological measures of cognitive load to improve
training,” in International Conference on Foundations of Augmented Cognition
(San Diego, CA: Springer), 469–478. Bell, A. J., and Sejnowski, T. J. (1995). An information-maximization approach
to blind separation and blind deconvolution. Neural Comput. 7, 1129–1159. doi: 10.1162/neco.1995.7.6.1129 (San Diego, CA: Springer), 469–478. 4. DISCUSSION Incorporating individual features should be addressed
in future research. FIGURE 7 | Difference of r2 score using an ANN before and after
individualizing the last layer for each subject. Error bars reflect standard error
of the mean. Paired t-test results are shown. ***p = 0.001. As part of our analysis, we checked the impact of additional
EEG measures, specifically metrics of connectivity and metrics
of neural complexity. Our results suggest that connectivity
measures do add information regarding cognitive load beyond
the simple spectral features. On the other hand, it seems difficulty was ordered on a scale from 1 to 36 (Forbes, 1964;
Arthur et al., 1999) and was treated here as a continuous value. Our results show that even when considering cognitive load in
a continuous manner, a reasonable prediction accuracy can be June 2019 | Volume 13 | Article 191 Frontiers in Human Neuroscience | www.frontiersin.org 7 EEG-Based Prediction of Cognitive Load Friedman et al. ETHICS STATEMENT that complexity features, while holding information regarding
cognitive load, do not afford additional information over and
above that found in connectivity and PS features. that complexity features, while holding information regarding
cognitive load, do not afford additional information over and
above that found in connectivity and PS features. The protocol was approved by the Ben-Gurion University
ethics committee. All subjects gave written informed consent
in accordance with the Declaration of Helsinki and received
compensation for participating. Lastly, we found that prediction quality did not deteriorate,
and even improved, when using a limited number of channels
(∼12), which is important for practical applications. This is
most probably due to better generalization, resulting from a less
complex model, as opposed to one utilizing all channels. ACKNOWLEDGMENTS The authors thank Dr. Nir Getter for helpful discussions on the
manuscript and Ms. Koral Regev for valuable help in running
the experiments. FUNDING This research was supported in part by the Helmsley Charitable
Trust through the Agricultural, Biological and Cognitive (ABC)
Robotics Initiative and by the Marcus Endowment Fund both at
Ben-Gurion University of the Negev. AUTHOR CONTRIBUTIONS We chose to utilize the advanced progressive matrices test in
this study because of the high validity of its operationalization of
difficulty levels. However, to extend our findings further toward
applicability, future studies should examine the utility of our
EEG-based metrics for cognitive load in real-life settings such
as control tower operator performance as aerial traffic ebbs and
flows. Since our results indicate the feasibility of employing an
array comprising as little as eight electrodes, potentially such
studies could be carried out in parallel using portable dry EEG
systems. The added benefit would be the feasibility of amassing
the expansive datasets necessary for utilizing elaborate neural
network models, which in this scenario are expected to improve
predictive ability. In addition, it would be useful to identify EEG
markers for different dimensions of cognitive workload. Such
markers would pave the way for optimizing and personalizing
learning processes from e-learning to military training (Ikehara
and Crosby, 2005; Mills et al., 2017). NF, KG, and OS: experiment planning and design. NF: data
acquisition. NF and TF: data analysis. NF, TF, KG, and OS:
writing the manuscript. REFERENCES What’s new in psychtoolbox-3. Perception 36:1. Rubin, D., Fekete, T., and Mujica-Parodi, L. R. (2013). Optimizing complexity
measures for fmri data: algorithm, artifact, and sensitivity. PLoS ONE 8:e63448. doi: 10.1371/journal.pone.0063448 Logan, G. D. (1985). Skill and automaticity: Relations, implications, and future
directions. Can. J. Psychol./Revue Can. Psychol. 39:367. doi: 10.1037/h0080066 Stevens, R., Galloway, T., and Berka, C. (2006). “Integrating EEG models
of cognitive load with machine learning models of scientific problem
solving,” in Augmented Cognition: Past, Present and Future, eds D. Schmorrow, K. Stanney, and L. Reeves (Arlington, VA: Strategic Analysis),
55–65. Mak, J. N., Chan, R. H., and Wong, S. W. (2013). “Spectral modulation of
frontal eeg activities during motor skill acquisition: task familiarity monitoring
using single-channel eeg,” in Engineering in Medicine and Biology Society
(embc), 2013 35th Annual International Conference of the IEEE (Osaka: IEEE),
5638–5641. Sweller, J.,
Van
Merrienboer,
J. J.,
and
Paas,
F. G. (1998). Cognitive
architecture and instructional design. Educ. Psychol. Rev. 10, 251–296. doi: 10.1023/A:1022193728205 McDonald, N. J., and Soussou, W. (2011). “Quasar’s qstates cognitive gauge
performance in the cognitive state assessment competition 2011,” in
Engineering in Medicine and Biology Society, EMBC, 2011 Annual International
Conference of the IEEE (Boston, MA: IEEE), 6542–6546. Tononi, G., and Edelman, G. M. (1998). Consciousness and complexity. science
282, 1846–1851. doi: 10.1126/science.282.5395.1846 Mills, C., Fridman, I., Soussou, W., Waghray, D., Olney, A. M., and D’Mello,
S. K. (2017). “Put your thinking cap on: detecting cognitive load using eeg
during learning,” in Proceedings of the Seventh International Learning Analytics
& Knowledge Conference (Vancouver, BC: ACM), 80–89. Zhang, X.-S., Roy, R. J., and Jensen, E. W. (2001). Eeg complexity as a measure
of depth of anesthesia for patients. IEEE Trans. Biomed. Eng. 48, 1424–1433. doi: 10.1109/10.966601 Zhou, Z.-H. (2012). Ensemble Methods: Foundations and Algorithms. Raton, FL:
Chapman and Hall/CRC. doi: 10.1201/b12207 Mullen, T. R., Kothe, C. A., Chi, Y. M., Ojeda, A., Kerth, T., Makeig, S., et al. (2015). Real-time neuroimaging and cognitive monitoring using wearable dry eeg. IEEE Trans. Biomed. Eng. 62, 2553–2567. doi: 10.1109/TBME.2015.2481482 Conflict of Interest Statement: The authors declare that the research was
conducted in the absence of any commercial or financial relationships that could
be construed as a potential conflict of interest. Conflict of Interest Statement: The authors declare that the research was
conducted in the absence of any commercial or financial relationships that could
be construed as a potential conflict of interest. REFERENCES Borghini, G., Aricò, P., Di Flumeri, G., Cartocci, G., Colosimo, A., Bonelli,
S., et al. (2017). Eeg-based cognitive control behaviour assessment: an
ecological study with professional air traffic controllers. Sci. Rep. 7, 547. doi: 10.1038/s41598-017-00633-7 Delorme, A., and Makeig, S. (2004). Eeglab: an open source toolbox for analysis
of single-trial eeg dynamics including independent component analysis. J. Neurosci. Methods 134, 9–21. doi: 10.1016/j.jneumeth.2003.10.009 Fekete, T., Beacher, F. D., Cha, J., Rubin, D., and Mujica-Parodi, L. R. (2014). Small-world network properties in prefrontal cortex correlate with predictors of
psychopathology risk in young children: a nirs study. NeuroImage 85, 345–353. doi: 10.1016/j.neuroimage.2013.07.022 Brainard, D. H., and Vision, S. (1997). The psychophysics toolbox. Spat. Vis. 10,
433–436. doi: 10.1163/156856897X00357 Brouwer, A.-M., Hogervorst, M. A., Van Erp, J. B., Heffelaar, T., Zimmerman,
P. H.,
and
Oostenveld,
R. (2012). Estimating
workload
using
eeg
spectral power and erps in the n-back task. J. Neural Eng. 9:045008. doi: 10.1088/1741-2560/9/4/045008 Forbes, A. (1964). An item analysis of the advanced matrices. Br. J. Educ. Psychol. 34, 223–236. doi: 10.1111/j.2044-8279.1964.tb00632.x Gerˇe, I., and Jaušcvec, N. (1999). Multimedia: Differences in cognitive processes
observed with eeg. Educ. Technol. Res. Dev. 47, 5–14. doi: 10.1007/BF022
99630 Buitinck, L., Louppe, G., Blondel, M., Pedregosa, F., Mueller, A., Grisel, O.,
et al. (2013). “API design for machine learning software: experiences from the June 2019 | Volume 13 | Article 191 Frontiers in Human Neuroscience | www.frontiersin.org 8 EEG-Based Prediction of Cognitive Load Friedman et al. Pope, A. T., Bogart, E. H., and Bartolome, D. S. (1995). Biocybernetic
system
evaluates
indices
of
operator
engagement
in
automated
task. Biol. Psychol. 40,
187–195. doi:
10.1016/0301-0511(95)
05116-3 Gruber, I., Hlaváˇc, M., Železn`y, M., and Karpov, A. (2017). “Facing face recognition
with resnet: round one,” in International Conference on Interactive Collaborative
Robotics (Hatfield: Springer), 67–74. Ho,
T. K. (1995). “Random
decision
forests,”
in
Proceedings
of
3rd
International Conference on Document Analysis and Recognition, Vol. 1,
(Montreal, QC: IEEE), 278–282. Raven, J. (2000). The raven’s progressive matrices: change and stability
over culture and time. Cogn. Psychol. 41, 1–48. doi: 10.1006/cogp.1999. 0735 Ikehara, C. S., and Crosby, M. E. (2005). “Assessing cognitive load with
physiological sensors,” in Proceedings of the 38th Annual Hawaii International
Conference on System Sciences, 295–303. Raven, J. C., and Court, J. H. (1998). Raven’s Progressive Matrices and Vocabulary
Scales. London: Psychological Corporation. Kleiner, M., Brainard, D., Pelli, D., Ingling, A., Murray, R., Broussard, C., et al. (2007). Frontiers in Human Neuroscience | www.frontiersin.org June 2019 | Volume 13 | Article 191 REFERENCES Palinko, O., Kun, A. L., Shyrokov, A., and Heeman, P. (2010). “Estimating cognitive
load using remote eye tracking in a driving simulator,´’ in Proceedings of the
2010 Symposium on Eye-Tracking Research & Applications (Austin, TX: ACM),
141–144. Copyright © 2019 Friedman, Fekete, Gal and Shriki. This is an open-access article
distributed under the terms of the Creative Commons Attribution License (CC BY). The use, distribution or reproduction in other forums is permitted, provided the
original author(s) and the copyright owner(s) are credited and that the original
publication in this journal is cited, in accordance with accepted academic practice. No use, distribution or reproduction is permitted which does not comply with these
terms. Pelli, D. G. (1997). The videotoolbox software for visual psychophysics:
transforming
numbers
into
movies. Spat. Vis. 10,
437–442. doi: 10.1163/156856897X00366 Peng, C.-K., Havlin, S., Stanley, H. E., and Goldberger, A. L. (1995). Quantification
of scaling exponents and crossover phenomena in nonstationary heartbeat time
series. Chaos 5, 82–87. doi: 10.1063/1.166141 June 2019 | Volume 13 | Article 191 Frontiers in Human Neuroscience | www.frontiersin.org 9
|
https://openalex.org/W2744465911
|
https://www.matec-conferences.org/articles/matecconf/pdf/2017/33/matecconf_imeti2017_01027.pdf
|
English
| null |
The application and practical benefits of “C theory” in project management
|
MATEC web of conferences
| 2,017
|
cc-by
| 15,673
|
© The Authors, published by EDP Sciences. This is an open access article distributed under the terms of the Creative
Commons Attribution License 4.0 (http://creativecommons.org/licenses/by/4.0/). The application and practical benefits of "C theory" in project
management I-Chan Kaoa
General Education Center, Open University of Kaohsiung, Kaohsiung 812, Taiwan General Education Center, Open University of Kaohsiung, Kaohsiung 812, Taiwan Abstract. This study regarded adult in-service students who are familiar with project
management courses as the subjects and collected data through questionnaires for confirmatory
factor analysis and reliability analysis. The aim is to construct a "C Theory" questionnaire scale
that encompasses the essence of Chinese management philosophy like decision-making of
Taoism, leadership of Legalism, tactics of School of Military Strategists, creativity of Mohism,
and coordination of Confucianism. Furthermore, management performance scales were
constructed for learning and growth, internal operation, customer satisfaction, and financial
control, while the questionnaires and statistical analyses were expected to probe into the impact
of "C Theory" on project management performance. This study found that the application and
practice of "C Theory" have a high and positive correlation with project management and a
significant influence on the improvement of performance. It is therefore suggested that
management methods in "C Theory" be appropriately used in project management in order to
enhance the efficiency of project management and facilitate the achievement of project
management targets. a Corresponding author : kic0929@ouk.edu.tw , 01027 (2017)
119
MATEC Web of Conferences
IMETI 2016 , 01027 (2017)
119
MATEC Web of Conferences
IMETI 2016 DOI: 10.1051/
7119010
matecconf/201
27 1.1 Research motives Chinese philosophy contains extensive knowledge and profound scholarship with a long history and
remarkable impact on human civilization. As early as the Spring and Autumn Period, there was the
grand occasion of contention of a hundred schools of thought that were passed down and concentrated
into five major scholars: Confucianism, Legalism, Mohism, School of Military Strategists, and
Taoism. Many scholars[20-22], [36, 43, 52] pointed out that Chinese style leadership and management
philosophy mainly adopt ethics of Confucianism, rule of law of Legalism, universal love of Mohism,
power tactics of School of Military Strategists, and letting things take their own course of Taoism as
the comprehensive framework. Kao (2014b), Kao (1999), and Tsai and Wu (2013) integrated the
previous five Chinese leadership thoughts into "Throne" tactics and leadership philosophy. Lu (2005)
pointed out that rule of law of Legalism was similar to X Theory of Douglas McGregor, while
"self-discipline and stabilizing others" of Confucianism was similar to Y Theory of Douglas
McGregor and Z Theory of William G. Ouchi, and letting things take their own course of Taoism was , 01027 (2017)
119
MATEC Web of Conferences
IMETI 2016 DOI: 10.1051/
7119010
matecconf/201
27 similar to L Theory of William D. Hitt. Furthermore, the Western military academic field spares no
effort to utilize the thoughts of School of Military Strategists. Specifically, Master Sun's Art of War
has become important material to educate military generals and leaders in the fields of industry,
government, and academic circle. Therefore, Chinese leadership and management and that of the West
are different in approach, but equally satisfactory in result. Cheng (1995) observed the differences of the East and the West in terms of cultures, social
organizations, philosophical way of thinking, value, and history experiences and concluded "C
Theory" from the research results of the Book of Changes, which could embrace all the valuable
management and philosophy ideas in the East and the West. Here, "C" refers to the creativity of the
Book of Changes of China. Cheng (1995) integrated the principle of mutual promotion of the Five
Elements to develop the management model of C Theory (as shown in Figure 1) and employed the
principle of mutual promotion and restraint between the Five Elements to elaborate the basic
leadership philosophy of the Book of Changes. 1.1 Research motives A decision-maker shall
have a driving force that must be a power in the center - that is, "earth" in the center. The central
power of "earth" shall be deep and sound with the potential of development. In other words, we call it
the driving force of decision-making. The wisdom to make decision produces leadership, and one with
such wisdom can be a leader. If explained by the principle of the Five Elements, it is "earth promotes
metal". "Metal" means being decisive. The strength to cut through metal is a power, or, leadership. The basis of leadership is the wisdom to make decision, or the wisdom of "earth". With the support of
vigorous leadership and decision-making, one can conquer "water". One with adaptability capacity
knows how to cope with the ever-changing external environment like water and come up with creative
plans. If explained by the principle of the Five Elements, it is "water promotes wood". "Wood" means
thriving and creative plans. With creative plans, one can combine the wills of members to inspire more
potential to form strong, vigorous, and high morale. If explained by the principle of the Five Elements,
it is "wood promotes fire". "Fire" has the function to fuse and aggregate. The best decisions can be
made by gathering fire and wisdom of all people and giving full play to the role of "fire promotes
earth" [53]. Tom Peters & Robert H. Waterman (2005) noted that, "In the future, more than 90% of
white-collar workers would face a crisis. All future white-collar works are project works. The same is
true for all the works with economic value." A project is not repeatable or routine, but unique. Once a
project is completed, it will never be repeated. A project refers to a specific program, recommendation,
decision, or matter [44]. According to the Project Management Quality System Guide established by
International Standard Organization (ISO) (ISO 10006: 2003), a project is a unique process that is
made up by a series of activities with start and end dates, which mutually coordinate and control and
which meet the constraint conditions or goals of time, costs, resources, and so on. According to IPMA
Competency baseline (ICB), a project refers to the temporary investment and efforts made by an
organization to reach a specific goal within a period of time [14]. 1.1 Research motives Control
Metal
Personnel
Decisions Adaptability
fire Earth water
Creativity
Wood
Figure 1. Management Model of C Theory, Source: Cited from Cheng, 38(1995). Adaptability
water Personnel
fire Decisions
Earth Creativity
Wood anagement Model of C Theory, Source: Cited from Cheng, 38(1995). Figure 1. Management Model of C Theory, Source: Cited from Cheng, 38(1995). According to Figure 1, the "Five Elements" refer to metal, wood, water, fire, and earth. Their key
implications are described as follows [5, 46]. 1. Metal: It is resolute, fortitude, firm, and flexible and not destroyed after severe training and
hammering. It represents "control". It is resolute, fortitude, firm, and flexible and not destroyed after severe training a
ring. It represents "control". g
p
2. Wood: It means prosperous and thriving and stands for "creative plans". g
p
It means prosperous and thriving and stands for "creative plans". It is changing and stands for the "adaptability capacity" of a leader in the face of
anging external environment. g g
4. Fire: It is strong and vigorous and stands for "personnel" with high morale. g g
is strong and vigorous and stands for "personnel" with high morale. 5. Earth: It is thick and sound and boundlessly inclusive. It can create life and guard the center. It
stands for "the wisdom of decision-making" of a leader. Cheng (1995) compared leadership to "gold". Gold stands for strength, firmness, and flexibility,
which are the characteristics required for a leader. A good leader shall be like gold rather than iron and
shall be loyal, wise, and flexible. Gold surpasses iron and copper, because it is soft, thoroughly
tempered, elastic, and resilient. On the contrary, iron rusts and is not plastic, while copper can easily 2 , 01027 (2017)
119
MATEC Web of Conferences , 01027 (2017)
119
MATEC Web of Conferences
IMETI 2016 DOI: 10.1051/
7119010
matecconf/201
27 IMETI 2016 be destroyed. If a leader is like iron or copper, then he/she may be unfaithful and not flexible. Hence,
the best leader shall be like pure gold. The order of mutual promotion among the Five Elements is that
wood promotes metal, metal promotes water, water promotes wood, wood promotes fire, and fire
promotes earth. It implies that decision-making is the core of management. , 01027 (2017)
119
MATEC Web of Conferences
IMETI 2016 DOI: 10.1051/
7119010
matecconf/201
27 and educational circles. BSC was first proposed by Robert Kaplan and David Norton in 1992. It is a
set of strategic management tools to assess the management system and strategy indexation of
organizational performance, stressing the spirit of "balance" and effectively converting strategies into
action to guide an organization to achieve its goals [18, 65]. and educational circles. BSC was first proposed by Robert Kaplan and David Norton in 1992. It is a
set of strategic management tools to assess the management system and strategy indexation of
organizational performance, stressing the spirit of "balance" and effectively converting strategies into
action to guide an organization to achieve its goals [18, 65]. The goals and indicators of BSC are derived from the visions and strategies of an organization. The dimensions of evaluation include finance control, customer identity, internal operation, and
learning and growth [7, 13]. BSC focuses on balance and seeks a balance between short- and
long-term goals, between financial and non-financial metrics, between delayed and advanced
assessment indicators, and between internal and external performances of an enterprise. According to
Fortune, among the top 1000 enterprises in the U.S., up to 40% have adopted BSC. A Bain &
Company survey also pointed out that over 50% of North American companies have adopted BSC as
the method to evaluate their internal performances. Harvard Business Review commented that BSC is
the most influential strategy management tool since 1975. The strategic and action considerations and
management skills of BSC have been stressed and applied by many enterprises and organizations
[58-61]. This researcher is a member of teaching staff engaged in in-service education at an adult education
institution and a teacher of courses related to project management. Hence, in terms of its research
motive, this study regarded adult in-service students with the experience of project management
courses as its samples in order to understand their outcomes of the application of C Theory in project
management in an organization and the organizational performances in BSC learning and growth,
internal operation, customer satisfaction, and finance control. , 01027 (2017)
119
MATEC Web of Conferences
IMETI 2016 This study learned the influences of the
application of "C Theory" in project management by samples with different social variables on
organizational performances and the strategies proposed by them to improve the performances of
project management, strengthen project management knowledge and system and organizational
training to introduce the procedures and methods, and to enhance the project management capacity of
members, organizational maturity of project management, and efficiency of practical application. 1.2 Research purposes Based on the aforementioned research motives and backgrounds, the research goals are described as
follows. 1.2.1 To understand the connotation of C Theory and organizational performance. 1.2.2 To develop Application of C Theory and Organizational Performance Questionnaire. 1.2.2 To develop Application of C Theory and Organizational Performance Questionnair 1.2.3 To explore the influence of the application of C Theory in project management on organizational
performance. 1 2 4 T
f
ibl
i
i
i
i
l
f
h
h h
li
i
f 1.2.3 To explore the influence of the application of C Theory in project management on organizational
performance. p
1.2.4 To propose feasible strategies to improve organizational performance through the application of
C Theory in project management based on the research results. 1.2.4 To propose feasible strategies to improve organizational performance through the application of
C Theory in project management based on the research results. 1.1 Research motives Project Management Institute (PMI)
deemed that a project refers to a temporary effort to create a unique product, service, or result [47]. Based on the above definitions of professional organizations and scholars [23, 30, 32, 35, 37, 41, 44]
this study assumed that a project refers to a one-time event, not repeated, with start and end dates,
pre-set and clear goals, and limited resources, which shall produce expected deliverables. The connotation and framework of this study’s project management was based on Project
Management Quality System Guide (ISO 10006: 2003) established by ISO, IPMA Competency
Baseline (ICB), A Guide to the Project Management Body of Knowledge of Project Management
Institute (PMI), and Project Management Procedures of British PRINCE2. To develop the methods
and strategies of project management, in principle it followed five major procedures: initiation,
planning, execution, control, and end [16, 25, 27, 31, 35, 37], [38, 41, 44, 47, 57, 64]. Project
management must organize, deploy, and recruit personnel, adopt effective communication and regime
development teams, integrate human resources, materials, and financial resources inside and outside
an organization, comply with the verification standards to produce deliverables, and meet the needs of
the project stakeholders[14, 24, 28, 30], [31, 33], [34, 47, 57, 64]. Modern organizational works have the characteristics of a project. Hence, they can no longer be
done by routine management methods. On the contrary, one shall employ the techniques and processes
of project management, set and use organizational key performance indicators to evaluate the effect of
management, and continue to improve in order to achieve the organization's business objectives. To
avoid unsatisfactory performance due to cost overruns, delay, improper range, and poor quality, one
shall utilize the methods and strategies of project management, carefully consider the feasibility to
initiate a project, rationality of planning, effectiveness of implementation, monitoring accuracy, and
completeness, and improve gradually to achieve goals and visions of the organization. However, how
can we tell that the goals and visions are reached? It is necessary to apply appropriate assessment
methods and indicators to assess whether an organization has reached its desired objectives and
visions. Balanced Scorecard (BSC) is a common way used for such a purpose, especially in industry 3 3 , 01027 (2017)
119
MATEC Web of Conferences
IMETI 2016 , 01027 (2017)
119
MATEC Web of Conferences
IMETI 2016 DOI: 10.1051/
7119010
matecconf/201
27 1.4 Research structure Figure 2 illustrates the statistical analysis framework of this study. Based on the empirical research in
recent years by this researcher [25, 27, 32], [33], this study explored the social variables of gender,
age, education, occupation, position, organizational history, and organizational scale and then adopted
the decision-making performance of Taoism, control performance of Legalism, tactics performance of
School of Military Strategists, creativity performance of Mohism, and coordination performance of
Confucianism as the independent variables for the application of C Theory in project management. Lastly, it adopted learning and growth, internal operation, customer satisfaction, and finance control as
the dependent variables of performance. Through empirical research and inferential statistical analysis,
it explored the relationships among the variables. Figure 2. Study Chart, Source: Drawn by this researcher. Statistical analysis:
Social
variables
Gender
Age
Education
Occupation
Position
Organizationa
l history
Organizationa
l scale
(1)
(1)
(3)
(2)
The application of C Theory
in project management
Decision-Making performance
of Taoism
Control performance of
Legalism
Tactics performance of School
of Military Strategists
Creative performance of
Mohism
Coordination performance of
Confucianism
Organizational
performance
Learning and growth
Internal operation
Customer satisfaction
Finance control The application of C Theory
in project management (3) Figure 2. Study Chart, Source: Drawn by this researcher. (1) T test and One-way ANOVA were adopted to learn the cognitive differences in the application of
C Theory in project management and organizational performance of the subjects with different
social backgrounds. (1) T test and One-way ANOVA were adopted to learn the cognitive differences in the application of
C Theory in project management and organizational performance of the subjects with different
social backgrounds. g
(2) Pearson correlation was used to analyze the correlation in the application of C Theory in project
management and organizational performance. g
(2) Pearson correlation was used to analyze the correlation in the application of C Theory in project
management and organizational performance. g
g
p
(3) Regression analysis was employed to understand the predictability of the application of C Theory
in project management for organizational performance. (3) Regression analysis was employed to understand the predictability of the application of C Theory
in project management for organizational performance. 1.3 Research method and process This study first learned the connotation of C Theory and BSC through a literature review and analysis
to develop a first draft of "Questionnaire of Application of C Theory and Organizational Performance". It then distributed the pre-run questionnaire to the adult in-service students with the experience of
project management courses to develop the reliability and validity of the questionnaire. It conducted a
formal questionnaire survey and collected feedbacks for statistical analysis. The methods of
descriptive and inferential statistics helped this study understand the application of C Theory and
organizational performances of the subjects during project management. It also learned the differences
in the application of C Theory in project management and organizational performance of the subjects
with different social variables and analyzed the correlation between the application of C Theory and
organizational performance. Lastly, this study shall provide a conclusion and propose feasible
strategies to improve organizational performance and efficiency of project management. 4 4 , 01027 (2017)
119
MATEC Web of Conferences
IMETI 2016 2 The connotation of C theory and the description of relevant leadership
and management Cheng (2015) proposed C Theory and pointed out in his book on the theory that management factors
could be classified into five categories: adjustment, practice of thoughts and ideals, business creation,
and creativity transformation. The five categories were the independent and coordinative application
of mechanisms, including skills such as operational ability, application of system knowledge, 5 5 , 01027 (2017)
119
MATEC Web of Conferences , 01027 (2017)
119
MATEC Web of Conferences
IMETI 2016 DOI: 10.1051/
7119010
matecconf/201
27 principles and values of principles, and the practice of "Taoism" (the basic principles of formation and
transformation) with intelligence. There is no doubt that, as ethical principles, they could also be used
in management as guiding principles, because ethics and management are inherently associated in
terms of content and functionality. The five categories serve as a reference for the levels of
management, which is the transformation of ethical levels. According to the management level
structure, C Theory for the first time described how a normative management theory was developed. Cheng (2015) assumed that modern leadership and management meant institutional reform in the face
of change and optimized structure. Hence, an appropriate form of leadership was identified and
expected. The most important factors for political leadership were listed from historical experience,
modern knowledge, and future expectations to discover important experiences and preservation of
memories. Learning from competitors and the development of modernization and globalization are
equally important. principles and values of principles, and the practice of "Taoism" (the basic principles of formation and
transformation) with intelligence. There is no doubt that, as ethical principles, they could also be used
in management as guiding principles, because ethics and management are inherently associated in
terms of content and functionality. The five categories serve as a reference for the levels of
management, which is the transformation of ethical levels. According to the management level
structure, C Theory for the first time described how a normative management theory was developed. Cheng (2015) assumed that modern leadership and management meant institutional reform in the face
of change and optimized structure. Hence, an appropriate form of leadership was identified and
expected. The most important factors for political leadership were listed from historical experience,
modern knowledge, and future expectations to discover important experiences and preservation of
memories. Learning from competitors and the development of modernization and globalization are
equally important. 2 The connotation of C theory and the description of relevant leadership
and management From the systematic operation of business organizations, the operational framework of the Five
Elements shown in Figure 1 could be applied in the system of modern business organizations and the
integration and circulation related to management and leadership in order to obtain creative
management functions of C Theory. Figure 3 shows the mutual promotion and restriction of each
system in an organization. C1 refers to planning and commitment (Chairman of the Board of
Directors), C2 to constitution and commands (General Manager, CEO), C3 to competition and
cooperation (Marketing Director, Chief Marketing Officer), C4 to creativity and transformation
(Product Operation), C5 to communication and coordination (HR), C6 to understanding and
communication (Shareholders' Meeting), and C7 to termination and continuity (re-election of
Shareholders' Meeting). Based on the operation of the Five Elements management system of an
organization shown in Figure 1, icon " " symbolizes the promotion of one function to another. In
other words, Ci Cj means that Ci produces Cj [5-6]. Thus, C1 representing commitments produces
C2 representing commands. Why is this so? In an organization, the commitment to goals leads to the
actual participation in commands. Accordingly, when commands enter the market, they become C3
representing competition. What the market needs and produces is C4. C3 follows the needs of
customers. When C4 develops in the market, it returns to C5, which is the situation realized through
education and development. C5 then needs to return to the beginning to circulate again. The process
implies a continuous and mutual promotion relationship, similar to the theory of separation of five
powers - namely, wood, metal, water wood, and fire [6, 41, 42]. Figure 3. Integration and Circulation of the Creative Management Function of C Theory, Source: Cited from
Cheng, 2015, 36. Figure 3. Integration and Circulation of the Creative Management Function of C Theory, Source: Cited from
Cheng, 2015, 36. Figure 3. Integration and Circulation of the Creative Management Function of C Theory, Source: Cited from
Cheng, 2015, 36. 2 The connotation of C theory and the description of relevant leadership
and management Kao (2014b) believed that after introduction of a leadership
philosophy like Confucianism, Legalism, Mohism, School of Military Strategists, and Taoism to the
project management and operation of an organization, the feasible strategies include strengthening the
professional function of a leader, establishment of the system and regulation of HR management and
development, establishment of the authorization system of project management to promptly finish
each task, setting up and implementing a project change control system to enhance the efficiency and
value of the project, and establishment of a reward and punishment system to enhance the oral and
honor of the project members. Tsai and Wu (2013) pointed out that Confucianism mainly employed the governance of morality,
etiquette, and music and developed five major leadership strategies of action by etiquette, governance
by morality, consistent cognition and behavior, and govern by doing nothing that goes against nature. Legalism mainly advocated the leadership strategies of reward and punishment, right person for the
right position, performance assessment, and letting things take their own course. Mohism stressed
hard practice, leading by example, strict self-discipline, virtue, unity, universal love, and strict rules. "According to the rules of Mohism, the murderer shall be sentenced to death, while the person who
hurt others shall be punished." This is the way to maintain justice and equality. School of Military
Strategists revealed the subjective law of human competition, emphasizing that understanding others
and oneself, avoiding the solid and striking the weak, plotting and creating situation, and adaptability
capacity could bring about successes. Taoism focused on the leadership strategies of being wise to
choose a person for a job, following procedures, letting things take their own course of Taoism, giving
authorization to employees, being employ-oriented, recognizing and adopting correct statements, and
being inclusive. Kao (2014b) used "Chinese Throne" (see Figure 4) to explain leadership. The left and
right handles are like the left and right tokens of a leader and the guarding laws of an organization. The right token is the ethics of Confucianism, while the left is the strategies to govern subordinates of
"rules", "strategies", and "trends". The base plate is like the strength of a leader's legs or the military
strategies of a strategist. The cushion in the middle balances among the philosophies of Confucianism,
Legalism, School of Military Strategists, and Mohism. 2 The connotation of C theory and the description of relevant leadership
and management Cheng (1995) asserted that aside from the integration framework of the Book of Changes, the
leadership and management philosophy of C Theory, in terms of leadership and management practices,
reflects the decision-making function of "earth" of Taoism, control function of "metal" of Legalism,
adaptability capacity of "water" of School of Military Strategists, creativity function of "wood" of
Mohism, and coordination function of "fire" of Confucianism, serving as the references and rules of
organizational leadership and management. Though each industry promotes different practices in the
tasks of project management from initiation, planning, execution, control, to end, the chronological
order of all the stages is called "project management procedures". The beginning and end of project
management together make up the period of a project. The project management procedures can be
classified into six stages: (1) Receipt of a project task; (2) Analysis of the project task; (3) Project
planning and action; (4) Distribution of the project task; (5) Execution of the project task; and (6) End 6 , 01027 (2017)
119
MATEC Web of Conferences , 01027 (2017)
119
MATEC Web of Conferences
IMETI 2016 DOI: 10.1051/
7119010
matecconf/201
27 of the project task [16, 14], [28-34], [57]. Kao (2014b) believed that after introduction of a leadership
philosophy like Confucianism, Legalism, Mohism, School of Military Strategists, and Taoism to the
project management and operation of an organization, the feasible strategies include strengthening the
professional function of a leader, establishment of the system and regulation of HR management and
development, establishment of the authorization system of project management to promptly finish
each task, setting up and implementing a project change control system to enhance the efficiency and
value of the project, and establishment of a reward and punishment system to enhance the oral and
honor of the project members. of the project task [16, 14], [28-34], [57]. 2 The connotation of C theory and the description of relevant leadership
and management When the four schools help an organization
enter into the right track, the leader can rest assured to implement the ideas of Taoism of "nature and
letting things take their own course of Taoism", "following the nature and people", and "adaption to
different circumstances". Figure 4. Leadership Philosophy of Chinese Throne, Source: Kao, 42(2014b). Figure 4. Leadership Philosophy of Chinese Throne, Source: Kao, 42(2014b). The aforementioned leadership philosophies of the different Chinese schools were experienced
without scientific evidence or objective and universal efficiency. Kang, Chang, Yang, and Huang
(2010) deemed that under the influence of traditional Chinese culture, Chinese had a unique
perception on the concept of leadership, and the leadership behavior of a leader would reflect some
characteristics of Chinese culture. Chen (2014) pointed out that Chinese believed that to complete one
thing smoothly, one must need "good timing, geographical convenience, and good human relations". Kang, Su, Chang, and Hsu (2001) believed that both Chinese and Western leadership had the
connotations of the trait theory, behavior theory, and contingency theory. However, the behavior
model of Chinese contingency leadership was based on "affection", "reason", and "law". A leader 7 7 , 01027 (2017)
119
MATEC Web of Conferences , 01027 (2017)
119
MATEC Web of Conferences
IMETI 2016 DOI: 10.1051/
7119010
matecconf/201
27 considered the respective weight of the leader roles of "sovereign", "parent", and "teacher" and
switched among them to realize the true functions of a leader [36]. Chinese often say "being fair and
reasonable" or "being reasonable and legal". When talking about "fairness" and "law", "reason" is
inseparable, indicating the true leadership philosophy of Chinese: Everything shall be fair,
reasonable, and legal [55]. From the perspective of leadership and management, national culture is an
important situational factor. Therefore, to understand the ideas or assumptions about Chinese
leadership, one must take "culture", "society", and "history" into consideration [63]. considered the respective weight of the leader roles of "sovereign", "parent", and "teacher" and
switched among them to realize the true functions of a leader [36]. Chinese often say "being fair and
reasonable" or "being reasonable and legal". When talking about "fairness" and "law", "reason" is
inseparable, indicating the true leadership philosophy of Chinese: Everything shall be fair,
reasonable, and legal [55]. From the perspective of leadership and management, national culture is an
important situational factor. 2 The connotation of C theory and the description of relevant leadership
and management Therefore, to understand the ideas or assumptions about Chinese
leadership, one must take "culture", "society", and "history" into consideration [63]. p
y
y
No matter for management theory or organization operating practices, the way of thinking of "The
West leads the East" has gradually turned into "The West meets the East". The idea of governance by
morality advocated by Chinese organizations has gradually attracted the attention of Eastern and
Western academia and industry. The interpretation of "the relationship competition and cooperation"
with
Eastern
and
Western
philosophies
could
serve
as
insightful
guidance
to
the
competition-cooperation strategy of leadership and management of an organization [3]. Though
Chinese leadership philosophy does not stress scientific evidence, it stresses situations, rule by man,
and authority and introduces the comprehensive and harmonious beauty to leadership so as to make up
for Western leadership philosophy, which focuses on system, law, and power and lacks a cultural
atmosphere. In fact, the humanization of science and the scientization of culture might be an
alternative solution to Eastern and Western leadership philosophies [49, 52]. Tseng (1987) and Tsai
(2008)deemed that Taiwan has long advocated Chinese style management. According to
Confucianism, the essence of management was to "human act", while to Legalism, it was "utilitarian
act", to Taoism, "nature act", to Mohism, "altruism act", to the Book of Changes, "humanity act", and
to Neo-confucianism, "reason act" [39]. Hsu (1992), Chu (1994), Tseng and Liu (1999), Tseng (2002),
Huang (1999), et al. all believed that Confucianism advocated "people"-oriented ideas, used
appropriate methods to gain the support of subordinates, promoted influencing others by one's own
morality, adhered to benevolence, upheld the principle of kindness to the people, and used acceptable
ideas and concepts to maintain justice and adaptability composition. This study integrated the aforementioned description on C Theory and relevant leadership and
management, used the operational procedures and characteristics of project management to explore
the operation framework and items of the application of C Theory in project management, and
summarized five major dimensions of project directors (namely, decision-making performance,
leadership performance, adaptability performance, creativity performance, and coordination
performance) and 25 management performance items of project directors. The items were adopted in
the Scale of Application of C Theory for the "Application of C Theory and Organizational
Performance Questionnaire" of this study. See below for a detailed description. 2 The connotation of C theory and the description of relevant leadership
and management y
p
ecision-making performance of project directors (decision-making function of Taoism g p
p
j
(
g
. Consideration of the advantages and disadvantages of the organization to make a decisio g
g
g
A-2. Consideration of the opportunities and risks outside the organization to make a de pp
g
A-3. Consideration of the instruction of senior management of the organization to mak sideration of the instruction of senior management of the organization to make a decision g
A-4. Consideration of the needs of customers to make a decision. A-5. Consideration of the suggestions of project members to make a decision. B C
l
f
f
j
di
(
l f
i
f L
li
) A-5. Consideration of the suggestions of project members to make a decision. B. Control performance of project directors (control function of Legalism) gg
p
j
B. Control performance of project directors (control function of Legalism) p
p
j
g
B-1. Establish a system and regulation for the implementation of the project. B-2. Emphasize the power and responsibility performances of project members. B-3. Implement a reward and punishment system for project members of the project. B-4. Actively establish the prestige of project directors B-5. Comprehensive supervision and control of the actions of project members. B-5. Comprehensive supervision and control of the actions of project members. C. Adaptability performance of project directors (adaptability function of School of Military
Strategists) C-1. Make execution guidelines for the project according to situations and conditions. C 2 Understand the needs and characteristics of project members to adjust managemen C-2. Understand the needs and characteristics of project members to adjust management approaches C 2. Understand the needs and characteristics of project members to adjust management
C-3. Promote the recognition of the visions and goals of the project to project members. C-4. Promote the cooperation of project members to the instruction of directors. p
p
j
C-5. Change the control of the project according to changes of internal and external co . Change the control of the project according to changes of internal and external contexts 8 , 01027 (2017)
119
MATEC Web of Conferences
IMETI 2016 DOI: 10.1051/
7119010
matecconf/201
27 D. Creativity performance of project directors (creativity function of Mohism D. Creativity performance of project directors (creativity function of Mohism). D. Creativity performance of project directors (creativity function of Mohism). D-1. Encourage project members to strengthen project efficiency by creative methods. D-2. 2 The connotation of C theory and the description of relevant leadership
and management Award given to project members who develop patents. D-3. Emphasize mutual support and cooperation during the implementation of tasks. D-4. Enhance efficiency and effectiveness by creative management approaches. D-5. Actively promote the maximization of project results. E. Coordination performance of project directors (coordination function of Confucianism)
E-1. Active communication and coordination with project stakeholders. E-2. Serving as role models for project members. E-3. Emphasize the moral and ethical performance of project members. E-4. Promote mutual respect among project stakeholders. E-5. Promote harmony among project stakeholders. D-1. Encourage project members to strengthen project efficiency by creative met . Emphasize mutual support and cooperation during the implementation of tasks. D-3. Emphasize mutual support and cooperation during the implementation of tasks D-5. Actively promote the maximization of project results. E. Coordination performance of project directors (coordination function of Confucianism)
E 1 A i
i
i
d
di
i
i h
j
k h ld E. Coordination performance of project directors (coordination function of Confucianism)
E 1 Active communication and coordination with project stakeholders E-3. Emphasize the moral and ethical performance of project members. E-3. Emphasize the moral and ethical performance of project members. E-5. Promote harmony among project stakeholders. 3. The connotation and relevant description of organizational
performance There are diversified studies on an organizational performance system in the literature. This study
regarded the BSC proposed by David P. Norton and Robert S. Kaplan as its assessment system of
organizational performance. BSC was derived from the research project co-held in 1990 by David P. Norton, the CEO of Norlan Norton Institute, and Robert S. Kaplan, a professor of Harvard University,
which had invited 12 companies to participate. Their team collected and analyzed many cases on a
creativity performance assessment system and proposed a performance assessment, BSC, covering
four major dimensions: finance, customer, internal processes, and learning and growth [18]. In
response to the eagerness to learn BSC, David P. Norton and Robert S. Kaplan published The
Balanced Scorecard: Translating Strategy into Action in 1996 [12, 59, 65]. David P. Norton and
Robert S. Kaplan believed that in the information age, there were more opportunities to apply BSC
creativity strategies than that of performance assessment. Hence, they wrote a second book on BSC:
The Strategy-focused Organization: How Balanced Scorecard Companies Thrive in the New Business
Environment [13]. After the publication of the research results related to BSC by David P. Norton and Robert S. Kaplan in 1992, BSC had become a hot topic of organizational performance evaluation. Since 2000,
non-profit organizations have also introduced BSC and achieved considerable successes. BSC has
developed into a widely-used policy management system [2, 48, 60]. BSC makes organizational tasks
and strategies more specific, translates them into actions so as to create competitive advantages,
employs the linkage of the four dimensions to convert the organizational tasks and strategies into
strategic objectives and strategic performance measurement, facilitates an organization to focus on
strategic topics in terms of performance evaluation, integrates organizational resources, effectively
implements strategies, and accelerates the realization of the visions of an organization. Described as
follows, BSC contains four major dimensions: finance, customer, internal processes, and learning
and growth [12]. g
2-1 Financial dimension: Organizational profitability and investment costs. Its assessment indicators
include operating income, return on capital, and added value. 2-1 Financial dimension: Organizational profitability and investment costs. Its assessment indicators
include operating income, return on capital, and added value. 2-2 Customer dimension: Other organizations or individuals who have contact with an organization. Its indicators include customer satisfaction, profitability ratios, and increase rate of new
customers. 2-3 Process dimension: Its assessment indicators include manufacturing, marketing, after-sales
service, product development, and creativity. 3. The connotation and relevant description of organizational
performance 2-3 Process dimension: Its assessment indicators include manufacturing, marketing, after-sales
service, product development, and creativity. 2-4 Learning and growth dimension: Its assessment indicators include investments in the growth of
skills and information capacity of staff. 2-4 Learning and growth dimension: Its assessment indicators include investments in the growth of
skills and information capacity of staff. BSC was designed and implemented by an organization to enhance its management performance. However, David P. Norton and Robert S. Kaplan (2001) pointed out that BSC could achieve better
results in the improvement of governmental or non-profit organizations [48]. As for the establishment BSC was designed and implemented by an organization to enhance its management performance. However, David P. Norton and Robert S. Kaplan (2001) pointed out that BSC could achieve better
results in the improvement of governmental or non-profit organizations [48]. As for the establishment 9 , 01027 (2017)
119
MATEC Web of Conferences , 01027 (2017)
119
MATEC Web of Conferences
IMETI 2016 DOI: 10.1051/
7119010
matecconf/201
27 of the overall "performance management system" of an organization, BSC also included the visions,
goals, and mid- and long-term development plans of an organization [7]. According to the
experiences of enterprises that had introduced BSC, BSC was an arduous and time-consuming task. Its difficulty lied in how to find truly proactive management strategies. It did not easily identify
unique, "distinct, and excellence elements" superior to the existing ones of an organization [61]. In
other words, the competitiveness of an organization could be enhanced by identifying the uniqueness
of each of the four dimensions of BSC. Therefore, an organization must re-examine and adjust its
organizational structure, processes, and organizational culture and translate its visions and goals into
concrete and feasible strategic plans [1, 8, 40]. It is thus necessary to establish a good performance
assessment system and appropriate performance assessment indicators to sustain the quality of project
management and operations of an organization. Therefore, while introducing BSC to project
management, an organization must consider its characteristics, internal and external resources, and
measurable indicators inside the organization. g
Based on technical ability index 1.06 of IPMA ICB 3.0, Project Management Organization is made
up of humans and a relevant basic structure. Project management includes project design and
maintenance of appropriate roles, organization, structure, responsibility, and ability. The organization
of a project and program is unique and temporary. 3. The connotation and relevant description of organizational
performance Its stages or conditions are adapted to conform to
the project life cycle. A portfolio project organization is similar to a permanent organization and can
be regarded as a permanent sector. Even so, a project organization usually has a shorter circle and
changes faster than other permanent organizational sectors [19]. The performance of a project management organization depends on the capacity of all the staff. The capacity of the staff assigned to participate in the project must be tested, and the executive
performance of members shall be verified by line management. The optimal operation model is to
provide an appropriate way for line managers to discuss with project members together so as to make
an individual fulfill a specific role on the team. The knowledge, skills, and experiences of an
individual should be reviewed. A line manager shall also consider whether the personality of an
individual can match with those of other members. When they do not participate in the selection of
team members, line managers must overcome obstacles to develop a project team made up of team
members with highly different characteristics [27]. Kao (2012) pointed out that each organization, after a period of operation, evolves into a certain
structure that might not necessarily be most suitable for the task and the environment. Different
organizational structures have different characteristics. For instance, a centralized and hierarchical
structure reduces coordination costs, but hinders communication. Democratic and equal structures
encourage communication, but have higher coordination costs. In addition, the size of a team also has
an effect on its efficiency, but small organizations tend to have better cohesion, low additional cost,
and higher morale. Before a team is formed, organization and structure factors should all be
considered. For long-term projects, job satisfaction should also be considered so as to reduce staff
turnover. Based on the above BSC connotation and the organizational characteristics of project management,
the organizational performance assessment system of project management includes learning and
growth, internal operation, customer satisfaction, and financial control and their items. The twenty
items of organizational performance of project management proposed were adopted in the Scale of
Application of C Theory from the "Application of C Theory and Organizational Performance
Questionnaire" of this study. See below for a detailed description. D. Finance control D-1. The organization could make an accurate budget based on real business needs. D 1. The organization could make an accurate budget based on real business needs. D-2. The organization could establish an appropriate budget implementation, supervision, and
evaluation mechanism. g
g
D-2. The organization could establish an appropriate budget implementation, supervision, and
evaluation mechanism. D-3. The organization could assess the efficiency of budget implementation on a regular basis. D-4. The organization could develop a variety of R&D plans to seek funds. 5 D-3. The organization could assess the efficiency of budget implementation on a regular basis. D-4. The organization could develop a variety of R&D plans to seek funds. D-5. ROI gradually increases. 4. Establishment of application of C theory and organizationa
performance questionnaire This study summarized and analyzed the relevant literature, edited questions on the application of C
Theory and organizational performance, and employed statistical analysis methods like expert validity,
configuration validity, and reliability in order to establish the Application of C Theory and
Organizational Performance Questionnaire. A. Learning and growth Customer satisfaction to the organization's services gradually increases. C-4. Customers’ re-visit rate gradually increases. C-4. Customers’ re-visit rate gradually increases. C-5. Customers actively introduce new customers. C-5. Customers actively introduce new customers. A. Learning and growth A. Learning and growth A. Learning and growth rganization members could actively participate in each advanced study activity. A-1. Organization members could actively participate in each advanced study activity. A-2. Organization members could actively analyze issues to solve and improve them. A-3. Organization members could actively publish research results and share succ A-3. Organization members could actively publish research results and share success experien
A 4 Th
i
ti
ld
ti
l
t ff t i
t t
h
k
f rganization members could actively publish research results and share success experience
he organization could actively encourage staff to innovate to enhance work performance. h
i
i
ld
bli h
d
l
d
ll
ff
l l A-5. The organization could establish a reward system to openly award excellent staff regularly. B. Internal operation B-1. Members could identify the visions, goals, and business strategies of the organization. 10 , 01027 (2017)
119
MATEC Web of Conferences
IMETI 2016 DOI: 10.1051/
7119010
matecconf/201
27 B-2. The organization could set up appropriate and feasible business development plans. B-3. The organization could establish clear rules, regulations, and work norms. B-4. The organization could arrange tasks among staff according to their expertise and
responsibilities. B-5. The organization could use diversified approaches to enhance work performance. C. Customer satisfaction
C-1. The organization could establish good channels to communicate with customers. C-2. The organization could respond positively to the reactions and requirements of custome
C-3. Customer satisfaction to the organization's services gradually increases. C-4. Customers’ re-visit rate gradually increases. C-5. Customers actively introduce new customers. D. Finance control
D-1. The organization could make an accurate budget based on real business needs. D-2. The organization could establish an appropriate budget implementation, supervision, an
evaluation mechanism. D-3. The organization could assess the efficiency of budget implementation on a regular bas
D-4. The organization could develop a variety of R&D plans to seek funds. D-5. ROI gradually increases. B-2. The organization could set up appropriate and feasible business development plan B-3. The organization could establish clear rules, regulations, and work norms. B-4. The organization could arrange tasks among staff according to their expertise and
responsibilities. B-5. The organization could use diversified approaches to enhance work performance. C. Customer satisfaction C-1. The organization could establish good channels to communicate with customers. . The organization could respond positively to the reactions and requirements of custome C-3. Customer satisfaction to the organization's services gradually increases. C-3. 4.2.1.1 SEM analysis of the scale of application of C Theory (5) The factor loadings from questions E-1 to E-5 of coordination performance of project directors
(coordination function of Confucianism) in order are: 0.60, 0.53, 0.54, 0.58, and 0.56. 4.2.1 Scale of application of C Theory Through the results of expert validity analysis, this study kept the five major dimensions (namely,
decision-making performance, control performance, adaptability performance, creativity performance,
and coordination performance of project directors) and 25 items in the first draft. Some text of the
questions was edited, while some questions were converged. The aforementioned questionnaire was
then formed and SEM analysis and Cronbach's analysis were conducted. The results are as follows. 4.1 Method to establish the application of C Theory and organizational performance
questionnaire Based on the aforementioned literature discussion and analysis, this study summarized and edited "a
first draft of Application of C Theory and Organizational Performance Questionnaire" and invited
project management practitioners, university professors, and scholars and experts in the field of
project management as auditors of the first draft of the questionnaire. In terms of project management
practitioners, it invited 2 project managers and 3 project members. In terms of university professors or
scholars in the field of project management, it invited a total of 10 scholars and researchers from
National Pingtung University, National Kaohsiung First University of Science and Technology,
Republic of China Military Academy, Chang Jung Christian University, Aletheia University, Cheng
Shiu University, Tajen University, and Open University of Kaohsiung through the help of Taiwan
Project Management Association. After the audit of the questionnaire’s first draft, a pre-test of the questionnaire should be conducted. A total of 160 copies were sent, while 142 valid feedbacks were collected to establish its validity and
reliability. For validity, this study adopted Structural Equation Modeling (SEM) to analyze factors. One of the most important characteristics of SEM is to be established on certain theories. In other
words, SEM is a statistical technique that verifies the fitness of a priori theoretical model [9]. The
content and nature of latent variables should be clearly described at the beginning of the test or have a
specific theoretical basis, and the corresponding composition model used to observe variables should
be determined. The purpose of analysis was to verify the fitness of the priori factor structure. The
procedures could be used to develop test tools as well as the effectiveness of the structure and
theoretical framework. Hence, it is also called confirmatory factor analysis (CFA)[10, 62]. As for the "Scale of Application of C Theory", this study referred to project management
knowledge systems like ISO10006, IPMA, PMI, and PRINCE2, edited the theoretical framework of 11 DOI: 10.1051/
7119010
matecconf/201
27 , 01027 (2017)
119
MATEC Web of Conferences
IMETI 2016 , 01027 (2017)
119
MATEC Web of Conferences IMETI 2016 the decision function of Taoism, control function of Legalism, adaptability function of School of
Military Strategists, creativity function of Mohism, and coordination function of Confucianism of C
Theory to the five major dimensions of project directors - decision-making performance, leadership
performance, adaptability performance, creativity performance, and coordination performance - and
proposed 25 performance items of project directors. 4.2 Results of the scales of application of C Theory and organizational performance The results of expert validity and statistical analysis are described as follows from the two aspects of
the Scale of Application of C Theory and the Scale of Organizational Performance. 4.2.1 Scale of application of C Theory 4.2.1.1 SEM analysis of the scale of application of C Theory Based on SEM analysis with LISREL, as the scale contained many questions, it could not reflect the
relationship road map, but the degrees of freedom (DOF) and chi-square statistics of the overall model
were obtained, wherein DOF were 2043 while the chi-square value was 11171.83 and the P value was
0.0, indicating that there were significant differences between the hypothetical model and the
observed values. In addition, other adapter indices like NFI, NNFI, and CFI were more than 0.95,
indicating that each question and dimension had good structure. The factor loading of the questions of
each level is as follows. (1) The factor loadings from questions A-1 to A-5 of decision-making performance of project
directors (decision-making function of Taoism) in order are: 0.53, 0.49, 0.53, 0.57, and 0.53. (1) The factor loadings from questions A-1 to A-5 of decision-making performance of project
directors (decision-making function of Taoism) in order are: 0.53, 0.49, 0.53, 0.57, and 0.53. g
(2) The factor loadings from questions B-1 to B-5 of control performance of project directors (control
function of Legalism) in order are: 0.51, 0.57, 0.50, 0.53, and 0.50. (2) The factor loadings from questions B-1 to B-5 of control performance of project directors (control
function of Legalism) in order are: 0.51, 0.57, 0.50, 0.53, and 0.50. g
(3) The factor loadings from questions C-1 to C-5 of adaptability performance of project directors
(adaptability function of School of Military Strategists) in order are: 0.53, 0.58, 0.52, 0.55, and
0.55. g
(3) The factor loadings from questions C-1 to C-5 of adaptability performance of project directors
(adaptability function of School of Military Strategists) in order are: 0.53, 0.58, 0.52, 0.55, and
0.55. (3) The factor loadings from questions C-1 to C-5 of adaptability performance of project directors
(adaptability function of School of Military Strategists) in order are: 0.53, 0.58, 0.52, 0.55, and
0.55. (4) The factor loadings from questions D-1 to D-5 of creativity performance of project directors
(creativity function of Mohism) in order are: 0.61, 0.55, 0.64, 0.62, and 0.61. (4) The factor loadings from questions D-1 to D-5 of creativity performance of project directors
(creativity function of Mohism) in order are: 0.61, 0.55, 0.64, 0.62, and 0.61. (5) The factor loadings from questions E-1 to E-5 of coordination performance of project directors
(coordination function of Confucianism) in order are: 0.60, 0.53, 0.54, 0.58, and 0.56. 4.1 Method to establish the application of C Theory and organizational performance
questionnaire As for the "Organizational Performance Scale",
this study then referred to the BSC theoretical framework of David P. Norton and Robert S. Kaplan,
adapted it to four major dimensions - finance control, customer identity, internal operation, and
learning and growth - and proposed 20 assessment items. The above two scales had clear theoretical
bases and models. Hence, in terms of the validity analysis of the questionnaire, LISREL was adopted
for use of CFA on the structure equation in order to understand the internal factor model. As for
reliability, this study adopted SPSS to conduct Cronbach's analysis to understand the internal
consistency of the questions of each level. the decision function of Taoism, control function of Legalism, adaptability function of School of
Military Strategists, creativity function of Mohism, and coordination function of Confucianism of C
Theory to the five major dimensions of project directors - decision-making performance, leadership
performance, adaptability performance, creativity performance, and coordination performance - and
proposed 25 performance items of project directors. As for the "Organizational Performance Scale",
this study then referred to the BSC theoretical framework of David P. Norton and Robert S. Kaplan,
adapted it to four major dimensions - finance control, customer identity, internal operation, and
learning and growth - and proposed 20 assessment items. The above two scales had clear theoretical
bases and models. Hence, in terms of the validity analysis of the questionnaire, LISREL was adopted
for use of CFA on the structure equation in order to understand the internal factor model. As for
reliability, this study adopted SPSS to conduct Cronbach's analysis to understand the internal
consistency of the questions of each level. 4.2.1.2 Cronbach's analysis of the scale of application of C Theory (1) As for the reliability of decision-making performance of project directors (decision-making
function of Taoism), was 0.85, indicating that their decision-making performance had a
certain degree of internal consistency. The correlation between each question and the total 12 , 01027 (2017)
119
MATEC Web of Conferences
IMETI 2016 DOI: 10.1051/
7119010
matecconf/201
27 questions was bigger than 0.57, indicating that each dimension had high internal consis questions was bigger than 0.57, indicating that each dimension had high internal consistency. questions was bigger than 0.57, indicating that each dimension had high internal consistency. (2) As for the reliability of control performance of project directors (control function of Legalism),
was 0.86, indicating that their control performance had a certain degree of internal
consistency. The correlation between each question and the total questions was bigger than
0.61, indicating that each dimension had high internal consistency. (3) As for reliability of adaptability performance of project directors (adaptability function of
School of Military Strategists), was 0.86, indicating that their adaptability performance had a
certain degree of internal consistency. The correlation between each question and the total
questions was bigger than 0.65, indicating that each dimension had high internal consistency. (4) As for the reliability of creativity performance of project directors (creativity function of
Mohism), was 0.89, indicating that their creativity performance had a certain degree of
internal consistency. The correlation between each question and the total questions was bigger
than 0.69, indicating that each dimension had high internal consistency. (5) As for the reliability of coordination performance of project directors (coordination function of
Confucianism), was 0.88, indicating that their coordination performance had a certain degree
of internal consistency. The correlation between each question and the total questions was
bigger than 0.66, indicating that each dimension had high internal consistency. Nunnally (1978) believed that more than 0.7 indicated high reliability cited from [62]. Therefore, from the Cronbach reliability analysis results of this study, the scale had a high degree
of reliability and consistency. Nunnally (1978) believed that more than 0.7 indicated high reliability cited from [62]. Therefore, from the Cronbach reliability analysis results of this study, the scale had a high degree
of reliability and consistency. 4.2.2 Scale of organizational performance Through the results of expert validity analysis, this study kept the four major dimensions (namely,
learning and growth, internal operation, customer identity, and finance control) and 20 items in the
first draft. Some text of the questions was edited, while some questions were converged, and the
aforementioned questionnaire was formed. Lastly, SEM analysis and Cronbach's analysis were
conducted. The results are as follows. 4.2.1.2 Cronbach's analysis of the scale of application of C Theory In summary, the five dimensions and their respective questions of the Scale of Application of C
Theory were all kept for the formal questionnaire in order to understand the five performances of the
application of C Theory: decision-making performance, control performance, adaptability
performance, creativity performance, and coordination performance of project directors. 4.2.2.2 Cronbach's analysis of the scale of organizational performance (1)
As for the reliability of learning and growth, was 0.87, indicating that learning and growth had
a certain degree of internal consistency. The correlation between each question and the total
questions was bigger than 0.65, indicating that each dimension had high internal consistency. (1)
As for the reliability of learning and growth, was 0.87, indicating that learning and growth had
a certain degree of internal consistency. The correlation between each question and the total
questions was bigger than 0.65, indicating that each dimension had high internal consistency. (2)
As for the reliability of internal operation, was 0.86, indicating that internal operation had a
certain degree of internal consistency. The correlation between each question and the total
questions was bigger than 0.64, indicating that each dimension had high internal consistency. (2)
As for the reliability of internal operation, was 0.86, indicating that internal operation had a
certain degree of internal consistency. The correlation between each question and the total
questions was bigger than 0.64, indicating that each dimension had high internal consistency. (3)
As for the reliability of customer satisfaction, was 0.87, indicating that customer satisfaction
had a certain degree of internal consistency. The correlation between each question and the total
questions was bigger than 0.68, indicating that each dimension had high internal consistency. (4)
As for the reliability of finance control, was 0.85, indicating that finance control had a certain
degree of internal consistency. The correlation between each question and the total questions was
bigger than 0.62, indicating that each dimension had high internal consistency. (4)
As for the reliability of finance control, was 0.85, indicating that finance control had a certain
degree of internal consistency. The correlation between each question and the total questions was
bigger than 0.62, indicating that each dimension had high internal consistency. Nunnally (1978) believed that more than 0.7 indicated high reliability cited from [62]. Therefore,
from the Cronbach reliability analysis results of this study, the scale had a high degree of reliability
and consistency. In summary, the four dimensions and their respective questions of the Scale of Organizational
Performance were all kept for the formal questionnaire in order to understand the four organizational
performances: learning and growth, internal operation, customer identity, and finance control. 5. Research results After the formal Questionnaire of Application of C Theory and Organizational Performance was
completed, the cluster random sampling method was adopted to select 800 subjects to receive the
survey, with 713 valid feedbacks collected. After statistical analyses like descriptive statistics, t test,
One-way ANOVA, and regression analysis, the results are described as follows. 5.1.1 Gender Among the valid samples, there were 314 males, accounting for 44.0% of all respondents; while there
were 399 females, accounting for 56.0%. The proportion of women was higher than that of men. 5.1.3 Education Most of them (403) graduated from vocational or senior high schools, accounting for 56.5% of all
respondents, followed by college and university graduates (269), accounting for 37.7%. 5.1.2 Age Most of the respondents (522) were between 41 and 60 years old, accounting for 73.2% of all
respondents, followed by the respondents (109) between 21 and 40 years old, accounting for 15.3%. 4.2.2.1 SEM analysis of the scale of organizational performance Based on the SEM analysis with LISREL, as the scale contained many questions, it could not reflect
the relationship road map, but the degrees of freedom (DOF) and chi-square statistics of the overall
model were obtained, wherein DOF were 395 while the chi-square value was 2930.37 and the P value
was 0.0, indicating that there were significant differences between the hypothetical model and the
observed values. In addition, other adapter indices like NFI, NNFI, and CFI were more than 0.95,
indicating that each question and dimension had good structure. The factor loading of the questions of
each level is as follows. (1) The factor loadings from questions A-1 to A-5 of learning and growth in order are: 0.60, 0.62,
0.58, 0.59, and 0.57. (1) The factor loadings from questions A-1 to A-5 of learning and growth in order are: 0.60, 0.62,
0.58, 0.59, and 0.57. (2) The factor loadings from questions B-1 to B-5 of internal operation in order are: 0.58, 0.61, 0.61,
0.63, and 0.56. (3) The factor loadings from questions C-1 to C-5 of customer identity in order are: 0.64, 0.63, 0.60,
0.59, and 0.57. (4) The factor loadings from questions D-1 to D-5 of finance control in order are: 0.58, 0.55, 0.56,
0.56, and 0.51. (4) The factor loadings from questions D-1 to D-5 of finance control in order are: 0.58, 0.55, 0.56,
0.56, and 0.51. 13 , 01027 (2017)
119
MATEC Web of Conferences
IMETI 2016 DOI: 10.1051/
7119010
matecconf/201
27 , 01027 (2017)
119
MATEC Web of Conferences
IMETI 2016 DOI: 10.1051/
7119010
matecconf/201
27 5.2.1 Difference analysis of gender The T test was used to compare the differences between men and women in terms of their application
of C Theory and organizational performance. Significant differences were found in terms of
adaptability performance, creativity performance, and coordination performance (T value = 2.79, P
value = 0.01; T value = 2.68, P value = 0.01; T value = 2.54, P value = 0.01). The values of men were
higher than those of women. There were also differences in terms of learning and growth and
customer satisfaction (T value = 2.06, P value = 0.04; T value = 2.06, P value = 0.04). The values of
men were higher than those of women too. 5.1.5 Position Most of them were employees (525), accounting for 73.6% of all respondents, followed by persons in
charge (96), accounting for 13.5% of all respondents. In addition, department heads (92) accounted for
12.9%. 5.1.7 Organizational scale Most of the organizations (333) had less than 50 employees, accounting for 46.7%, followed by those
(236) of 201 employees, accounting for 33.1%. 5.1.6 Organizational history Most of the organizations (271) had a history of over 21 years, accounting for 38.0%, followed by
those of 5 years (inclusive) (183), accounting for 25.7%. In addition, those organizations between 11
and 20 years (158) accounted for 22.2%, while those between 6 and 10 years (101) accounted for
14.2%. 5.1.4 Occupation Most of them (350) were engaged in the industry of business and service, accounting for 49.1% of all
respondents, followed by civil servants (including those of military, government, education, police,
and fire departments) (240), accounting for all samples 33.7%. 14 5.2.2 Difference analysis of age One-way ANOVA was used to compare the differences among the respondents of different ages in
terms of the application of C Theory and organizational performance. Significant differences were
found for different ages in adaptability performance, creativity performance, and coordination
performance (F value = 2.56, P value = 0.03; F value = 3.05, P value = 0.01; F value = 2.33, P value =
0.04). After further multiple comparisons, it was found that the average factor values of those between
21 and 40 years old, between 41 and 60 years old, and above 61 years old were higher than the values
of those younger than 20 years old in terms of adaptability performance, creativity performance, and
coordination performance, indicating that the former three age groups had better performances than
the latter one in the three aspects. p
Based on the statistical results of learning and growth and customer satisfaction (F value = 2.55, P
value = 0.04; F value = 3.06, P value = 0.01; F value = 2.32, P value = 0.04), it was also found that the
average factor values of those between 21 and 40 years old, between 41 and 60 years old, and above
61 years old were higher than the values of those younger than 20 years old in terms of learning and
growth and customer satisfaction, indicating that the former three age groups had better performances
than the latter one in the two aspects. 15 , 01027 (2017)
119
MATEC Web of Conferences
IMETI 2016 DOI: 10.1051/
7119010
matecconf/201
27 5.2.3 Difference analysis of education One-way ANOVA was used to compare the differences among the respondents of different education
backgrounds. Significant differences were found in decision-making performance, control
performance, adaptability performance, creativity performance, coordination performance, and overall
project management (F value = 5.04, P value = 0.00; F value = 2.92, P value = 0.03; F value = 3.39, P
value = 0.02; F value = 5.46, P value = 0.00; F value = 4.16, P value = 0.00; F value = 5.02, P value =
0.00). There were also significant differences in learning and growth, internal operation, customer
satisfaction, finance control, and overall organizational performance (F value = 5.07, P value = 0.00; F
value = 9.39, P value = 0.00; F value = 5.22, P value = 0.00; F value = 3.35, P value = 0.02; F value =
4.12, P value = 0.00). After further multiple comparisons, it was found that the average factor values of those graduated
from colleges or universities or graduate institutes were higher than those graduated from senior
middle schools or vocational schools in all the dimensions of the application of C Theory and
organizational performance, indicating that the former had better cognitive perception than the latter
in the aspects. Hence, enhancement of project leaders’ education backgrounds and strengthening of
in-service learning and growth had significant benefits for enhancing the application of C Theory and
organizational performances. 5.2.4 Difference analysis of occupation One-way ANOVA was used to compare the differences among the respondents of different
occupations. Significant differences were found in decision-making performance (F value = 5.8, P
value = 0.00), control performance (F value = 5.95, P value = 0.00), adaptability performance (F value
= 6.66, P value = 0.00), creativity performance (F value = 6.72, P value = 0.00), coordination
performance (F value = 6.72, P value = 0.00), learning and growth (F value = 5.74, P value = 0.00),
internal operation (F value = 3.18, P value = 0.01), customer satisfaction (F value = 4.07, P value =
0.00), finance control (F value = 5.36, P value = 0.00), and overall organizational performance (F
value = 3.88, P value = 0.00). ,
)
From further multiple analysis, it has been found that the average factor values of the respondents
engaged in agriculture, forestry, fishery, and animal husbandry were the lowest, while those engaged
in business and services or civil servants (including those of military, government, education, police,
and fire departments) had high factor average values. The above results show that currently Taiwanese
business and services practitioners and civil servants attach great importance to the introduction of
project management and organizational performance. Taiwanese industry is currently in the active
transition from being manufacturing-oriented to service-oriented, and the government and public
sectors are stressing services quality and improvement of administration service satisfaction as their
policy objectives. Hence, the perceptions of the respondents in project management performance and
organizational performance are significantly enhanced. In addition, the respondents engaged in
agriculture, forestry, animal husbandry, fishery, and manufacturing had low scores in the recognition
of project management and organizational performance, mainly because the operations of systems and
processes of primary and secondary industries are more stable than those of other industries. Hence,
the practitioners in the former industries find it more difficult to perceive the improvement of project
management and organizational performance. 5.2.6 Difference analysis of organizational history One-way ANOVA was used to compare the differences among organizations with different history. Significant differences were found in decision-making performance (F value = 11.21, P value = 0.00),
control performance (F value = 8.65, P value = 0.00), adaptability performance (F value = 8.57, P
value = 0.00), overall project management (F value = 10.53, P value = 0.00), learning and growth (F
value = 10.34, P value = 0.00), internal operation (F value = 10.51, P value = 0.00), customer
satisfaction (F value = 5.68, P value = 0.00), finance control (F value = 2.57, P value = 0.01), and
overall organizational performance (F value = 8.92, P value = 0.00). After further multiple comparisons, it was found that the average factor values of organizations
with a history of over 21 years were higher than those with a history of lower than 5 years (inclusive),
between 6 and 10 years, and between 11 and 20 years in all the dimensions of the application of C
Theory and organizational performance, indicating that the longer history an organization has, the
better project management and organizational performance it has. Hence, skillful application of C
Theory in project management, introduction of project management methods and procedures, and
enhancement of organizational maturity can enhance operation and management performances and
bring about sustainable operation and development opportunities and space. 5.2.5 Difference analysis of position One-way ANOVA was used to compare the differences among the respondents of different positions. Significant differences were found in decision-making performance (F value = 5.65, P value = 0.00),
control performance (F value = 5.94, P value = 0.00), adaptability performance (F value = 6.67, P
value = 0.00), creativity performance (F value = 6.65, P value = 0.00), coordination performance (F
value = 6.68, P value = 0.00), overall project management (F value = 5.75, P value = 0.00), learning 16 , 01027 (2017)
119
MATEC Web of Conferences
IMETI 2016 , 01027 (2017)
119
MATEC Web of Conferences
IMETI 2016 DOI: 10.1051/
7119010
matecconf/201
27 and growth (F value = 3.18, P value = 0.01), internal operation (F value = 4.07, P value = 0.00),
customer satisfaction (F value = 5.36, P value = 0.00), finance control (F value = 3.88, P value = 0.00),
and overall organizational performance (F value = 3.88, P value = 0.00). and growth (F value = 3.18, P value = 0.01), internal operation (F value = 4.07, P value = 0.00),
customer satisfaction (F value = 5.36, P value = 0.00), finance control (F value = 3.88, P value = 0.00),
and overall organizational performance (F value = 3.88, P value = 0.00). After further multiple comparisons, it was found that the average factor values of persons in
charge and department directors were higher than employers in all the dimensions of the application
of C Theory and organizational performance, indicating that the former have better cognitive
perception than the latter in the aspects. Hence, persons in charge and department directors took full
advantages of C Theory, actively introduced project management procedures and methods,
strengthened staff's ability of project management, and established team common sense so as to
enhance project management and organizational performance. 5.2.7 Difference analysis of organizational scale One-way ANOVA was used to compare the differences among organizations with different scales. Significant differences were found in decision-making performance (F value = 10.23, P value = 0.00),
control performance (F value = 8.72, P value = 0.00), adaptability performance (F value = 8.54, P
value = 0.00), creativity performance (F value = 8.43, P value = 0.00), overall project management (F
value = 10.53, P value = 0.00), learning and growth (F value = 10.25, P value = 0.00), internal
operation (F value = 10.64, P value = 0.00), customer satisfaction (F value = 5.71, P value = 0.00),
finance control (F value = 2.58, P value = 0.01), and overall organizational performance (F value =
8.85, P value = 0.00). )
After further multiple comparisons, it was found that the average factor values of organizations
with over 201 employees were higher than those with lower than 50 employees (inclusive), between
51 and 100 employees, and between 101 and 200 employees in decision, control, adaptability,
creativity, and overall project management, and all the dimensions of organizational performance,
indicating that the bigger an organization is, the better project management and organizational
performance it has. Hence, SMEs could apply C Theory in project management, introduce project
management methods and procedures, and enhance organizational maturity so as to enhance operation
and management performances. , 01027 (2017)
119
MATEC Web of Conferences
IMETI 2016 DOI: 10.1051/
7119010
matecconf/201
27 correlation of project management and organizational performance. It can be seen that each level and
overall dimensions of the application of "C Theory" in project management and organizational
performance have high and positive correlation (Correlation coefficients were above .70), indicating
that the higher the application of "C Theory" in project management is, the better the organizational
performance is. correlation of project management and organizational performance. It can be seen that each level and
overall dimensions of the application of "C Theory" in project management and organizational
performance have high and positive correlation (Correlation coefficients were above .70), indicating
that the higher the application of "C Theory" in project management is, the better the organizational
performance is. Table 1. Product Moment Correlation Analysis between Project Management and Organizational Performance. Correlation Coefficient
Organizational Performance
Learning and
growth
Internal
operation
Customer
satisfaction
Finance
control
Overall
organizational
performance
Application of
"C Theory" in
project
management
Decision-making
performance
.710***
.725***
.703***
.710***
.762***
Control
performance
.797***
.770***
.778***
.705***
.710***
Adaptability
performance
.844***
.705***
.722***
.785***
.774***
Creativity
performance
.829***
.853***
.856***
.797***
.844***
Coordination
performance
.845***
.714***
.752***
.792***
.760***
Overall
project
management
.760***
.760***
.762***
.778***
.799***
*** p<.001 uct Moment Correlation Analysis between Project Management and Organizational Performance. Table 1. Product Moment Correlation Analysis between Project Management and Organizationa 5.4 Regression and predictive power analysis of project management for
organizational performance 5.4.1 Predictive power for learning and growth
Table 2. Predictive Power of Project Management in Learning and Growth. Unstandardized
Coefficient
Standardized
Coefficient
t
Significance Multicollinearity
Statistics
Estimated
value of B
Standard
error
Beta
distribution
Tolerance VIF
(Constant)
0.38
0.09
4.00
0.00
Decision-making
performance
0.27
0.05
0.28
5.66
0.00
0.31
3.15
Adaptability
performance
0.28
0.06
0.23
3.44
0.00
0.18
4.14
Creativity
performance
0.33
0.06
0.36
5.46
0.00
0.21
4.04
Dependent variable: Learning and growth
Rsquare: 0.65
Stepwise multiple regression analysis method was adopted to analyze the predictive power of each
factor of the application of "C Theory" in project management for learning and growth. Table 2 shows 5.4.1 Predictive power for learning and growth Stepwise multiple regression analysis method was adopted to analyze the predictive power of each
factor of the application of "C Theory" in project management for learning and growth. 5.3 Correlation analysis of the application of C Theory and organizational performance Pearson product moment correlation was adopted to analyze the correlation between project
management and organizational performance. Table 1 lists the date related to product moment 17 , 01027 (2017)
119
MATEC Web of Conferences
IMETI 2016 Stepwise multiple regression analysis method was adopted to analyze the predictive power of each
factor of the application of "C Theory" in project management for learning and growth. Table 2 shows
that in the criterion variable of learning and growth, decision-making performance, adaptability
performance, and creativity performance reach significant levels (T value = 5.66, P value = 0.00; T
value = 3.44, P value = 0.00; T value = 5.46, P value = 0.00), wherein the influence of creativity , 01027 (2017)
119
MATEC Web of Conferences
IMETI 2016 Table 2 shows
that in the criterion variable of learning and growth, decision-making performance, adaptability
performance, and creativity performance reach significant levels (T value = 5.66, P value = 0.00; T
value = 3.44, P value = 0.00; T value = 5.46, P value = 0.00), wherein the influence of creativity 18 , 01027 (2017)
119
MATEC Web of Conferences
IMETI 2016 DOI: 10.1051/
7119010
matecconf/201
27 performance on learning and growth is the biggest, followed by that of decision-making performance. The standardized regression coefficients of the three independent variables are positive, indicating that
their influences on learning and growth are positive. The degrees of tolerance are about between 2 and
3. The expansion coefficients of variation are below 5, lower than the evaluation index value of 10,
indicating that the issue of multicollinearity among the independent variables is not obvious. performance on learning and growth is the biggest, followed by that of decision-making performance. The standardized regression coefficients of the three independent variables are positive, indicating that
their influences on learning and growth are positive. The degrees of tolerance are about between 2 and
3. The expansion coefficients of variation are below 5, lower than the evaluation index value of 10,
indicating that the issue of multicollinearity among the independent variables is not obvious. 5.4.2 Predictive power for internal operation Stepwise multiple regression analysis method was adopted to analyze the predictive power of each
factor of the application of "C Theory" in project management for internal operation. Table 3 shows
that in the criterion variable of internal operation, control performance, adaptability performance, and
coordination performance reach significant levels (T value = 4.85, P value = 5.92; T value = 5.92, P
value = 0.00; T value = 3.31, P value = 0.00), wherein the influence of adaptability performance on
internal operation is the biggest, followed by that of control performance. The standardized regression
coefficients of the three independent variables are positive, indicating that their influences on learning
and growth are positive. The degrees of tolerance are about between 2 and 3. The expansion
coefficients of variation are lower than the evaluation index value of 5 and thus less than the
evaluation index value of 10, indicating that the issue of multicollinearity among the independent
variables is not obvious. Table 3. Predictive Power of Project Management in Internal Operation. Unstandardized
Coefficient
Standardized
Coefficient
t
Significance Multicollinearity
Statistics
Estimated
value of B
Standard
error
Beta
distribution
Tolerance VIF
(Constant)
0.33
0.09
3.51 0.00
Decision-making
performance
0.30
0.06
0.30
4.85 0.00
0.23
4.33
Adaptability
performance
0.36
0.06
0.38
5.92 0.00
0.21
4.86
Coordination
performance
0.27
0.07
0.27
3.31 0.00
0.19
4.34
Dependent variable: Internal operation
Rsquare: 0.63 Table 3. Predictive Power of Project Management in Internal Operation. , 01027 (2017)
119
MATEC Web of Conferences
IMETI 2016 DOI: 10.1051/
7119010
matecconf/201
27 Table 4. Predictive Power of Project Management in Customer Satisfaction. Unstandardized
Coefficient
Standardized
Coefficient
t
Significance Multicollinearity
Statistics
Estimated
value of B
Standard
error
Beta
distribution
Tolerance VIF
(Constant)
0.02
0.07
0.32 0.75
Decision-making
performance
0.27
0.05
0.25
5.86 0.00
0.22
4.65
Adaptability
performance
0.13
0.05
0.12
2.57 0.01
0.18
4.77
Creativity
performance
0.31
0.05
0.31
6.05 0.00
0.16
4.85
Coordination
performance
0.28
0.05
0.28
6.17 0.00
0.19
4.87
Dependent variable: Customer satisfaction
Rsquare: 0.83 Table 4. Predictive Power of Project Management in Customer Satisfaction. 5.4.4 Predictive power for finance control Stepwise multiple regression analysis method was adopted to analyze the predictive power of each
factor of the application of "C Theory" in project management for finance control. Table 4 shows that
in the criterion variable of finance control, decision-making performance, control performance,
creativity performance, and coordination performance reach significant levels (T value = 5.42, P value
= 0.00; T value = 11.18, P value = 0.00; T value = 4.52, P value = 0.00), wherein the influence of
control performance on learning and growth is the biggest, followed by that of decision-making
performance. The standardized regression coefficients of the three independent variables are positive,
indicating that their influences on learning and growth are positive. The degrees of tolerance are about
between 2 and 3. The expansion coefficients of variation are below 5, lower than the evaluation index
value of 10, indicating that the issue of multicollinearity among the independent variables is not
obvious. Table 5. Predictive Power of Project Management in Finance Control. Unstandardized
Coefficient
Standardized
Coefficient
t
Significance Multicollinearity
Statistics
Estimated
value of B
Standard
error
Beta
distribution
Tolerance VIF
(Constant)
0.00
0.07
0.05 0.96
Decision-making
performance
0.25
0.04
0.22
5.42 0.00
0.26
3.87
Control performance 0.56
0.05
0.53
11.18 0.00
0.18
4.44
Creativity
performance
0.23
0.05
0.21
4.52 0.00
0.19
4.15
Dependent variable: Finance control
Rsquare: 0.82 Table 5. Predictive Power of Project Management in Finance Control. 5.4.3 Predictive power for customer satisfaction Stepwise multiple regression analysis method was adopted to analyze the predictive power of each
factor of the application of "C Theory" in project management for customer satisfaction. Table 4
shows that in the criterion variable of customer satisfaction, decision-making performance,
adaptability performance, creativity performance, and coordination performance reach significant
levels (T value = 5.86, P value = 0.00; T value = 2.57, P value = 0.01; T value = 6.05, P value = 0.00;
T value = 6.17, P value = 0.00), wherein the influence of creativity performance on customer
satisfaction is the biggest, followed by that of coordination performance. The standardized regression
coefficients of the four independent variables are positive, indicating that their influences on learning
and growth are positive. The degrees of tolerance are about between 2 and 3. The expansion
coefficients of variation are below 5, lower than the evaluation index value of 10, indicating that the
issue of multicollinearity among the independent variables is not obvious. 19 , 01027 (2017)
119
MATEC Web of Conferences
IMETI 2016 6. Conclusion and suggestions In summary, this study suggests that in order to promote
organizational performance, during project management the decision-making performance of Taoism,
control performance of Legalism, tactics performance of School of Military Strategists, creative
performance of Mohism, and coordination performance of Confucianism of "C Theory" can all be
fully used to give full play to the benefits of decision-making function, control function, adaptability
function, creativity function, and coordination function and to improve the practical benefits of project
management so as to enhance organizational performance, such as learning and growth, internal
operation, customer satisfaction, and finance control.
1. M.H. Chang, The application of balanced scorecard in school performance management, Journal
of Educational Research and Development, 123, 19-25 (2004) 1. M.H. Chang, The application of balanced scorecard in school performance management, Journal
of Educational Research and Development, 123, 19-25 (2004) 2. S.C. Chang, A probing design of the balanced scorecard application in nonprofit organizationsa
case of national vocational high school of commerce, Unpublished Master’s Thesis of Chung
Yuan University (Taoyuan, Taiwan, 2002) 3. M.J. Chen, Dynamic competition (Taipei, Taiwan, 2010) 4. M.T. Chen, The secret to operation and management in the book of changes (Taipei, Taiwan,
2014) heng, C Theory—management philosophy of the book of changes (Taipei, Taiwan, 1995) 5. C.Y. Cheng, C Theory—management philosophy of the book of changes (Taipei, 6. C.Y. Cheng, Chinese business ethics and management based on the book of changes, Chinese
Journal of Management, 12 (1), 29-37 (2015) 6. C.Y. Cheng, Chinese business ethics and management based on the book of changes, Chinese
Journal of Management, 12 (1), 29-37 (2015) 7. S.T. Chiang, Study on the relationship between teacher efficiency and school efficiency of senior
middle schools—A case study of total performance scorecard, Unpublished Master’s thesis of
National Kaohsiung Normal University (Kaohsiung, Taiwan, 2008) 7. S.T. Chiang, Study on the relationship between teacher efficiency and school efficiency of senior
middle schools—A case study of total performance scorecard, Unpublished Master’s thesis of
National Kaohsiung Normal University (Kaohsiung, Taiwan, 2008) 8. H.C. Chien, Inspiration of balanced scorecard theory to the establishment of the performance
assessment indicators of primary and middle schools, Journal of Yu-Da College, 4, 89-98 (2003)
9. H.C. Chiu, Structural Equation Modeling (SEM): LISREL theories, technologies, and
applications (Taipei, Taiwan, 2003) 8. H.C. Chien, Inspiration of balanced scorecard theory to the establishment of the
assessment indicators of primary and middle schools, Journal of Yu-Da College, 4, 8 assessment indicators of primary and middle schools, Journal of Yu-Da College, 4, 89-98 (2003)
9. H.C. Chiu, Structural Equation Modeling (SEM): LISREL theories, technologies, and
applications (Taipei, Taiwan, 2003) 9. H.C. Chiu, Structural Equation Modeling (SEM): LISREL theories, techn
applications (Taipei, Taiwan, 2003) 10. H.C. Chiu, Quantitative Research and Statistical Analysis: Quantitative Research and Statistical
Analysis in Social & Behavioral Sciences (3rd Edition) (Taipei, Taiwan, 2008) 1. C.M. Chu, Management Philosophy of Confucianism: Leaders of Utopia (Taipei, Taiw 11. C.M. 6. Conclusion and suggestions According to the results of this study, it can be seen that for the differences of social variables and in
terms of female, young employees, low education backgrounds, agriculture, forestry, animal
husbandry, fishery, and production and manufacturing practitioners, average employees, organizations 20 , 01027 (2017)
119
MATEC Web of Conferences , 01027 (2017)
119
MATEC Web of Conferences
IMETI 2016 DOI: 10.1051/
7119010
matecconf/201
27 with short history and middle and small scales, measures like training courses of project management,
introduction of the knowledge system, procedures, and methods of project management, enhancement
of the function of project management and organizational maturity, that enhancement of the
application of "C Theory" in project management shall be adopted to enhance the efficiency and
performances of organizations. Furthermore, there is a highly positive correlation between the
application of "C Theory" in project management and organizational performance. There are also
highly positive correlations among creativity performance, learning and growth, and customer
satisfaction, between adaptability performance and internal operation, and between control
performance and finance control. In summary, this study suggests that in order to promote
organizational performance, during project management the decision-making performance of Taoism,
control performance of Legalism, tactics performance of School of Military Strategists, creative
performance of Mohism, and coordination performance of Confucianism of "C Theory" can all be
fully used to give full play to the benefits of decision-making function, control function, adaptability
function, creativity function, and coordination function and to improve the practical benefits of project
management so as to enhance organizational performance, such as learning and growth, internal
operation, customer satisfaction, and finance control. with short history and middle and small scales, measures like training courses of project management,
introduction of the knowledge system, procedures, and methods of project management, enhancement
of the function of project management and organizational maturity, that enhancement of the
application of "C Theory" in project management shall be adopted to enhance the efficiency and
performances of organizations. Furthermore, there is a highly positive correlation between the
application of "C Theory" in project management and organizational performance. There are also
highly positive correlations among creativity performance, learning and growth, and customer
satisfaction, between adaptability performance and internal operation, and between control
performance and finance control.
Chu, Management Philosophy of Confucianism: Leaders of Utopia (Taipei, Taiwan, 1994)
12 T K Chu (trans) Balanced scorecard (Taipei Taiwan 2008) 12. T.K. Chu (trans), Balanced scorecard (Taipei, Taiwan, 2008) 13. P.D. Norton R.S. Kaplan, The Strategy-Focused Organization: How Balanced Scorecard
Companies Thrive in the New Business Environment (Taipei, Taiwan, 2001) 14. The international project management knowledge system, Editorial Board of Taiwan Project
Management Association (2014) 15. C.Y. Hsu, Leadership from the perspective of history (Taipei, Taiwan, 1992) 16. H.Y. Hsu, P.L. Hsiung, Y.F. Chu, Y.H. Chang, M. Fan, C.M. Huang, H.T. Chou, and C.P. Lu,
Basic Project Management Knowledge and Applications in Practice (5th Edition) (New Taipei,
Taiwan, 2014) 17. K.K. Huang, Chinese corporate culture and productivity, Applied Psychology Research, 1,
163-185 (1999) 18. T.H. Huang, Application of basic concepts of balanced scorecard in creative operation of school,
Journal of Educational Research and Development, 124, 52-58 (2004) 21 , 01027 (2017)
119
MATEC Web of Conferences , 01027 (2017)
119
MATEC Web of Conferences
IMETI 2016 DOI: 10.1051/
7119010
matecconf/201
27 IMETI 2016 19. ICB: IPMA Competence Baseline, http://ipma.ch/resources/ipma-publications/ipma-competence-baseline/
(2009) 19. ICB: IPMA Competence Baseline, http://ipma.ch/resources/ipma-publications/ipma-competence-baseline/
(2009) 19. ICB: IPMA Competence Baseline, http://ipma.ch/resources/ipma-publications/ipma-competence-baseline/
(2009) 20. T.L. Kang, S.H. Chang, L.J. Yang, and S.K. Huang, Development of chinese leadership model:
An empirically-based theory, Education Policy Forum, 13 (1), 115-142 (2010) 21. T.L. Kang, K.J. Su, S.H. Chang, and S.C. Hsu, The development of the chinese leadership model
and its application in vocational schools, Bulletin of Educational Research, 47, 133-154 (2001) 22. I.C. Kao, Study on Chinese style contingency leadership-Preliminary structure of "throne”
leadership philosophy, Journal of Educational Research and Development, 7, 45-56(1999) 23. I.C. Kao, Management of Civic Knowledge: Feasible Strategies to Establish Knowledge
Management System of Educational Project of School, Cited from Wu, Y.M. and Tsai, T.C. (ed.),
Civic Urbanology (18-35) (Kaohsiung, Taiwan, 2009) 24. I.C. Kao and Y.S. Chen, A study on model of project management maturity and criteria of
assessment, Journal of Project Management, 2 (1), 73-103 (2009) 25. I.C. Kao, Construction and application of the evaluation indications for the practical operation
and organizational performance of project management in primary schools, Journal of Research
on Elementary Education, 8, 109-136 (2011) 26. I.C. Kao, Feasible strategies to apply restriction theory, knowledge management, and project
management to enhance school performance, Journal of Knowledge Economy, 8 (2), 59-82
(2011) 27. I.C.
Kao, The application of the environment analysis and organizational pattern of project
management in school organizational development, Journal of Knowledge Economy, 9 (2),
170-199 (2012) 28. I.C. Kao, The research on construction of project management indicators, Storage Management
Solutions, 1, 154-173 (2013) 29. I.C. Kao, Feasible strategies of project management processes, team formation and
institutionalized operations applied to school governance, Storage Management Solutions, 1,
198-229 (2013) 30. I.C. Kao, Application of definition and role analysis of project stakeholders in the human
resource development of school project management strategies, Storage Management Solutions, 6,
158-190 (2013) 31. I.C. Kao, Construction of human resource development strategy of community empowerment by
project management, TOC and TTQS, Storage Management Solutions, 2, 42-65 (2014) 32. I.C. Kao, Control of power and projects by “Sitting on the dragon throne”: Feasible school
project management strategies based on contingent chinese leadership, Advances in Engineering
Education, 5 (4), 38-80 (2014) 33. I.C. Kao, The Correlation between Project Management and Organizational Effectiveness at
Taiwan’s Public and Private Organizations—Students Who Receive In-Service Continuing
Education at the Open University of Kaohsiung as Survey Participants, Advances in Engineering
Education, 89-123 (2015) 34. I.C. Kao, Feasible strategies for setting the course of “Community development project
management” in adult education institutions, Storage Management Solutions, 2, 35-62 (2016) 35. H. Kerzner, Project Management: A System Approach to Planning, Scheduling, and Controlling
(9th Edition) (New York, USA, 2006) 36. C.C. Ko, Study on the correlation among chinese style leadership, organizational culture, and
organizational creativity—A case study of SMEs in Taiwan, Unpublished Doctoral thesis,
National Changhua University of Education (Changhua, Taiwan, 2012) 37. J. Lewis, Fundamentals of Project Management (2nd Edition) (New York, USA, 20
38 J Lewis Mastering Project Management (New York USA 2008) 37. J. Lewis, Fundamentals of Project Management (2nd Edition) (New York, 37. J. Lewis, Fundamentals of Project Management (2nd Edition) (New York, USA, 2002)
38. J. Lewis, Mastering Project Management (New York, USA, 2008) 38. J. Lewis, Mastering Project Management (New York, USA, 2008) 39. C.C. Li and H.C. Yu, Master sun’s art of war and the book of changes—Decision-making theory
and decision-making management (Taipei, Taiwan, 2012) 40. C.W. Liao, Study on the universality of balanced scorecard in school performance assessment,
Bulletin of Research on Elementary Education, 12, 1-17 (2004) y
41. Y.J. Lin, Study on the Book of Changes, T.J. Pan, H.T.
Lai, (ed.) Eastern and Weste y
Study on the Book of Changes, T.J. Pan, H.T. Lai, (ed.) Eastern and Western Philosophy 22 , 01027 (2017)
119
MATEC Web of Conferences , 01027 (2017)
119
MATEC Web of Conferences
IMETI 2016 DOI: 10.1051/
7119010
matecconf/201
27 IMETI 2016 and Interpretation of Ontology, Kangde Publishing Press October , 63-102(2005) 42. C.T. Liu, Breathtaking heaven, amazing humanity—The first lesson of the book of changes
(Taipei, Taiwan, 2015) 42. C.T. Liu, Breathtaking heaven, amazing humanity—The first lesson of the book of changes
(Taipei, Taiwan, 2015) 43. J.C. Lu, Wisdom of Chinese management philosophy of the pre-qin periodXYZ and L type
management, T&D Feixun, 34, 1-42 (2005) 44. K.M. Lu, Software project management (Taipei, Taiwan, 2009) 45. J.C. Nunnally, Psychometric Theory (New York, USA, 1978) 46. C.C. Peng and C. Liu, Study on contingency management theory and the philosophy of the book
of changes, Study on the Book of Changes, 79, 88-91 (2006) 47. PMI, A Guide to the Project Management Body of Knowledge (5th Edition) (Pennsylvania, USA,
2012) 48. C.W. Shih, Exploration on the performance assessment system of non-profit organizations from
the perspective of balanced scorecard—A cast study of student affairs divisions of colleges and
universities in Kaohsiung, Unpublished Master’s thesis of I-Shou University (Kaohsiung, Taiwan,
2004) 49. R.M. Si, H.E. Yin, C.H. Chen, and K.T. Hsieh, Read the Books of Changes, Master Management
(Hong Kong, China, 2010) 50. T. Peters and R.H. Waterman, In Search of Excellence (Taipei, Taiwan, 2005) 51. C.H. Tsai, Reflection and inspiration of school leadership from the perspective of Chinese style
leadership, Journal of Educational Research and Development, 168, 105-111 (2008) 52. P.T. Tsai and W.Y. Wu, Leadership Science: Theory, Practice, and Study (2nd Edition)
(Kaohsiung, Taiwan, 2013) 53. C.S. Tsao, “Easy” Mode of Thinking in Systems Thinking and its Reference Value of Modern
Management, Management Science Research, 6, 26-31 (2009) 54. S.C. Tseng, Modern Chinese style management (Taipei, Taiwan, 1987) 54. S.C. Tseng, Modern Chinese style management (Taipei, Taiwan, 1987)
55 S C Tseng Chinese style management: Management science with Chinese characteristics (Tai 55. S.C. Tseng, Chinese style management: Management science with Chinese characteristics (Taipei,
Taiwan, 2002) 56. S.C. Tseng and C.C. Liu, Interaction of three ways of management (Taipei, Taiwan, 57. C.J. Wei, The general project management knowledge system (Taipei, Taiwan, 2013 58. A.N.
Wu, The essence, scope, and integration of balanced scorecard, Accoun
Monthly, 211, 45–54 (2003) y,
,
(
)
59. A.N. Wu, Promotion of strategy and performance management by balanced scorecard, Data
Systems Consulting Co., http://www.dsc.com.tw/bsc/bsc_a/bsc_a1.htm (2004) 60. A.N. Wu, Characteristics and functions of balanced scorecard, Data Systems C
http://www.dsc.com.tw/bsc/bsc_a/bsc_a2.htm(2004) 61. H. Wu, Balanced scorecard and performance management, Accounting Research Monthly, 12 (2),
35-40 (2000) 62. M.L. Wu, Structural Equation Modeling (SEM)–The application of SIMPLIS (Taipei, Taiwan,
2006) 63. C.F. Yang, How to Study Chinese (Taipei, Taiwan, 1996) 64. W.D. Yu, Practical project management (Taipei, Taiwan, 2015) 65. Y. H. Yu and I.P. Chen, Complete guidance on balanced scorecard (Taipei, Taiwan, 2004) 23
|
https://openalex.org/W2054668300
|
https://www.ams.org/proc/2008-136-02/S0002-9939-07-09269-6/S0002-9939-07-09269-6.pdf
|
English
| null |
Simultaneous Surface Resolution in Cyclic Galois Extensions
|
Proceedings of the American Mathematical Society
| 2,007
|
public-domain
| 2,573
|
Received by the editors August 28, 2006 and, in revised form, November 9, 2006.
2000 Mathematics Subject Classification. Primary 14A05. (Communicated by Ted Chinburg) Abstract. We show that simultaneous surface resolution is not always possi-
ble in a cyclic extension whose degree is greater than three and is not divisible
by the characteristic. This answers a recent question of Ted Chinburg. c
⃝2007 American Mathematical Society
Reverts to public domain 28 years from publication PROCEEDINGS OF THE
AMERICAN MATHEMATICAL SOCIETY
Volume 136, Number 2, February 2008, Pages 449–452
S 0002-9939(07)09269-6
Article electronically published on November 1, 2007 PROCEEDINGS OF THE
AMERICAN MATHEMATICAL SOCIETY
Volume 136, Number 2, February 2008, Pages 449–452
S 0002-9939(07)09269-6
Article electronically published on November 1, 2007 PROCEEDINGS OF THE
AMERICAN MATHEMATICAL SOCIETY
Volume 136, Number 2, February 2008, Pages 449–452
S 0002-9939(07)09269-6
Article electronically published on November 1, 2007 2. Two lemmas Let M(R) denote the maximal ideal of a local ring R. In Lemma (2.1) we
summarize some properties of the integral closure of a normal noetherian domain
in a cyclic extension. In Lemma (2.2) we give a consequence of the Harbater-Pop
theorem. Lemma 2.1. Let R be a normal noetherian domain with quotient field K, let S
be the integral closure of R in a finite algebraic field extension L of K, and let
[L : K] = q. Assume that q is a unit in R and that L contains a nonzero element
z such that L = K(z) and zq = u
d
j=1
xa(j)
j where u is a unit in R, d is a nonnegative integer, a(j) is an integer such that
GCD(a(j), q) = 1 for 1 ≤j ≤d, and x1, . . . , xd are elements in R such that
x1R, . . . , xdR are pairwise distinct minimal (= height one) prime ideals in R. Let
b(i, j) and c(i, j) be the unique integers such that b(i, j) = a(j)i + c(i, j)q
and
0 ≤b(i, j) < q. b(i, j) = a(j)i + c(i, j)q
and
0 ≤b(i, j) < q. Let zi = zi
d
j=1
xc(i,j)
j
. Then we have the following: (1) (z0, · · · , zq−1) is a free R-basis of S. (1) (z0, · · · , zq−1) is a free R-basis of S. (2) If R is a local domain and d ≥1, then S is a local domain and for
its maximal ideal M(S) we have M(S) = M(R)S + (z1, · · · , zq−1)S with
S/M(S) = R/M(R). (3) If R is a regular local domain and d ≥2, then S is a nonregular local
domain. (3) If R is a regular local domain and d ≥2, then S is a nonregular local
domain. Proof. For (1) and (2) see Theorem 7 [Ab3]. For (3) see Theorem 6 [Ab2] with the
observation that, although in the context of this theorem q is a prime number, the
primeness of q was never used in its proof. A different version of (1) and (2) can
also be found in Theorems 4 and 5 [Ab2]; see Remark 2 on page 28 of [Ab3]. □ Lemma 2.2. 1. Introduction They are used in the proof of
Theorem (3.1) of Section 3, which gives a sufficient condition for a two dimensional 449 SHREERAM S. ABHYANKAR AND NAN GU 450 local domain to be nonregular. Theorem (3.1) is used in proving the special case
of Theorem (3.2) of Section 3, which corresponds to our extended result, i.e., the
H = 1 case of our generalized extended result. The general case of Theorem (3.2),
which corresponds to our generalized extended result, then follows by using Lemma
(2.2). 1. Introduction Let K be a two dimensional algebraic function field over an algebraically closed
ground field k. Recall that K/k has a minimal model means that amongst all the
nonsingular projective models of K/k there is one which is dominated by all others
(basic reference [Ab4] or [Ab5]). Also recall that K/k has a minimal model if and
only if it is not a ruled function field, i.e., K is not a simple transcendental field
extension of a one dimensional algebraic function field over k (see [Zar]). A finite
algebraic field extension L/K is said to have a simultaneous resolution if there exist
nonsingular projective models V and W of K/k and L/k, respectively, such that
W is the normalization of V in L. Given any positive integer q that is not divisible
by the characteristic char(K) of K and letting Zq denote a cyclic group of order
q, in [Ab2] it was shown that if q ≤3 and L/K is a Zq extension, i.e., a Galois
extension whose Galois group is a cyclic group of order q, then it has a simultaneous
resolution, whereas if K/k has a minimal model and q > 3 with q being a prime
number, then there exists a Zq extension L/K that has no simultaneous resolution. Here we shall extend this second result to those nonprimes q that are divisible by
the square of some prime p. By taking q = 4, this answers a question raised by Ted
Chinburg at the March 2006 AMS Meeting in New Hampshire to the effect whether
every Z2 by Z2 extension L/K, i.e., a Z2 extension L/J of a Z2 extension J/K,
has a simultaneous resolution. By using a theorem of David Harbater and Florian
Pop, we generalize our extended result by replacing Zq by its direct sum H ⊕Zq
with any finite group H. For a related matter, see [AbK]. In Lemma (2.2) of Section 2 we shall give a consequence of the Harbater-Pop
Theorem to be used in proving our generalized extended result. In Lemma (2.1) of
Section 2 we shall summarize some technical results from our previous papers [Ab2]
and [Ab3]. These technical results deal with the structure of the integral closure of
a normal noetherian domain in a cyclic extension. SIMULTANEOUS SURFACE RESOLUTION IN CYCLIC GALOIS EXTENSIONS SIMULTANEOUS SURFACE RESOLUTION IN CYCLIC GALOIS EXTENSIONS 451 Given a two dimensional algebraic function field K over k, choose a separating
transcendence basis x, y for K over k. So K is a finite separable field extension of
k(x, y). Let E be the algebraic closure of k(x) in K. E is finite over k(x), since K
is finite over k(x, y) and since k(x) is algebraically closed in k(x, y). Thus E is a
one dimensional algebraic function field over k and so, by the one-variable existence
theorem, H is the Galois group of a finite extension F of E. Since E is algebraically
closed in K and since F is algebraic over E, it follows that F and K are linearly
disjoint over E. So the compositum L = KF (in an algebraic closure of K) is a
Galois extension of K with Galois group H, completing the proof. □ 3. Two theorems In Theorem (3.1) we give a sufficient condition for a local domain to be nonreg-
ular. In Theorem (3.2) we construct our examples of simultaneous nonresolvability. Theorem 3.1. Let R be a two dimensional regular local domain, let (X, Y ) be
generators of its maximal ideal M(R), and let K be its quotient field. Let R0 = R. For all n > 0, let Yn = Y/Xn and let Rn be the localization of the ring Rn−1[Yn] at
the maximal ideal in it generated by (X, Yn). Note that then Rn is a two dimensional
regular local domain with quotient field K such that Rn dominates Rn−1 and (X, Yn)
are generators of M(Rn). Theorem 3.1. Let R be a two dimensional regular local domain, let (X, Y ) be
generators of its maximal ideal M(R), and let K be its quotient field. Let R0 = R. For all n > 0, let Yn = Y/Xn and let Rn be the localization of the ring Rn−1[Yn] at
the maximal ideal in it generated by (X, Yn). Note that then Rn is a two dimensional
regular local domain with quotient field K such that Rn dominates Rn−1 and (X, Yn)
are generators of M(Rn). Let q be a positive integer that is a unit in R. Assume that q = pm where p is a
prime number and m is a positive integer divisible by p. Assume that K contains
q distinct q-th roots of 1. Let L be a splitting field over K of the polynomial of
Zq −XY m. Let Sn be the integral closure of Rn in L. Let q be a positive integer that is a unit in R. Assume that q = pm where p is a
prime number and m is a positive integer divisible by p. Assume that K contains
q distinct q-th roots of 1. Let L be a splitting field over K of the polynomial of
Zq −XY m. Let Sn be the integral closure of Rn in L. Then L/K is a Zq extension and for every nonnegative integer n, the ring Sn is
a two dimensional nonregular local domain. Then L/K is a Zq extension and for every nonnegative integer n, the ring Sn i
a two dimensional nonregular local domain. Proof. 2. Two lemmas Let K/k be a two dimensional algebraic function field over an al-
gebraically closed ground field k. For any finite group H, there exists a Galois
extension L/K with Galois group H. Proof. It follows from Theorem 4.4 [Har] or the Corollary to Theorem A [Pop] that
given any finite group H and any one dimensional algebraic function field E over
an algebraically closed ground field k, there exists a Galois extension F/E whose
Galois group is H. The following argument, provided by Harbater and Pop, shows
how the desired two-variable existence follows from this. 3. Two theorems Let w be the discrete valuation whose valuation ring is the one dimensional
regular local domain obtained by localizing the ring R at the prime ideal in it
generated by X. Then w(XY m) = 1 and hence the polynomial Zq −XY m is
irreducible in K[Z] and L/K is a Zq extension. Let z ∈L be a root of the said
polynomial. Then zq = XY m and L = K(z). Let X = zp/Y and J = K(X). Then
X
m = X and hence J/K is a Zm extension. By (2.1)(2) the integral closure Tn of
Rn in J is a two dimensional regular local domain whose maximal ideal M(Tn) is
generated by (X, Yn). Also zp = XY = X
1+nmYn and, since m is assumed divisible
by p, upon letting ζ = z/X
nm/p we get L = J(ζ) with ζp = XYn. Now L/J is a
Zp extension with L = J(ζ), and Sn is the integral closure of Tn in L. Therefore
by (2.1)(3) we see that Sn is a two dimensional nonregular local domain. □ Theorem 3.2. Let K/k be a two dimensional algebraic function field over an
algebraically closed ground field k. Assume that K/k has a minimal model V ∗. Let
q be a positive integer that is not divisible by char(K). Assume that q = pm where
p is a prime number and m is a positive integer divisible by p. Then, given any
finite group H, there exists a Galois extension L′/K with Galois group H ⊕Zq such
that L′/K has no simultaneous resolution. Proof. By (2.2) there exists a Galois extension L/K with Galois group H. Take R
in (3.1) to be the local ring of a point of V ∗that is not ramified in L. Let L′ be
a compositum of L and L. It is easy to see that L′/K is a Galois extension whose
Galois group is H ⊕Zq. SHREERAM S ABHY SHREERAM S. ABHYANKAR AND NAN GU 452 By [Ab1, Lemma 12], there exists a unique valuation v of K dominating Rn for
all n ≥0. By construction each Rn+1 is the immediate quadratic transform of Rn
along v. Let v be an extension of v to L. Let, if possible, V and W be nonsingular
projective models of K/k and L′/k, respectively, such that W is the normalization
of V in L′. 3. Two theorems Then by the minimality of V ∗, V must dominate V ∗. Consequently by
[Ab1, Theorem 3] the local ring of the center P of v on V must equal Rn for some
nonnegative integer n. Since Rn dominates R and R is not ramified in L, Rn is not
ramified in L. Let V be the normalization of V in L, and Rn be the local ring of
the center P of v on V . Then P lies above P in V and Rn is a two dimensional
regular local ring whose maximal ideal M( Rn) is generated by (X, Yn). Now L′ is
a Zq extension of L constructed from L in the same way as L is constructed from
K in (3.1), and W is the normalization of V in L′. By (3.1), the point of W lying
above P is not a simple point, which is a contradiction. □ Remark 3.3. The construction of a Zq extension L/K having no simultaneous res-
olution does not use the results of Harbater and Pop. Their results plus the fact
that a regular system of parameters lifts to a regular system of parameters through
an unramified local ring extension allow us to mimic such a construction to get an
H ⊕Zq extension. Similar arguments will show that the statement of (3.2) remains
true if q > 3 is a prime number; see [Ab2, Theorem 11] for details. Remark 3.3. The construction of a Zq extension L/K having no simultaneous res-
olution does not use the results of Harbater and Pop. Their results plus the fact
that a regular system of parameters lifts to a regular system of parameters through
an unramified local ring extension allow us to mimic such a construction to get an
H ⊕Zq extension. Similar arguments will show that the statement of (3.2) remains
true if q > 3 is a prime number; see [Ab2, Theorem 11] for details. References [Ab1]
S. S. Abhyankar, On the valuations centered in a local domain, American Journal o
Mathematics, 78 (1956), 321-348. MR0082477 (18:556b) [Ab2]
S. S. Abhyankar, Simultaneous resolution for algebraic surfaces, American Journal o
Mathematics, 78 (1956), 761-790. MR0082722 (18:600b) [Ab3]
S. S. Abhyankar, Uniformization of Jungian local domains, Mathematische Annalen, 159
(1965), 1-43. MR0177989 (31:2247) [Ab4]
S. S. Abhyankar, Resolution of Singularities of Embedded Algebraic Surfaces, Springe
Verlag (1998). MR1617523 (99c:14021) [Ab5]
S. S. Abhyankar, Lectures on Algebra I, World Scientific, 2006. [Ab5]
S. S. Abhyankar, Lectures on Algebra I, [AbK] S. S. Abhyankar and M. Kumar, Simultaneous surface resolution in quadratic and bi-
quadratic Galois extensions, Contemporary Mathematics, 390 (2005), 1-8. MR2187320
(2006h:14017) [Har]
D. Harbater, Fundamental groups and embedding problems in characteristic p, Contem
porary Mathematics, 186 (1995), 353-369. MR1352282 (97b:14035) [Pop]
F. Pop, ´Etale Galois covers of affine smooth curves, Invent. Math., 120 (1995), 555-578
MR1334484 (96k:14011) [Zar]
O. Zariski, The problem of minimal models in the theory of algebraic surfaces, American
Journal of Mathematics, 80 (1958), 146-184. MR0097404 (20:3873) Department of Mathematics, Purdue University, West Lafayette, Indiana 47907
E-mail address: ram@cs.purdue.edu E-mail address: ram@cs.purdue.edu Department of Mathematics, Purdue University, West Lafayette, Indiana 47907
E-mail address: ngu@math.purdue.edu
|
https://openalex.org/W3213121584
|
https://jurnal.yudharta.ac.id/v2/index.php/Teknologi-Pangan/article/download/2458/1925
|
Indonesian
| null |
Produksi, kualitas dan cita rasa tempe biji labu kuning, biji bunga matahari dan kacang adzuki
|
Teknologi Pangan : Media Informasi dan Komunikasi Ilmiah Teknologi Pertanian/Teknologi pangan : media informasi dan komunikasi ilmiah teknologi pertanian
| 2,021
|
cc-by
| 5,045
|
TEKNOLOGI PANGAN : Media Informasi dan Komunikasi Ilmiah Teknologi Pertanian
Website: https://jurnal.yudharta.ac.id/v2/index.php/Teknologi-Pangan
Licensed : Creative Commons Attribution 4.0 International License. (CC-BY) TEKNOLOGI PANGAN : Media Informasi dan Komunikasi Ilmiah Teknologi Pertanian
Website: https://jurnal.yudharta.ac.id/v2/index.php/Teknologi-Pangan
Licensed : Creative Commons Attribution 4.0 International License. (CC-BY) Terakreditasi No. 36/E/KPT/2019
Volume 12, No. 2, (2021), Halaman 209-219
p-ISSN: 2087-9679, e-ISSN: 2597-436X ABSTRACT Tempeh is a traditional Indonesian fermented food that can be processed from various
types of raw materials. However, the most widely produced is soybean tempeh, so the
available types of tempeh are limited. It is necessary to study the alternative raw
material for tempeh apart from soybeans. Therefore, this study aims to examine the
taste, antioxidants, and composition of tempeh made from pumpkin seeds, sunflower
seeds, and adzuki beans. Tempeh is processed in the laboratory using soybean/ TK,
pumpkin seeds/ TL, sunflower seeds/ TBM, adzuki seeds/ TA, and a mixture of all (1: 1:
1: 1) / TC. The results showed that all of the tempeh had a white surface, a compact
texture that did not fall off when cut, and a distinctive aroma of tempeh without the
smell of ammonia. These parameters are in line with soybean tempeh regulated in SNI
3144: 2015. The protein content and fat content of all tempeh are higher than the
quality requirements of soybean tempeh which are regulated in SNI 3144: 2015. However, only the taste of TK, TBM, and TC Tempeh were well received by the
panelists. The antioxidant activity of the three types of tempeh was not significantly
different, but the highest was found in TBM Tempeh. Thus, pumpkin seeds, sunflower
seeds, and a mixture of soybeans, pumpkin seeds, sunflower seeds, adzuki beans have
the potential to be developed as raw materials for tempeh. However, further research is
still needed to examine the functional properties of the TK, TBM, and TC Tempeh and
their effects on the health and balance of the intestinal microbiota. d
h
b
k
d
d
k b
l
d Keywords: tempeh, soybean, pumpkin seeds, adzuki beans, sunflower seeds Tati Barus1)*, Jazzieca Yokota1), Rory Anthony Hutagalung1)
1Fakultas Teknobiologi, Universitas Katolik Indonesia Atma Jaya, Jakarta 12930, Indonesia
*Email korespondensi: tati.barus@atmajaya.ac.id Tati Barus1)*, Jazzieca Yokota1), Rory Anthony Hutagalung1)
1Fakultas Teknobiologi, Universitas Katolik Indonesia Atma Jaya, Jakarta 12930, Indonesia
*Email korespondensi: tati.barus@atmajaya.ac.id Informasi Artikel:
Dikirim: 16/02/2021; disetujui: 15/07/2021; diterbitkan: 25/09/2021 ABSTRAK Salah satu makanan fermentasi khas Indonesia adalah tempe. Tempe dapat diolah dari
berbagai jenis bahan baku. Namun tempe yang banyak diproduksi dan dikonsumsi
adalah tempe kedelai sehingga ragam jenis tempe yang ada terbatas. Dengan demikian
perlu dikaji terus alternatif bahan baku tempe selain kedelai. Oleh sebab itu, penelitian
ini bertujuan membandingkan cita rasa, antioksidan dan komposisi tempe dari bahan
kedelai dengan tempe bahan baku biji bunga matahari, kacang adzuki, biji labu kuning
dan campuran semua. Pembuatan tempe dilakukan di laboratorium menggunakan bahan
baku: kedelai/TK, biji bunga matahari/Tempe TBM, biji adzuki /Tempe TA, biji labu
kuning/Tempe TL dan campuran semua (1:1:1:1)/ Tempe TC. Hasil penelitian
menunjukkan bahwa semua tempe memiliki permukaan yang warna putih, tekstur
kompak sehingga saat dipotong tidak rontok, dan aromanya berbau khas tempe tanpa
bau amonia. Parameter tersebut sesuai dengan syarat mutu tempe kedelai yang diatur
pada SNI 3144:2015. Berdasarkan hasil uji organoleptik, semua tempe, kecuali Tempe 209 DOI: https://doi.org/10.35891/tp.v12i2.2458 Barus, Yokota, dan Hutagalung Volume 12, No.2, (2021), Halaman 209-219 TA, memiliki cita rasa yang diterima dengan baik oleh panelis. Semua tempe, kecuali
Tempe TA, memiliki kadar protein dan kadar lemak memenuhi syarat mutu tempe
kedelai pada SNI 3144:2015. Semua tempe memiliki aktivitas antioksidan namun tidak
berbeda nyata, dan yang tertinggi ditemukan pada Tempe TBM. Dengan demikian maka
biji labu kuning, biji bunga matahari, dan campuran kedelai, biji labu kuning, biji bunga
matahari, kacang adzuki (1:1:1:1:1) berpotensi dikembangkan sebagai bahan baku
tempe. Penelitian lebih lanjut masih diperlukan untuk mengkaji sifat fungsional Tempe:
TK, TBM dan TC tersebut dan mengkaji pengaruhnya terhadap kesehatan dan
keseimbangan mikrobiota usus. Kata kunci: tempe, kedelai, biji labu kuning, kacang adzuki, biji bunga matahari PENDAHULUAN protein dari bahan pangan lain, seperti telur,
ikan dan daging. Handajani et al., 2020
melaporkan mengkonsumsi tempe dapat
meningkatkan kemampuan kognitif pada
para lanjut usia. Makanan fermentasi dapat berperan
menjaga
keseimbangan
komunitas
mikrobiota usus yang baik. Publikasi yang
melaporkan pentingnya peran mikrobiota
usus terhadap kesehatan terus meningkat. Sejak tahun 1998 hingga 2018 telah tercatat
2,891 publikasi penelitian ilmiah terkait
mikrobiota usus (Huang et al., 2019). Telah
dilaporkan mikrobiota usus berpengaruh
secara langsung maupun tidak langsung
terhadap kesehatan mental (Clapp et al.,
2017), autoimmun (Zhang, Zhao dan Li,
2020), alergi (Berni et al., 2019) dan
hipertensi (Daliri et al., 2020). Mikrobiota
usus berhubungan serat dengan fungsi otak
(Collins, Surette dan Bercik, 2012) sehingga
ada istilah usus merupakan otak kedua
manusia
sehingga
dapat
mempengaruhi
perilaku dan mood seseorang. Oleh sebab itu
sangat penting untuk menjaga kesehatan dan
keseimbangan mikrobiota usus yang baik. Salah
satu
jenis
makanan
yang
baik
dikonsumsi untuk untuk menjaga kesehatan
mikrobiota usus adalah makanan fermentasi. Tempe
yang
diproduksi
dan
dikonsumsi pada umumnya adalah tempe
kedelai. Namun, pembuatan tempe dapat
dilakukan dari bahan baku selain kedelai,
seperti biji labu kuning dan biji bunga
matahari. Kedua jenis biji-bijian ini
memiliki kandungan essential oil, vitamin E
dan senyawa asam folat yang baik. Biji labu
kuning telah dilaporkan memiliki kadar
fenolik yang tinggi (Gumolung, 2018). Selain
biji labu kuning dan biji bunga matahari, ada
juga kacang adzuki yang memiliki aktivitas
antioksidan yang tinggi (Amarowicz et al.,
2007). Penelitian ini bertujuan untuk mengkaji
sifat sensori, aktivitas antioksidan dan
komposisi kimia tempe yang diolah dari
bahan baku: biji bunga matahari, kacang
adzuki, biji labu kuning, dan campuran
semuanya dengan perbandingan 1:1:1:1. Implikasi dari hasil penelitian ini adalah
untuk mengurangi ketergantungan terhadap
impor kedelai dan menambah ragam jenis
tempe yang diproduksi dan dikonsumsi di
Indonesia. Tempe
makanan
fermentasi
khas
Indonesia. Mengkonsumsi
tempe
dapat
meningkatkan
meningkatkan
jumlah
Akkermansia
muciniphila pada
sistem
pencernaan
(Stephanie
et
al.,
2017). Keberadaan
A. muciniphila pada
usus
penting untuk kesehatan (Dao et al., 2016;
Depommier et al., 2019). Tempe juga telah
dilaporkan mengandung vitamin B12 (Keuth
dan Bisping, 1994), mengandung antioksidan
(Barus, Maya dan Hartanti, 2019) dan
merupakan sumber protein penting yang
harganya relatif murah dibandingkan sumber Analisis organoleptik tempe Untuk uji organoleptik masing masing
tempe
dilakukan
secara
uji
hedonik
mengikuti metode Lawless dan Heymann
(1999). Organoleptik dilakukan oleh 30
orang panelis tidak terlatih. Kriteria penilaian
meliputi warna, aroma, tekstur, dan rasa,
dengan skala kesukaan: 1 (sangat tidak suka),
2 (tidak suka), 3 (netral), 4 (suka), 5 (sangat
suka). Sampel tempe yang diberikan kepada
panelis berukuran 1 cm x 2 cm x 2 cm yang
telah dikukus selama 15 menit. Pembuatan tepung tempe Setelah semua bahan baku dikering
anginkan maka diinokulasi dengan inokulum
tempe (2 g/kg bahan baku) dan diinkubasi
pada suhu 30 °C selama 42 jam pada
kemasan plastik yang telah diberi lubang. Proses pembuatan tempe diulang sebanyak 3
kali. Selanjutnya semua tempe dievaluasi
dalam hal: penerimaan panelis melalui uji
organoleptik (3 kali ulangan), aktivitas
antioksidan dengan DPPH (3 kali ulangan)
dan analisis proksimat. Analisis proksimat
dilakukan 1 kali dengan mencampur masing-
masing tepung tempe yang dilakukan 3 kali
menjadi satu. Masing-masing
sebanyak
250
g
sampel tempe dikeringkan dengan pengering
beku (freeze dryer) kemudian dihaluskan
dengan food processor. Selanjutnya disaring
dengan saringan berukuran 300 µm dan
disimpan pada suhu 4 °C sebelum digunakan
untuk
analisis
proksimat
dan
analisis
aktivitas antioksidan. Ekstraksi antioksidan Ekstraksi
antioksidan
pada
tempe
dilakukan dengan mengikuti metode Ningsih,
Siswanto
dan
Winarsa
(2018)
dengan
modifikasi. Sebanyak 1 g tepung tempe
dicampur dengan 10 mL etanol absolut
dalam conical tube 25 mL. Suspensi
dimaserasi pada waterbath (Memmert WB-
10, Jerman) dengan suhu 65 °C selama 1 jam. Selanjutnya suspensi disentrifugasi dengan
kecepatan 5.000 rpm selama 15 menit hingga
terpisah menjadi dua fase. Supernatan yang
diperoleh akan digunakan untuk pengukuran
aktivitas antioksidan dengan metode DPPH. Alat Alat yang digunakan adalah pengering
beku
(freeze
dryer),
food
processor
(Panasonic, Indonesia), waterbath (Memmert
WB-10, Jerman), conical tube 25 mL,
sentrifugasi, dan spektrofotometer (Thermo
Fisher Scientific Genesys-20, United State). Pembuatan tempe Proses pembuatan tempe dilakukan Tabel 1. Waktu perendaman dan perebusan bahan baku tempe
Jenis bahan baku tempe
Waktu perendaman (jam)
Waktu perebusan (menit)
I
II
I
II
Kedelai
2
24
20
15
Biji labu kuning
2
24
20
15
Biji bunga matahari
2
24
20
15
Kacang adzuki
48
24
90
15 Tabel 1. Waktu perendaman dan perebusan bahan baku tempe Bahan Bahan yang digunakan adalah kedelai
(Glycine max (L) Merrill), biji labu kuning
(Cucurbita moschata), biji bunga matahari
(Helianthus annuus), kacang adzuki (Vigna
angularis)
yang
dibeli
dari
beberapa 210 DOI: https://doi.org/10.35891/tp.v12i2.2458 Barus, Yokota, dan Hutagalung Volume 12, No.2, (2021), Halaman 209-219 pedagang di Jakarta, inokulum tempe, etanol
absolut (Smart Lab, Indonesia), dan larutan
2,2-diphenyl-1-picrylhydrazyl
(DPPH)
(Sigma Aldrich, Jerman). menggunakan bahan kedelai (Tempe K), biji
labu kuning (Tempe TK), biji bunga matahari
(Tempe TBM), kacang adzuki (Tempe TA)
dan campuran semua dengan perbandingan
1:1:1:1) sesuai dengan metode Barus net al.,
(2008) dengan modifikasi. Masing-masing
bahan baku dibersihkan dari pengotor lalu
dicuci. Selanjutnya dilakukan perendaman I,
perebusan
I,
perendaman
II
(setelah
pengupasan kulit khusus pada kedelai) dan
perebusan II dengan masing masing waktu
seperti Tabel 1. Analisis statistik Data uji organoleptik dianalisis dengan
metode analysis of variance (ANOVA) satu
arah program IBM SPSS 25.0 Signifikansi
diterima pada taraf kepercayaan P ≤ 0,05. Data uji organoleptik dianalisis dengan
metode analysis of variance (ANOVA) satu
arah program IBM SPSS 25.0 Signifikansi
diterima pada taraf kepercayaan P ≤ 0,05. Pengukuran aktivitas antioksidan Pengukuran
aktivitas
antioksidan
mengikuti metode Ningsih, Siswanto dan
Winarsa (2018). Sebanyak 1 mL supernatan
sampel dimasukkan ke dalam tabung reaksi 211 DOI: https://doi.org/10.35891/tp.v12i2.2458 Barus, Yokota, dan Hutagalung Volume 12, No.2, (2021), Halaman 209-219 yang ditutup dengan aluminium foil. Setelah
itu ekstrak tersebut ditambahkan 2 mL
larutan
DPPH
0,06
mM,
kemudian
dihomogenisasi. Suspensi diinkubasi selama
30 menit pada suhu 37 °C lalu dilakukan
pengukuran absorbansi dengan menggunakan
spektrofotometer pada panjang gelombang
517 nm. Blanko yang digunakan yaitu etanol,
sedangkan kontrol menggunakan larutan
DPPH dalam etanol. Aktivitas antioksidan
ditunjukkan dengan adanya perubahan warna
dari ungu menjadi kuning dan dapat dihitung
dengan rumus: yang ditutup dengan aluminium foil. Setelah
itu ekstrak tersebut ditambahkan 2 mL
larutan
DPPH
0,06
mM,
kemudian
dihomogenisasi. Suspensi diinkubasi selama
30 menit pada suhu 37 °C lalu dilakukan
pengukuran absorbansi dengan menggunakan
spektrofotometer pada panjang gelombang
517 nm. Blanko yang digunakan yaitu etanol,
sedangkan kontrol menggunakan larutan
DPPH dalam etanol. Aktivitas antioksidan
ditunjukkan dengan adanya perubahan warna
dari ungu menjadi kuning dan dapat dihitung
dengan rumus: Pembuatan tempe Tempe kedelai (Tempe TK), tempe biji
labu kuning (Tempe TL), tempe biji bunga
matahari (Tempe TBM), tempe kacang
adzuki (Tempe TA), dan tempe campuran
semua (1:1:1:1) (Tempe TC) telah berhasil
dibuat
dengan
menggunakan
inokulum
komersial (2 g/kg bahan baku) (Gambar 1). Miselium kapang R. microsporus dapat
tumbuh dengan baik pada masing-masing
bahan baku sehingga permukaan semua
tempe tampak berwarna putih. Gambar 1
menunjukkan
miselium
R. microsporus
tersebut mengikat bahan baku satu sama lain
sehingga tempe bersifat kompak. Hal ini
mengakibatkan saat tempe dipotong tetap
utuh (tidak rontok). Aroma semua tempe
tersebut memiliki bau khas tempe tidak
ditemukan adanya bau amoniak. Aktivitas antioksidan (%)
= 1
−absorbansi sampel
absorbansi kontrol
× 100% Aktivitas antioksidan (%)
= 1
−absorbansi sampel
absorbansi kontrol
× 100% Aktivitas antioksidan (%)
1 Aktivitas antioksidan (%)
1 −absorbansi sampel Analisis komposisi kimia Sampel
tepung
tempe
dianalisis
menggunakan standar Internasional AOAC
edisi ke-19 tahun 2012 (AOAC, 2012) di PT. Saraswanti
Indo
Genetech. Pengukuran
proksimat yang dilakukan terdiri atas kadar
protein,
abu,
lemak
total,
air,
dan
karbohidrat. Gambar 1. Tempe berbahan baku kedelai (TK), biji labu kuning (TL), biji bunga
matahari(TBM), kacang adzuki (TA), dan campuran keempatnya (1:1:1:1) (TC) Gambar 1. Tempe berbahan baku kedelai (TK), biji labu kuning (TL), biji bunga
matahari(TBM), kacang adzuki (TA), dan campuran keempatnya (1:1:1:1) (TC) tempe harus memiliki tekstur yang kompak
sehingga saat diiris tetap utuh (tidak mudah
rontok),
berwarna
putih
pada
seluruh
permukaan tempe dan memiliki aroma bau
khas tempe tanpa adanya bau amoniak. Berdasarkan warna, tekstur dan aroma
(bau) semua tempe tersebut memenuhi syarat
mutu tempe kedelai yang diatur pada SNI
3144:2015. Dimana menurut SNI 3144:2015
tersebut menyatakan bahwa syarat mutu 212 DOI: https://doi.org/10.35891/tp.v12i2.2458 Volume 12, No.2, (2021), Halaman 209-219 Barus, Yokota, dan Hutagalung 3144:2015). Aroma tempe muncul karena
aktivitas enzimatik yang memecah berbagai
macam makromolekul bahan baku tempe,
seperti protease yang memecah protein serta
lipase
yang
memecah
lemak
sehingga
berukuran lebih sederhana dan menghasilkan
senyawa yang bersifat volatil. Pada umumnya
senyawa
3-octanone
dan
1-octen-3-ol
merupakan aroma Rhizopus spp. yang tercium
(Feng, Larsen dan Schnürer, 2007). Perbedaan
aroma pada masing-masing tempe disebabkan
karena perbedaan komponen yang terkandung
pada setiap substrat sehingga saat proses
fermentasi komponen yang dipecah akan
menghasilkan senyawa volatil yang berbeda-
beda. Mikroorganisme
tersebut
juga
berpengaruh terhadap terbentuknya berbagai
senyawa yang bersifat volatil yang berpengaruh
terhadap aroma. Namun setelah tempe dipotong tampak
perbedaan ragam warna pada tempe tersebut
sesuai dengan bahan baku biji bijian yang
digunakan (Gambar 1). Perbedaan ragam
warna pada biji-bijian tanaman akibat adanya
kandungan pigmen pada masing masing
bahan baku tersebut, seperti: pigmen xantofil
pada kedelai (Aptesia dan Rasyid, 2013),
foto klorofil pada biji labu kuning (Schoefs,
2002), fito melanin pada biji bunga matahari
(Keles dan Ozdemir, 2018), dan antosianin
pada kacang adzuki (Takahama, Yamauchi
dan Hirota, 2013). Saat ini, pengaturan syarat mutu hanya
diperuntukkan untuk tempe kedelai, yaitu SNI
3144:2015. Hal ini karena tempe yang paling
banyak diproduksi dan dikonsumsi di Indonesia
adalah tempe kedelai. Namun perlu terus
dikembangkan variasi jenis tempe dengan
menggunakan berbagai jenis bahan baku tempe
selain kedelai sehingga menambah ragam jenis
tempe, seperti biji labu kuning, biji bunga
matahari, kacang adzuki dan campurannya. Hal
ini sudah dilakukan oleh peneliti dari berbagai
negara, dimana tempe diolah dari bahan baku
Phaseolus vulgaris L. Analisis komposisi kimia (Reyes-Bastidas et al.,
2010), white bean (Vital et al., 2018) dan red-
bean (Chen, Hsieh dan Hu, 2020). Sifat sensori tempe p
Tabel 2 menunjukkan data hasil uji
organoleptik tempe dalam hal rasa, tekstur,
aroma, dan warna. Penerimaan panelis terhadap
atribut rasa, tekstur dan aroma Tempe TL,
Tempe TBM, dan Tempe TC cenderung tidak
berbeda secara signifikan dibandingkan dengan
Tempe TK. Hal ini merupakan hal yang surprise
dan yang kita harapkan. Dengan demikian biji
labu kuning, bunga biji matahari dan campuran
antara biji labu kuning, bunga biji matahari dan
kacang
adzuki
dapat
digunakan
sebagai
alternatif bahan baku tempe. Dengan demikian
dapat bertambah ragam jenis tempe yang dapat
diproduksi yang selama ini didominasi oleh
tempe kedelai. Dalam hal warna panelis paling
menyukai
warna
Tempe
TK. Hal
ini
kemungkinan dipengaruhi oleh warna bahan
baku. Variasi warna kedelai yang kuning muda
dengan miselium kapang yang berwarna putih
menjadi warna yang lebih cerah sehingga
menarik bagi panelis. Tabel 1 menunjukkan
bahwa panelis tidak menyukai Tempe TA
terhadap semua parameter yang diuji. Dengan
demikian penggunaan kacang adzuki saja tidak
berpotensi dikembangkan sebagai bahan baku
tempe. Namun bila dicampur dengan bahan
baku lain (Tempe C) menghasilkan tempe yang
dapat diterima dengan baik. Penelitian tentang tempe dilakukan di
beberapa negara karena tempe merupakan
pangan fermentasi yang bersifat fungsional dan
dapat diolah menjadi berbagai jenis bahan
pangan turunannya. Seperti dapat membangun
mikrobiota usus yang baik (Stephanie et al. 2017a), dapat meningkatkan imun sistem
(Stephanie
et
al.,
2017),
mengandung
antioksidan (Barus, Maya dan Hartanti, 2019;
Liao, et al., 2013), dan mengandung the γ-
aminobutyric
acid
(GABA)
yang
telah
dilaporkan baik untuk kesehatan (Handoyo dan
Morita, 2006). Tempe juga dapat diolah menjadi
bahan
baku
pembuatan
makanan
dari
turunannya,
seperti
tepung
tempe
untuk
pembuatan beberapa jenis makanan (Reyes-
Bastidas et al., 2010), burger tempe (Vital et al.,
2018),
dan
sausage
tempeh
(Syamsuri,
Dewayani dan Septianti, 2020). Aroma tempe yang baik adalah berbau
khas tempe dan tidak ada bau amonia (SNI 213 DOI: https://doi.org/10.35891/tp.v12i2.2458 Volume 12, No.2, (2021), Halaman 209-219 Barus, Yokota, dan Hutagalung Tabel 2. Sifat sensori tempe Hasil organoleptik Tempe TK (kedelai), TB (biji labu kuning), TBM (biji bunga
matahari), TA (kacang adzuki), TC (campuran semua dengan perbandingan 1:1:1:1)
Jenis tempe
Rasa
Tekstur
Aroma
Warna
TK
3,7 ± 0,8b
3,9 ± 0,7b
3,9 ± 0,8c
4,2 ± 0,7b
TL
3,4 ± 1,1b
3,7 ± 0,9b
3,6 ± 0,8bc
3,0 ± 1,0a
TBM
3,4 ± 1,3b
3,6 ± 0,9b
3,4 ± 1,0b
2,9 ±1,0a
TA
2,5 ± 1,1a
2,4 ± 1,0a
2,9 ± 1,0a
3,2 ± 1,1a
TC
3,5 ± 0,9b
3,8 ± 0,9b
3,5 ± 0,9bc
3,0 ± 1,2a
Keterangan: Hurup superscript yang berbeda pada kolom yang sama menunjukkan perbedaan
yang signifikan (P ≤ 0.05) Cita rasa tempe pada penelitian ini
terutama ditentukan oleh karena jenis bahan
baku
yang
digunakan
karena
semua
dikondisikan sama kecuali bahan baku. Namun banyak faktor lain yang dapat
mempengaruhi cita rasa tempe. Diantaranya
dapat karena adanya jenis mikroorganisme
yang berbeda selama proses fermentasi
berlangsung. Telah dilaporkan bahwa cita
rasa tempe yang disenangi panelis dan yang
tidak disenangi memiliki jumlah dan jenis
mikroorganisme yang sangat berbeda (Barus
et al., 2008). Aktivitas mikroorganisme
berbeda dapat menghasilkan peptida, asam
amino bebas dan senyawa lainnya yang
berbeda. Degradasi makromolekul bahan
baku secara enzimatik, seperti protein oleh
enzim protease berpengaruh terhadap cita
rasa tempe (Zhao et al., 2018). Dimana mikroorganisme yang menghasilkan enzim,
seperti Bacillus spp. yang menghasilkan
enzim protease ditemukan berlimpah selama
fermentasi tempe berlangsung (Barus, Wati
dan Suwanto, 2017). Jenis Rhizopus berbeda
dapat juga menghasilkan cita rasa berbeda
(Barus, Maya dan Hartanti, 2019). Aktivitas antioksidan Hasil pengukuran aktivitas antioksidan
melalui persen penghambatan radikal bebas
dengan metode DPPH ditunjukkan pada
Gambar 2. Semua tempe ditemukan memiliki
aktivitas antioksidan, namun secara statistik
menunjukkan
perbedaan
yang
tidak
signifikan satu dengan yang lain. Aktivitas
antioksidan tertinggi ditemukan pada Tempe
TBM lalu diikuti secara berurutan oleh oleh
Tempe: TK, TC, TA dan TL. Gambar 2. Aktivitas antioksidan Tempe: TK (kedelai), TL (biji labu kuning), TBM (biji
bunga matahari), TA (kacang adzuki), TC (campuran semua dengan perbandingan
1:1:1:1) Gambar 2. Aktivitas antioksidan Tempe: TK (kedelai), TL (biji labu kuning), TBM (biji
bunga matahari), TA (kacang adzuki), TC (campuran semua dengan perbandingan
1:1:1:1) paparan radikal bebas. Radikal bebas
dapat terbentuk di dalam tubuh sebagai
produk metabolisme atau dari luar tubuh
seperti polusi dan sisa pestisida pada Antioksidan adalah zat atau senyawa
yang sangat dibutuhkan oleh tubuh untuk
menjaga kesehatan karena dapat mencegah
atau memperlambat kerusakan sel akibat 212 DOI: https://doi.org/10.35891/tp.v12i2.2458 Volume 12, No.2, (2021), Halaman 209-219 Barus, Yokota, dan Hutagalung 2
menunjukkan
bahwa
Tempe
TK
memiliki
kadar
protein
tertinggi
dibandingkan dengan Tempe: TL, TBM,
dan TC. Hal ini karena kandungan protein
kedelai yang tinggi, yaitu sekitar 45%
(Liu, 1997) dan merupakan kandungan
protein tertinggi diantara semua jenis
kacang kacangan. Dengan demikian maka
pengembangan tempe yang diperuntukkan
sebagai sumber protein maka tempe
berbahan
baku
kedelai
adalah
yang
terbaik. Kadar protein Tempe: TK, TL,
TBM, dan TC berkisar antara 21,83% -
43,51% (Tabel 1). Kadar protein semua
tempe tersebut berada di atas kadar protein
yang ditetapkan pada SNI 3144:2015,
yaitu minimal 15%. makanan. Metode DPPH merupakan salah
satu metode yang banyak digunakan untuk
mengukur penghambatan radikal bebas
dari suatu bahan makanan (Musa et al.,
2013). Aktivitas antioksidan berdasarkan
reduksi radikal DPPH yang berwarna
violet oleh antioksidan dari makanan
melalui mekanisme transfer atom hidrogen
yang
menyebabkan
perubahan
warna
DPPH menjadi kuning pucat yang stabil. Lalu perubahan warna diukur dengan
spektrofotometer UV-Vis pada sekitar 515
- 520 nm (Sirivibulkovit, Nouanthavong
dan Sameenoi, 2018). Keberadaan
aktivitas
antioksidan
pada biji labu kuning ini sejalan dengan
yang telah dilaporkan oleh Junita et al. (2017) dan pada tempe kedelai (Barus,
Maya dan Hartanti, 2019). Namun, kadar
aktivitas antioksidannya bervariasi yang
kemungkinan disebabkan oleh jenis atau
varietas
bahan
baku
yang
berbeda,
pengolahan tempe yang berbeda, preparasi
pengukuran aktivitas antioksidan yang
berbeda dan jenis mikroorganisme yang
berbeda
selama
proses
fermentasi
berlangsung. Perbedaan
jenis
mikroorganisme yang digunakan pada saat
proses
pembuatan
tempe
dapat
menghasilkan aktivitas antioksidan yang
berbeda (Barus, Maya dan Hartanti, 2019). Aktivitas antioksidan Namun,
sejauh
ini
kami
belum
menemukan informasi tentang aktivitas
antioksidan dari tempe yang terbuat dari
kacang adzuki, biji bunga matahari dan
tempe yang dibuat dari campuran kacang
adzuki, biji bunga matahari, biji labu
kuning dan kedelai. Dengan demikian
maka hasil penelitian ini yang pertama
melaporkan tentang aktivitas antioksidan
tempe tersebut. y
Kadar protein Tempe LB (34,62%)
lebih rendah dibandingkan Tempe TK. Namun pemanfaatan biji labu kuning
tersebut
sebagai
bahan
baku
tempe
berpotensi dikembangkan sebab biji labu
kuning mengandung zat besi yang tinggi
(290.0 ppm) dan zinc yang tinggi (39.9
ppm). Kedua unsur tersebut penting bagi
kesehatan. Kadar abu mewakili nilai kadar
mineral yang terkandung pada bahan
pangan. Tempe TL memiliki kadar abu
yang lebih tinggi dibanding tempe lain hal
ini kemungkinan karena kadar besi dan
zinc pada biji labu kuning tersebut. Demikian juga kadar protein Tempe TBM
(24,68%)
lebih
rendah
dibandingkan
Tempe
TK. Walaupun
demikian,
pemanfaatan biji bunga matahari tersebut
berpotensi
dikembangkan
sebab
mengandung E (alpha-tocopherol) yang
tinggi dan kadar lemak jenuh yang tinggi
(Škrbic dan Cvejanov 2011; Franco, Iseppi
dan Taverna, 2018). Dengan demikian
maka
pengembangan
tempe
yang
diperuntukkan
sebagai
sumber
nutrisi
lemak yang sehat maka tempe berbahan
baku biji bunga matahari adalah yang
terbaik. DAFTAR PUSTAKA Amarowicz, R., Estrella, I., Hernandez, T., &
Troszyńska, A. (2008). Antioxidant
activity of extract of adzuki bean and
its
fractions. Journal
of
Food
Lipids, 15(1), 119-136. https://doi.org/
10.1111/j.1745-4522.2007.00106.x p
p
Tabel 3 menunjukkan bahwa kadar
lemak tertinggi ditemukan pada Tempe TBM
(45,522%). Tingginya kadar lemak tempe
TBM tersebut karena kadar lemak pada biji
matahari yang tinggi, yaitu sekitar 46%. Minyak biji bunga matahari merupakan
sumber asam lemak tidak jenuh yang paling
banyak dikonsumsi di dunia. Mengandung
berbagai essential oil dan vitamin E penting
yang baik bagi kesehatan (Škrbic dan
Cvejanov 2011; Franco, Iseppi dan Taverna,
2018). Vitamin E merupakan salah satu jenis
vitamin penting untuk antioksidan yang
berperan sebagai penangkal radikal bebas
sehingga
antioksidan
Tempe
TBM
ditemukan paling tinggi pada penelitian ini
(Gambar 2). Dengan demikian maka tempe
berbahan baku biji bunga matahari adalah
yang terbaik. AOAC (Association of Official Analytical
Chemist). 2012. Official methods of
analysis, 19th edition. Arlington Aptesia, L. T., & Al Rasyid, H. (2013). Pemanfaatan Lactobacillus casei dan
tapioka dalam upaya menghambat
kerusakan
tempe
kedelai. Jurnal
Teknologi
&
Industri
Hasil
Pertanian, 18(2), 175-184. http://dx. doi.org/10.23960/jtihp.v18i2.175%20-
%20184 Barus, T., Maya, F., & Hartanti, A. T. (2019). Peran beberapa galur Rhizopus
microsporus yang berasal dari “laru
tradisional” dalam menentukan kualitas
Tempe. Jurnal
Aplikasi
Teknologi
Pangan, 8(1), 17-22. https://doi.org/
10.17728/jatp.3761 Komposisi kimia tempe Tempe merupakan sumber protein
penting
masyarakat
Indonesia. Kadar
protein tempe ditentukan oleh kadar
protein bahan baku yang digunakan. Tabel 215 DOI: https://doi.org/10.35891/tp.v12i2.2458 Volume 12, No.2, (2021), Halaman 209-219 Barus, Yokota, dan Hutagalung Tabel 3. Uji proksimat tempe TK (kedelai), TL (biji labu kuning), TBM (biji bunga
matahari), TA (kacang adzuki), TC (campuran semua). Kadar (%)
Tempe TK
Tempe TL
Tempe TBM
Tempe TA
Tempe TC
Protein
43,5
34,6
24,6
21,8
32,1
Kadar abu
2,9
3,9
3,0
1,6
2,4
Lemak
23,1
45,5
45,0
1,5
31,8
Kadar air
7,6
5,9
7,9
17,1
7,7
Karbohidrat
22,7
9,9
19,3
57,7
25,7 . Uji proksimat tempe TK (kedelai), TL (biji labu kuning), TBM (biji bunga
matahari), TA (kacang adzuki), TC (campuran semua). UCAPAN TERIMA KASIH Tabel 3 menunjukkan kadar lemak
Tempe: TK, TL, TBM dan TC bervariasi
mulai dari 1,59% hingga 45,52%. Dengan
demikian, kadar lemak Tempe: TK, TL,
TBM dan TC hasil penelitian ini masih di
atas kadar lemak yang ditetapkan pada SNI
3144:2015, yaitu minimal 7%. Sebaliknya,
Tempe TA memiliki kadar lemak hanya
1,59% jauh di bawah kadar lemak yang
ditetapkan pada SNI 3144:2015. Terima
kasih
kepada
Universitas
Katolik Indonesia Atma Jaya yang telah
mendanai
penelitian
ini
melalui hibah
Fakultas. KESIMPULAN Biji labu kuning, biji bunga matahari,
dan campuran kedelai, biji labu kuning, biji
bunga matahari, kacang adzuki (1:1:1:1:1)
berpotensi dikembangkan sebagai bahan
baku tempe. Namun penelitian lebih lanjut
masih dibutuhkan untuk mengkaji sifat
fungsional tempe yang diolah dari semua
semua bahan baku tersebut dan mengkaji
pengaruhnya
terhadap
kesehatan
dan
kesimbangan mikrobiota usus. Barus, T., Suwanto, A., Wahyudi, A. T., &
Wijaya, H. (2008). Role of bacteria in
tempe
bitter
taste
formation:
microbiological
and
molecular
biological analysis based on 16S rRNA
gene. Microbiology
Indonesia, 2(1),
4-4. https://doi.org/10.5454/mi.2.1.4 DOI: https://doi.org/10.35891/tp.v12i2.2458 216 Volume 12, No.2, (2021), Halaman 209-219 Barus, Yokota, dan Hutagalung 208. https://doi.org/10.12938/bmfh. 2020-001 Barus, T., Titarsole, N. N., Mulyono, N., &
Prasasty,
V. D. (2019). Tempeh
antioxidant
activity
using
DPPH
method:
effects
of
fermentation,
processing,
and
microorga-
nisms. Journal of Food Engineering
and Technology, 8(2), 75-80. https:/
/doi.org/10.32732/jfet.2019.8.2.75 Dao , M. C., Everard, A., Aron-Wisnewsky,
J., Sokolovska, N., Prifti, E., Verger, E. O., ... & MICRO-Obes Consortium. (2016). Akkermansia muciniphila and
improved metabolic health during a
dietary
intervention
in
obesity:
relationship
with
gut
microbiome
richness and ecology. Gut, 65(3), 426-
436. 10.1136/gutjnl-2014-308778 Barus, T., Wati, L., & Suwanto, A. (2017). Diversity
of
protease-producing
Bacillus spp. from fresh Indonesian
tempeh based on 16S rRNA gene
sequence. HAYATI
Journal
of
Biosciences, 24(1), 35-40. https://doi. org/10.1016/j.hjb.2017.05.001 Depommier, C., Everard, A., Druart, C.,
Plovier, H., Van Hul, M., Vieira-Silva,
S.,
... &
Cani,
P. D. (2019). Supplementation
with
Akkermansia
muciniphila in overweight and obese
human volunteers: a proof-of-concept
exploratory
study. Nature
medicine, 25(7), 1096-1103. https://doi. org/10.1038/s41591-019-0495-2 Berni, C. R., Paparo, L., Nocerino, R., Di
Scala, C., Della, G., Maddalena, Y.,
Buono, A., Bruno, C., Voto, L., &
Ercolini, D. (2019). Gut microbiome as
target for innovative strategies against
food
allergy. Frontiers
in
immu-
nology, 10,
1-40. https://doi.org/10. 3389/fimmu.2019.00191 g
Feng, X. M., Larsen, T. O., & Schnürer, J. (2007). Production
of
volatile
compounds by Rhizopus oligosporus
during soybean and barley tempeh
fermentation. International journal of
food microbiology, 113(2), 133-141. https://doi.org/10.1016/j.ijfoodmicro.20
06.06.025 Chen, Y. C., Hsieh, S. L., & Hu, C. Y. (2020). Effects of red-bean tempeh
with various strains of Rhizopus on
GABA content and cortisol level in
zebrafish. Microorganisms, 8(9), 330. https://doi.org/10.3390/microorganisms
8091330 Franco, R., Iseppi, L., & Taverna, M. (2018). Sunflower oil functional properties for
specialty food. Nutrition and Food
Science
International Journal, 5(4),
4-7. http://dx.doi.org/10.19080/NFSIJ. 2018.05.555668 Clapp, M., Aurora, N., Herrera, L., Bhatia,
M., Wilen, E., & Wakefield, S. (2017). KESIMPULAN Gut microbiota’s effect on mental
health: the gut-brain axis. Clinics and
practice, 7(4), 131-136. https://doi.org/
10.4081/cp.2017.987 Gumolung D. 2018. Analisis kandungan total
fenolik pada jonjot buah labu kuning
(Cucurbita
moschata). Fullerene
Journal
of
Chemistry, 3(1),
1-4. https://doi.org/10.37033/fjc.v3i1.25 Collins, S. M., Surette, M., & Bercik, P. (2012). The interplay between the
intestinal microbiota and the brain. Nature Reviews Microbiology, 10(11),
735-742. https://doi.org/10.1038/
nrmicro2876 Handajani, Y. S., Turana, Y., Yogiara, Y.,
Widjaja, N. T., Sani, T. P., Christianto,
G. A. M., & Suwanto, A. (2020). Tempeh consumption and cognitive
improvement
in
mild
cognitive
impairment. Dementia and geriatric
cognitive
disorders,
49,
1-6. https://doi.org/10.1159/000510563 Daliri, E. B. M., Ofosu, F. K., Chelliah, R.,
Lee, B. H., An, H., Elahi, F., ... & Oh,
D. H. (2020). Influence of fermented
soy
protein
consumption
on
hypertension
and
gut
microbial
modulation
in
spontaneous
hypertensive
rats. Bioscience
of
microbiota, food and health 39(4), 199- Handoyo, T., & Morita, N. (2006). Structural
and functional properties of fermented
soybean (tempeh) by using Rhizopus 217 DOI: https://doi.org/10.35891/tp.v12i2.2458 Volume 12, No.2, (2021), Halaman 209-219 Barus, Yokota, dan Hutagalung oligosporus. International Journal of
Food
Properties, 9(2),
347-355. https://doi.org/10.1080/1094291050022
4746 Musa, K. H., Abdullah, A., Kuswandi, B., &
Hidayat, M. A. (2013). A novel high
throughput method based on the DPPH
dry reagent array for determination of
antioxidant
activity. Food
chemist-
ry, 141(4), 4102-4106. Huang, X., Fan, X., Ying, J., & Chen, S. (2019). Emerging trends and research
foci
in
gastrointestinal
microbiome. Journal of translational
medicine, 17(1), 1-11. Ningsih, T. E., Siswanto, S., & Winarsa, R. (2018). Aktivitas antioksidan kedelai
edamame
hasil
fermentasi
kultur
campuran oleh Rhizopus oligosporus
dan
Bacillus
subtilis. Berkala
Sainstek, 6(1), 17-21. https://doi.org/
10.19184/bst.v6i1.7556 Junita, D., Setiawan, B., Anwar, F., &
Muhandri, T. (2017). Komponen gizi,
aktivitas antioksidan dan karakteristik
sensori bubuk fungsional labu kuning
(Cucurbita
moschata)
dan
tempe. Jurnal Gizi dan Pangan, 12(2),
109-116. https://doi.org/10.25182
/jgp.2017.12.2.109-116 Reyes-Bastidas, M., Reyes-Fernández, E. Z.,
López-Cervantes, J., Milán-Carrillo, J.,
Loarca-Piña, G. F., & Reyes-Moreno,
C. (2010). Physicochemical, nutritional
and antioxidant properties of tempeh
flour from common bean (Phaseolus
vulgaris
L.). Food
Science
and
Technology International, 16(5), 427-
434. https://doi.org/10.1177/
1082013210367559 Keles, Y., & Özdemir, Ö. (2018). Extraction,
purification, antioxidant properties and
stability conditions of phytomelanin
pigment
on
the
sunflower
seeds. International
Journal
of
Secondary Metabolite, 5(2), 140-148. https://doi.org/10.1016/j.foodchem.201
3.06.112 Schoefs,
B. (2002). Chlorophyll
and
carotenoid analysis in food products. Properties of the pigments and methods
of analysis. KESIMPULAN Trends in food science &
technology, 13(11), 361-371. https://
doi.org/10.1016/S0924-244(02)00182-6 Keuth, S., & Bisping, B. (1994). Vitamin
B12 production by Citrobacter freundii
or
Klebsiella
pneumoniae
during
tempeh fermentation and proof of
enterotoxin absence by PCR. Applied
and
environmental
microbiology,
60(5), 1495-1499. Sirivibulkovit, K., Nouanthavong, S., &
Sameenoi, Y. (2018). based DPPH
assay
for
antioxidant
activity
analysis. Analytical
sciences, 34(7),
795-800. https://doi.org/10.2116
/analsci.18P014 Lawless, H. T., & Heymann, H. (1999). Discrimination
testing. In Sensory
evaluation
of
food (pp. 116-139). Springer, Boston, MA. Škrbić, B., & Cvejanov, J. (2011). The
enrichment of wheat cookies with high-
oleic sunflower seed and hull-less
barley flour: Impact on nutritional
composition, content of heavy elements
and
physical
properties. Food
Chemistry, 124(4),
1416-1422. https://doi.org/10.1016/j.foodchem.201
0.07.101 Liao, W. C., Wang, C. Y., Shyu, Y. T., Yu,
R. C., & Ho, K. C. (2013). Influence of
preprocessing
methods
and
fermentation of adzuki beans on γ-
aminobutyric
acid
(GABA)
accumulation
by
lactic
acid
bacteria. Journal
of
Functional
Foods, 5(3), 1108-1115. https://doi.org/
10.1016/j.jff.2013.03.006 10.1016/j.jff.2013.03.006 SNI (Standar Nasional Indonesia) 3144-2015
Tempe
Kedelai. 2015. Badan
Standarisasi Nasional. Jakarta. Liu, K. (1997). Chemistry and nutritional
value
of
soybean
components. In Soybeans (pp. 25-113). Springer,
Boston, MA. Stephanie, S., Ratih, N. K., Soka, S., &
Suwanto, A. (2017). Effect of tempeh
supplementation on the profiles of 218 DOI: https://doi.org/10.35891/tp.v12i2.2458 Volume 12, No.2, (2021), Halaman 209-219 Barus, Yokota, dan Hutagalung human intestinal immune system and
gut
microbiota. Microbiology
Indonesia, 11(1), 2. https://doi.org/10. 5454/mi.11.1.2 Zhang, X., Zhao, L. D., & Li, H. (2020). The
gut microbiota: emerging evidence in
autoimmune
diseases. Trends
in
molecular medicine 16(9), 862-873. https://doi.org/10.1016/j.molmed.2020. 04.001 Syamsuri, R., Dewayani, W., & Septianti, E. (2020,
October). Chemical
characteristic and sensory of tempeh
sausage on different soybean varieties
and cooking methods variation. In IOP
Conference
Series:
Earth
and
Environmental Science, 575(1). IOP
Publishing. https://doi.org/10.1088/
1755-1315/575/1/012012 Zhao, G., Ding, L. L., Yao, Y., Cao, Y., Pan,
Z. H., & Kong, D. H. (2018). Extracellular proteome analysis and
flavor formation during soy sauce
fermentation. Frontiers
in
micro-
biology, 9,
1-7. https://doi.org/
10.3389/fmicb.2018.01872 Takahama, U., Yamauchi, R., & Hirota, S. (2013). Isolation and characterization
of a cyanidin-catechin pigment from
adzuki bean (Vigna angularis). Food
chemistry, 141(1), 282-288. https://doi. org/10.1016/j.foodchem.2013.02.113 Vital, R. J., Bassinello, P. Z., Cruz, Q. A.,
Carvalho, R. N., De Paiva, J., &
Colombo, A. O. (2018). Production,
quality, and acceptance of tempeh and
white
bean
tempeh
burgers. Foods, 7(9),
136. https://doi. org/10.3390/foods7090136 219 DOI: https://doi.org/10.35891/tp.v12i2.2458
|
https://openalex.org/W2000298025
|
https://hal.science/hal-01191070/document
|
English
| null |
Action versus Result-Oriented Schemes in a Grassland Agroecosystem: A Dynamic Modelling Approach
|
PloS one
| 2,012
|
cc-by
| 10,193
|
Introduction efficient either on the ecological or productive performance
without decreasing performance on the other dimension. Second,
it may give more flexibility to the farming system [2]. Due to the
difficulties
of
implementing
and
monitoring
result-oriented
schemes, a third kind of scheme has been created. These schemes
aim at producing suitable habitat for biodiversity [7]. Their
evaluation is based on indicators of habitat quality and not directly
on biodiversity levels [8]. Hereafter, we will call these schemes
habitat-oriented schemes. By providing suitable habitats for target
species, such schemes are expected to lead to better ecological
performance than action-oriented ones. However, the potential of
innovation may be limited by the constraints applied on the
habitat instead of on biodiversity levels. For example, result-
oriented schemes allow inter-annual variability and strategies with
successions of ecology-oriented and production-oriented years may
appear. Moreover, these schemes may not systematically ensure
good ecological performance whereas result-oriented ones should
always lead, by definition, to good levels of biodiversity. After 15 years of implementation, the effectiveness of agri-
environment schemes (AES) is still under debate [1]. Result-
oriented AES have been proposed to improve the efficiency of
conservation policies [2]. They rely on payment for effective
biodiversity conservation (e.g. abundance, richness) independently
from the management practices implemented by farmers. Such
AES have been studied in the case of carnivores [3], grassland
flora [4] or grassland birds [5]. If quite a large number of result-
oriented
schemes
already
exist,
most
of
them
are
either
experimental or have been run for too short a term and on too
small a scale to be properly evaluated [2]. This situation could
explain why few comparisons between result-oriented and action-
oriented schemes are available and why no clear difference has
been found in their effects on population sizes [6]. One of the main advantages of result-oriented schemes is to
allow farmers to develop innovative management practices that
would be efficient on both productive and ecological performance. By relaxing constraints on management, these schemes make it
possible to implement a wider set of management strategies (i.e. sequences of management practices over time). Widening the
range of management strategies may offer two advantages. First,
out of the new management strategies some of them may be more The objective of this study was to compare three scenarios
corresponding to the different kinds of agri-environment schemes:
action-oriented, habitat-oriented and result-oriented schemes. Abstract This is an open-access article distributed under the terms of the Creative Commons Attribut
unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited. Funding: This work was carried out with the financial support of the ‘‘ANR - Agence Nationale de la Recherche - The French National Research Agency’’ under the
‘‘SYSTERRA program - Ecosystems and Sustainable Development,’’ project ‘‘ANR-08-STRA-007, FARMBIRD - Coviability models of FARMing and BIRD biodiversity.’’
The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript. Competing Interests: The authors have declared that no competing interests exist. * E-mail: rodolphe.sabatier@forst.uni-goettingen.de Abstract Effects of agri-environment schemes (AES) on biodiversity remain controversial. While most AES are action-oriented, result-
oriented and habitat-oriented schemes have recently been proposed as a solution to improve AES efficiency. The objective
of this study was to compare action-oriented, habitat-oriented and result-oriented schemes in terms of ecological and
productive performance as well as in terms of management flexibility. We developed a dynamic modelling approach based
on the viable control framework to carry out a long term assessment of the three schemes in a grassland agroecosystem. The model explicitly links grazed grassland dynamics to bird population dynamics. It is applied to lapwing conservation in
wet grasslands in France. We ran the model to assess the three AES scenarios. The model revealed the grazing strategies
respecting ecological and productive constraints specific to each scheme. Grazing strategies were assessed by both their
ecological and productive performance. The viable control approach made it possible to obtain the whole set of viable
grazing strategies and therefore to quantify the management flexibility of the grassland agroecosystem. Our results showed
that habitat and result-oriented scenarios led to much higher ecological performance than the action-oriented one. Differences in both ecological and productive performance between the habitat and result-oriented scenarios were limited. Flexibility of the grassland agroecosystem in the result-oriented scenario was much higher than in that of habitat-oriented
scenario. Our model confirms the higher flexibility as well as the better ecological and productive performance of result-
oriented schemes. A larger use of result-oriented schemes in conservation may also allow farmers to adapt their
management to local conditions and to climatic variations. Citation: Sabatier R, Doyen L, Tichit M (2012) Action versus Result-Oriented Schemes in a Grassland Agroecosystem: A Dynamic Modelling Approach. PLoS
ONE 7(4): e33257. doi:10.1371/journal.pone.0033257
Editor: Raphae¨l Arlettaz University of Bern Switzerland Citation: Sabatier R, Doyen L, Tichit M (2012) Action versus Result-Oriented Schemes in a Grassland Agroecosystem: A Dynamic Modelling Approach. PLoS
ONE 7(4): e33257. doi:10.1371/journal.pone.0033257 Editor: Raphae¨l Arlettaz, University of Bern, Switzerland Editor: Raphae¨l Arlettaz, University of Bern, Switzerland Received August 24, 2011; Accepted February 7, 2012; Published April 5, 2012 Received August 24, 2011; Accepted February 7, 2012; Published April 5, 2012 atier et al. This is an open-access article distributed under the terms of the Creative Commons Attribution License, which permits
tion, and reproduction in any medium, provided the original author and source are credited. Copyright: 2012 Sabatier et al. Rodolphe Sabatier1,2*, Luc Doyen3, Muriel Tichit1,2
1 INRA, UMR 1048 SADAPT, Paris, France, 2 AgroParisTech, UMR 1048 SADAPT, Paris, France, 3 CNRS, UMR 7204 CERSP, MNHN, Paris, France phe Sabatier1,2*, Luc Doyen3, Muriel Tichit1,2
MR 1048 SADAPT, Paris, France, 2 AgroParisTech, UMR 1048 SADAPT, Paris, France, 3 CNRS, UMR 7204 CERSP, MNHN, Paris, France Rodolphe Sabatier1,2*, Luc Doyen3, Muriel Tichit1,2
1 INRA, UMR 1048 SADAPT, Paris, France, 2 AgroParisTech, UMR 1048 SADAPT, Paris, France, 3 CNRS, UMR 7204 CERSP, MNHN, Paris, France PLoS ONE | www.plosone.org April 2012 | Volume 7 | Issue 4 | e33257 Action versus Result-Oriented Schemes Secondly, we explore the management flexibility linked with each
scenario. A scenario will have a higher flexibility if it allows more
management option than another. Finally, we illustrate the
importance of management flexibility in the face of climate shock. The overall comparison of the three scenarios is based on two
hypotheses: effects of grazing on bird life traits. Even if other factors than
grazing may also play a role e.g. field wetness or predation, grazing
indisputably remains a major factor driving the life cycle of waders
(review in [9]). We therefore focus on the effects of grazing on
wader dynamics. Grazing intensity has a direct effect on clutch size
through nest trampling by cattle [16]. Grass height (i.e. habitat
quality), generated by grazing is a key factor for foraging [17] and
impacts juvenile survival. Grass height is also an important
predictor of habitat nest selection [9]; however, in the absence of
spatial dimension in our model, we did not model this process. The model computes two indicators summarizing the ecological
and productive performance of each grazing strategy. Hypothesis 1: For a given result-oriented scenario, there is no
habitat-oriented one that leads to better perfor-
mances and for a given habitat-oriented scenario,
there is no action oriented one that lead to better
performances. Hypothesis 2: For a given result-oriented scenario, there is no
habitat oriented one that leads to a higher flexibility. Hypothesis 2: For a given result-oriented scenario, there is no
habitat oriented one that leads to a higher flexibility. We studied the co-viability of the grassland agroecosystem in
three scenarios (action-oriented, habitat-oriented and result-
oriented scenario) by looking for viable management strategies
that satisfy both ecological and productive constraints. The type of
ecological constraints applied to the system differs from one
scenario to the other and reflects their specificities. In the action-
oriented scenario, constraints correspond to thresholds on minimal
and maximal stocking density during the nesting period. Such
management requirements aim at limiting the effects of trampling,
while ensuring a minimum level of grazing so as to reduce grass Formal definitions of these two hypotheses will be given in the
core of the text. As a case study, we focused on the conservation of lapwings
Vanellus vanellus in wet grasslands of the French Atlantic coast
(46u229N, 1u259W). Action versus Result-Oriented Schemes Due to their high position in trophic networks
and their close connection with wet grasslands, wader species give
good information about the health of the ecosystem. The lapwing
life cycle is deeply linked to the management of grassland [9] and
lapwing was one of the first species to benefit from result-oriented
schemes [5,6]. Wet grasslands were the first habitats targeted by
agri-environment schemes in France during the early 90’s and the
conservation of lapwings in these agroecosystem has long been of
major concern. To compare different AES in their ability to
ensure productive and ecological performance in the long term, we
developed a dynamic model linking grazed grassland dynamics
and lapwing population dynamics. This model focuses on the
effect of AES and is thus limited to the impact of farming practices
on bird dynamics. The model is built under the viable control
approach [10] which is closely related to the viability theory [11]. This framework enables the satisfaction of production, socio-
economic and environmental constraints and is, in this respect, a
multi-criteria approach. It makes it possible to find the whole set of
viable management strategies that keep a system within some
constraints. As it focuses on a set of management strategies and not
on a single optimal one, it is of high interest to study management
flexibility, i.e. the system ability to adapt to internal or external
changes. Viability analysis has been applied to biodiversity
management [12], and the sustainability of agricultural systems
[13,14]. Figure 1. Conceptual model of the direct and indirect effects of
grazing on bird population dynamics. Dynamics of grass biomass
(black arrows) is controlled through timing and intensity of grazing;
double arrow represents cattle consumption of standing live and dead
biomass. The bird model is a single stage matrix model. doi:10.1371/journal.pone.0033257.g001 Introduction We
first assess their differences in productive and ecological perfor-
mance. A scenario will lead to better performance if it performs
better in one dimension without performing worse in the other. April 2012 | Volume 7 | Issue 4 | e33257 1 Action versus Result-Oriented Schemes PLoS ONE | www.plosone.org Model overview In line with the model of Sabatier, Doyen & Tichit [15], our
model relies on a state-control approach that represents a
grassland agroecosystem which is the breeding habitat of a bird
species, the lapwing, and the feeding resource for domestic cattle. It is a discrete time model linking grazed grass dynamics to bird
population dynamics (Fig. 1). Time step is defined on a monthly
basis, which is coherent with farmers’ management as most
farmers implement middle term grazing sequences (three weeks to
several months). In the grazed grass sub-model, biomass is
harvested through grazing. The biomass represents a single
grassland patch homogeneously managed without any spatial
dimension. The grassland patch is one of the feeding resources
available for cattle. We assumed that when cattle do not graze the
grassland patch, they are fed elsewhere with other resources (either
on temporary grasslands or indoor). The bird sub-model simulates
population changes over time in response to the direct and indirect Figure 1. Conceptual model of the direct and indirect effects of
grazing on bird population dynamics. Dynamics of grass biomass
(black arrows) is controlled through timing and intensity of grazing;
double arrow represents cattle consumption of standing live and dead
biomass. The bird model is a single stage matrix model. doi:10.1371/journal.pone.0033257.g001 PLoS ONE | www.plosone.org April 2012 | Volume 7 | Issue 4 | e33257 2 Action versus Result-Oriented Schemes cattle density u(t), s the primary sex ratio and s1 (h(F(B,u))) the chick
survival that depends on grass height h(B) at time t*+1. Grass height
depends on grass biomass B(t*+1) and therefore on F(B(t*),u(t*)). We
consider that breeding success is affected by an intra-specific
competition. We use a Beverton-Holt-like density dependence
function to model this competition in which c measures the strength
of competition. A full description of the bird model along with
parameter values are given in Appendix S2 and Table S2. height, heading toward better habitats. Habitat-oriented scenario
combines thresholds on bird fecundity and grass heights during the
chick rearing period. It ensures both fecundity and juvenile
survival to be maintained to a high level. In the result-oriented
scenario, constraints correspond to a threshold on minimal bird
population size throughout time. Management is free and any
management strategy maintaining the bird population through
time is considered as viable. Productive constraints do not differ
between the three scenarios. For each scenario, the model
computed the viable grazing strategies meeting the constraints. Grazed grass state and dynamics The first state of the system represents a grass biomass vector
B(t) considered monthly and partitioned into live and standing
dead grass (BL(t), BD(t)) both expressed in organic matter (g OM
ha21). Grass dynamics is controlled by the timing and intensity of
grazing u(t), expressed in livestock unit per ha (LU ha21). The
dynamics of the grass biomass B(t) controlled by grazing intensity
u(t) is summarized as follows: a
ð Þ
qu(t)ƒB tð Þ
for t~0,1,::,T:
ð6Þ ð6Þ This feeding requirement constraint limits stocking density
which cannot exceed the available biomass B*(t). It assumes that
cattle cannot graze below a minimal biomass threshold and
situations where insufficient grass availability could lead to a
poorer body condition of livestock are not considered. B tz1
ð
Þ~A t,B tð ÞB tð Þ
ð
Þ{G u tð Þ,B tð Þ
ð
Þ
for t~0, 1, . . . ,T
ð1Þ Productive
constraint. A
second
constraint
defines
a
minimal level of productive performance necessary for the
farmer. Productive
performance
P(u,T)
corresponds
to
the
number of grazing days (simplifying to 30 days per month)
associated with a grazing strategy u = [u(0),…,u(T)]. The model
does
not
incorporate
explicitly
any
spatial
scale
but
the
quantification of the productive performance is given for one
hectare. The productive constraint corresponds to a lower
threshold on the number of grazing days on the whole time
period studied. It does not imply any minimum time period or
upper threshold for grazing. It reads as follows: where matrix A is a time dependent matrix that encompasses the
transition rates from t to t+1. It includes grass growth, senescence
and decay rates that are time dependent on a monthly basis. G is a
vector representing biomass harvest through grazing. The state of
the biomass is linked to grass height through a linear function h(B). Databases from the Ouest-du-Lay marsh were used to parame-
terize the grazed grass dynamics ([15]; appendix S3). For further
details on biomass dynamics and parameter values, see Appendix
S1 and Table S1. Model overview The number of viable management strategies is used as an
indicator of the flexibility of the system. Due to their extremely
high number, viable strategies could not be counted directly and
the number of states reached by the viable strategies (or size of the
viability tube) was used as an index of the system flexibility. Viability constraints Three types of viability constraints formalize the multiple roles
played by the grazed grassland. Constraints applied to the three
scenarios are listed in Table 1. Cattle feeding requirement constraint. Given a monthly
biomass demand per livestock unit q, the feeding requirement
constraint is defined as follows: M t,u,B,N
ð
Þ~1 Bird state and dynamics b
ð Þ
P(u,T)~
X
T
t~0
30:u tð Þ§Pb
ð7Þ ð7Þ The second state of the system describes the lapwing life cycle. By
contrast with Sabatier, Doyen & Tichit [15], the bird model is
deterministic and represents the female portion of a single class
population. During the nesting period, cattle trampling impacts
clutch size and during the chick rearing period grass height is a
variation factor of juvenile survival. Assuming a pre-breeding census,
the monthly dynamics of birds N(t) from t to t+1 reads as follows: where Pb is the threshold of minimal productive performance. Its
value was defined by the 10% lower quantile of a dataset of 344
real grazing strategies recorded on our study site [18]. Ecological constraints. Ecological constraints are defined in
three different ways so as to capture the three kinds of scenarios. In the action-oriented scenario (AO), the model includes two
ecological constraints. The first one is related to trampling
mechanisms. An upper threshold u# is imposed on cattle density
during the nesting month t*: N tz1
ð
Þ~M t,u tð Þ,B tð Þ,N tð Þ
ð
ÞN tð Þ
ð2Þ ð2Þ N tz1
ð
Þ~M t,u tð Þ,B tð Þ,N tð Þ
ð
ÞN tð Þ where N(t) is the population size and M(t,u,B,N) the population
growth function: c
ð Þ
u t
ð
Þƒu#
ð8Þ ð8Þ M t,u,B,N
ð
Þ~1
if t=t ð3Þ if t=t ð3Þ M t,u,B,N
ð
Þ~1 M t,u,B,N
ð
Þ~1 The second constraint is related to grass height during the first
month following chick birth. It is represented by a lower threshold
on cattle density during nesting month to induce a minimum level
of grazing: M t,u,B,N
ð
Þ~s2za:f u
ð Þ:s:s1 h F B,u
ð
Þ
ð
Þ
ð
Þ= 1zc:N
ð
Þ
if t~t ð4Þ M t,u,B,N
ð
Þ~s2za:f u
ð Þ:s:s1 h F B,u
ð
Þ
ð
Þ
ð
Þ= 1zc:N
ð
Þ
if t~t ð4Þ with F B,u
ð
Þ~A t ,B
ð
ÞB{G u,B
ð
Þ
ð5Þ ð5Þ d
ð Þ
u t
ð
Þ§ub
ð9Þ ð9Þ where t* is the nesting month, s2 the annual adult survival, a the
proportion of breeding females, f (u) the clutch size depending on PLoS ONE | www.plosone.org PLoS ONE | www.plosone.org April 2012 | Volume 7 | Issue 4 | e33257 PLoS ONE | www.plosone.org 3 Action versus Result-Oriented Schemes Table 1. Constraint sets of the three scenarios. Scenarios
Constraints
Action-oriented
Habitat-oriented
Result-oriented
Productive performance P(u,T).Pb
X
X
X
Cattle feeding requirements q.u(t),B*(t)
X
X
X
Trampling u(t)#u#
X
Fecundity f(t)#fb
X
Grazing u(t)$ub
X
Habitat quality hb#h (t)#h#
X
Population size N(t)$Nb
X
Productive performance constraint imposes that productive performance P(u,T) stays over a minimal productive performance Pb (the minimal annual number of grazing
days per hectare associated with a grazing strategy u). Cattle feeding requirement constraint imposes that cattle demand q.u(t) is always lower than the available
biomass B*(t). Cattle density constraint is an upper threshold u# on cattle density u(t) during the nesting month. A habitat quality constraint imposes grass height to
remain within a minimal hb and maximal h# grass heights during chick rearing. Population size constraint imposes that populations size N(t) stays over a minimum
population size Nb throughout time. doi:10.1371/journal.pone.0033257.t001 Productive performance constraint imposes that productive performance P(u,T) stays over a minimal productive performance Pb (the minimal annual number of grazing
days per hectare associated with a grazing strategy u). Cattle feeding requirement constraint imposes that cattle demand q.u(t) is always lower than the available
biomass B*(t). Cattle density constraint is an upper threshold u# on cattle density u(t) during the nesting month. A habitat quality constraint imposes grass height to
remain within a minimal hb and maximal h# grass heights during chick rearing. if t=t ð3Þ ,T,
satisfying constraint g
ð Þ for any time tw12 and
satisfying constraints b
ð Þ at time Tg
ð15Þ In the result-oriented scenario (RO), it is defined as follows:
ViabRO t0
ð Þ~
B t0
ð Þ, N t0
ð Þ
ð
Þjthere exists grazing u tð Þ and
f
a sequence of states B tð Þ,N tð Þ
ð
Þstarting from B t0
ð Þ, N t0
ð Þ
ð
Þ
satisfying constraints a
ð Þ for any time t~t0, . . . ,T,
satisfying constraint g
ð Þ for any time tw12 and
satisfying constraints b
ð Þ at time Tg
ð15Þ if t=t ð3Þ ,T
ð12Þ ViabHO t0
ð Þ~
B t0
ð Þ, N t0
ð Þ
ð
Þjthere exists grazing u tð Þ and
f
a sequence of states B tð Þ,N tð Þ
ð
Þstarting from B t0
ð Þ, N t0
ð Þ
ð
Þ
satisfying constraint a
ð Þ for any time t~t0, . . . ,T,
satisfying constraints e
ð Þ and
f
ð Þ for any time tw12
and satisfying constraints b
ð Þ at time Tg
ð14Þ ViabHO t0
ð Þ~
B t0
ð Þ, N t0
ð Þ
ð
Þjthere exists grazing u tð Þ and
f
a sequence of states B tð Þ,N tð Þ
ð
Þstarting from B t0
ð Þ, N t0
ð Þ
ð
Þ
satisfying constraint a
ð Þ for any time t~t0, . . . ,T,
satisfying constraints e
ð Þ and
f
ð Þ for any time tw12
and satisfying constraints b
ð Þ at time Tg
ð14Þ ð12Þ In the action-oriented scenario, the ecological constraints bound
the control variable. In habitat-oriented scenario, ecological
constraints combine both control and state constraints. It still
limits cattle density to ensure a good clutch size and also focuses on
an intermediate management objective linked with grass height to
achieve a good juvenile survival. In the result-oriented scenario, no
constraint is set either on cattle density or grass height and the only
ecological constraint corresponds to a state constraint on the
management goal which is the maintenance of the bird population
size above a minimal threshold at any time step. Using such a state
constraint relaxes all management restriction on farmer’s decision. ð14Þ In the result-oriented scenario (RO), it is defined as follows:
ViabRO t0
ð Þ~
B t0
ð Þ, N t0
ð Þ
ð
Þjthere exists grazing u tð Þ and
f
a sequence of states B tð Þ,N tð Þ
ð
Þstarting from B t0
ð Þ, N t0
ð Þ
ð
Þ
satisfying constraints a
ð Þ for any time t~t0, . . . if t=t ð3Þ Population size constraint imposes that populations size N(t) stays over a minimum
population size Nb throughout time. doi:10.1371/journal.pone.0033257.t001 In the habitat-oriented scenario (HO), the model still includes
two ecological constraints. The first one is related to clutch size in
relation with trampling mechanisms. During the nesting month t*,
a lower threshold, f# is imposed on clutch size f(u). As f is a
decreasing function this constraint is similar to eqn 8: kernel depends on time and we prefer to speak of a viability corridor
Viab(t). In this section we will refer to three concepts: the viability
corridor, the viable grazing strategy and the viability tube. Viability corridor. The viability corridor Viab (t) is the set of
grass biomass conditions and bird population sizes (states, BL(t),
BD(t) and N(t)) from which at least one grazing strategy is viable. At
t = t0, the Viability corridor Viab(t0) is thus defined differently in
each scenario. e
ð Þ
f u t
ð
Þ
ð
Þ§f b
ð10Þ ð10Þ In the action-oriented scenario (AO), it is defined as follows: In addition to the previous constraint (eqn 10), the model also
includes a constraint on habitat quality. It is imposed on grass
height during the first month following chick birth (t*+1) in order
to ensure suitable habitat for chicks. It is bounded by minimal and
maximal grass heights as follows: ViabAO t0
ð Þ~
B t0
ð Þ, N t0
ð Þ
ð
Þjthere exists grazing u tð Þ and
f
a sequence of states B tð Þ,N tð Þ
ð
Þstarting from
B t0
ð Þ, N t0
ð Þ
ð
Þsatisfying constraint a
ð Þ for any time
t~t0, . . . ,T,satisfying constraints c
ð Þ and d
ð Þ for any time
tw12 and satisfying constraints b
ð Þ at time Tg
ð13Þ ð13Þ fð Þ
hbƒh t z1
ð
Þƒh#
ð11Þ ð11Þ In the result-oriented scenario (RO) the model involves a single
ecological constraint that imposes a minimum population size Nb
throughout time: In the habitat-oriented scenario (HO), it is defined as follows: g
ð Þ
N tð Þ§Nb
for t~12, . . . The two hypotheses can be formalized as follows: The two hypotheses can be formalized as follows: Hypothesis 1: For a given result-oriented scenario, there is no
habitat-oriented one that leads to better perfor-
mances and for a given habitat-oriented scenario,
there is no action oriented one that lead to better
performances. Hypothesis 1: E(NAO(T))vE(NHO(T))vE(NRO(T))
and
E(PAO(T))vE(PHO(T))vE(PRO(T))
8
>
<
>
:
ð22Þ ð22Þ In the action-oriented scenario (AO), it is defined as follows: UAO t,B,N
ð
Þ~ u(t)j u(t) satisfies (a), (b), (c) and (d)
(B tz1
ð
Þ,N(tz1))[ViabAO tz1
ð
Þ
ð16Þ whith E(N(T)) the average value of N(T) over a set of
10 000 random viable grazing strategies. ð16Þ whith E(N(T)) the average value of N(T) over a set of
10 000 random viable grazing strategies. E N T
ð Þ
½
~1=10000
X
10000
i~1
Ni(T)
where Ni(T),Bi(T)
ð
Þ[VT(T)
ð23Þ In the habitat-oriented scenario (HO), it is defined as follows: In the habitat-oriented scenario (HO), it is defined as follows: ð23Þ UHO t,B,N
ð
Þ~ u(t)j u(t) satisfies (a), (b), (e) and (f )
(B tz1
ð
Þ,N(tz1))[ViabHO tz1
ð
Þ
ð17Þ ð17Þ Similarly, E(P(T)) is the average value of P(T) over
the same set of 10 000 random viable grazing
strategies. In the result-oriented scenario (RO), it is defined as follows: In the result-oriented scenario (RO), it is defined as follows: Hypothesis 2: For a given result-oriented scenario, there is no
habitat oriented one that leads to a higher flexibility. URO t,B,N
ð
Þ~ uj
u(t) satisfies (a), (b) and (g)
(B tz1
ð
Þ,N(tz1))[ViabRO tz1
ð
Þ
ð18Þ ð18Þ H(VTHO)vH(VTRO)
ð24Þ ð24Þ Viability tube. Finally, we identify the Viability tube VT (t). It is the temporal succession of biomass conditions that are
reachable by viable grazing strategies. It takes into account the fact
that not every viable state can be reached by a viable grazing
strategy. Some states are viable (i.e. starting from them, there is at
least one viable grazing strategy) but they can only be reached by
grazing strategies that are not viable. The viability tube is defined
as follows: The volume of the viability tube is used as an index
of flexibility. A scenario leading to a bigger viability
tube will allow more management strategies, and
is considered being more flexible. Action versus Result-Oriented Schemes plot the viability tubes, we limited the tubes to two states (BL and
BD). The tubes therefore corresponded to projections of the 4
dimensional tubes on the three dimensional spaces defined by BL,
BD and t for a given initial abundance N(t0). For the three scenarios, constraints (c) to (g) were not taken into
account the first year (t,12) so as to enable a transition of the
grazed system toward AES. This choice reflects a conventionally
driven system in which AES would be introduced at the end of the
first year. For the three scenarios, constraints (c) to (g) were not taken into
account the first year (t,12) so as to enable a transition of the
grazed system toward AES. This choice reflects a conventionally
driven system in which AES would be introduced at the end of the
first year. Co-viability analysis ð15Þ The viability framework is used to identify combinations of
biomass B(.), population size N(.) and cattle density u(.) that satisfy
viability constraints throughout time. It relies on the computation of
the so called viability kernel [11]. In the present case, this viability PLoS ONE | www.plosone.org April 2012 | Volume 7 | Issue 4 | e33257 4 Action versus Result-Oriented Schemes Simulations To test hypotheses 1 and 2, we followed a two step approach. First we tested them for a given set of ecological constraints and
initial conditions (ub = 0.5; u# = 2; f b = 2.5; hb = 0; h# = 14; Nb = 30;
N(t0) = 30). Then we performed a sensitivity analysis to verify the
generality of our results under a wider range of ecological
constraints and initial conditions (ub = [0, 0.5, 1, 1.5, 2]; u# = [1,
1.5, 2, 3, 4, 5]; f b = [3.2, 2.5, 1.9, 1.5, 1.1]; hb = [0, 5, 7, 10, 12, 13,
14]; h# = [10, 12, 14, 17, 20, 30]; N(t0) = [25, 30, 35]). Constraint
values were chosen to explore the range of possible states and
controls observed in our study area on lapwing nesting fields
(0#h#30 and 0#u#5; [18]). As f(u) is a monotonous function of
u, values of f b were thus chosen to correspond to the different
thresholds on u#. VT 0
ð Þ~Viab 0
ð Þ
ð19Þ ð19Þ VT(tz1)~
B,N
ð
Þ
A ~B, ~N,~u
~B, ~N
[Viab(t)
~u[U(t,~B, ~N)
~B(tz1), ~N(tz1)
~(B,N)
8
>
>
<
>
>
:
9
>
>
=
>
>
;
ð20Þ ð20Þ As they differ among scenarios, we distinguished VTHO, VTRO and
VTAO. We characterized the Viability tubes by their volumes
H(VT). As they differ among scenarios, we distinguished VTHO, VTRO and
VTAO. We characterized the Viability tubes by their volumes
H(VT). A dynamic programming algorithm [10] was used to identify
viable initial conditions (B(t0), N(t0)), viable grazing strategy
U(t,B,N), grass state trajectories B(t) and bird population state
trajectories N(t) respecting the different constraints at each time
step over a period of T = 96 months. The numerical computations
were performed with Scilab 4.1.2 software (http://www.scilab. org/; Scilab Consortium 2007). Once viable grazing strategies and
state trajectories were found, their ecological and productive H(VT)~
X
T
t~t0
ð ð
VT(t)
dBDdBL
for a given N t0
ð Þ
ð21Þ H(VT) (expressed in g2.s.ha22) is a viability metric and an
indicator of the quantity of viable state trajectories. Our system
includes three state dimensions (BL, BD and N). Hypotheses Viable grazing strategies. Once the viability corridor has
been found, we compute the viable grazing strategies that verify
the different constraints over the period of time involved. Such U
exist as long as the state (B(t),N(t)) lies within the viability corridor
Viab(t). We thus consider the set of the viable grazing strategy at
time t for a given viable state (B(t), N(t)). A viable grazing strategy
is a temporal sequence of grazing intensities that keeps the whole
system within the constraint set. To each viable grazing strategy
corresponds a viable state trajectory defined in terms of grass
biomass and population size. These viable grazing strategies U (t,
B, N) are defined through a dynamic programming structure. Hypothesis 1: scenarios differ in performance Fig. 2 shows the ecological and productive performance of a
sample of 10 000 grazing strategies for each of the three scenarios. Comparison of both average ecological and productive perfor-
mance of the three scenarios showed significant differences
(permutation test, p-value = 0). However, differences between the
habitat and result-oriented scenarios were much lower than
differences between the action-oriented scenario and the other two
scenarios (Table 2). The result-oriented scenario led to better
performances than the habitat-oriented one and the latter scenario
led to much better ecological performance than the action-
oriented one and slightly better productive ones. However, it
should be kept in mind that the habitat and result-oriented
scenarios were very similar for both performance criteria. Discussion First, our results showed that in most cases the habitat and
result-oriented scenarios led to much better ecological perfor-
mance than the action-oriented scenario. Productive performance
was quite similar among the scenarios. Secondly, our results
showed that the result-oriented scenario had a higher flexibility
than the habitat-oriented one. This difference in flexibility was
even greater when the grazed grassland agroecosystem was
exposed to climatic variation. The inclusion of the two tubes means that the flexibility of the
result-oriented scenario at least as high as the flexibility of the
habitat-oriented one. For these two scenarios, ensuring similar
levels of performance (Table 2), tubes were bigger in the result-
oriented than in the habitat-oriented scenario. Indeed, the
calculation of H(VT) showed 1.5 more viable grass states in the
result-oriented
scenario
than
in
the
habitat-oriented
one
(H(VTRO) = 6842 versus H(VTHO) = 4997 g2.s.ha22). A larger
range of grass biomass conditions was thus available for farmers
throughout time. The shape of the Viability tube for both habitat
and result-oriented scenarios illustrates the couples of possible
viable states (BL, BD) throughout time and the higher flexibility of
the result-oriented scheme (Fig. 3.a and 3.b). ð25Þ ð25Þ Action versus Result-Oriented Schemes Action versus Result-Oriented Schemes Action versus Result-Oriented Schemes performances N(T) and P(T) were assessed. The performance of
the agroecosystem under the three scenarios AO, HO and RO
was compared with a permutation tests using Python 2.6 (http://
www.python.org/) so as to test Hypothesis 1. For a given
performance (ecological or productive one) and for a given pair
of scenarios, the test calculates a criterion (the difference of the
average performances) and compares it to the distribution of this
criterion for n = 10000 random permutations within the two sets of
trajectories tested. The p value of the test is the proportion of
permuted situations for which the criterion is larger (in absolute
value) than the criterion of the not permuted situation. More
details on permutation tests can, for example, be found in [19]. In
order to investigate the advantage of the improved flexibility of the
result-oriented scenario in facing climatic variations, we tested the
effect on the viability tubes of a shock in climatic conditions
represented by an increased grass growth in year 5. Parameters of
matrix A were modified so as to simulate an earlier grass onset in
the season (i.e. one month earlier) and a stronger grass growth (i.e. +25%). These results illustrate that more flexibility was given to the
grazing strategies in the result-oriented scenario. We have therefore
validated Hypothesis 2. In terms of management this means that the
farmer could implement a wider range of grazing strategies in the
result-oriented scenario than in the habitat-oriented one (appendix
S4). Especially, higher cattle densities can be implemented in spring
with the result-oriented scheme. Sensitivity analysis Results of the sensitivity analysis are presented in Appendix S5. Sensitivity analysis showed one limit case (h# = 30 cm) for which
Hypothesis 1 was falsified. In this situation both performances of the
action-oriented scenario were higher than those of the habitat-
oriented one. Apart from this case, when scenarios could be ranked,
action-oriented scenario always led to worse performances than
habitat-oriented one and both action and habitat-oriented scenarios
led to worse performances than result-oriented scenario. Hypothesis
1 was therefore acceptable for most constraint values. Whatever the
parameter settings, Hypothesis 2 was always true. Results We examined the interest of the improved flexibility of the result-
oriented scenario in facing environmental variations. It turned out
that the state of the system still lied within the result-oriented
viability tube VTRO despite the disturbance, while it left the habitat-
oriented viability tube VTHO. In other words, no couple of control
strategy and state trajectory respected all productive and habitat-
oriented ecological constraints. Thus it was not possible for the
farmer to produce a suitable grass height for birds every year with
low trampling while ensuring good productive performance and
satisfying cattle feeding requirements. However, the result-oriented
tube was not empty and it was possible to find viable state
trajectories and control strategies. As illustrated with one simulation
(Fig. 4), a viable result-oriented grazing strategy did not respect
habitat-oriented constraints every year but it did, however, maintain
bird populations throughout time due to inter-annual compensa-
tions. In this example, grazing intensity in spring was low in 2009
and 2010 (Fig. 4.a). It implied low levels of trampling and an
increase in bird population sizes (Fig. 4.c). In 2011, spring grazing
intensity was stronger and bird population decreased but still
remained above the population threshold. This result shows how, in
the result-oriented scenario, the farmer can adapt his management
to climatic shocks by implementing an inter-annual variation of
management strategies. Such inter-annual variation in management
was not available in the habitat-oriented scenario. This result again
emphasized the advantages of the increased flexibility provided by
the result-oriented scenario. Simulations So as to be able to PLoS ONE | www.plosone.org PLoS ONE | www.plosone.org April 2012 | Volume 7 | Issue 4 | e33257 April 2012 | Volume 7 | Issue 4 | e33257 5 Hypothesis 2: the result-oriented scenario improves
management flexibility We restricted the comparison of flexibility to the other two
scenarios since the action-oriented scenario did not maintain bird
populations. The inclusion of the tubes, their shape and their
volumes showed that more states and controls were viable in the
result-oriented scenario than in the habitat-oriented one. Numer-
ical computations showed that the habitat-oriented tube was
included in the result-oriented one: Vt,VTHO(t)5VTRO(t)
ð25Þ A modelling approach to compare management
schemes In such a context, biodiversity
becomes a joined-production that could be considered as a new
‘‘crop’’
and
the
capacity
of
farmers
(in
link
with
local
environmental managers and/or researchers) to produce the
empirical knowledge needed should not be underestimated. In
this
transition
phase,
the
modalities
of
the
compensation
payments may however be reconsidered and a form of payment
for knowledge production could replace the payment for results. In the model, such an imperfect knowledge could be integrated
by adding uncertainty on the key parameters in the form of
stochasticity. Using algorithms of stochastic viability [10] would
make it possible to maintain the viability approach in such a
context. mechanisms such as environmental stochasticity or landscape
source/sink mechanisms. These mechanisms are of high impor-
tance in the real world but management through grazing has low
(if any) impact on them and including them in the model would
only have blurred the simulation results. These considerations
have to be kept in mind when considering the results. As an
example, one of the main differences between the model and
reality is the absence of migration. Here, we considered a closed
population of birds to assess the effects of management practices. Using population size as an indicator of ecological performances
was therefore possible as well as very convenient and illustrative. In the field, such an indicator would raise questions. In the one
hand it does not only reflect mechanisms occurring at field scale
but in the other hand, this indicator is much closer to the final
objective of a conservation policy than a bird productivity index
would be. Our results showed several undetermined situations. They could
occur for two reasons: either the three scenarios could not be
ranked or it was impossible to find any viable grazing strategies. Changing the values of the constraints oriented the set of viable
strategies
to
either
better
ecological
or
better
productive
performance, illustrating the trade-off between production and
conservation in such agroecosystems [15,20]. It could lead to
extreme situations with very high performances on one dimension
and very low on the other dimension. These situations could not
be put in a hierarchy. In other cases, the constraint values tested
pushed the system too far and no viable grazing strategy could be
found. Consequently, nothing could be said on Hypothesis 1 since
no performance could be assessed. A modelling approach to compare management
schemes Using a modelling approach gave us the opportunity to
compare situations all other things being equal, as we would have
done in a controlled trial. We therefore did not include PLoS ONE | www.ploson April 2012 | Volume 7 | Issue 4 | e33257 PLoS ONE | www.plosone.org 6 Action versus Result-Oriented Schemes result-oriented schemes. Greater flexibility of management is one
of the major arguments in favour of result-oriented schemes [2]. First, it is expected to improve the resilience of the agroecosystem
as farmers may choose alternative management strategies to
adapt to inter-annual climatic variability. The agroecosystems we
studied are low input, extensively grazed grasslands. Such systems
are highly dependent on climatic conditions and flexibility in
grassland use is a major component in coping with unexpected
events [29]. In comparison with habitat-oriented schemes that
impose constraints on habitat and fecundity every year and forces
periodic management strategies, the result-oriented schemes
allow for inter-annual variability. It gives the possibility of
segregating ecological and productive objectives among years
(e.g. to adapt grazing strategies to climatic conditions). These new
strategies are the basis of the higher flexibility of the result
oriented schemes. Our study focussed on temporal flexibility of
grazing strategies but we conjecture that in the same way, spatial
flexibility would allow farmers to adapt their management to
variations in external conditions among several fields. Further
development of the model will take these spatial variations into
account. A second advantage of this greater flexibility would be to
allow farmers to look for innovative management strategies. Our
results suggest that loosening the ecological constraints of the
agroecosystem gives farmers a higher
degree
of
freedom. Matzdorf & Lorenz [30] indicate that this potential of innovation
is very well used by farmers involved in result-oriented schemes. It also leads the farmers to become more involved in conservation
and increases their willingness to improve ecological performance
of their fields [2]. In this study, we focussed on a well known
species. However, such detailed knowledge is not often available. In the absence of stabilized knowledge on the effects of farming
activities on biodiversity, the high potential of innovation,
associated to the willingness to improve ecological performance
that result-oriented schemes provide may help finding ecological
sound management strategies. A modelling approach to compare management
schemes p
Result-oriented schemes aim at protecting the whole agroe-
cosystem by targeting umbrella species. We could here focus on
management strategies that impact the whole agroecosystem and
offer
advantages
to
other
species
with
similar
ecological
requirements and similar sensitivity to management. However,
in the field, farmers may implement very specific measures only
benefiting the target species. For example, in the result-oriented
scheme implemented in the Netherlands, it happened that
farmers only build an electric fence around the nest [5]. If this
management leads to better hatching success for the target
species, it is
of
minor
interest for
other species
in the
agroecosystem. This measure has been strongly criticized for
its lack of cost effectiveness [21] and was cancelled in the new
scheme. To avoid it, the evaluation of management must be
done on an indicator as close to the final objective of
conservation as possible. Considering several species [8,22]
could be a powerful solution. Best effects are expected with
management options having broad effects on the agroecosystem. In this respect, management options at field scale include grazing
sequences, amount and timing of fertilization as well as mowing
techniques and dates. At upper scale, the proportion of land uses
[23,24] as well as their spatial arrangement [25,26,27] could also
be efficient management options that would impact the whole
agroecosystem. Result-oriented schemes have many advantages. They seem
moreover to be very well accepted by farmers since they do not
necessarily imply extra-costs and allow for more room for
manoeuvre in the management of their farm [8]. The set up of
such schemes in the field seems to be more limited by legal issue
than by acceptance by local stakeholders. Indeed, the Rural
Development Regulation, based on a strict interpretation of the
World Trade Organisation rules, restricts payments for farmers
to compensations of income losses or additional costs due to a
change of management practice. This rule fits well to Action-
Oriented Schemes but result-oriented ones are seen as distorting
measures
and
public
stakeholders
are
often
reluctant
to
implement them. This legal problem is one of the reasons for
the abandonment of the Dutch result-oriented scheme [31]. According to Schwartz et al. [2], a window of negotiation seems
however to be available in the WTO rules but would imply high
level negotiations. PLoS ONE | www.plosone.org Improving management flexibility Multi-criterion analysis mainly looks for optimal performance
but do not take into account the issue of flexibility in decision
making. Optimality is well adapted to static situations or stable
environments but flexibility is of major concern for systems
exposed to uncertainties [28]. The viable control approach makes
it possible to go beyond the search of optimum and to look for a
diversity
of
management
strategies. Although
management
strategies were quite similar in terms of performance, the number
of viable management strategies gave a strong advantage to PLoS ONE | www.plosone.org April 2012 | Volume 7 | Issue 4 | e33257 April 2012 | Volume 7 | Issue 4 | e33257 7 Action versus Result-Oriented Schemes Figure 2. Ecological performance N(T) and productive performance P(u,T) and histograms of distributions of action-oriented,
habitat-oriented and result-oriented. For each scenario, results are plotted for a sample of 10 000 couples of viable state trajectories and viable
grazing strategies. The action-oriented scenario (dark gray) is run with cattle density constraint (ub = 0.5; u# = 2 livestock units per hectare); the
habitat-oriented scenario (light gray) is run with fecundity and habitat quality constraints (fb = 2.5, hb = 0 cm and h# = 14 cm); the result-oriented
scenario (empty black) is run with minimum population size (Nb = 30); all, scenario involve constraints on productive performance and cattle feeding
requirement; all scenarii are run with initial population size N(t0) = 30. doi:10.1371/journal.pone.0033257.g002 Figure 2. Ecological performance N(T) and productive performance P(u,T) and histograms of distributions of action-oriented,
habitat-oriented and result-oriented. For each scenario, results are plotted for a sample of 10 000 couples of viable state trajectories and viable
grazing strategies. The action-oriented scenario (dark gray) is run with cattle density constraint (ub = 0.5; u# = 2 livestock units per hectare); the
habitat-oriented scenario (light gray) is run with fecundity and habitat quality constraints (fb = 2.5, hb = 0 cm and h# = 14 cm); the result-oriented
scenario (empty black) is run with minimum population size (Nb = 30); all, scenario involve constraints on productive performance and cattle feeding
requirement; all scenarii are run with initial population size N(t0) = 30. doi:10.1371/journal.pone.0033257.g002 Toward increased spatial scales ecological outcome. With result-oriented schemes both productive
and ecological performance would have to be taken into account as
the ecological outcome would be of major concern to farmers. Such
schemes could thus be expected to reach better levels of
effectiveness. The level of payment would however need to be
addressed with caution for the scheme to remain attractive. Our
model does not include economic incentives yet and development in
this direction should help defining these levels of payment. Other mechanisms may improve the effectiveness of result-
oriented schemes. For instance, farmers frequently allocate
schemes to fields with the lowest productivity so as to limit the
impact on the overall performance of the farm [2]. Therefore, the
localisation of AES fields is often defined regardless of its expected Table 2. Ecological and productive performance of action,
habitat and result-oriented scenarios. The two state d
biomass both expressed in organic matter (1024 g OM ha21). The viability tube corresponds to the vo
and the wireframe. Dark gray areas are the ones for which no viable state exists. doi:10.1371/journal.pone.0033257.g003 Figure 3. Zoom on three years of the viability tubes (VTRO and VTHO) for the result-oriented (fig a) and habitat-oriented (fig b)
scenarios. The tubes show the set of viable states throughout time (in months). The two state dimensions are the live biomass and the dead
biomass both expressed in organic matter (1024 g OM ha21). The viability tube corresponds to the volume (in blue) between the light gray surface
and the wireframe. Dark gray areas are the ones for which no viable state exists. doi:10.1371/journal.pone.0033257.g003
Action versus Result-Oriented Schemes Action versus Result-Oriented Schemes Figure 3. Zoom on three years of the viability tubes (VTRO and VTHO) for the result-oriented (fig a) and habitat-oriented (fig b)
scenarios. The tubes show the set of viable states throughout time (in months). The two state dimensions are the live biomass and the dead
biomass both expressed in organic matter (1024 g OM ha21). The viability tube corresponds to the volume (in blue) between the light gray surface
and the wireframe. Dark gray areas are the ones for which no viable state exists. doi:10.1371/journal.pone.0033257.g003 would have another major advantage. Groups of farmers could
both adapt their management practices at the field scale and
modify the spatial allocation of management practices at the
landscape scale in order to create habitat heterogeneity. Increased
landscape heterogeneity could improve ecological performance as
it makes spatial complementarities among habitats possible [26]. Improvement of the model presented here to account for these
spatial effects (nest site selection, landscape heterogeneity,…) is
another major perspective of this work that we are currently
handling [27,32]. would have another major advantage. Groups of farmers could
both adapt their management practices at the field scale and
modify the spatial allocation of management practices at the
landscape scale in order to create habitat heterogeneity. Increased
landscape heterogeneity could improve ecological performance as
it makes spatial complementarities among habitats possible [26]. Improvement of the model presented here to account for these
spatial effects (nest site selection, landscape heterogeneity,…) is
another major perspective of this work that we are currently
handling [27,32]. occurring at the field scale but also on processes occurring at a
larger scale (i.e. Table 2. Ecological and productive performance of action,
habitat and result-oriented scenarios. Table 2. Ecological and productive performance of action,
habitat and result-oriented scenarios. Table 2. Ecological and productive performance of action,
habitat and result-oriented scenarios. Table 2. Ecological and productive performance of action,
habitat and result-oriented scenarios. Action
oriented
Habitat
oriented
Result
oriented
Productive performance
(LU.days/ha)
1313
(95)
1321
(74)
1339
(58)
Ecological performance
(Population size)
4
(1)
29
(4)
31
(2)
Means and standard deviation () are given for three random samples of 10 000
viable state trajectories and viable grazing strategies. Productive performance
P(u,T) is the number of livestock unit.days ha21 (LU.days/ha) characterizing a
grazing strategy. The ecological performance N(T) is the bird population size at
time horizon (starting with N(t0) = 30). doi:10.1371/journal.pone.0033257.t002 p
g
p y
Beyond the legal issues mentioned at the end of the former
section, other limits of result-oriented schemes arise from the
possible difficulties to assess the ecological outputs. Schwarz et al. [2] recommend focusing in a first step on plant communities as
ecological and agricultural processes fit into the same scale: the
field. Methods that prove to be fair to the farmer have been
developed in Germany [22] and in France [8] to provide
assessments in the case of grassland flora. However, concerning
mobile species, such as birds, with larger home ranges, assessment
at field scale is more difficult. First, birds are not present in the
field all the time and accurate surveys imply heavy monitoring
protocols. A solution to this first problem was to focus on local
indicators such as breeding success but results were mitigated. [5,6]. The second difficulty, which is linked to the latter point, is
that bird population trends not only depend on processes Means and standard deviation () are given for three random samples of 10 000
viable state trajectories and viable grazing strategies. Productive performance
P(u,T) is the number of livestock unit.days ha21 (LU.days/ha) characterizing a
grazing strategy. The ecological performance N(T) is the bird population size at
time horizon (starting with N(t0) = 30). doi:10.1371/journal.pone.0033257.t002 April 2012 | Volume 7 | Issue 4 | e33257 April 2012 | Volume 7 | Issue 4 | e33257 PLoS ONE | www.plosone.org 8 occurring at the field scale but also on processes occurring at a
would have another m
Figure 3. Zoom on three years of the viability tubes (VTRO and VTHO) for the result-orie
scenarios. The tubes show the set of viable states throughout time (in months). Appendix S1
Discrete time dynamics of the grazed
grassland.
(DOC) We thank Donald White for proofreading the manuscript. We also thank
Dr Melman for fruitful discussion on Dutch agri-environment schemes. We thank Donald White for proofreading the manuscript. We also thank
Dr Melman for fruitful discussion on Dutch agri-environment schemes. References 1. Kleijn D, Baquero RA, Clough Y, Diaz M, De Esteban J, et al. (2006) Mixed
biodiversity benefits of agri-environment schemes in five European countries. Ecology Letters 9: 243–254. 15. Sabatier R, Doyen L, Tichit M (2010) Modelling trade-offs between livestock
grazing and wader conservation in a grassland agroecosystem. Ecological
Modelling 221: 1292–1300. 1. Kleijn D, Baquero RA, Clough Y, Diaz M, De Esteban J, et al. (2006) Mixed
biodiversity benefits of agri-environment schemes in five European countries. Ecology Letters 9: 243–254. g
16. Beintema AJ, Muskens GJDM (1987) Nesting success of birds breeding in Dutch
agricultural grassland. Journal of Applied Ecology 24: 743–758. 2. Schwarz G, Moxey A, McCracken DI, Huband S, Cummins R (2008) An analysis
of the potential effectiveness of a Payment-by-Results approach to the delivery of
environmental public goods and services supplied by Agri-Environment Schemes. 108 p. Available: http://www.lupg.org.uk/pdf/LUPG_Payment_by_Results_
Feb09.pdf. 17. Devereux CL, McKeever CU, Benton TG, Whittingham MJ (2004) The effect
of sward height and drainage on Common Starlings Sturnus vulgaris and Northern
Lapwings Vanellus vanellus foraging in grassland habitats. Ibis 146: 115–122. 3. Zabel A, Holm-Muller K (2008) Conservation performance payments for
carnivore conservation in Sweden. Conservation Biology 22: 247–251. 18. Durant D, Tichit M, Fritz H, Kerneis E (2008) Field occupancy by breeding
lapwings Vanellus vanellus and redshanks Tringa totanus in agricultural wet
grasslands. Agriculture Ecosystems & Environment 128: 146–150. 4. Klimek S, Kemmermann AR, Steinmann HH, Freese J, Isselstein J (2008)
Rewarding farmers for delivering vascular plant diversity in managed grasslands:
A transdisciplinary case-study approach. Biological Conservation 141:
2888–2897. 19. Sokal R, Rohlf F (1995) Biometry (3rd edn). WH Freman and company: New
York. 20. Groot JCJ, Rossing WAH, Jellema A, Stobbelaar DJ, Renting H, et al. (2007)
Exploring multi-scale trade-offs between nature conservation, agricultural profits
and landscape quality–A methodology to support discussions on land-use
perspectives. Agriculture, Ecosystems & Environment 120: 58–69. 5. Verhulst J, Kleijn D, Berendse F (2007) Direct and indirect effects of the most
widely implemented Dutch agri-environment schemes on breeding waders. Journal of Applied Ecology 44: 70–80. perspectives. Agriculture, Ecosystems & Environment 120: 58–69 6. Musters CJM, Kruk M, De Graaf HJ, Ter Keurs WJ (2001) Breeding birds as a
farm product. Conservation Biology 15: 363–369. 21. Brunner A, Huyton H (2005) Agri-environment schemes and biodiversity:
lessons learnt and examples from across Europe Bird Life International. 14 p. Available: http://www.birdlife.org/eu/pdfs/Agrienvironment_schemes_lesson_
learnt.pdf. 7. Appendix S2
Discrete time dynamics of the wader
population. Conceived and designed the experiments: RS MT LD. Performed the
experiments: RS. Analyzed the data: RS. Wrote the paper: RS LD MT. Conceived and designed the experiments: RS MT LD. Performed the
experiments: RS. Analyzed the data: RS. Wrote the paper: RS LD MT. Appendix S4
Different degrees of freedom in grazing
sequences.
(DOC) Appendix S5
Sensitivity analysis. (DOC) Table S1
Parameters used in the grazed grass model. (DOC) Table 2. Ecological and productive performance of action,
habitat and result-oriented scenarios. a set of neighbouring fields). A solution to this
problem could be to develop schemes at a scale matching the
home range of species under concern. However, management at
larger scales involving several land owners may lead to situations
where some land owners behave as free-riders and compromise
the success of the scheme. This issue has been taken into
consideration in Sweden in the case of carnivores with very large
home ranges [3]. In this case, payments by results were not given
directly to individuals but to the communities. The efficiency of
the conservation policy thus relied on collective action. Result-
oriented schemes at the landscape scale based on collective action April 2012 | Volume 7 | Issue 4 | e33257 PLoS ONE | www.plosone.org 9 Action versus Result-Oriented Schemes PLoS ONE | www.plosone.org
10
April 2 PLoS ONE | www.plosone.org
10
April 2012 | Volume 7 | Issue 4 | e33257 April 2012 | Volume 7 | Issue 4 | e33257 April 2012 | Volume 7 | Issue 4 | e33257 PLoS ONE | www.plosone.org PLoS ONE | www.plosone.org PLoS ONE | www.plosone.org Action versus Result-Oriented Schemes Figure 4. Example of one viable grazing strategies and state trajectories in the result-oriented scenario with a climatic perturbation
(zoom on the three years around the climatic perturbation). The different constraints are plotted to illustrate the fact that viable result-
oriented strategies would not respect action and habitat oriented constraints. Figure a, viable grazing strategy in the result oriented scheme. Red bars
represent the cattle density constraint u# applied in the habitat and action oriented schemes. Figure b, viable grass height trajectory h in the action
and result oriented schemes. Red bars represent the grass height constraint h# applied in the habitat and action oriented schemes. Figure c, viable
bird population trajectory N. The red dotted line stands for the population size constraint Nb. The green and red arrows highlight the mechanisms of
inter-annual compensation. doi:10.1371/journal.pone.0033257.g004 Conclusion Compared with action-oriented schemes, our study shows that
improvement of ecological performance is high when schemes are
habitat or result-oriented. Differences in performances between
habitat and result-oriented schemes remained limited. The main
advantage of result-oriented schemes is to increase the overall
management flexibility of the grassland agroecosystem. Such
improved flexibility may also allow farmers to adapt their
management to climatic variations. Further model developments
will focus on both the spatial and temporal dimensions of farming
flexibility. This next step will make it possible to better match
management and ecological processes. g
g
p
32. Sabatier R, Tichit M (2011) Does landscape heterogeneity modulate the trade-
off between production and biodiversity? Proceedings of the 3rd Farming system
design conference: Resilient Food systems for a Changing World, Brisbane. pp
54–55. Available: http://aciar.gov.au/files/node/13992/does_landscape_
heterogeneity_modulate_the_trade_of_55592.pdf. 31. NEAS (2007) Executive summary: Ecological Evaluation of Nature Conserva-
tion Schemes run under the Stewardship Programme and the Dutch National
Forest Service 2000–2006 Bilthoven. 46 p. Available: http://www.mnp.nl/
images/500410004%20Ecological%20Evaluation_tcm61-35640.pdf. 30. Matzdorf B, Lorenz J (2010) How cost-effective are result-oriented agri-
environmental measures? An empirical analysis in Germany. Land Use Policy
27: 535–544. p
gy
29. Martin G, Cruz P, Theau JP, Jouany C, Fleury P, et al. (2009) A multi-site study
to classify semi-natural grassland types. Agriculture Ecosystems & Environment
129: 508–515. p. Available: http://hal.archives-ouvertes.fr/docs/00/51/05/52/PDF/
Sabatier_Reconciling_production.pdf.
28. Gunderson L (1999) Resilience, flexibility and adaptive management - antidotes
for spurious certitude? Conservation Ecology 3.
29. Martin G, Cruz P, Theau JP, Jouany C, Fleury P, et al. (2009) A multi-site study
to classify semi-natural grassland types. Agriculture Ecosystems & Environment
129: 508–515.
30. Matzdorf B, Lorenz J (2010) How cost-effective are result-oriented agri-
environmental measures? An empirical analysis in Germany. Land Use Policy
27: 535–544. 28. Gunderson L (1999) Resilience, flexibility and ad
for spurious certitude? Conservation Ecology 3. p. Available: http://hal.archives-ouvertes.fr/docs/00/51/05/52/PDF/
Sabatier_Reconciling_production.pdf. p. Available: http://hal.archives-
Sabatier_Reconciling_production.pdf. References SNH (2005) East Scottland Grassland Management Scheme. 12 p. Available:
http://www.snh.org.uk/pdfs/NatCare/GrasslandScheme.pdf. 8. Mestelan P, Agreil C, Marie CdS, Meuret M, Mailland-Rosset S (2007)
Implementing agro-environmental measures based on ecological results. The
case of meadows and rangelands in the massif des Bauges regional Park. Rencontres Recherche Ruminants. Paris. Available: http://www.journees3r.fr/
IMG/pdf/2007_04_pastoralisme_06_Mestelan.pdf. 22. Wittig B, Kemmermann ARG, Zacharias D (2006) An indicator species
approach for result-orientated subsidies of ecological services in grasslands - A
study in Northwestern Germany. Biological Conservation 133: 186–197. 23. Sabatier R, Doyen L, Tichit M (2008) Assessing the effect of stocking density
thresholds on productive and ecological performances of livestock systems based
on grasslands with high biodiversity stakes Institut National de la Recherche
Agronomique (INRA). pp 213–216. Available: http://www.journees3r.fr/IMG/
pdf/2008_07_environnement_01_Sabatier.pdf. IMG/pdf/2007_04_pastoralisme_06_Mestelan.pdf. 9. Durant D, Tichit M, Kerneis E, Fritz H (2008) Management of agricultural wet
grasslands for breeding waders: integrating ecological and livestock system
perspectives - a review. Biodiversity and Conservation 17: 2275–2295. 10. De Lara M, Doyen L (2008) Sustainable management of natural resources;
Allan RUF, W. Salomons, eds. Berlin: Springer. 266 p. 24. Tichit M, Puillet L, Sabatier R, Teillard F (2011) Multicriteria performance and
sustainability in livestock farming systems: Functional diversity matters. Livestock Science 139: 161–171. 11. Aubin J-P, ed (1991) Viability theory. Boston. 542 p. 12. Tichit M, Doyen L, Lemel JY, Renault O, Durant D (2007) A co-viability model
of grazing and bird community management in farmland. Ecological Modelling
206: 277–293. 25. Melman TCP (2010) A web-based tool for tailor made management for meadow
birds. In: biologists Aoa, editor. Leicester. 26. Sabatier R (2010) Multiscale trade-offs between agricultural production and
biodiversity in a grassland agroecosystem. Paris: Agroparistech. 226 p. Available:
https://www.versailles-grignon.inra.fr/sadapt/content/download/4266/
40211/version/1/file/SABATIER_Manuscrit_THESE_+XIV.pdf. 13. Tichit M, Hubert B, Doyen L, Genin D (2004) A viability model to assess the
sustainability of mixed herds under climatic uncertainty. Animal Research 53:
405–417. 14. Baumgartner S, Quaas MF (2009) Ecological-economic viability as a criterion
of strong sustainability under uncertainty. Ecological Economics 68: 2008–
2020. 27. Sabatier R, Doyen L, Tichit M (2010) Reconciling production and conservation
in agrolandscape, does landscape heterogeneity help? Montpellier, France. 10p PLoS ONE | www.plosone.org April 2012 | Volume 7 | Issue 4 | e33257 11 April 2012 | Volume 7 | Issue 4 | e33257 Action versus Result-Oriented Schemes ious certitude? Conservation Ecology 3. Action versus Result-Oriented Schemes Action versus Result-Oriented Schemes PLoS ONE | www.plosone.org April 2012 | Volume 7 | Issue 4 | e33257 12
|
https://openalex.org/W2788394846
|
https://discovery.ucl.ac.uk/id/eprint/10045409/1/Hilt_Response%20of%20submerged.pdf
|
English
| null |
Response of Submerged Macrophyte Communities to External and Internal Restoration Measures in North Temperate Shallow Lakes
|
Frontiers in plant science
| 2,018
|
cc-by
| 18,525
|
1 Department of Ecosystem Research, Leibniz-Institute of Freshwater Ecology and Inland Fisheries, Berlin, Germany,
2 Departmnet of Aquatic Ecology, Netherlands Institute of Ecology (NIOO-KNAW), Wageningen, Netherlands, 3 Biological
Station of Hiddensee, University of Greifswald, Greifswald, Germany, 4 Department of Bioscience, Aarhus University,
Silkeborg, Denmark, 5 Department of Biology, Lund University, Lund, Sweden, 6 Netherlands Environmental Assessment
Agency (PBL), Den Haag, Netherlands, 7 Water Systems and Global Change Group, Wageningen University and Research,
Wageningen, Netherlands, 8 Sino-Danish Centre for Education and Research, University of Chinese Academy of Sciences,
Beijing, China, 9 Institute of Applied Freshwater Ecology, Seddiner See, Germany, 10 Broads Authority, Norwich,
United Kingdom, 11 Department of Aquatic Ecology and Water Quality Management, Wageningen University and Research,
Wageningen, Netherlands, 12 Deltares, Delft, Netherlands, 13 Biological and Environmental Sciences, University of Stirling,
Stirling, United Kingdom, 14 Department of Freshwater Conservation, Brandenburg University of Technology
Cottbus-Senftenberg, Senftenberg, Germany, 15 Niedersächsischer Landesbetrieb für Wasserwirtschaft, Küsten- und
Naturschutz, Sulingen, Germany, 16 Lanaplan, Nettetal, Germany, 17 Gesellschaft für Naturschutz und Landschaftsökologie,
Kratzeburg, Germany, 18 Department of Geography, Environmental Change Research Centre, University College London,
London, United Kingdom Edited by:
Janne Alahuhta,
University of Oulu, Finland University of Oulu, Finland
Reviewed by:
Rebecca Lester,
Deakin University, Australia
Ludwig Triest,
Vrije Universiteit Brussel, Belgium
*Correspondence:
Sabine Hilt
hilt@igb-berlin.de Reviewed by:
Rebecca Lester,
Deakin University, Australia
Ludwig Triest,
Vrije Universiteit Brussel, Belgium Specialty section:
This article was submitted to
Functional Plant Ecology,
a section of the journal
Frontiers in Plant Science Specialty section:
This article was submitted to
Functional Plant Ecology,
a section of the journal
Frontiers in Plant Science
Received: 28 September 2017
Accepted: 01 February 2018
Published: 19 February 2018 Submerged macrophytes play a key role in north temperate shallow lakes by stabilizing
clear-water conditions. Eutrophication has resulted in macrophyte loss and shifts to
turbid conditions in many lakes. Considerable efforts have been devoted to shallow lake
restoration in many countries, but long-term success depends on a stable recovery
of submerged macrophytes. However, recovery patterns vary widely and remain to
be fully understood. We hypothesize that reduced external nutrient loading leads
to an intermediate recovery state with clear spring and turbid summer conditions
similar to the pattern described for eutrophication. In contrast, lake internal restoration
measures can result in transient clear-water conditions both in spring and summer
and reversals to turbid conditions. ORIGINAL RESEARCH
published: 19 February 2018
doi: 10.3389/fpls.2018.00194 Response of Submerged Macrophyte
Communities to External and Internal
Restoration Measures in North
Temperate Shallow Lakes Sabine Hilt 1*, Marta M. Alirangues Nuñez 1, Elisabeth S. Bakker 2, Irmgard Blindow 3,
Thomas A. Davidson 4, Mikael Gillefalk 1, Lars-Anders Hansson 5, Jan H. Janse 2,6,
Annette B. G. Janssen 2,7, Erik Jeppesen 4,8, Timm Kabus 9, Andrea Kelly 10, Jan Köhler 1,
Torben L. Lauridsen 4,8, Wolf M. Mooij 2,11, Ruurd Noordhuis 12, Geoff Phillips 13,
Jacqueline Rücker 14, Hans-Heinrich Schuster 15, Martin Søndergaard 4,8, Sven Teurlincx 2,
Klaus van de Weyer 16, Ellen van Donk 2, Arno Waterstraat 17, Nigel Willby 13 and
Carl D. Sayer 18 Furthermore, we hypothesize that these contrasting
restoration measures result in different macrophyte species composition, with added
implications for seasonal dynamics due to differences in plant traits. To test these
hypotheses, we analyzed data on water quality and submerged macrophytes from 49
north temperate shallow lakes that were in a turbid state and subjected to restoration
measures. To study the dynamics of macrophytes during nutrient load reduction, we
adapted the ecosystem model PCLake. Our survey and model simulations revealed
the existence of an intermediate recovery state upon reduced external nutrient loading,
characterized by spring clear-water phases and turbid summers, whereas internal lake Received: 28 September 2017
Accepted: 01 February 2018
Published: 19 February 2018 INTRODUCTION critical threshold or a significant reduction in the abundance of
planktivorous and benthivorous fish (e.g., by biomanipulation or
natural fish kills) will lead to a recovery of clear-water conditions
and a return of macrophytes (Scheffer et al., 1993). In practice,
reductions in the external nutrient load to shallow lakes often fail
to deliver macrophyte recovery (Jeppesen et al., 2005). Similarly,
biomanipulation of the fish community in turbid shallow lakes
has produced variable effects on macrophytes in shallow lakes
(Hansson et al., 1998; Bergman et al., 1999; Søndergaard et al.,
2008; Jeppesen et al., 2012; Bernes et al., 2015; Sayer et al., 2016). Overall, the response of macrophyte communities to different
types of lake restoration measures remains to be fully understood
(Jeppesen et al., 2005; Bakker et al., 2013). Shallow lakes are the most abundant freshwater ecosystems on
earth (Verpoorter et al., 2014). In their pristine state, they are
often characterized by abundant submerged vegetation which can
stabilize clear-water conditions (Scheffer et al., 1993) and plays a
key role in the functioning of the ecosystem (Hilt et al., 2017). Several mechanisms contribute to a positive feedback between
macrophytes and clear water conditions. As a consequence,
shallow lakes are resistant to increasing nutrient loading up
to a critical threshold, above which their macrophytes collapse
and the lakes shift into a turbid, phytoplankton-dominated state
(Scheffer et al., 1993). In recent centuries, excessive nutrient
loading has resulted in a loss of macrophytes and shift to this
turbid state in many temperate shallow lakes (e.g., Körner, 2002;
Phillips et al., 2016). We hypothesize that (1) external lake restoration measures
leading to nutrient load reduction in turbid temperate shallow
lakes result in macrophyte re-establishment in a reversed
sequence to the one described by Sayer et al. (2010a,b) for
advancing eutrophication. An intermediate recovery state should
occur where the water is clear in spring but dominated by
phytoplankton and thus turbid in late summer, until, eventually,
seasonally stable conditions characterized by high water clarity in
both spring and summer would dominate (Figure 1). In contrast,
lake internal measures such as biomanipulation or phosphorus
precipitation are expected to result in transient clear-water
conditions in spring and summer if either zooplankton is
sufficiently released from fish predation or internal phosphorus
loading from sediments is reduced enough to control summer
phytoplankton. Such conditions are supposed to occur only
temporarily in the absence of additional external nutrient load
reduction (Figure 1). Frontiers in Plant Science | www.frontiersin.org Citation: Hilt S, Alirangues Nuñez MM,
Bakker ES, Blindow I, Davidson TA,
Gillefalk M, Hansson L-A, Janse JH,
Janssen ABG, Jeppesen E, Kabus T,
Kelly A, Köhler J, Lauridsen TL,
Mooij WM, Noordhuis R, Phillips G,
Rücker J, Schuster H-H,
Søndergaard M, Teurlincx S,
van de Weyer K, van Donk E,
Waterstraat A, Willby N and Sayer CD
(2018) Response of Submerged
Macrophyte Communities to External
and Internal Restoration Measures in
North Temperate Shallow Lakes. Front. Plant Sci. 9:194. doi: 10.3389/fpls.2018.00194 February 2018 | Volume 9 | Article 194 Frontiers in Plant Science | www.frontiersin.org 1 Hilt et al. Response of Macrophytes to Restoration restoration measures often resulted in clear-water conditions in spring and summer
with returns to turbid conditions after some years. External and internal lake restoration
measures resulted in different macrophyte communities. The intermediate recovery state
following reduced nutrient loading is characterized by a few macrophyte species (mainly
pondweeds) that can resist wave action allowing survival in shallow areas, germinate early
in spring, have energy-rich vegetative propagules facilitating rapid initial growth and that
can complete their life cycle by early summer. Later in the growing season these plants
are, according to our simulations, outcompeted by periphyton, leading to late-summer
phytoplankton blooms. Internal lake restoration measures often coincide with a rapid
but transient colonization by hornworts, waterweeds or charophytes. Stable clear-water
conditions and a diverse macrophyte flora only occurred decades after external nutrient
load reduction or when measures were combined. Keywords: aquatic plants, biomanipulation, eutrophication, lake restoration. nutrient load reduction, PCLake, plant
traits, regime shift INTRODUCTION (2) Lake-internal measures (biomanipulation, sediment suction dredging) leading to unstable clear-water conditions with
specific macrophyte communities that may collapse resulting in a shift back to turbid conditions unless nutrient loading is reduced, or (3) a combination of external
and internal restoration leading to stable clear-water conditions with an abundant and diverse macrophyte community. FIGURE 1 | Response patterns of turbid north temperate shallow lakes to different restoration measures: (1) External restoration measures (reduction of external
nutrient loading) are expected to lead to an intermediate recovery state with clear-water conditions in spring and turbid water in summer and specific macrophyte
communities with short growth seasons and eventually stable clear conditions with a diverse macrophyte flora if nutrient loading is reduced sufficiently or additional
internal measures are applied (reversed order as suggested for eutrophication by Sayer et al., 2010a,b). Thresholds in phosphorus (P) loading are based on
simulations using PCLake (see Figure 5). (2) Lake-internal measures (biomanipulation, sediment suction dredging) leading to unstable clear-water conditions with
specific macrophyte communities that may collapse resulting in a shift back to turbid conditions unless nutrient loading is reduced, or (3) a combination of external
and internal restoration leading to stable clear-water conditions with an abundant and diverse macrophyte community. FIGURE 1 | Response patterns of turbid north temperate shallow lakes to different restoration measures: (1) External restoration measures (reduction of external
nutrient loading) are expected to lead to an intermediate recovery state with clear-water conditions in spring and turbid water in summer and specific macrophyte
communities with short growth seasons and eventually stable clear conditions with a diverse macrophyte flora if nutrient loading is reduced sufficiently or additional
internal measures are applied (reversed order as suggested for eutrophication by Sayer et al., 2010a,b). Thresholds in phosphorus (P) loading are based on
simulations using PCLake (see Figure 5). (2) Lake-internal measures (biomanipulation, sediment suction dredging) leading to unstable clear-water conditions with
specific macrophyte communities that may collapse resulting in a shift back to turbid conditions unless nutrient loading is reduced, or (3) a combination of external
and internal restoration leading to stable clear-water conditions with an abundant and diverse macrophyte community. UK where shallow lakes are abundant and experience similar
eutrophication problems and climatic conditions. INTRODUCTION We hypothesize that (2) these contrasting
types of restoration measures result in different macrophyte
community
composition
and
seasonal
patterns
in
plant
abundance. Specific macrophyte communities with short growth
seasons should dominate during the intermediate recovery
state following external nutrient loading reduction, while
species with longer growing season requirements are predicted
to
temporarily
establish
following
lake-internal
measures
(Figure 1). The establishment of stable clear conditions with a Sayer et al. (2010a) suggested a typical pattern of lake
macrophyte loss, defining a so-called “crashing” state lying
between the stable clear-water state featuring a diverse plant
community and the final turbid state lacking in macrophytes. This crashing state is characterized by the occurrence of only a
few macrophyte species that can complete their life cycle during
clear-water conditions in spring and early summer while later
in summer, cyanobacteria blooms often occur. Eventually, the
remaining macrophyte stands are also lost and give way to year-
round phytoplankton dominance (Sayer et al., 2010a,b, 2016). Under these conditions, several ecosystem functions and services
deteriorate, including biodiversity support, nutrient retention,
provision of water of drinking or swimming quality (Hilt et al.,
2017). Hence, considerable efforts and financial resources have been
devoted to the restoration of shallow lakes in many countries
in recent decades (Jeppesen et al., 2005). The success of lake
restoration in the long-term depends critically on the stable
recovery of submerged macrophytes (Hilt et al., 2006). However,
the turbid state is stabilized by feedback mechanisms that can
prevent macrophyte re-colonization even at reduced nutrient
loading. In theory, only the reduction of nutrient levels below a February 2018 | Volume 9 | Article 194 2 Hilt et al. Response of Macrophytes to Restoration FIGURE 1 | Response patterns of turbid north temperate shallow lakes to different restoration measures: (1) External restoration measures (reduction of external
nutrient loading) are expected to lead to an intermediate recovery state with clear-water conditions in spring and turbid water in summer and specific macrophyte
communities with short growth seasons and eventually stable clear conditions with a diverse macrophyte flora if nutrient loading is reduced sufficiently or additional
internal measures are applied (reversed order as suggested for eutrophication by Sayer et al., 2010a,b). Thresholds in phosphorus (P) loading are based on
simulations using PCLake (see Figure 5). INTRODUCTION We selected
lakes that had lost most of their submerged macrophytes during
a turbid phase and subsequently had been subjected to either
external restoration via nutrient load reduction (summarized
in Table 1), or internal restoration using biomanipulation or
sediment dredging (Table 2). We did not carry out a full
systematic review of available data, but instead focussed on
known lakes within the research network of the authors where
at least partial recovery through restoration measures was
evident. diverse macrophyte community is thus assumed to require both,
external and internal measures (Figure 1). To test these hypotheses, we analyse existing data on the water
quality and submerged macrophytes of 49 temperate shallow
lakes that had deteriorated to a turbid state and subsequently
were subject to either external or internal restoration measures
or both. In addition, we use an adapted version of the ecosystem
model PCLake (Janse et al., 2008) to simulate the response of
water clarity and macrophyte biomass to external nutrient load
reduction and to detect any thresholds in nutrient loading for
macrophyte recovery. Traits of the typical macrophyte species
found after external and internal lake restoration measures
are compared to provide a mechanistic understanding of the
observed re-colonization patterns. For all lakes, we retrieved information from the turbid
period and its macrophyte assemblage, macrophyte composition
after
external/internal
restoration,
total
phosphorus
(TP)
concentrations in the water and Secchi depth in spring (April-
June) and summer (July–September) using published studies or
questionnaire responses provided by co-authors. As is commonly
the case, data on TP concentrations and Secchi depth were very
diverse, ranging from multi-year weekly measurements to
single values. The data were merged into a single value for
each season and lake by using means (Figures 2A,B) and
raw data are available as a supplement. We also analyzed
the occurrence of dominant macrophyte species during the
recovery period. To visualize potentially typical recovery
patterns, the long-term changes in nutrient concentrations,
water transparency and macrophyte occurrence are shown
in more detail for three lakes restored only by reductions in
external nutrient loading (Müggelsee, Veluwemeer, Eemmeer,
for details see no. 1, 19 and 20 in Table 1) and for three lakes Literature and Data Search on Macrophyte
Species Recovery We started our literature review with the 22 shallow lakes
described in detail in the study by Jeppesen et al. (2005) on the
response of lakes to reduced external nutrient loading. However,
only two of these 22 lakes were turbid before the nutrient
load reduction (Müggelsee, Veluwemeer) and both showed an
increase in macrophyte coverage following the intervention. The
rest showed no change, macrophyte declines or lacked suitable
data (Jeppesen et al., 2005). Therefore, we searched for more
examples in published and unpublished studies on lakes in
Germany, The Netherlands, Denmark, southern Sweden and Frontiers in Plant Science | www.frontiersin.org February 2018 | Volume 9 | Article 194 3 Hilt et al. Hilt et al. Response of Macrophytes to Restoration TABLE 1 | The response of north temperate shallow lakes in Germany (DE), United Kingdom (UK), The Netherlands (NL), and Denmark (DK) to external nutrient load reduction (x: data on Secchi disk transparency and
total phosphorus concentrations were available and used in Figure 2A). No. Lake
Country
Size
(ha)
Depth
(max/mean)
(m)
Measures of nutrient load
reduction
Period of turbid
conditions, major
remaining macrophyte
species and coverage (%
lake area)
Period of intermediate recovery
conditions, major macrophyte
species and coverage (% lake
area)
Period of stable
clear conditions
and macrophyte
species
Data
Figure 2A
References
1
Großer
Müggelsee
DE
750
8/4.9
Since 1989: improved
wastewater treatment in
catchment
1970–1989
Sparse stands of
P. pectinatus
1990–2013
P. pectinatus, P. perfoliatus,
P. crispus, Z. palustris
ca. 5% (Figure 4)
2014-now
P. pectinatus,
P. perfoliatus,
Najas marina,
E. nuttallii
ca. 25% (Figure 4)
x
Hilt et al., 2013, S. Hilt
unpubl. data
2
Großer
Wannsee
282
9.8/5.5
Nutrient load reduction in
catchment
?-ongoing
P. pectinatus, P. crispus
<5%
not yet reached
Hilt and Grünert, 2008;
Van de Weyer, 2011
3
Galen-becker
See
590
1.8/0.8
Since 1972: reduction of P
loading by treatment in
upstream reservoir, since
2007 by regulation of whole
water supply to lake
1995–2002
2003–2006
P. pectinatus
2008–now
Chara contraria,
C. tomentosa,
C. globularis,
C. intermedia,
Nitellopsis obtusa,
N. marina f. intermedia,
P. pectinatus,
Z. palustris
x
Waterstraat, 2008,
unpubl. data
4
Dümmer
1350
4/1.1
Reduction of P load by
wastewater treatment plants
by 95%
1960–2011
P. pectinatus (rare)
2011–ongoing
P. pectinatus, P. crispus, Z. palustris,
P. pusillus
25-50%
not yet reached
x
Blüml et al., 2008;
Schuster, unpubl. Frontiers in Plant Science | www.frontiersin.org Literature and Data Search on Macrophyte
Species Recovery data
5
Schwielow-
see
786
9.1/2.8
Nutrient load reduction in
catchment area (River Havel)
?−2005
2006–now
P. pectinatus
not yet reached
x
Kabus et al., 2007
6
Wuster-witzer
See
172
9.2/3.4
Nutrient load reduction in
catchment area
?−2005
2006–ongoing
P. pectinatus
not yet reached
x
Kabus et al., 2007
7
Grimnitz-see
777
10.3/4.5
Since 1994: sewage
treatment plant in operation
1970–1990
P. pectinatus
1991–? P. pectinatus
?–now
2008: C. contraria,
N. obtusa,
P. pectinatus,
P. perfoliatus,
C. demersum, L. trisulca
2001: C. contraria,
C. globularis,
N. obtusa,
P. pectinatus
x
Mauersberger and
Mauersberger, 1996;
Gervais et al., 1999;
Kabus and
Mauersberger, 2011
(Continued) February 2018 | Volume 9 | Article 194 Hilt et al. Response of Macrophytes to Restoration TABLE 1 | Continued
No. Lake
Country
Size
(ha)
Depth
(max/mean)
(m)
Measures of nutrient load
reduction
Period of turbid
conditions, major
remaining macrophyte
species and coverage (%
lake area)
Period of intermediate recovery
conditions, major macrophyte
species and coverage (% lake
area)
Period of stable
clear conditions
and macrophyte
species
Data
Figure 2A
References
8
Wardersee
357
10.8/3.7
Wastewater treatment
plants
? 1996–2006
P. pectinatus, P. perfoliatus,
P. crispus, Z. palustris
? x
Landesamt für Natur
und Umwelt des
Landes
Schleswig-Holstein,
1997; Heinzel and
Martin, 2006
9
Hemmels-
dorfer
See
450
6 (northern
part)
1998: P load reduction
?-1978-? ?-2006-? P. pectinatus, Z.palustris,
P. perfoliatus, R. circinatus
? Heinzel and Martin,
2006
10
Großer
Varchen-tiner
See
182
1.7/? Lowered nutrient input from
agricultural catchment since
1990
?–? ?-2012-? P. pectinatus,
M. spicatum
90%
2016
M. spicatum,
Najas marina ssp. intermedia
x
Kabus unpubl. data
11
Großer
Dambecker
See
94
2.1/0.8
Lowered nutrient input from
agricultural catchment since
1990
?- 2007
few P. pectinatus
2010-ongoing
P. pectinatus,
C. globularis
70%
Not yet reached
x
Kabus unpubl. data
12
Langer See
130
3.8/2.2
Since 1990: catchment
restoration, improved
wastewater treatment
?-1997-2001-? ?-2011-2015-ongoing
M. spicatum, C. demersum,
N. marina
not yet reached
x
Rücker et al., 2015,
unpubl. data
13
Nonnensee
76
2.2/? Re-flooded area (mid
1990’s)
?- 2012
very few
P. pusillus
2012, 2016
P. pectinatus,
M. spicatum, Ceratophyllum spp. not yet reached
Kabus unpubl. data
14
De Wittsee
24
2.1/1.4
Improvement of wastewater
treatment
1970-2009
Nuphar lutea
2009-ongoing
E. nuttallii, Lemna
not yet reached
x
Van de Weyer, unpubl. data
15
Steinhuder
Meer
3000
2.9/1.35
Improvement of wastewater
treatment
1960-1998 none
1999-2001:
2002-03: E. nuttallii
2003-08: P. perfoliatus, P. February 2018 | Volume 9 | Article 194 Frontiers in Plant Science | www.frontiersin.org February 2018 | Volume 9 | Article 194 Frontiers in Plant Science | www.frontiersin.org Literature and Data Search on Macrophyte
Species Recovery crispus
2009-?: shift back to turbid
not yet reached
Hussner et al., 2014
16
Felbrigg Lake
UK
2.7
1.3/0.9
Creation of pre-lake wetland
in 2012 resulting in N-limited
conditions. Cormorant
predation on rudd
1960-2013
P. pectinatus, P. pusillus,
P. crispus, Z. palustris
2014-ongoing
C. demersum, Chara spp. P. pectinatus, P. crispus, P. pusillus
unclear whether
already reached
Sayer et al., 2010a,b;
Sayer et al. unpubl. data
17
Barton Broad
75
2/1
Progressive increase in
number of effluents with P
removal
1974-1990
none
1990-2000
C. demersum, P. crispus
<1%
reached after
sediment removal
1996 (Table 2, no. 41)
x
Phillips et al., 2005,
2015 February 2018 | Volume 9 | Article 194 5 Hilt et al. Hilt et al. Response of Macrophytes to Restoration No. Lake
Country
Size
(ha)
Depth
(max/mean)
(m)
Measures of nutrient load
reduction
Period of turbid
conditions, major
remaining macrophyte
species and coverage (%
lake area)
Period of intermediate recovery
conditions, major macrophyte
species and coverage (% lake
area)
Period of stable
clear conditions
and macrophyte
species
Data
Figure 2A
References
18
Wolderwijd
NL
2650
2.5/1.5
1982-89: Flushing with
nutrient-poor water
1969-1975
P. pectinatus, P. perfoliatus
(Figure 5)
1976-1995
P. pectinatus,
P. perfoliatus, P. pusillus (Figure 5)
reached after
biomanipulation
carried out since
1990 (Table 2, no. 25, Figure 5)
x
Scheffer et al., 1992;
Noordhuis et al., 2016
19
Veluwe-meer
3400
5/1.55
1982-89: Flushing with
nutrient-poor water
1975-76
P. pectinatus
0-5%
(Figure 4)
1977-1995
P. pectinatus, P. perfoliatus,
P. pusillus, Characeae
10-15% (Figure 4)
1996-now
P. pectinatus,
C. aspera,
C. contraria,
N. obtusa
30-85% (Figure 4)
x
Scheffer et al., 1992;
Van den Berg et al.,
1998, 1999; Noordhuis
et al., 2016
20
Eemmeer
1520
?/2.1
Since 1995: improved
sewage treatment and
closure of treatment plant
1970-1999 P. pectinatus
1–5% (Figure 4)
2000-ongoing P. crispus,
P. pectinatus, P. pusillus 10-40%
(Figure 4)
not yet reached,
but first Characeae
visible since 2010
(Figure 4)
x
Noordhuis et al., 2016
21
Arresø
DK
3987
5.6/3.1
Improved sewage
treatment, artificial lakes on
the main inlet stream,
reduced
catchment.fertilization
1989-1996
none
?- 2011: C. globularis, C. vulgaris,
M. spicatum, P. perfoliatus,
P. berchtoldii, P. cripus, P. pectinatus
unclear whether
already reached
x
Jeppesen et al.,
2007a,b Søndergaard,
unpubl. data February 2018 | Volume 9 | Article 194 6 Hilt et al. Hilt et al. Response of Macrophytes to Restoration Hilt et al. Literature and Data Search on Macrophyte
Species Recovery Response of Macrophytes to Re
TABLE 2 | The response of north temperate shallow lakes in The Netherlands (NL), Sweden (SE), Denmark (DK), United Kingdom (UK) ,and Germany (DE) to biomanipulation, natural fish kills or other lake-internal
measures (x: data on Secchi disk transparency and total phosphorus concentrations were available and used in Figure 2B). No. Lake
Country
Size
(ha)
Depth
(max/mean)
(m)
Measures applied
Period of turbid
conditions and major
remaining macrophyte
species
Periods of clear conditions,
macrophyte species and
coverage (% lake area)
Period of return
to turbid
conditions
Data
Figure 2B
References
22
Duiniger-meer
NL
30
?/1
1992,1993,1994: Fish
removal
?-1992
1992-ongoing
Chara globularis, C. vulgaris
Nitellopsis obtusa (2000-04)
Varying
macrophyte cover,
but no shift back
to turbid
x
Van Berkum et al.,
1995; Meijer et al.,
1999; Riegman, 2007;
Verhofstad et al., 2017;
Van Donk unpubl. data
23
Ijzeren Man
11
2.3/2.3
1989: Complete removal of
fish biomass by pumping
dry, restocking with pike
fingerlings, roach, rudd, ide
and tench, sediments
removed
1960s−1989
Within 2 months of fish removal,
macrophytes covered 50% of lake,
Characeae
1995-?,
varying
macrophyte cover
x
Meijer et al., 1999; Van
Donk unpubl. data
24
Noorddiep
4.5
?/1.5
1988: Biomanipulation
?-1989
1989-? E. canadensis,
C. demersum
clear for at least 8
years despite TP
250 µg L−1
x
Meijer et al., 1999; Van
Donk unpubl. data
25
Wolderwijd
2650
5/1.5
1990: Biomanipulation
See Table 1 (lake no. 18)
1992-
C. contraria, C. vulgaris
Figure 5
none
x
Meijer and Hosper,
1997; Noordhuis et al.,
2016
26
Zwemlust
1.5
2.5/1.5
1987: Lake drained empty,
fish completely removed,
restocked with pike and
rudd
1999: Temporary lowering
of water level, fish removal
?-1987
1988–1996
1988–1989: E. nuttallii,
1990–1991: C. demersum
1992–1994: P. berchtoldii
1995–1996: E. Nuttallii
1997–1999
x
Van de Bund and Van
Donk, 2002;
Verhofstad et al., 2017
27
Terra Nova
85
?/1.4
2003: (Removal of
planktivorous and
benthivorous fish)
1987–2003
1994: sparse stands of
C. demersum, P. lucens,
P. obtusifolius, P. pectinatus,
M. spicatum
2004-? C. demersum, E. nuttallii,
P. obtusifolius, Nitella mucronata,
N. marina, Utricularia vulgaris
Van de Haterd and Ter
Heerdt, 2007
28
Galgje
3.1
?/1.1
1987: Removal of all
planktivorous and 85% of
benthivorous fish in 1987
?-1988
1988: Within 2 months of fish removal
macrophyte covered lake, Characeae
2014: C. demersum, L. minor,
P. crispus, S. February 2018 | Volume 9 | Article 194 Frontiers in Plant Science | www.frontiersin.org Literature and Data Search on Macrophyte
Species Recovery polyrhiza
Meijer et al., 1999;
Verhofstad et al., 2017
29
Loender-
veense
Plas
270
?/2.7
2004/05: Removal of 95%
of fish stock
1980s−2004
P. perfoliatus
2005-present
E. nuttallii, N. marina, C. globularis,
C. connivens
none
Pot and Ter Heerdt,
2014; Verhofstad et al.,
2017
30
Naarder-meer
1,042
?/1.0
1993–1996: Sediment
dredging
1980–1989
Z. palustris, P. pectinatus
1990–1995-? Najas marina, M. spicatum,
C. globularis, R. circinatus,
C. demersum, N. obtusa
Bootsma et al., 1999 K), United Kingdom (UK) ,and Germany (DE) to biomanipulation, natural fish kills or other lake-internal
n Figure 2B). February 2018 | Volume 9 | Article 194 7 Hilt et al. Response of Macrophytes to Restoration TABLE 2 | Continued
No. Lake
Country
Size
(ha)
Depth
(max/mean)
(m)
Measures applied
Period of turbid
conditions and major
remaining macrophyte
species
Periods of clear conditions,
macrophyte species and
coverage (% lake area)
Period of return
to turbid
conditions
Data
Figure 2B
References
31
Finjasjön
SE
1,100
12/2.7
since 1970: nutrient load
reduction, 1987:
suction-dredging of
sediments, 1992-2014. removal of cyprinids
(Abramis brama, Rutilus
rutilus)
?-1994
M. spicatum, E. canadensis,
P. perfoliatus
1995-? E. canadensis (disappeared 1997),
M. spicatum
P. perfoliatus
? Annadotter et al., 1999;
Strand and Weisner,
2001; Lage et al., 2015
32
Ringsjön
(Western Bay)
1,480
5.4/3.1
1992: Removal of about
50% of cyprinid fish
?-1992
P. crispus, P. perfoliatus,
P. pectinatus,
M. spicatum,
P. lucens, E. canadensis
2000–2005
Strand, 1999; Hansson
unpubl. data
33
Vasatorp-
dammen
2.1
1.4/1.1
1992: Fish removal by
rotenone
1989–1992 none
1993–1996
P. natans, P. obtusifolius,
C. demersum, C. globularis
? Blindow et al., 2000
34
Væng
DK
15
1.9/1.2
1986-88 and 2007-09: Fish
removal
?-1986
1989–1996 and
2010-now
E. canadensis
1997–2009
x
Jeppesen et al., 1991;
Søndergaard et al.,
2017
35
Arreskov
317
3.7/1.9
1991: Fish removal
?-1991
1992–1998
Characeae,
P. pectinatus, P. pusillus, P. crispus,
Z. palustris,C. demersum,
Z. pedunculata
x
Lauridsen et al., 2003b;
Søndergaard et al.,
unpubl. data
36
Alderfen
UK
5.2
1.2/1
1979: Isolation from inflow,
1990: natural fish kill,
1992–1993: sediment
removal, 1995, 2000: fish
removal
Several turbid periods
C. demersum
20–65%
several phases of
macrophyte
decline (e.g.,
1994, 1999–2000)
x
Moss et al., 1986,
1990; Perrow et al.,
1997; Hoare et al.,
2008; Phillips et al.,
2015
37
Cockshoot
Broad
5.5
1.2/1
1992: Isolation from river,
sediment removal;
1989/90, 1996-2002,
2004-08:
fish removal
1970–1980
none
1990-2012
C. demersum, N. marina,
Z. Frontiers in Plant Science | www.frontiersin.org Literature and Data Search on Macrophyte
Species Recovery palustris
40-58%
none
x
Moss et al., 1986;
Hoare et al., 2008;
Phillips et al., 2015
38
Hoveton Little
Broad Pound
End
15.5
1.5/1.0
1990: Suction dredging,
1990-1999: several fish
removals from isolated bay
(Pound End)
1970s−1991
1995-2006
C. demersum, N. marina
Low macrophyte
abundance since
2001
x
Hoare et al., 2008;
Phillips et al., 2015
39
Ormesby
Great Broad
40
1.5/0.9
1995: Fish removal
1970–1989
Chara, Z. palustris,
P. pectinatus, C. demersum,
M. spicatum, P. pusillus,
P. crispus
1995–2010
C demersum, E. canadensis,
Z. palustris, P. pectinatus, Chara,
P. friesii
67%
none
x
Phillips et al., 2015 February 2018 | Volume 9 | Article 194 8 Hilt et al. Hilt et al. Response of Macrophytes to Restoration No. Lake
Country
Size
(ha)
Depth
(max/mean)
(m)
Measures applied
Period of turbid
conditions and major
remaining macrophyte
species
Periods of clear conditions,
macrophyte species and
coverage (% lake area)
Period of return
to turbid
conditions
Data
Figure 2B
References
40
Cromes
4.3
1.2/1
1988: Sediment removal,
1992: Barrier to isolate from
river,1999: natural fish kill,
2004: Sediment removal
? 1995-ongoing
C. demersum
80%
x
Perrow et al., 1997;
Phillips et al., 2015
41
Barton Broad
75
2/1
1980: P reduction from
effluents upstream 1996:
sediment removal
see Table 1 (lake no. 17)
2000-2012
C. demersum, P. pectinatus,
E. canadensis
12%
x
Phillips et al., 2005,
2015
42
Schollener
See
DE
95
1/<1
2002: Natural fish kill during
summer flood
∼1980 to 2003 none
2004
C. demersum, N. marina,
P. berchtoldii, P. crispus
2005-ongoing? x
Knösche, 2008
43
Rangsdorfer
See
272
2.5/1.5
2009/10: Natural winter fish
kill
?-2009
(probably several decades)
2010–2011
C. demersum, M. spicatum,
P. crispus
2012- ongoing
Hussner et al., 2014
44
Schwandter
See
16.5
2.5/1.6
2002: P-precipitation with
aluminum sulfate
2009/10: Natural winter fish
kill
?-1995-2002
sparse stands of
C. demersum, P. crispus
2003-2010
C. demersum, P. crsipus
100%
2011–2015-?:
shift to turbid state
after carp stocking
x
Mathes, 2007; Nixdorf
et al., 2013; Hussner
et al., 2014
5
Ivenacker See
73.3
1.9/1.1
2009/10: Winter fish kill
2012/13: sediment dredging
?-2009
2010-? Characeae
? x
Nixdorf et al., 2013
46
Schloßsee
Buggen-
hagen
9.8
2.9/0.8
1990-96:
Improved wastewater
treatment,
1997: Sediment dredging
?-1997
1998-?, 2006-? C. demersum, C. submersum. C. hispida
2003
x
Mathes, 2007
47
Möllener See
18.7
2.2/2
2006: P-precipitation with
aluminum sulfate
? 2007–2009-? Characeae (occurred 3 years after
treatment)
? February 2018 | Volume 9 | Article 194 Frontiers in Plant Science | www.frontiersin.org Literature and Data Search on Macrophyte
Species Recovery Hussner et al., 2014
48
Bachtel-
weiher
4.8
2/1.6
2002: Lake drained empty,
fish completely removed,
restocked with pike
?-2002
2003–2006-? M. spicatum, R. trichophyllus,
C. globularis, N. mucronata
50-90%
? Hussner et al., 2014
49
Herren-wieser
Weiher
6.7
4.7/1.8
2001: Lake drained empty,
partial sediment removal,
fish completely removed,
restocked with pikeperch
?-2001
2002
Mainly R. circinatus, P. berchtoldii,
P. crispus, P. lucens, P. pectinatus,
C. contraria, C. globularis
75%
2003–2005-? (after illegal carp
stocking)
Hussner et al., 2014
ers in Plant Science | www.frontiersin.org
9
February 2018 | Vo February 2018 | Volume 9 | Article 194 9 Hilt et al. Hilt et al. Response of Macrophytes to Restoration nutrient loading for either intermediate recovery or clear state
required adaptation of the established ecosystem model PCLake. This model has previously been used to estimate threshold
responses of shallow lakes to nutrient loading (Janse et al.,
2008; Janssen et al., 2017), and to simulate the response of
temperate shallow lakes to climate warming (Mooij et al., 2007),
to mowing of macrophytes (Kuiper et al., 2017) and—in a variant
of the model with three plant species—to biomanipulation and
herbivory (Janse et al., 1998). PCLake consists of a number
of coupled ordinary differential equations that describe the
most important biotic (submerged macrophytes, phytoplankton,
detritivorous macrozoobenthos, zooplankton, zooplanktivorous
fish, benthivorous fish, and piscivorous fish) and abiotic (detritus,
inorganic material, dissolved phosphorus, ammonium, and
nitrate) components of both the water column and the top-
layer of the sediment in a non-stratifying shallow lake (Janse,
1997, 2005). All organic components (apart from predatory
fish) are modeled in terms of dry weight (DW), nitrogen
(N), and phosphorus (P), and hence the nutrient-to-dry-weight
ratios of the organic components are variable. Internal fluxes
of nutrients between the sediment layer and the pelagic zone,
including internal loading, are accounted for and modeled
dynamically. FIGURE 2 | Total phosphorus (TP) concentrations and Secchi depth in spring
(April–June) and summer (July–September) of different north temperate
shallow lakes (A) before and after external nutrient load reductions during the
turbid, the intermediate recovery and the clear-water state (for details see
Table 1) and (B) before (turbid) and after (clear) biomanipulation or other
lake-internal measures (for details see Table 2). y
y
For our simulations we used the default settings of a lake in
PCLake. Literature and Data Search on Macrophyte
Species Recovery This default lake represents a relatively shallow lake with
an average depth of 2 m and is relatively small with a maximum
fetch of 1,000 m, an areal hydraulic loading of 20 mm days−1
(= 7.2 m year−1), no infiltration or seepage, no surrounding
wetland zone, and a lightly clayish sediment (30% dry matter, of
which 10% organic matter, and 10% fine mineral material) (Janse
2005). Due to small size and shallowness, the lake is dominated
by macrophytes when nutrient loads are sufficiently low, but as
nutrient loads increase the lake switches to a turbid state. This
switch occurs rather suddenly due to the positive feedbacks in the
model (Janse 2005) that lead to a critical transition (e.g., Scheffer
and Carpenter, 2003). A common method to determine critical
transitions is bifurcation analysis. In this approach the model is
run to equilibrium several times, each with a different nutrient
load. For each run, the yearly average phytoplankton chlorophyll-
a concentration and macrophyte biomass are calculated. To
assess the presence of hysteresis this procedure is repeated twice
for each level of nutrient load, the first starting from a clear
lake and the second from a turbid lake. Where the equilibrium
outcomes of these two runs with identical nutrient load differ,
hysteresis is inferred. Here we ran the model for nutrient loads
ranging from 0.1 to 2.5 mg m−2 days−1 (0.4–9.0 kg ha−1 year−1)
to cover a wide range of the eutrophication axis. The output of
the bifurcation analysis is a load-response curve (or bifurcation
plot) showing the effect of nutrient load on the biomass of
primary producers. The point of a sudden switch marks the
critical transition(s). FIGURE 2 | Total phosphorus (TP) concentrations and Secchi depth in spring
(April–June) and summer (July–September) of different north temperate
shallow lakes (A) before and after external nutrient load reductions during the
turbid, the intermediate recovery and the clear-water state (for details see
Table 1) and (B) before (turbid) and after (clear) biomanipulation or other
lake-internal measures (for details see Table 2). restored through biomanipulation of the fish community
(Noorddiep, Wolderwijd, Zwemlust, for details see no. 24–26 in
Table 2). Mann-Whitney U-tests were performed to compare lake
size, maximum and mean depths between lakes with different
restoration measures (external nutrient load reduction vs. internal measures). Literature and Data Search on Macrophyte
Species Recovery Total phosphorus concentrations and Secchi
disk transparency were compared between the different states
(turbid, intermediate, clear) in lakes with external nutrient load
reduction) using Kruskal-Wallis tests and subsequent posthoc
comparisons (separately for different seasons). The same was
done for lakes with internal measures using Mann-Whitney U
tests. All statistical tests were run in SPSS. To simulate the influence of temperature and light on
the response of different macrophyte species to nutrient load
reduction we had to make two adjustments to the original
formulations of PCLake, while maintaining the modeling of
macrophytes as one functional group. First, the original power
function for temperature limitation of macrophytes was replaced February 2018 | Volume 9 | Article 194 Lake Water Quality following External and
Internal Restoration Internal Restoration
Our literature review provided information on water quality
and macrophyte development in 21 turbid lakes that were
subject
to
external
nutrient
loading
reduction
without
additional in-lake measures (Table 1) and 28 lakes with in-
lake restorative measures. Some of these measures were preceded
or accompanied by external nutrient loading reduction (Table 2). Lakes with internal measures were on average smaller
than lakes with external nutrient load reduction alone, while
maximum depths were higher but mean depths were similar
(Table 4). Turbid conditions lasted from 1 year (Lake Veluwe)
to 51 years (Dümmer). Often, however, exact timing and
duration of the turbid period are unknown (Tables 1, 2). During
the turbid phase, spring and summer TP concentrations were
high (∼
>0.15 mg L−1), while Secchi disk transparencies
were low (∼0.4 m), with considerable differences between lakes
(Figures 2A,B, Table 4). Our literature review provided information on water quality
and macrophyte development in 21 turbid lakes that were
subject
to
external
nutrient
loading
reduction
without
additional in-lake measures (Table 1) and 28 lakes with in-
lake restorative measures. Some of these measures were preceded
or accompanied by external nutrient loading reduction (Table 2). PAR = PAR0 ∗D −
LN
PAR0
PARbot
D
∗
1 −εperip hyton
Lakes with internal measures were on average smaller
than lakes with external nutrient load reduction alone, while
maximum depths were higher but mean depths were similar
(Table 4). Turbid conditions lasted from 1 year (Lake Veluwe)
to 51 years (Dümmer). Often, however, exact timing and
duration of the turbid period are unknown (Tables 1, 2). During
the turbid phase, spring and summer TP concentrations were
high (∼
>0.15 mg L−1), while Secchi disk transparencies
were low (∼0.4 m), with considerable differences between lakes
(Figures 2A,B, Table 4). In
which
PAR
(W
m−2)
is
an
approximation
of
the
average Photosynthetic Available Radiation (PAR) for plant
photosynthesis, PAR0 (W m−2) is the PAR available at the top of
the macrophyte layer, PARbot (W m−2) is the PAR available at the
bottom of the macrophyte layer, D (m) is the depth and εperiphyton
is the shading by periphyton. In order to restrict complexity,
we refrained from adding periphyton as an extra compartment
to the model, but instead used the empirical relationship of
Vadeboncoeur et al. (2006) to estimate periphyton chlorophyll-a
biomass: TABLE 3 | Parameter settings for PCLake. PCLake Simulations With this function the model is flexible to simulate
macrophytes with different temperature optima. The second
adjustment is the timing of root allocation which occurs in
the autumn when macrophytes store energy to overwinter, for
instance, in propagules. In the original PCLake, this timing was
linked to a specific day in the year while submerged plants
are known to respond to physiological and environmental cues,
such as light availability, to determine timing of root allocation
(Madsen, 1991). Hence, we decided to link the timing of root
allocation to a minimum daily light availability for macrophytes,
following Madsen (1991), Van Dijk and Van Vierssen (1991), and
Van Dijk and Janse (1993). The available light for macrophytes is
based on the light availability over the depth of the water column
corrected for periphyton shading which is estimated by: PCLake Simulations Simulating the response of water clarity and macrophyte biomass
to external nutrient load reduction and detecting thresholds of February 2018 | Volume 9 | Article 194 Frontiers in Plant Science | www.frontiersin.org 10 Response of Macrophytes to Restoration Hilt et al. by a temperature optimum curve LT (-): integrate this effect indirectly (Kuiper, 2016). By adding the effect
of periphyton to PCLake implicitly, we thus have to technically
recalibrate the model. This implies adjusting the parameter
settings of the model, such that given the same boundary
conditions, the model produces the same output. Therefore we
have calibrated the adjusted model manually by lowering the
half saturation light constant of vegetation and decreasing the
parameter for dark respiration of vegetation (see Table 3 for new
and calibrated parameter settings). PCLake is implemented in
DATM (Mooij et al., 2014). For the full overview of parameter
settings and model formulations, please see the DATM-file in the
Supplementary Material. LT = e
−0.5
σ2
h
(T−Topt)2 −(Tref −Topt)
i Here, σ (◦C) is the temperature constant based on a Gaussian
curve, T (◦C) is the water temperature, Topt (◦C) is the optimum
temperature for macrophytes and Tref (◦C) is the reference
temperature used to normalize the limitation function to 1
(Janse, 2005). With this function the model is flexible to simulate
macrophytes with different temperature optima. The second
adjustment is the timing of root allocation which occurs in
the autumn when macrophytes store energy to overwinter, for
instance, in propagules. In the original PCLake, this timing was
linked to a specific day in the year while submerged plants
are known to respond to physiological and environmental cues,
such as light availability, to determine timing of root allocation
(Madsen, 1991). Hence, we decided to link the timing of root
allocation to a minimum daily light availability for macrophytes,
following Madsen (1991), Van Dijk and Van Vierssen (1991), and
Van Dijk and Janse (1993). The available light for macrophytes is
based on the light availability over the depth of the water column
corrected for periphyton shading which is estimated by: Here, σ (◦C) is the temperature constant based on a Gaussian
curve, T (◦C) is the water temperature, Topt (◦C) is the optimum
temperature for macrophytes and Tref (◦C) is the reference
temperature used to normalize the limitation function to 1
(Janse, 2005). Frontiers in Plant Science | www.frontiersin.org Lake Water Quality following External and
Internal Restoration Parameter Description
Unit
Valuea
Source
σ
Temperature constant
based on a Gaussian curve
◦C
20
-
Topt
Optimum temperature for
macrophytes
◦C
20
-
Tref
Reference temperature
◦C
20
–
c1
Slope of logistic curve
periphyton
–
1.79
Vadeboncoeur
et al., 2006
c2
Intercept logistic curve
periphyton
–
−0.85
Vadeboncoeur
et al., 2006
S
Light attenuation by
periphyton
m g−1
0.03
Van Dijk, 1993
Lmin
Minimal light availability cue
needed for plants to initiate
increased root allocation
W m−2
91.2
Calibrated
hLveg
Half saturation light constant
of vegetation at 20◦C
W m−2
12 (17)
Calibrated
kDresp
Dark respiration rate of
vegetation
day−1
0.015 (0.02) Calibrated
aValues between brackets are the original value. TABLE 3 | Parameter settings for PCLake. LOG10
Chlaperip hyton
= c1 ∗LOG10 (TP) + c2 where c1 = 1.79 and c2 = 0.85 and TP is the in-lake
total phosphorus concentration (mg m−3). The periphyton
chlorophyll-a biomass is then used to estimate the shading effect: εperi phyton = Chlaperip hyton ∗S ∗f (L) ∗f (T) ∗f (α) where S is the specific light attenuation by periphyton (m g−1)
and correction factors for light limitation f(L), temperature
limitation f(T) and available plant surface area for periphyton
growth f(α). The shading effect of periphyton affects the timing
of root allocation through light availability to macrophytes as well
as the light limitation of macrophyte shoots. PCLake has previously been calibrated following a Bayesian
approach to parameter estimation and uncertainty analysis using
data from nearly 40 temperate shallow lakes (Janse et al., 2010). Although this calibration did not account for the specific effect
of periphyton, the data used for model calibration most likely Frontiers in Plant Science | www.frontiersin.org February 2018 | Volume 9 | Article 194 11 Hilt et al. Response of Macrophytes to Restoration External nutrient loads were usually reduced following
improved sewage treatment in the catchment. Lake Veluwe and
Wolderwijd were flushed with nutrient-poor water (Table 1). The
most commonly applied internal measure was biomanipulation
in the form of removal of benthivorous and planktivorous fish
(in some cases by draining/pumping the lake dry). In nine
lakes, sediment removal was applied as an additional or the sole
measure and in five lakes natural fish kills occurred during severe
winters or as a result of a summer flood (Table 2), emulating
the effects of planned biomanipulation. Lake Water Quality following External and
Internal Restoration During the first years
after nutrient load reduction, TP concentrations were lower than
during the turbid period, but still about twice as high in summer
as in spring. Spring water transparency was higher than during
the turbid period, while summer values were still low (Table 4,
Figure 2A). This phase has been found to last up to 20 years
(e.g., Müggelsee), but often its start has not been recorded and/or
lakes have not yet reached stable clear conditions (Table 1). Lakes Müggelsee, Veluwemeer, and Eemmeer show a similar
intermediate recovery state with spring water transparencies
being higher than during turbid summer conditions which lasted
for about 20 years (Figure 3). A switch back from intermediate
to turbid conditions has only been observed in Lake Steinhuder
Meer, in this case about 10 years after macrophytes returned
(Table 1). TABLE 4 | Size, maximum and mean depths, total phosphorus concentrations and Secchi disk transparency in lakes after external nutrient load reduction or implementation of internal measures. Lakes
State
Size (ha)
Maximum depth (m)
Mean depth (m)
Total phosphorus concentrations (µg L−1)
Secchi disk transparency (m)
spring
summer
spring
summer
mean
median
mean
median
mean
median
mean
median
mean
median
mean
median
mean
median
External nutrient
load reduction
Turbid
984
450
5.0
3.9
2.4
1.8
281
163 A
317
244 A
0.66
0.45 A
0.46
0.4 A
Intermediate
97
94 A
206
206 A
1.23
1.11 AB
0.75
0.6 AB
Clear
53
51 B
59
49 B
1.74
2.17 B
1.17
1.12 B
P (Kruskal-Wallis test)
0.02
0.007
0.042
0.02
Internal measures
Turbid
273
18
2.8
2.1
1.5
1.3
229
156 a
256
150 a
0.62
0.45 a
0.45
0.40 a
Clear
168
81 a
170
103 b
1.12
0.96 b
1.10
0.99 b
P (Mann-Whitney U test)
<0.001
0.016
0.076
0.051
0.011
0.026
<0.001
Size and depths include all lakes, while TP and transparency data were only available for selected lakes (see Figure 2, Tables 1, 2). Different letters indicate significant differences between values of different states (external nutrient load
reduction: Kruskal Wallis test capital letters; internal measures: Mann Whitney U test small letters)
rontiers in Plant Science | www.frontiersin.org After implementation of in-lake measures, TP concentrations
were at the same level as before, in both spring and
summer, whereas water transparency in spring and summer was
significantly higher than before restoration (Table 4, Figure 2B). Lake Water Quality following External and
Internal Restoration These unstable clear conditions often only lasted for a few
years and many lakes shifted back to turbid conditions (e.g.,
no. 23, 26, 32, 34, 36, 38, 42, 43, 44, 46, 49 in Table 2). Fish
stock reductions in Wolderwijd, Zwemlust and Noorddiep led
to clear conditions both in spring and summer (Figure 4). Lake
Zwemlust shifted back to turbid conditions after 9 years, while
Noorddiep stayed clear for at least 8 years with no further
information on subsequent periods. Only six of the lakes with external nutrient load reduction
(∼25%, no. 1, 3, 7, 10, 18, and 19 in Table 1) reached stable
clear-water conditions with lower ambient TP concentrations
and higher Secchi depths than during the intermediate recovery
state both in spring and summer (Figure 2A, Table 4). In
four of the 28 lakes (14%, no. 25, 29, 37, 39 in Table 2,
Lake Wolderwijd in Figure 4), stable, longer-term clear-water
conditions with a diverse macrophyte flora were obtained after
the application of in-lake measures. For several lakes, their
longer-term development is not known (Table 2). Model Simulations on Lake Response to
External Nutrient Load Reduction The results of the simulations using the adjusted PCLake model
revealed three stages for lakes that undergo external nutrient
load reduction: a turbid state, an intermediate recovery state and
a clear state (Figure 5A). For the default lake in PCLake, the
turbid state occurs if the P load exceeds 1.3 mg P m−2 days−1. The intermediate recovery state occurs between the two critical
transitions that appear at a P load of 1.06 and 1.3 mg P m−2 February 2018 | Volume 9 | Article 194 12 Hilt et al. Hilt et al. Response of Macrophytes to Restoration FIGURE 3 | Total phosphorus (TP) concentrations and Secchi depth in spring (April–June) and summer (July–September) and macrophyte coverage in Lake
Müggelsee; Lake Veluwe and Lake Eem during the turbid (green), the intermediate recovery (brown) and the clear-water (blue) state (for lake details see Table 1). FIGURE 3 | Total phosphorus (TP) concentrations and Secchi depth in spring (April–June) and summer (July–September) and macrophyte coverage in Lake
Müggelsee; Lake Veluwe and Lake Eem during the turbid (green), the intermediate recovery (brown) and the clear-water (blue) state (for lake details see Table 1). days−1. If the P load drops below 1.06 mg P m−2 days−1 the lake
turns into a clear state. The critical P loads for the intermediate
recovery state are smaller in the adjusted model including
periphyton than with the original formulation of PCLake (critical
transition between approximately 1 and 2 mg P m−2 days−1,
Janse et al., 2008). The periphyton effect thus reduces the
threshold for intermediate recovery state of the default lake by
about a quarter of what it would be without periphyton. The
reason for this is that in case of hysteresis, the position of the
highest critical nutrient load is mainly determined by macrophyte
characteristics, while the position of the lowest critical nutrient
load is mainly determined by phytoplankton characteristics. This
is in agreement with the results of the sensitivity analysis of
PCLake (Janse et al., 2010). Since periphyton negatively affects
the performance of macrophytes, the highest critical nutrient
load was reduced, while phytoplankton characteristics were
unaltered and the lowest critical nutrient load (1 mg P m−2
days−1) thus did not change. nutrient loading, the phytoplankton summer peak shifted to
earlier dates of the year and this advancement of phytoplankton
was mirrored by an abbreviated macrophyte growing season
(Figure 5A). Frontiers in Plant Science | www.frontiersin.org Model Simulations on Lake Response to
External Nutrient Load Reduction The shorter growing season was a direct effect
of the inclusion of periphyton shading in our adapted version
of the model. The shading of macrophytes by periphyton was
most severe at the peak of summer when the light input and
water temperature are high. As a result, macrophyte growth was
limited to the period just after the clear water phase in spring
until the start of the summer phytoplankton bloom. The clear
state was characterized by an increase of macrophyte biomass
with increasing nutrient loading (Figure 5A). Phytoplankton
production was restricted to the spring bloom peak and there was
no summer bloom. The bifurcation plot (Figure 5B) shows a less sudden
transition of macrophytes and phytoplankton compared to the
abrupt transition between the phytoplankton-dominated turbid
state and the macrophyte-dominated clear state, often seen in
bifurcation plots in literature (e.g., Janse et al., 2008). The
gradual course of the bifurcation plot is due to the inclusion
of periphyton shading of macrophytes. This shading permits
high biomass of both macrophytes and phytoplankton within The turbid state was characterized by lack of macrophytes
and enhanced phytoplankton biomass with increasing nutrient
loading (Figure 5A). The intermediate state was characterized by
a changed phenology of the primary producers. With increasing February 2018 | Volume 9 | Article 194 Frontiers in Plant Science | www.frontiersin.org 13 Hilt et al. Response of Macrophytes to Restoration FIGURE 4 | Total phosphorus (TP) concentrations and Secchi depth in spring (April–June) and summer (July–September) and macrophyte coverage in Lake
Wolderwijd, Lake Zwemlust and Lake Noorddiep before (green and brown) and after (blue) biomanipulation (for lake details see Table 2). In Lake Zwemlust,
P. berchtoldii occurred instead of P. pectinatus, and the coverage in Lake Noorddiep was only estimated based on the information that it was higher than 25% (Gulati
and Van Donk, 2002). FIGURE 4 | Total phosphorus (TP) concentrations and Secchi depth in spring (April–June) and summer (July–September) and macrophyte coverage in Lake
Wolderwijd, Lake Zwemlust and Lake Noorddiep before (green and brown) and after (blue) biomanipulation (for lake details see Table 2). In Lake Zwemlust,
P. berchtoldii occurred instead of P. pectinatus, and the coverage in Lake Noorddiep was only estimated based on the information that it was higher than 25% (Gul
and Van Donk, 2002). with reduced external nutrient loading. Other pondweed species
such as P. perfoliatus, P. crispus and P. Model Simulations on Lake Response to
External Nutrient Load Reduction pusillus or Zannichellia
palustris were also found in several lakes during the intermediate
recovery state, while other groups such as Characeae or Elodea
species were much less common (Table 1, Figure 6). Lakes
Müggelsee, Veluwemeer and Eemmeer were all dominated by
P. pectinatus during the intermediate recovery state which lasted
for about 20 years (Figure 3). Müggelsee and Veluwemeer seem
to have entered a stable clear state with more diverse submerged
vegetation (including Characeae in Veluwemeer) in 2011 and
1996, respectively, while Eemmeer has not yet reached that
phase despite the recent detection of Characeae (Figure 3). Three
other lakes also reached a stable clear state and were colonized
by different species of Characeae and/or Najas marina, Elodea
species and C. demersum (Table 1). the same year during the crashing or intermediate recovery
phase. Furthermore, the region of hysteresis is tilted, leading
to a less abrupt and thus more realistic critical transition from
phytoplankton dominance to macrophyte dominance and back
again. Frontiers in Plant Science | www.frontiersin.org February 2018 | Volume 9 | Article 194 Macrophyte Species Recovery following
External and Internal Restoration During the turbid phase, three lakes with subsequent external
nutrient load reductions were reported to lack macrophyte
stands altogether and six and one lakes had sparse stands of
Potamogeton pectinatus (also known as Stuckenia pectinata) and
P. pusillus, respectively, while no information was available
for the remaining lakes (Table 1). For lakes with internal
measures, information on macrophyte species present during the
turbid phase is available for 10 lakes. Apart from P. pectinatus
and P. perfoliatus, plants such as Ceratophyllum demersum,
M. spicatum or E. canadensis are mentioned, if indeed plant
stands were present at all (Table 2). The dominant macrophytes occurring after lake internal
measures were Characeae, C. demersum, Elodea species or
N. marina (Table 2, Figure 6). Often, lakes had only one or two
dominant species, and in at least 10 cases, lakes switched back
to turbid conditions and lost their macrophytes again (Table 2,
Figure 4). The response of macrophytes occurred gradually in During the intermediate recovery state P. pectinatus was the
dominant macrophyte species in two thirds of the analyzed lakes February 2018 | Volume 9 | Article 194 Frontiers in Plant Science | www.frontiersin.org 14 Hilt et al. Response of Macrophytes to Restoration FIGURE 5 | (A) Time series of simulation using PCLake of chlorophyll-a concentrations and macrophyte shoot biomass for different phosphorus (P) loadings within
the clear, intermediate recovery and turbid states. Darker colors are associated with higher P loading simulations. (B) Hysteresis plots showing yearly mean simulated
values of chlorophyll-a concentrations and macrophyte shoot biomass for different phosphorus (P) loadings within the clear, intermediate recovery and turbid states. Arrows denote the directions of the hysteresis effects. FIGURE 5 | (A) Time series of simulation using PCLake of chlorophyll-a concentrations and macrophyte shoot biomass for different phosphorus (P) loadings within
the clear, intermediate recovery and turbid states. Darker colors are associated with higher P loading simulations. (B) Hysteresis plots showing yearly mean simulated
values of chlorophyll-a concentrations and macrophyte shoot biomass for different phosphorus (P) loadings within the clear, intermediate recovery and turbid states. Arrows denote the directions of the hysteresis effects. FIGURE 5 | (A) Time series of simulation using PCLake of chlorophyll-a concentrations and macrophyte shoot biomass for different phosphorus (P) loadings within
the clear, intermediate recovery and turbid states. Darker colors are associated with higher P loading simulations. Macrophyte Species Recovery following
External and Internal Restoration (B) Hysteresis plots showing yearly mean simulated
values of chlorophyll-a concentrations and macrophyte shoot biomass for different phosphorus (P) loadings within the clear, intermediate recovery and turbid states. Arrows denote the directions of the hysteresis effects. model simulations (Figure 5). As hypothesized, macrophyte re-
establishment following nutrient load reduction occurs in a
reversed sequence to the one described for eutrophication by
Sayer et al. (2010a,b). In contrast, lake internal measures such as
fish or sediment removal often result in clear-water conditions
during spring and summer. This clear-water state is, however,
often only temporary and lakes frequently shift back to turbid
conditions one or a few years after the restoration. Likely, the
duration of the clear-water conditions is related to nutrient
loading and the intensity of the restoration effort (Hansson et al.,
1998). Only in a few examples have longer lasting clear-water
conditions been observed. These required spring and summer TP
concentrations below 0.05 mg L−1. Wolderwijd, showing an increase in Characeae coverage, while
in the much smaller Zwemlust and Noorddiep, macrophytes
immediately covered large areas, E. nuttallii, E. canadensis and
C. demersum dominating (Figure 4, Table 2). In four lakes, stable
clear-water conditions were obtained after in-lake measures, all
being characterized by Characeae dominance (Table 2, Figure 4). Frontiers in Plant Science | www.frontiersin.org DISCUSSION Our analysis of long-term data from 49 temperate shallow
lakes during their recovery from a turbid phase reveals
that both a reduction of the external nutrient loading and
implementation of lake-internal measures often result in the
occurrence of an intermediate state (Figure 1) that can last
for several decades. External nutrient load reductions are often
followed by the re-occurrence of a spring clear-water phase
that opens a “window of opportunity” for macrophyte re-
colonization, but with only a short growth period due to turbid
conditions during summer. This pattern was confirmed by our We also have evidence for our second hypothesis, namely
that different types of restoration measures influence the
macrophyte community composition. P. pectinatus and a few
other pondweeds most often recolonise temperate shallow
lakes with reduced external nutrient loading and dominate
during the intermediate recovery state. The implementation of Frontiers in Plant Science | www.frontiersin.org February 2018 | Volume 9 | Article 194 15 Hilt et al. Response of Macrophytes to Restoration apical growth form may allow it to survive at greater depths
in turbid water. The shallow littoral, especially in larger lakes,
is strongly disturbed by wave action and only species with
high anchorage and breaking strength can survive under these
conditions (Schutten et al., 2005). Potamogeton pectinatus has
a high breaking strength (Brewer and Parker, 1990) and its
phenotypic plasticity allows it to form short plants in shallower
water (Idestam-Almquist and Kautsky, 1995). In contrast, other
common species in eutrophic lakes such as E. canadensis or
M. spicatum have been described as deep water species with lower
tensile strength (Brewer and Parker, 1990), while C. demersum
has no roots for anchorage. These species are thus less likely
to persist through severely turbid states in very shallow littoral
areas and to serve as remnant populations for re-colonization, at
least in larger lakes. Nevertheless, they have been reported during
turbid phases in four lakes included in our survey, most probably
in wind-protected areas or bays. FIGURE 6 | The 10 most common macrophyte taxa in north temperate
shallow lakes during re-colonization after reduction of external nutrient loading
(light gray) or implementation of internal measures (dark gray; for more details
see Tables 1, 2). y
Knowledge of the survival of propagules in sediments during
turbid phases is limited. Submerged Macrophyte Survival during
the Turbid Phase Shallow temperate lakes often show a relatively rapid response
to reductions in external phosphorus loading, characterized by
a reduction in phytoplankton biomass during spring and early
summer (Jeppesen et al., 2005). During late summer, however,
the response is delayed because of sustained remobilisation of
phosphorus from the sediment (Søndergaard et al., 2013). As a
consequence, high phytoplankton and cyanobacterial abundance
are often reasserted in summer (Sommer et al., 2012) resulting in
turbid water and preventing macrophyte growth. In our survey,
such conditions occurred at spring TP concentrations of around
0.1 mg L−1, while summer concentrations were twice as high. Whether and which macrophyte stands or propagules survive
during the turbid phase depends on the occurrence of
macrophyte species before the shift to turbid conditions and on
the length and severity of the turbid phase (e.g., Vari and Toth
2017). Seed banks in shallow lake sediments have often been
assumed to be insufficient for recovery of submerged vegetation
by germination, due to low numbers of viable seedlings, strong
seed dormancy, strict germination cues and the reliance of many
species upon vegetative reproduction (Haag, 1983; Kautsky, 1990;
Rodrigo et al., 2013; Baldridge and Lodge, 2014). In contrast,
De Winton et al. (2000) and Verhofstad et al. (2017) have
shown that seed banks from even the most degraded lakes are
capable of an emergence response and thus offer a potential
means to restore vegetation. In our survey, the duration of
complete macrophyte loss was often unrecorded, but periods of
several decades are common (Tables 1, 2). If macrophyte stands
survived during this period and were recorded, these were often
sparse stands of P. pectinatus, a species commonly associated
with the crashing phase during eutrophication (Sayer et al.,
2010a). This species survives in very shallow water even under
phytoplankton dominance (Hilt et al., 2013) and its strongly The increased water transparency in spring and early summer
seems to be exploited by certain macrophyte species, in our
survey mainly P. pectinatus along with P. perfoliatus, P. crispus
and Z. palustris. These macrophyte species are characterized by
specific traits that may explain their prevalence. Firstly, they can
compress their whole life cycle into the short clear-water period
in spring and early summer due to early germination from tubers,
turions or seeds, shortening time to peak biomass and allowing
early formation of overwintering tubers and seeds (Table 5). DISCUSSION In general, charophyte oospores and
macrophyte seeds have been found to survive up to 150 years
(Kaplan and Muer, 1990; De Winton et al., 2000; Alderton et al.,
2017). Germination tests with sediments have been suggested
before implementing lake restoration measures to forecast the
potential for macrophyte recovery from internal sources (Hilt
et al., 2006), however, these are not routinely applied. Re-
colonizing macrophyte clones of several species that originated
from periods before eutrophication have been found (Sand-
Jensen et al., 2008), but knowledge about the origin of re-
colonizing macrophytes post-restoration remains scarce (Bakker
et al., 2013). FIGURE 6 | The 10 most common macrophyte taxa in north temperate
shallow lakes during re-colonization after reduction of external nutrient loading
(light gray) or implementation of internal measures (dark gray; for more details
see Tables 1, 2). internal restoration measures results in the establishment of
a different community, often consisting of a small selection
from either hornwort (C. demersum), charophytes, water weeds
(E. canadensis, E. nuttallii) or naiad (N. marina). Only in a
few cases have lakes reached a state of clear-water conditions
during spring and summer and with a more diverse macrophyte
community. Frontiers in Plant Science | www.frontiersin.org February 2018 | Volume 9 | Article 194 Submerged Macrophyte Survival during
the Turbid Phase Secondly, they can have short growth forms that can establish
in very shallow habitats (e.g., Van Vierssen, 1982a). Thus, they
are often the species that survive in turbid conditions in shallow February 2018 | Volume 9 | Article 194 Frontiers in Plant Science | www.frontiersin.org Frontiers in Plant Science | www.frontiersin.org 16 Hilt et al. Response of Macrophytes to Restoration TABLE 5 | Characteristics of macrophyte species/groups typically recolonising temperate shallow lakes after external nutrient load reduction or following implementation of lake-internal measures such as
biomanipulation. Restoration
measure
Macrophyte
species
Re-production
mode
Germination
Colonization
mode
Timing of peak
biomass
Specific features
Epiphyton
density
Allelopathic
activity
Susceptibility to
herbivory
External nutrient
load reduction
Potamogeton
pectinatus
Mainly by tubers,
although seeds
are formed1
Between
10-15◦C7
Rhizomatic
growth11
June19
Concentration of
biomass below
water surface1
Tubers rich in
carbohydrates32
Low13
Low2
High3,14,28
P. perfoliatus
Turions29, seeds
Rhizomatic
growth11
High20
P. crispus
Mainly by
turions26
Turions can sprout
in late summer, but
shoot elongation
after winter above
10◦C26
Turions
Late spring to
early summer25
Low2
P. pusillus
Turions4
Rhizomatic
growth4,12
Low13
Low2
High14
Zannichellia
palustris
Seeds5
Mainly
temperature-
dependent, above
12-16◦C5,27
Seeds12
Short, but
numerous shoots
allow development
in very shallow
water27
Tolerant to
disturbance by
wave action27
Low13
Low2
Internal measures
Elodea canadensis
Vegetatively6
May-June22, can
be evergreen
Fragments,
vegetative growth
(peripheral
propagation)4,12
Sudden biomass
collapses6,8,9
High13
Medium2
Medium3
E. nuttallii
Vegetatively17
Can be evergreen
Fragments,
vegetative
growth12
Aug.-Oct.21
Extremely high
reproductive
capacity16,17
Sudden biomass
collapses
High13
Low15
Medium2
Medium18
Ceratophyllum
demersum
Vegetatively
(dormant apices)
April22, can be
evergreen
Fragments,
vegetative
growth12
August31
Sudden biomass
collapses10
Low15
High2
Low
Najas marina
Annual, seeds
Late in (temperate)
season (>20◦C)30
Seeds11
Mid August11
Medium2
Low18
Characeae
Oospores
Can be evergreen,
overwintering by
shoot apices
Oospores
Late summer23
Low15
Medium2
Medium3
1Van Wijk (1989), 2Hilt and Gross (2008), 3Dorenbosch and Bakker (2011), 4Barrat-Segretain and Bornette (2000), 5Bytnerowicz and Carruthers (2014), 6Simberloff and Gibbons (2004), 7Madsen and Adams (1988), 8Rørslett et al. (1985), 9Strand and Weisner (2001), 10Van de Bund and Van Donk (2002), 11Vari and Toth (2017), 12Capers (2003), 13Lalonde and Downing (1991), 14Hidding et al. (2010), 15Grutters et al. (2017), 16Simpson (1990), 17Josefsson
(2011), 18Pot and Ter Heerdt (2014), 19Körner (2001), 20Choi et al. (2002), 21Best and Dassen (1987), 22Best (1977), 23Talling and Parker (2002), 24Lillie et al. (1997), 25Woolf and Madsen (2003), 26Tobiessen and Snow (1984), 27Van
Vierssen (1982b), 28Weisner et al. Frontiers in Plant Science | www.frontiersin.org Submerged Macrophyte Survival during
the Turbid Phase Most of the dominant macrophyte species
during intermediate recovery after nutrient load reductions are
also highly susceptible to herbivory due to their low content of
polyphenols, low carbon to nitrogen ratio and low dry matter
content (Elger and Willby, 2003; Dorenbosch and Bakker,
2011, Table 5). Periphyton shading may further increase the
sensitivity of macrophytes to herbivory (Hidding et al., 2016). Finally, fine-leaved species such as P. pectinatus, P. pusillus and
Z. palustris also suffer from leaf plucking by omnivorous fish
during periods of low zooplankton abundance when those fish
switch to macroinvertebrate prey found in the periphyton of
macrophytes (Körner and Dugdale, 2003). Such leaf plucking by
fish can lead to a considerable leakage of nutrients from injured
macrophyte tissue, thereby further stimulating phytoplankton
growth (Hansson et al., 1987). Overall, while being well-suited
for survival during turbid phases and for exploiting the clear-
water conditions in spring for re-colonization, other traits of
macrophyte species typical of the intermediate recovery phase
following nutrient load reduction prevent their survival during
later summer (Table 5). Usually, the maximum colonization depth of macrophytes
during the intermediate recovery phase is low (around 1 m)
and consequently, depending on the lake morphometry, only
small parts of the lake bed might be covered. In contrast,
very shallow lakes may reach over 50% cover (Table 1). This
suggests that in “deeper” shallow lakes, macrophyte coverage
during the intermediate phase may be insufficient to stabilize
clear-water conditions during later summer. Based on the
findings of Søndergaard et al. (2016) submerged macrophyte
coverage on average needs to pass a threshold of 20% of lake
area to markedly lower phytoplankton densities. In principle,
small stands can be sufficient as a refuge for phytoplankton-
grazing zooplankton against fish predation (Lauridsen et al.,
1996; Portielje and Van der Molen, 1999). However, abundant
colonial and filamentous cyanobacteria which often dominate
the summer phytoplankton communities during lake recovery
cannot be effectively controlled by zooplankton grazers (Wang
et al., 2010 and references therein). Bottom-up stabilizing
mechanisms of macrophytes on water clarity such as nutrient
competition, increased sedimentation within stands and reduced
sediment resuspension will be inefficient at low plant coverage
(Blindow et al., 2014). Submerged Macrophyte Survival during
the Turbid Phase (1997), 29Wolfer and Straile (2004), 30Van Vierssen (1982c), 31Best (1982), 32Spencer (1986). Frontiers in Plant Science | www.frontiersin.org
17
February 2018 | Volume 9 | Article 194 February 2018 | Volume 9 | Article 194 17 Response of Macrophytes to Restoration Hilt et al. margins (see Submerged Macrophyte Survival during the Turbid
Phase) and then expand into deeper water with improvements in
clarity during nutrient load reduction. Rhizomatic growth from
remaining P. pectinatus stands has been shown, via microsatellite
analyses, to be the dominant re-colonization mode in Lake
Müggelsee; more recently established P. pectinatus stands had
lower genotype diversity and were comprised of only a small
subset of genotypes from shallower areas (Hilt et al., 2013). Thirdly, energy reserves in vegetative propagules such as tubers
of P. pectinatus allow early onset of growth independent of
light availability (Spencer, 1986). In addition, P. pectinatus can
concentrate large parts of its biomass just under the water surface
and thus survive in relatively turbid water (Van Wijk, 1988). An
initial colonization of formerly turbid lakes with P. pectinatus
during recovery has also been observed in deeper, stratifying
lakes. Thus, in Lake Tegel (Germany), this species dominated for
more than 20 years after the start of phosphorus stripping in the
major inflow (Hilt et al., 2010). and TP in the water column (Vadeboncoeur et al., 2006). In
eutrophic shallow, temperate lakes, periphyton is often top-
down controlled by a cascading effect from omnivorous fish
that feed on periphyton grazers such as snails and chironomid
larvae (Jones and Sayer, 2003). Thus, nutrient load reductions
will only reduce periphyton shading after the fish biomass built
up during the turbid period has also been reduced, which
may take 10–15 years (Jeppesen et al., 2005). Furthermore,
the observed dominant macrophyte species in the intermediate
recovery phase after nutrient load reduction (Table 1) show little
or no allelopathic activity that might hamper periphyton growth
(Table 5), thus making them more susceptible to shading by
periphyton. Cyanobacteria have been shown to potentially inhibit
submerged
macrophyte
growth
via
allelopathy
(Zheng
et al., 2013), but whether this mechanism contributes to
the disappearance of macrophytes during the recovery phase in
summer is unknown. Frontiers in Plant Science | www.frontiersin.org Submerged Macrophyte Survival during
the Turbid Phase pectinatus in Lake Wolderwijd after
biomanipulation (Figure 4), while in Swedish Lake Krankesjön
a similar development has been observed, the reasons for which
are unknown (Blindow, 1992; Hargeby et al., 1994; Hansson et al.,
2010). In Lake Müggelsee and Eemmeer, the additional influence
of a sudden invasion of the quagga mussel (Dreissena rostriformis
bugensis) in around 2013 might have contributed to a decline in
TP concentrations and increased water transparencies (Figure 4,
S. Hilt, unpublished, Noordhuis et al., 2016). This species can
colonize soft substrates and thus cover much larger areas than
those previously occupied by the zebra mussel (D. polymorpha)
(Karatayev et al., 2015). In Lake Eemmeer, quagga mussels
filtered the lake volume about five times a day in 2013 (Noordhuis
et al., 2016). macrophyte survival after nutrient load reduction (see Response
of Macrophytes to Nutrient Load Reductions). The relevance
of this process for macrophyte recovery after fish removal
has not yet been directly tested, although mesocosm trials
in which the density of periphyton grazers are manipulated
produce predictable outcomes in terms of periphyton biomass
and macrophyte composition (Elger et al., 2009). Excretion of
allelopathic substances, which has been detected for many of the
typical species that colonize after biomanipulation (Table 5), may
also contribute to lower periphyton densities. In general, macrophyte species typically occurring after
introduction of in-lake restoration measures allow for a longer
period of high macrophyte cover and dampen seasonal changes
in phytoplankton abundance as described for “stable” lakes prior
to major eutrophication (Sayer et al., 2010b). Both, Elodea and
charophyte species can remain evergreen in temperate lakes (e.g.,
Søndergaard et al., 2017), thus extending their positive influence
on water quality to seasons outside the influence of annual
species. Macrophyte recovery in Steinhuder Meer and Langer See
deviates from the suggested pattern in that both are dominated
by species more typical of lakes having undergone restoration
with internal measures (Table 1). In Steinhuder Meer, a
strong reduction of the fish population has been observed
which was attributed to cormorant activities (Niedersächsischer
Landesbetrieb für Wasserwirtschaft, Küsten- und Naturschutz,
2011). Cormorant effects on fish populations are also suggested
for Felbrigg Lake (C. Sayer, unpublished). Through a natural
increase in cormorants, the lake food web configuration
was likely affected in ways comparable to those of lakes
undergoing biomanipulation. Submerged Macrophyte Survival during
the Turbid Phase Low coverage is, however, not the
only reason why macrophytes in the intermediate recovery
phase cannot stabilize clear-water conditions in late summer,
as shown in the case of the very shallow Lake Dümmer, where
cyanobacteria blooms still occurred in summer despite high
macrophyte coverage. Stable clear-water conditions in spring and summer with more
diverse macrophyte vegetation were observed when both spring
and summer TP concentrations reached about 0.05 mg L−1. This
value corresponds well with a threshold for low cyanobacterial
abundance in shallow lakes found by Jeppesen et al. (2005) and
Triest et al. (2016) and the average critical loading for shifts
from turbid to clear conditions estimated for Dutch shallow
lakes by Janse (2005) and in eastern England by Phillips et al. (2015). Whether external nutrient load reductions alone were
responsible for the observed low in-lake TP concentrations
in the lakes in our survey that reached stable clear-water
conditions, however, remains questionable. It seems that in
most cases additional changes in either the fish community
(Lake Veluwe, Galenbecker See) and/or exotic mussel invasions
(Lake Müggelsee, Eemmeer) contributed to the observed trend. In Lake Veluwe, several severe winters, an increase in bream
(Abramis brama) fisheries between 1993 and 1997 and the
increase in zebra mussel densities are all thought to have
contributed to a break in the dominance of cyanobacteria, thus
allowing for the prevalence of stable clear-water conditions Our model simulations suggest that high periphyton shading
triggers macrophyte disappearance in summer. Periphyton
shading, often accompanied by herbivory (Hidding et al., 2016),
has been shown to impair macrophyte development in empirical
studies (e.g., Jones et al., 2002; Jones and Sayer, 2003; Roberts
et al., 2003) and is argued to be a major factor in the failure
of macrophytes to establish even decades after the start of
nutrient loading reduction, despite suitable water clarity for
plant re-establishment in spring (Phillips et al., 2005). In our
adapted PCLake model, periphyton biomass was dependent on
TP concentrations in the water, based on the positive correlation
between chlorophyll content of periphyton on hard substrata February 2018 | Volume 9 | Article 194 18 Response of Macrophytes to Restoration Hilt et al. with charophyte dominance since 1996 (Noordhuis et al., 2016,
Figure 4). Once established, these dense charophyte beds provide
more efficient stabilizing mechanisms for clear-water conditions
than rooted angiosperms (Blindow et al., 2014). Characeae also
successfully replaced P. Submerged Macrophyte Survival during
the Turbid Phase This makes Steinhuder Meer and
Felbrigg Lake cases of external load reduction with added
unintentional
internal
measures
(natural
biomanipulation)
possibly accelerating recovery. Therefore, these cases show closer
correspondence to macrophyte community patterns of lakes
having undergone internal measures. They also illustrate that
parallel biological processes may be at play in recovering lakes
that need to be considered in unison to understand the speed and
trajectory of macrophyte recovery. p
In many cases, however, mass developments of monocultures
occur. Monocultures of Elodea or Ceratophyllum species are often
unstable in terms of interannual persistence (Table 5) and can
collapse leading to a shift back to turbid conditions as in, for
instance, Lakes Zwemlust, Væng, and Alderfen Broad (Table 2). Characeae seem less often involved in sudden collapses, although,
exceptions are known, for example Schlosssee Buggenhagen
(Table 2) or Lake Botshol (Rip et al., 2007). If lakes remain
clear for several consecutive years, which is usually only the case
at lower nutrient concentrations, a more diverse macrophyte
community develops (Table 2, Lauridsen et al., 2003a). Lauridsen
et al. (2003b) assumed that differences in the success of
biomanipulation in Danish and Dutch shallow lakes might be
attributable to variation in pioneer macrophyte species; thus,
Elodea and Potamogeton species, typical for Danish lakes, were
preferred over charophytes by macrophyte-grazing waterfowl
(Weisner et al., 1997). Indeed, increasing top-down control
of periphyton-grazing invertebrates by omnivorous fish, which
increase in abundance in the period after a biomanipulation,
may render macrophytes more susceptible to herbivory (Hidding
et al., 2016). Frontiers in Plant Science | www.frontiersin.org Conclusions and Implications for Lake
Management In contrast to lakes undergoing only reduced external nutrient
loading, submerged macrophytes often respond very quickly
in shallow lakes subjected to biomanipulation by fish removal
(Hansson et al., 1998; Bakker et al., 2013), even at rather
high nutrient concentrations (Figure 2). Macrophytes colonizing
these lakes are often “pioneer” species, such as Elodea or
Ceratophyllum, re-colonizing from either seeds, oospores or
fragments and characterized by high growth rates (Tables 2, 4). Similar species have also been recorded in lakes following
natural fish kills (Sayer et al., 2016) or implementation of
other in-lake restoration measures such as sediment dredging
and phosphorus precipitation (Table 2). Fish removal may
indirectly (due to more periphyton grazing invertebrates) reduce
periphyton shading in summer, a major mechanism preventing Our analyses suggest that the composition of the macrophyte
community and their seasonal abundance in shallow lakes
during recovery from turbid, highly eutrophic conditions often
depends on remnant macrophyte stands, the specific restoration
measure applied and additional stochastic influences on water
clarity such as winter fish kills, cormorant predation on fish
or introduction of invasive filter-feeding mussel populations. In
turn, the prevailing macrophyte community can influence lake
water quality. Reductions in external nutrient loading often result in
the re-occurrence of spring clear-water phases exploitable by
a few macrophyte species (mainly pondweeds) with specific
traits. Resistance to wave action permits survival during the February 2018 | Volume 9 | Article 194 19 Hilt et al. Response of Macrophytes to Restoration turbid phase in very shallow areas, in particular in larger
lakes. During recovery these plants germinate early in spring
from energy-rich vegetative propagules and complete their
life cycle in early summer, when phytoplankton takes over. This intermediate recovery phase may, in some cases, last for
several decades before a more diverse and abundant submerged
macrophyte community develops that stabilizes clear-water
conditions during the entire potential growing season (Figure 1,
Table 1). Our model simulations suggest that, if the premature
termination of macrophyte growth can be prevented, the summer
phytoplankton peak responsible for turbid water and potentially
harmful algae blooms will also be reduced. Simulations also
revealed that at high periphyton shading, the intermediate
recovery phase is shifted to lower nutrient loads compared with a
scenario with lower periphyton shading. Therefore, if periphyton
shading can be reduced external restoration measures could
potentially be effective at a higher nutrient load. ACKNOWLEDGMENTS In contrast, fish stock reductions, natural fish kills and
sediment removal via suction-dredging can, by themselves,
temporarily restore clear-water conditions in spring and summer,
even at high nutrient concentrations and then allow rapid
colonization by pioneer macrophyte species from in situ seeds,
oospores or vegetative fragments. Fish stock reductions might
thus be a suitable short-term management strategy, but fish
removal needs to be frequently repeated. We thank all technicians of IGB Berlin responsible for the
long-term measuring program in Lake Müggelsee. Antje Barsch,
Nadine Baadke (Landesumweltamt Brandenburg) and Antje
Köhler (Senat Berlin) provided data for selected German lakes. The Environment Agency provided water quality data and the
Broads Authority macrophyte data for the lakes in the Broads
National Park (UK). We acknowledge linguistic improvements
by Anne Mette Poulsen. We thank two reviewers for their helpful
comments. Lasting macrophyte recovery can only be achieved in
combination with reduced nutrient loading (Figure 1), a need
that is further accentuated under future climate change scenarios,
where cyanobacterial shading of macrophytes will likely be more
severe (Kosten et al., 2012). Although global, political measures
against ongoing climate warming are slow, local restoration
efforts may reduce the combined stress from e.g. eutrophication
and climate warming (Moss et al., 2011; Scheffer et al., 2015), FUNDING MA and MG were supported by the German Research
Foundation (DFG, grant no. SU 623/1-1 and GRK 2032/1,
respectively). AJ is supported by the Netherlands Environmental
Assessment Agency (PBL) and ST by STOWA (grant no. 443.269). L-AH was supported by the BiodivERsA ERA-
net LIMNOTIP. EJ, MS, TD, and TL were supported by
MARS (Managing Aquatic ecosystems and water Resources
under multiple Stress) funded under the 7th EU Framework
Programme (Contract No.: 603378). JR was supported by
the German Ministry of Education and Research (project
NITROLIMIT, grant no. 033L041 A). MA and MG were supported by the German Research
Foundation (DFG, grant no. SU 623/1-1 and GRK 2032/1,
respectively). AJ is supported by the Netherlands Environmental
Assessment Agency (PBL) and ST by STOWA (grant no. 443.269). L-AH was supported by the BiodivERsA ERA-
net LIMNOTIP. EJ, MS, TD, and TL were supported by
MARS (Managing Aquatic ecosystems and water Resources
under multiple Stress) funded under the 7th EU Framework
Programme (Contract No.: 603378). JR was supported by
the German Ministry of Education and Research (project
NITROLIMIT, grant no. 033L041 A). SUPPLEMENTARY MATERIAL The Supplementary Material for this article can be found
online at: https://www.frontiersin.org/articles/10.3389/fpls.2018. 00194/full#supplementary-material AUTHOR CONTRIBUTIONS SH conceived the presented idea, wrote the manuscript and
performed the literature research. MA, EB, IB, TD, L-AH, EJ, TK,
AK, JK, TL, RN, GP, JR, H-HS, MS, KvdW, EvD, AW, NW, and
CS provided lake data. MG, JJ, AJ, WM, and ST performed the
modeling. All authors contributed to discussions and the writing
of different parts of the text. SH conceived the presented idea, wrote the manuscript and
performed the literature research. MA, EB, IB, TD, L-AH, EJ, TK,
AK, JK, TL, RN, GP, JR, H-HS, MS, KvdW, EvD, AW, NW, and
CS provided lake data. MG, JJ, AJ, WM, and ST performed the
modeling. All authors contributed to discussions and the writing
of different parts of the text. Conclusions and Implications for Lake
Management Macrophyte
recovery during the intermediate recovery state might be
facilitated by establishing exclosures to protect certain areas from
herbivory by birds and/or predation of periphyton grazers by
omnivorous fish, a lake-wide biomanipulation of fish, or internal
measures, such as TP precipitation to lower water column
TP concentrations in summer. Additional, usually unintended
internal changes, such as reductions in fish abundance by
commercial fisheries, natural fish kills or exotic mussel invasions
can facilitate a shift to clearer conditions in summer and
further aid the establishment of a more diverse macrophyte
community. and thereby serve as a buffer against further deterioration of
macrophyte beds and the ecosystem services that derive from
lakes and reservoirs (Urrutia-Cordero et al., 2016). REFERENCES Baldridge, A. K., and Lodge, D. M. (2014). Long-term studies of crayfish-invaded
lakes reveal limited potential for macrophyte recovery from the seed bank. Freshw. Sci. 33, 788–797. doi: 10.1086/677070 Alderton, E., Sayer, C. D., Davies, R., Lambert, S. J., and Axmacher, J. C. (2017). Buried alive: Aquatic plants survive in “ghost ponds” under agricultural fields. Biol. Conserv. 212, 105–110. doi: 10.1016/j.biocon.2017.06.004 Barrat-Segretain, M. H., and Bornette, G. (2000). Regeneration and colonization
abilities of aquatic plant fragments: effect of disturbance seasonality. Hydrobiologia 421, 31–39. doi: 10.1023/A:1003980927853 Annadotter, H., Cronberg, G., Aagren, R., Lundstedt, B., Nilsson, P. A., and
Ströbäck, S. (1999). Multiple techniques for lake restoration. Hydrobiologia 395,
77–85. doi: 10.1023/A:1017011132649 Bergman, E., Hansson, L.-A., Persson, A., Strand, J., Romare, P., Enell, M.,
et al. (1999). Synthesis of theoretical and empirical experiences from nutrient
and cyprinid reductions in Lake Ringsjön. Hydrobiologia 404, 145–156. doi: 10.1023/A:1003788900521 Bakker, E. S., Sarneel, S. M., Gulati, R. D., Liu, Z., and Van Donk, E. (2013). Restoring macrophyte diversity in shallow temperate lakes: biotic
versus abiotic constraints. Hydrobiologia 710, 23–37. doi: 10.1007/s10750-012-
1142-9 Bernes, C., Carpenter, S. R., Gårdmark, A., Larsson, P., Persson, L., Skov, C.,
et al. (2015). What is the influence of a reduction of planktivorous and February 2018 | Volume 9 | Article 194 Frontiers in Plant Science | www.frontiersin.org 20 Hilt et al. Response of Macrophytes to Restoration benthivorous fish on water quality in temperate eutrophic lakes? A systematic
review. Environ. Evidence 4, 7. doi: 10.1186/s13750-015-0032-9 constraints, synthesis, and recommendations for temperate lakes. Ecosystems
1, 558–574. doi: 10.1007/s100219900051 benthivorous fish on water quality in temperate eutrophic lakes? A systematic
review. Environ. Evidence 4, 7. doi: 10.1186/s13750-015-0032-9 constraints, synthesis, and recommendations for temperate lakes. Ecosystems
1, 558–574. doi: 10.1007/s100219900051 Best, E. P. H. (1977). Seasonal changes in mineral and organic components
of Ceratophyllum demersum and Elodea canadensis. Aquat. Bot. 3, 337–348. doi: 10.1016/0304-3770(77)90038-9 Hansson, L.-A., Johansson, L., and Persson, L. (1987). Effects of fish grazing on
nutrient release and succession of primary producers. Limnol. Oceanogr. 32,
723–729. doi: 10.4319/lo.1987.32.3.0723 Best, E. P. H. (1982). The aquatic macrophytes of Lake Vechten: species
composition, spatial distribution and production. Hydrobiologia 95, 65–77. doi: 10.1007/BF00044477 Hansson, L.-A., Nicolle, A., Brönmark, C., Hargeby, A., Lindström, Å., and
Andersson, G. (2010). Waterfowl, macrophytes, and the clear water state of
shallow lakes. Hydrobiologia 646, 101–109. doi: 10.1007/s10750-010-0169-z Hargeby, A., Andersson, G., Blindow, I., and Johansson, S. (1994). REFERENCES Adaptations of macrophytes to life in moving
water: upslope limits and mechanical properties of stems. Hydrobiologia 194,
133–142. doi: 10.1007/BF00028414 Hilt, S., and Gross, E. M. (2008). Can allelopathically active submerged
macrophytes stabilise clear-water states in shallow eutrophic lakes? Bas. Appl. Ecol. 9, 422–432. doi: 10.1016/j.baae.2007.04.003 Bytnerowicz, T. A., and Carruthers, R. I. (2014). Temperature-dependent models
of Zannichellia palustris seed germination for application in aquatic systems. Environ. Exp. Bot. 104, 44–53. doi: 10.1016/j.envexpbot.2014.03.006 Hilt, S., Gross, E. M., Hupfer, M., Morscheid, H., Mählmann, J., Melzer, A.,
et al. (2006). Restoration of submerged vegetation in shallow eutrophic lakes
– guideline and state of the art in Germany. Limnologica 36, 155–171. doi: 10.1016/j.limno.2006.06.001 Capers, R. S. (2003). Macrophyte colonization in a freshwater tidal wetland (Lyme,
CT, USA). Aquat. Bot. 77, 325–338. doi: 10.1016/j.aquabot.2003.08.001 Hilt, S., and Grünert, U. (2008). Praxistest zur Bewertung von Makrophyten
in Berliner Seen im Rahmen der Umsetzung der EU-Wasserrahmenrichtlinie. Report for Senate (Berlin), 1–58. Choi, C., Bareiss, C., Walenciak, O., and Gross, E. M. (2002). Impact of polyphenols
on growth of the aquatic herbivore Acentria ephemerella. J. Chem. Ecol. 28,
2223–2238. doi: 10.1023/A:1021049332410 De Winton, M. D., Clayton, J. S., and Champion, P. D. (2000). Seedling emergence
from seed banks of 15 New Zealand lakes with contrasting vegetation histories. Aquat. Bot. 66, 181–194. doi: 10.1016/S0304-3770(99)00074-1 Hilt, S., Van de Weyer, K., Köhler, A., and Chorus, I. (2010). Submerged
macrophyte responses to reduced phosphorus concentrations in two peri-
urban lakes. Rest. Ecol. 18, 452–461. doi: 10.1111/j.1526-100X.2009.00
577.x Dorenbosch, M., and Bakker, E. S. (2011). Herbivory in omnivorous fishes:
effect of plant secondary metabolites and prey stoichiometry. Freshw. Biol. 56,
1783–1797. doi: 10.1111/j.1365-2427.2011.02618.x Hoare, D., Phillips, G., and Perrow, M. (2008). Broads Lake Restoration Strategy. Appendix 4. Review of Biomanipulaton. Norwich: Broads Authority, 1–59. Elger, A., and Willby, N. J. (2003). Leaf dry matter content as an integrative
expression of plant palatability: the case of freshwater macrophytes. Func. Ecol. 17, 58–65. doi: 10.1046/j.1365-2435.2003.00700.x Hussner,
A.,
Gross,
E. M.,
Van
de
Weyer,
K.,
and
Hilt,
S. (2014). Handlungsempfehlung zur Abschätzung der Chancen einer Wiederansiedlung
von Wasserpflanzen bei der Restaurierung von Flachseen Deutschlands. Arbeitsblatt: Deutsche Gesellschaft für Limnologie (DGL), Elger, A., Willby, N. J., and Cabello-Martinez, M. (2009). Invertebrate
grazing
during
the
regenerative
phase
affects
the
ultimate
structure
of
macrophyte
communities. Freshw. Biol. 54,
1246–1255
doi: 10.1111/j.1365-2427.2009.02171.x Idestam-Almquist, J., and Kautsky, L. (1995). Plastic responses in morphology
of Potamogeton pectinatus L. REFERENCES Trophic
web structure in a shallow eutrophic lake during a dominance shift
from phytoplankton to submerged macrophytes. Hydrobiologia 279, 83–90. doi: 10.1007/BF00027843 Best, E. P. H., and Dassen, J. H. A. (1987). A seasonal study of growth
characteristics and the levels of carbohydrates and proteins in Elodea nuttallii,
Polygonum amphibium and Phragmites australis. Aquat. Bot. 28, 353–372. doi: 10.1016/0304-3770(87)90011-8 Heinzel, K., and Martin, C. (2006). Monitoring der Qualitätskomponente
Makrophyten für WRRL und FFH-RL in schleswig-holsteinischen Seen;
Gutachten im Auftrag des Landesamt für Natur und Umwelt des Landes
Schleswig-Holstein. Available online at: http://www.umweltdaten.landsh.de/
nuis/wafis/seen/Berichte_Gutachten/Ufer_Unterwasservegetation/Bericht_
Makrophyten 2006 WRRL Heinzel GFN.pdf. Blindow,
I. (1992). Long-
and
short-term
dynamics
of
submerged
macrophytes in two shallow eutrophic lakes. Freshw. Biol. 28, 15–27. doi: 10.1111/j.1365-2427.1992.tb00558.x j
Blindow, I., Hargeby, A., and Hilt, S. (2014). Facilitation of clear-water
conditions
in
shallow
lakes
by
macrophytes:
differences
between
charophyte
and
angiosperm
dominance. Hydrobiologia
737,
99–110. doi: 10.1007/s10750-013-1687-2 Hidding, B., Bakker, E. S., Hootsmans, M. J. M., and Hilt, S. (2016). Synergy
between shading and herbivory triggers plant loss and regime shifts in aquatic
systems. Oikos 125, 1489–1495. doi: 10.1111/oik.03104 Blindow, I., Hargeby, A., Wagner, B. M. A., and Andersson, G. (2000). How
important is the crustacean plankton for the maintenance of water clarity in
shallow lakes with abundant submerged vegetation? Freshw. Biol. 44, 185–197. doi: 10.1046/j.1365-2427.2000.00552.x Hidding, B., Bakker, E. S., Keuper, F., De Boer, T., De Vries, P. P., and Nolet,
B. A. (2010). Differences in tolerance of pondweeds and charophytes to
vertebrate herbivores in a shallow Baltic estuary. Aquat. Bot. 93, 123–128. doi: 10.1016/j.aquabot.2010.04.002 Blüml, V., Körner, F., Marxmeier, U., Richter, M., and Schönheim, A. (2008). Entwicklung und aktuelle Situation der Verlandungsvegetation des Dümmers
(Niedersachsen). Osnabr. Naturwiss. Mitteil. 33/34, 19–46. Hilt, S., Adrian, R., Köhler, J., Monaghan, M. T., and Sayer, C. (2013). Clear,
crashing, turbid and back – long-term changes of macrophyte assemblages in a
shallow lake. Freshw. Biol. 58, 2027–2036. doi: 10.1111/fwb.12188 Bootsma, M. C., Barendregt, A., and Van Alphen, J. C. A. (1999). Effectiveness
of reducing external nutrient load entering a eutrophicated shallow lake
ecosystem in the Naardermeer nature reserve, The Netherlands. Biol. Conserv. 90, 193–201. doi: 10.1016/S0006-3207(99)00027-0 Hilt, S., Brothers, S., Jeppesen, E., Veraart, A., and Kosten, S. (2017). Translating
regime shifts in shallow lakes into changes in ecosystem functions and services. Bioscience 67, 928–936. doi: 10.1093/biosci/bix106 Brewer, C. A., and Parker, M. (1990). REFERENCES Plankt. Res. 18, 2283–2294. doi: 10.1093/plankt/18.12.2283 Jeppesen, E., Søndergaard, M., Lauridsen, T. L., Davidson, T. A., Liu, Z.,
Mazzeo, N., et al. (2012). Biomanipulation as a response tool to combat
eutrophication: recent advances and future challenges. Adv. Ecol. Res. 47,
411–488. doi: 10.1016/B978-0-12-398315-2.00006-5 Lillie, R. A., Budd, J., and Rasmussen, P. W. (1997). Spatial and temporal variability
in biomass density of Myriophyllum spicatum L. in a northern temperate lake. Hydrobiologia 347, 69–74. doi: 10.1023/A:1003059000277 Jones, J. I., and Sayer, C. D. (2003). Does fish-invertebrate-periphyton
cascade precipitate plant loss in shallow lakes? Ecology 84, 2155–2167. doi: 10.1890/02-0422 Madsen, J. D. (1991). Resource allocation at the individual plant level. Aquat. Bot. 41, 67–86. doi: 10.1016/0304-3770(91)90039-8 Jones, J. I., Young, J. O., Eaton, J. W., and Moss, B. (2002). The influence of
nutrient loading, dissolved inorganic carbon and higher trophic levels on
the interaction between submerged plants and periphyton. J. Ecol. 90, 12–24. doi: 10.1046/j.0022-0477.2001.00620.x Madsen, J. D., and Adams, M. S. (1988). The germination of Potamogeton
pectinatus tubers: environmental control by temperature and light. Can. J. Bot. 66, 2523–2526. doi: 10.1139/b88-343 Josefsson, M., (2011). NOBANIS - Invasive Species Fact Sheet - Elodea canadensis,
Elodea nuttallii and Elodea callitrichoides – NOBANIS. Available online at:
www.nobanis.org. Date of Access (Accessed Nov 28, 2012). Mathes, J. (2007). Erfahrungen mit Seentherapien in Mecklenburg-Vorpommern. Available
online
at:
http://www.lake-jewel.net/download/Dateien/
Seesanierung%20Mecklenburg-Vorpommern.pdf Mauersberger, H., and Mauersberger, R. (1996). Die Seen im Biosphärenreservat
Schorfheide-Chorin:
eine
ökologische
Studie. Dissertation,
Universität
Greifswald. Kabus, T., Schumann, M., Weiß, G., and Kalhoff, M. (2007). Beitrag zur Flora der
Havel und angrenzender Flächen zwischen Potsdam und Pritzerbe. Verh. Bot. Ver. Berlin Brandenburg 140, 25–55. Kabus,
T.,
and
Mauersberger,
R.,
(2011). Liste
und
Rote
Liste
der
Armleuchteralgengewächse (Characeae) des Landes Brandenburg. Natursch. Landschaftspfl. Brbg. 20, 1–32. Meijer, M. L., De Boois, I., Scheffer, M., Portielje, R., and Hosper, H. (1999). Biomanipulation in shallow lakes in The Netherlands: an evaluation of 18 case
studies. Hydrobiologia 408, 13–30. doi: 10.1023/A:1017045518813 Kaplan, K., and Muer, T. (1990). Beobachtungen zum diasporenreservoir im
Bereich ehemaliger Heideweiher. Florist. Rundbriefe 24, 38–45. Meijer, M. L., and Hosper, H. (1997). Effects of biomanipulation in the large
and shallow Lake Wolderwijd, The Netherlands. Hydrobiologia 342, 335–349. doi: 10.1023/A:1017088701621 Karatayev, A. Y., Burlakova, L. E., and Padilla, D. K. (2015). Zebra versus quagga
mussels: a review of their spread, population dynamics, and ecosystem impacts. Hydrobiologia 746, 97–112. doi: 10.1007/s10750-014-1901-x Mooij, W. M., Brederveld, R. REFERENCES J., Van
Gerven, L. P. A., et al. (2017). Mowing submerged macrophytes in shallow
lakes with alternative stable states: battling the good guys? Environ. Manage. 59, 619–634. doi: 10.1007/s00267-016-0811-2 Janssen, A. B. G., de Jager, V. C. L., Janse, J. H., Kong, X., Liu, S., Ye,
Q., et al. (2017). Spatial identification of critical nutrient loads of large
shallow lakes: implications for Lake Taihu (China). Wat. Res. 119, 276–287. doi: 10.1016/j.watres.2017.04.045 Lage, S., Annadotter, H., Rasmussen, U., and Rydberg, S. (2015). Biotransfer of
β-N-methylamino-L-alanine (BMAA) in a eutrophicated freshwater lake. Mar. Drugs 13, 1185–1201 doi: 10.3390/md13031185 Jeppesen, E., Kristensen, P., Jensen, J. P., Søndergaard, M., Mortensen, E., and
Lauridsen, T. (1991). Recovery resilience following a reduction in external
phosphorus loading of shallow, eutrophic Danish lakes: duration, regulating
factors and methods for overcoming resilience. Mem. Ist. Ital. Idrobiol. 48,
127–148. Lalonde, S., and Downing, J. A. (1991). Epiphyton biomass is related to lake
trophic status, depth, and macrophyte architecture. Can. J. Fish. Aquat. Sci. 48,
2285–2291. doi: 10.1139/f91-268 Landesamt für Natur und Umwelt des Landes Schleswig-Holstein (1997). Zustand
und Belastungsquellen des Warder Sees. Flintbek: Landesamt für Natur und
Umwelt des Landes Schleswig-Holstein. Jeppesen, E., Søndergaard, M., Meerhoff, M., Lauridsen, T. L., and Jensen,
J. P. (2007a). Shallow lake restoration by nutrient loading reduction –
some recent findings and challenges ahead. Hydrobiologia 584, 239–252. doi: 10.1007/s10750-007-0596-7 Lauridsen, T. L., Jensen, J. P., Jeppesen, E., and Sondergaard, M. (2003a). Response
of
submerged
macrophytes
in
Danish
lakes
to
nutrient
loading reductions and biomanipulation. Hydrobiologia 506, 641–649. doi: 10.1023/B:HYDR.0000008633.17385.70 Jeppesen, E., Meerhoff, M., Jacobsen, B. A., Hansen, R. S., Søndergaard, M.,
Jensen, J. P., et al. (2007b). Restoration of shallow lakes by nutrient control
and biomanipulation - the successful strategy varies with lake size and climate. Hydrobiologia 581, 269–285. doi: 10.1007/s10750-006-0507-3 Lauridsen, T. L., Sandsten, H., and Hald Møller, P. (2003b). Restoration
of a shallow lake by introducing Potamogeton spp.: impact of waterfowl
grazing. Lakes Reserv. Res. Manage. 8, 177–187. doi: 10.1111/j.1440-1770.2003. 00224.x Jeppesen, E., Søndergaard, M., Jensen, J. P., Havens, K., Anneville, O., Carvalho,
L., et al. (2005). Lake responses to reduced nutrient loading – an analysis of
contemporary data from 35 European and North American long term studies. Freshw. Biol. 50, 1747–1771. doi: 10.1111/j.1365-2427.2005.01415.x Lauridsen, T., Pedersen, L. J., Jeppesen, E., and Søndergaard, M. (1996). The importance of macrophyte bed size for composition and horizontal
migration of cladocerans in a shallow lake. J. REFERENCES to sediment and above-sediment conditions
at two sites in the northern Baltic proper. Aquat. Bot. 52, 205–216. doi: 10.1016/0304-3770(95)00499-8 Gervais, F., Berger, S., Schönfelder, T., and Rusche, R. (1999). Basic limnological
characteristics of the shallow eutrophic lake Grimnitzsee (Brandenburg,
Germany). Limnologica
29,
105–119. doi:
10.1016/S0075-9511(99)80
058-9 Janse, J. H. (1997). A model of nutrient dynamics in shallow lakes in relation to
multiple stable states. Hydrobiologia 342, 1–8. doi: 10.1023/A:1017018812215 Janse, J. H. (1997). A model of nutrient dynamics in shallow lakes in relation to
multiple stable states. Hydrobiologia 342, 1–8. doi: 10.1023/A:1017018812215
Janse, J. H. (2005). Model Studies on the Eutrophication of Shallow Lakes and
Ditches, Dissertation, Wageningen University. Janse, J. H. (2005). Model Studies on the Eutrophication of Shallow Lakes and
Ditches, Dissertation, Wageningen University. Grutters, B. M. C., Gross, E. M., Van Donk, E., and Bakker, E. S. (2017). Periphyton
density is similar on native and non-native plant species. Freshw. Biol. 62,
906–915. doi: 10.1111/fwb.12911 Janse, J. H., De Senerpont Domis, L. N., Scheffer, M., Lijklema, L., Van
Liere, L., Klinge, M., et al. (2008). Critical
phosphorus loading of
different types of shallow lakes and the consequences for management
estimated with the ecosystem model PCLake. Limnologica 38, 203–219. doi: 10.1016/j.limno.2008.06.001 Gulati, R. D., and Van Donk, E. (2002). Lakes in the Netherlands, their origin,
eutrophication and restoration: state-of-the-art review. Hydrobiologia 478,
73–106. doi: 10.1023/A:1021092427559 Haag, R. W. (1983). Emergence of seedlings of aquatic macrophytes from lake
sediments. Can. J. Bot. 61, 148–156. doi: 10.1139/b83-014 Janse, J. H., Scheffer, M., Lijklema, L., Van Liere, L., Sloot, J. S., and Mooij, W. M. (2010). Estimating the critical phosphorus loading of shallow lakes with the
ecosystem model PCLake: sensitivity, calibration and uncertainty. Ecol. Model. 221, 654–665. doi: 10.1016/j.ecolmodel.2009.07.023 Hansson, L.-A., Annadotter, H., Bergman, E., Hamrin, S. F., Jeppesen, E., Kairesalo,
T., et al. (1998). Biomanipulation as an application of food-chain theory: February 2018 | Volume 9 | Article 194 Frontiers in Plant Science | www.frontiersin.org 21 Hilt et al. Response of Macrophytes to Restoration Janse, J. H., Van Donk, E., and Aldenberg, T. (1998). A model study on the stability
of the macrophyte-dominated state as affected by biological factors. Wat. Res. 32, 2696–2706. doi: 10.1016/S0043-1354(98)00049-9 Kuiper, J. J. (2016). Making Eco Logic and Models work: An Integrative Approach to
Lake Ecosystem Modelling, Dissertation, Wageningen University. Kuiper, J. J., Verhofstad, M. J., Louwers, E. L., Bakker, E. S., Brederveld, R. REFERENCES J., De Klein, J. J. M., DeAngelis, D. L., Downing,
A. S., Faber, M., et al. (2014). Serving many at once: how a database approach
can create unity in dynamical ecosystem modelling. Environ. Mod. Softw. 61,
266–273. doi: 10.1016/j.envsoft.2014.04.004 Kautsky,
L. (1990). Seed
and
tuber
banks
of
aquatic
macrophytes
in
thE
Askö
area,
northern
Baltic
proper. Holarct. Ecol. 13,
143–148. doi: 10.1111/j.1600-0587.1990.tb00600.x Mooij, W. M., Janse, J. H., De Senerpont Domis, L. N., Hülsmann, S., and
Ibelings, B. W. (2007). Predicting the effect of climate change on temperate
shallow lakes with the ecosystem model PCLake. Hydrobiologia 584, 443–454. doi: 10.1007/s10750-007-0600-2 Knösche, R. (2008). Wiederfund von Najas marina L. spp. marina im Schollener
See (Elbe-Havel-Winkel, Sachsen-Anhalt). Mitteil. Florist. Kart. Sachsen Anhalt
13, 41–51. Moss, B., Balls, H., Irvine, K., and Stansfield, J. (1986). Restoration of two lowland
lakes by isolation from nutrient-rich water sources with and without removal Körner,
S. (2001). Development
of
submerged
macrophytes
in
shallow
Lake Müggelsee (Berlin, Germany) before and after its switch to the
phytoplankton-dominated
state. Arch. Hydrobiologia
152,
395–409. doi: 10.1127/archiv-hydrobiol/152/2001/395 Moss, B., Balls, H., Irvine, K., and Stansfield, J. (1986). Restoration of two lowland
lakes by isolation from nutrient-rich water sources with and without removal
of sediment. J. Appl. Ecol. 23, 391–414. doi: 10.2307/2404025 of sediment. J. Appl. Ecol. 23, 391–414. doi: 10.2307/2404025 Moss, B., Kosten, S., Meerhoff, M., Battarbee, R. W., Jeppesen, E., Mazzeo, N.,
et al. (2011). Allied attack: climate change and nutrient pollution. Inland Wat. 1, 101–105. doi: 10.5268/IW-1.2.359 Körner, S. (2002). Loss of submerged macrophytes in shallow lakes in
North-Eastern
Germany. Intern. Rev. Hydrobiologia
87,
375–384. doi: 10.1002/1522-2632(200207)87:4<375::AID-IROH375>3.0.CO;2-7 Moss, B., Stansfield, J., and Irvine, K. (1990). Problems in the restoration of a
hypertrophic lake by diversion of a nutrient-rich inflow. Verh. Intern. Ver. Limnol. 24, 568–572. doi: 10.1080/03680770.1989.11898801 Körner, S., and Dugdale, T. (2003). Is roach herbivory preventing re-colonization
of a shallow lake with submerged macrophytes? Hydrobiologia 506, 497–501. doi: 10.1023/B:HYDR.0000008561.67513.ec Nixdorf, B., Rücker, J., Dolman, A. M., Wiedner, C., Hilt, S., Kasprzak, P., et al. (2013). Prozessverständnis als Grundlage für die Gewässerbewirtschaftung –
Fallbeispiele
für
Limitation,
Konkurrenz,
Gewässerstruktur
und
Nahrungsnetzsteuerung. Korr. Wasserwirtschaft 60, 693–701. Kosten, S., Huszar, V. L. M., Bécares, E., Costa, L., Van Donk, E., Hansson, L.-A.,
et al. (2012). Warmer climates boost cyanobacterial dominance in shallow lakes. Glob. Chang. Biol. 18, 118–126. REFERENCES doi: 10.1146/annurev-ecolsys-110411-160251 Phillips, G. L., Willby, N., and Moss, B. (2016). Submerged macrophyte decline in
shallow lakes: what have we learnt in the last forty years? Aquat. Bot. 135, 37–45. doi: 10.1016/j.aquabot.2016.04.004 Søndergaard, M., Bjerring, R., and Jeppesen, E. (2013). Persistent internal
phosphorus loading in shallow eutrophic lakes. Hydrobiologia 710, 95–107. doi: 10.1007/s10750-012-1091-3 Portielje, R., and Van der Molen, D. T. (1999). Relationships between
eutrophication variables: from nutrient loading to transparency. Hydrobiologia
408, 375–387. doi: 10.1023/A:1017090931476 Søndergaard, M., Larsen, S. E., Johansson, L. S., Lauridsen, T. L., and Jeppesen,
E. (2016). Ecological classification of lakes: uncertainty and the influence of
year-to-year variability. Ecol. Indic. 61, 248–257. doi: 10.1016/j.ecolind.2015. 09.024 Pot, R., and Ter Heerdt, G. N. J. (2014). Succession dynamics of aquatic
lake vegetation after restoration measures: increased stability after 6 years
of development. Hydrobiologia 737, 333–345. doi: 10.1007/s10750-014-
1835-3 Søndergaard, M., Lauridsen, T. L., Johansson, L. S., and Jeppesen, E. (2017). Repeated fish removal to restore lakes: case study of Lake Væng, Denmark-
two biomanipulations during 30 years of monitoring. Water
9, 43. doi: 10.3390/w9010043 Riegman, R. (2007). Evaluatie Rapport Vijftien jaar Actief Biologisch Beheer (ABB)
in het Duinigermeer (in Dutch). Reest aen Wieden Waterschap. Søndergaard, M., Liboriussen, L., Pedersen, A. R., and Jeppesen, E. (2008). Lake
restoration by fish removal: short- and long-term effects in 36 Danish lakes. Ecosystems 11, 1291–1305. doi: 10.1007/s10021-008-9193-5 Rip, W. J., Ouboter, R. L., and Los, H. L. (2007). Impact of climatic fluctuations
on Characeae biomass in a shallow, restored lake in The Netherlands. Hydrobiologia 584, 415–424. doi: 10.1007/s10750-007-0608-7 Spencer, D. F. (1986). Early growth of Potamogeton pectinatus L. in response
to temperature and irradiance: morphology and pigment composition. Aquat. Bot. 26, 1–8. doi: 10.1016/0304-3770(86)90002-1 Roberts, E., Kroker, J., Körner, S., and Nicklisch, A. (2003). The role of periphyton
during the re-colonization of a shallow lake with submerged macrophytes. Hydrobiologia 506, 525–530. doi: 10.1023/B:HYDR.0000008560.73832.1c Strand,
J. (1999). The
development
of
submerged
macrophytes
in
Lake
Ringsjön
after
biomanipulation. Hydrobiologia
404,
113–121. doi: 10.1023/A:1003728730563 Rodrigo, M. A., Rojo, C., Alonso-Guillén, J. L., and Vera, P. (2013). Restoration
of two small Mediterranean lagoons: the dynamics of submerged macrophytes
and factors that affect the success of revegetation. Ecol. Eng. 54, 1–15. doi: 10.1016/j.ecoleng.2013.01.022 Strand, J. A., and Weisner, S. E. B. (2001). Dynamics of submerged macrophyte
populations in response to biomanipulation. Freshw. Biol. 46, 1397–1408. doi: 10.1046/j.1365-2427.2001.00746.x Rørslett,
B.,
Berge,
D.,
and
Johansen,
S. W. (1985). REFERENCES doi: 10.1111/j.1365-2486.2011.02488.x February 2018 | Volume 9 | Article 194 Frontiers in Plant Science | www.frontiersin.org 22 Hilt et al. Response of Macrophytes to Restoration Niedersächsischer Landesbetrieb für Wasserwirtschaft, Küsten- und Naturschutz,
NLWKN. (2011). Leitfaden
Maßnahmenplanung
Oberflächengewässer. Wasserrahmenrichtlinie Band 3 Steinhuder Meer. 1–33. Scheffer, M., and Carpenter, S. R. (2003). Catastrophic regime shifts in
ecosystems: linking theory to observation. Trends Ecol. Evol. 18, 648–656. doi: 10.1016/j.tree.2003.09.002 Scheffer, M., De Redelijkheid, M. R., and Noppert, F. (1992). Distribution and
dynamics of submerged vegetation in a chain of shallow eutrophic lakes. Aquat. Bot. 42, 199–216. doi: 10.1016/0304-3770(92)90022-B Noordhuis, R., Van Zuidam, B. G., Peeters, E. T. H. M., and Van Geest, G. J. (2016). Further improvements in water quality of the Dutch Border lakes:
two types of clear states at different nutrient levels. Aquatic Ecol. 50, 521–539. doi: 10.1007/s10452-015-9521-8 Scheffer, M., Hosper, S. H., Meijer, M. L., Moss, B., and Jeppesen, E. (1993). Alternative equilibria in shallow lakes. Trends Ecol. Evol. 8, 275–279. doi: 10.1016/0169-5347(93)90254-M Perrow, M. R., Schutten, J. H., Howes, J. R., Holzer, T., Madgwick, F. J., and
Jowitt, A. J. D. (1997). Interactions between coot (Fulica atra) and submerged
macrophytes: the role of birds in the restoration process. Hydrobiologia 342,
241–255. doi: 10.1023/A:1017007911190 Schutten, J., Dainty, J., and Davy, A. J. (2005). Root anchorage and its
significance for submerged plants in shallow lakes. J. Ecol. 93, 556–571. doi: 10.1111/j.1365-2745.2005.00980.x Phillips, G., Bennion, H., Perrow, M., Sayer, C. D., Spears, B., and Willby, N. (2015). A Review of Lake Restoration Practices and their Performance in the
Broads National Park, 1980-2013. Report for the Broads Authority, Norwich
and Natural England. Available online at: http://www.broads-authority.gov.uk/
news-and-publications/publications-and-reports/conservation-publications-
and-reports/water-conservation-reports Simberloff, D., and Gibbons, L. (2004). Now you see them, now you don’t! –
population crashes of established introduced species. Biol. Invasions 6, 161–172. doi: 10.1023/B:BINV.0000022133.49752.46 Simpson, D. A. (1990). Displacement of Elodea canadensis Michx. by Elodea
nuttallii (Planch.) St. John in the British Isles. Watsonia 18, 173–177. Phillips, G., Kelly, A., Pitt, J. A., Sanderson, R., and Taylor, E. (2005). The recovery of a very shallow eutrophic lake, 20 years after the
control of effluent derived phosphorus. Freshw. Biol. 50, 1628–1638. doi: 10.1111/j.1365-2427.2005.01434.x Sommer, U., Adrian, R., De Senerpont Domis, L. N., Elser, J. J., Gaedke, U.,
Ibelings, B. W., et al. (2012). Beyond the Plankton Ecology Group (PEG)
Model: mechanisms driving plankton succession. Ann. Rev. Ecol. Evol. Syst. 43,
429–448. REFERENCES Mass
invasion
of
Elodea
canadensis
in
a
mesotrophic,
South
Norwegian
lake
–
impact
of
water
quality. Verh. Intern. Ver. Limnol. 22,
2920–2926. doi: 10.1080/03680770.1983.11897803 Talling, J. F., and Parker, J. E. (2002). Seasonal dynamics of phytoplankton
and phytobenthos, and associated chemical interactions, in a shallow
upland lake (Malham Tarn, northern England). Hydrobiologia 487, 167–181. doi: 10.1023/A:1022934705359 Rücker, J., Barsch, A., and Nixdorf, B. (2015). Besser, aber noch nicht gut“ –
Ökologischer Zustand der Seen in Brandenburg 2014. Wasserwirtschaft 12,
41–47. Tobiessen, P., and Snow, P. D. (1984). Temperature and light effects on the
growth of Potamogeton crispus in Collins Lake, New York State. Can. J. Bot. 62, 2822–2826. doi: 10.1139/b84-376 Sand-Jensen, K., Pedersen, N. L., Thorsgaard, I., Moeslund, B., Borum, J., and
Brodersen, K. P. (2008). 100 years of vegetation decline and recovery in Lake
Fure, Denmark. J. Ecol. 96, 206–271. doi: 10.1111/j.1365-2745.2007.01339.x Triest, L., Stiers, I., and Van Onsem, S. (2016). Biomanipulation as a nature-
based solution to reduce cyanobacterial blooms. Aquat. Ecol. 50, 461–483. doi: 10.1007/s10452-015-9548-x Sayer, C. D., Burgess, A., Kari, K., Davidson, T. A., and Rose, N. (2010a). Long-
term dynamics of submerged macrophytes and algae in a small and shallow,
eutrophic lake: implications for the stability of macrophyte-dominance. Freshw. Biol. 55, 565–583. doi: 10.1111/j.1365-2427.2009.02353.x Urrutia-Cordero, P., Ekvall, M. K., and Hansson, L. A. (2016). Local food web
management increases resilience and buffers against global change effects on
freshwaters. Sci. Rep. 6:29542. doi: 10.1038/srep29542 Vadeboncoeur, Y., Kalff, J., Christoffersen, K., and Jeppesen, E. (2006). Substratum
as
a
driver
of
variation
in
periphyton
chlorophyll
and
productivity in lakes. J. N. Am. Benth. Soc. 25, 379–392. doi: 10.1899/0887-
3593(2006)25[379:SAADOV]2.0.CO;2 Sayer, C. D., Davidson, T. A., and Jones, J. I. (2010b). Seasonal dynamics
of macrophytes and phytoplankton in shallow lakes: a eutrophication-
driven pathway from plants to plankton? Freshw. Biol. 55, 500–513. doi: 10.1111/j.1365-2427.2009.02365.x Van Berkum, J. A., Klinge, M., and Grimm, M. P. (1995). Biomanipulation
in
the
duinigermeer,
first
results. Neth. J. Aquat. Ecol. 29,
81–90. doi: 10.1007/BF02061791 Sayer, C. D., Davidson, T. A., Rawcliffe, R., Langdon, P. G., Leavitt, P. R.,
Cockerton, G., et al. (2016). Consequences of fish kills for long-term trophic
structure in shallow lakes: implications for theory and restoration. Ecosystems
16, 1289–1309. doi: 10.1007/s10021-016-0005-z Van de Bund, W. J., and Van Donk, E. (2002). REFERENCES Short-term and long-term
effects of zooplanktivorous fish removal in a shallow lake: a synthesis
of 15 years of data from Lake Zwemlust. Freshw. Biol. 47, 2380–2387. doi: 10.1046/j.1365-2427.2002.01006.x Scheffer, M., Barrett, S., Carpenter, S. R., Folke, C., Green, A. J., Holmgren, M.,
et al. (2015). Creating a safe operating space for iconic ecosystems. Science 347,
1317–1319. doi: 10.1126/science.aaa3769 February 2018 | Volume 9 | Article 194 Frontiers in Plant Science | www.frontiersin.org 23 Hilt et al. Response of Macrophytes to Restoration Van de Haterd, R. J. W., and Ter Heerdt, G. N. J. (2007). Potential for
the development of submerged macrophytes in eutrophicated shallow
peaty
lakes
after
restoration
measures. Hydrobiologia
584,
277–290. doi: 10.1007/s10750-007-0593-x Verhofstad, M. J. J. M., Alirangues Núñez, M. M., Reichman, E. P., van Donk, E.,
Lamers, L. P. M., and Bakker, E. S. (2017). Mass development of monospecific
submerged macrophyte vegetation after the restoration of shallow lakes: roles of
light, sediment nutrient levels, and propagule density. Aquat. Bot. 141, 29–38. doi: 10.1016/j.aquabot.2017.04.004 Van den Berg, M. S., Scheffer, M., Coops, H., and Simons, J. (1998). The role of
characean algae in the management of eutrophic shallow lakes. J. Phycol. 34,
750–756. doi: 10.1046/j.1529-8817.1998.340750.x Verpoorter, C., Kutser, T., Seekell, D. A., and Tranvik, L. J. (2014). A global
inventory of lakes based on high-resolution satellite imagery. Geophys. Res. Lett. 41, 6396–6402. doi: 10.1002/2014GL060641 Van den Berg, M. S., Scheffer, M., Van Nes, E., and Coops, H. (1999). Dynamics and stability of Chara sp. and Potamogeton pectinatus in a
shallow lake changing in eutrophication level. Hydrobiologia 408, 335–342. doi: 10.1023/A:1017074211970 Wang, X., Qin, B., Gao, G., and Paerl, H. W. (2010). Nutrient enrichment and
selective predation by zooplankton promote Microcystis (Cyanobacteria) bloom
formation. J. Plankt. Res. 32, 457–470. doi: 10.1093/plankt/fbp143 formation. J. Plankt. Res. 32, 457–470. doi: 10.1093/plankt/fbp1 Van de Weyer, K. (2011). Erfassung und Bewertung der Makrophytenvegetation in
der Unteren Havel und im Wannsee im Jahr 2011. Report for Senate (Berlin). 1–111. Waterstraat, A. (2008). Die Fischfauna des Galenbecker Sees und seiner Zu- und
Abflüsse. In: 70 Jahre Naturschutzgebiet Galenbecker See. Staatliches Amt für
Umwelt und Natur Ueckermünde. 27–34. Weisner, S. E., Strand, J. A., and Sandsten, H. (1997). Mechanisms regulating
abundance of submerged vegetation in shallow eutrophic lakes. Oecologia 109,
592–599. doi: 10.1007/s004420050121 Van Dijk, G. M. (1993). REFERENCES Dynamics and attenuation characteristics of periphyton
upon artificial substratum under various light conditions and some additional
observations on periphyton upon Potamogeton pectinatus L. Hydrobiologia 252,
143–161. doi: 10.1007/BF00008152 Wolfer, S. R., and Straile, D. (2004). Spatio-temporal dynamics and plasticity
of clonal architecture in Potamogeton perfoliatus. Aquat. Bot. 78, 307–318. doi: 10.1016/j.aquabot.2003.11.005 Van Dijk, G. M., and Janse, J. H. (1993). Modeling resource allocation in
Potamogeton pectinatus L. J. Aquat. Plant Manage. 31, 128–134. Woolf, T. E., and Madsen, J. D. (2003). Seasonal biomass and carbohydrate
allocation patterns in southern Minnesota curlyleaf pondweed populations. J. Aquat. Plant Manage. 41, 113–118. Van Dijk, G. M., and Van Vierssen, W. (1991). Survival of a Potamogeton
pectinatus
L.population
under
various
light
conditions
in
a
shallow
eutrophic lake (Lake Veluwe) in The Netherlands. Aquat. Bot. 39, 121–129. doi: 10.1016/0304-3770(91)90027-3 Zheng, G., Xu, R., Chang, X., Hilt, S., and Wu, C. (2013). Cyanobacteria
can allelopathically inhibit submerged macrophytes: effects of Microcystis
aeruginosa extracts and exudates on Potamogeton malaianus. Aquat. Bot. 109,
1–7. doi: 10.1016/j.aquabot.2013.02.004 Van Vierssen, W. (1982a). The ecology of communities dominated by
Zannichellia taxa in Western Europe. 1. Characterization and autecology of
the Zannichellia taxa. Aquat. Bot. 12, 103–155. doi: 10.1016/0304-3770(82)90
010-9 Conflict of Interest Statement: The authors declare that the research was
conducted in the absence of any commercial or financial relationships that could
be construed as a potential conflict of interest. Conflict of Interest Statement: The authors declare that the research was
conducted in the absence of any commercial or financial relationships that could
be construed as a potential conflict of interest. Van Vierssen, W. (1982b). The ecology of communities dominated by Zannichellia
taxa in Western Europe. II. Distribution, synecology and productivity
aspects in relation to environmental factors. Aquat. Bot. 13, 385–483. doi: 10.1016/0304-3770(82)90073-0 The handling Editor is currently co-organizing a Research Topic with one of
the authors EB, and confirms the absence of any other collaboration. The handling Editor is currently co-organizing a Research Topic with one of
the authors EB, and confirms the absence of any other collaboration. Van Vierssen, W. (1982c). Some notes on the germination of seeds of
Najas marina L. Aquat. Bot. 12, 201–203. doi: 10.1016/0304-3770(82)
90015-8 Copyright © 2018 Hilt, Alirangues Nuñez, Bakker, Blindow, Davidson, Gillefalk,
Hansson, Janse, Janssen, Jeppesen, Kabus, Kelly, Köhler, Lauridsen, Mooij,
Noordhuis, Phillips, Rücker, Schuster, Søndergaard, Teurlincx, van de Weyer, van
Donk, Waterstraat, Willby and Sayer. Frontiers in Plant Science | www.frontiersin.org Copyright © 2018 Hilt, Alirangues Nuñez, Bakker, Blindow, Davidson, Gillefalk,
Hansson, Janse, Janssen, Jeppesen, Kabus, Kelly, Köhler, Lauridsen, Mooij,
Noordhuis, Phillips, Rücker, Schuster, Søndergaard, Teurlincx, van de Weyer, van
Donk, Waterstraat, Willby and Sayer. This is an open-access article distributed
under the terms of the Creative Commons Attribution License (CC BY). The use,
distribution or reproduction in other forums is permitted, provided the original
author(s) and the copyright owner are credited and that the original publication
in this journal is cited, in accordance with accepted academic practice. No use,
distribution or reproduction is permitted which does not comply with these terms. February 2018 | Volume 9 | Article 194 REFERENCES This is an open-access article distributed
under the terms of the Creative Commons Attribution License (CC BY). The use,
distribution or reproduction in other forums is permitted, provided the original
author(s) and the copyright owner are credited and that the original publication
in this journal is cited, in accordance with accepted academic practice. No use,
distribution or reproduction is permitted which does not comply with these terms. Van Wijk, R. J. (1988). Ecological studies on Potamogeton pectinatus L. I. General
characteristics, biomass production and life cycles under field conditions. Aquat. Bot. 31, 211–258. doi: 10.1016/0304-3770(88)90015-0 Van Wijk, R. J. (1989). Ecological studies on Potamogeton pectinatus L. 3. Reproductive strategies and germination ecology. Aquat. Bot. 33, 271–299. doi: 10.1016/0304-3770(89)90042-9 Vari, A., and Toth, V. R. (2017). Quantifying macrophyte colonisation strategies -
A field experiment in a shallow lake (Lake Balaton, Hungary). Aquat. Bot. 136,
56–60. doi: 10.1016/j.aquabot.2016.09.006 February 2018 | Volume 9 | Article 194 Frontiers in Plant Science | www.frontiersin.org 24
|
W2976064566.txt
|
https://europepmc.org/articles/pmc6855679?pdf=render
|
en
|
Diversity of Ocular Dominance Patterns in Visual Cortex Originates from Variations in Local Cortical Retinotopy
|
The journal of neuroscience/The Journal of neuroscience
| 2,019
|
public-domain
| 13,133
|
The Journal of Neuroscience, November 13, 2019 • 39(46):9145–9163 • 9145
Systems/Circuits
Diversity of Ocular Dominance Patterns in Visual Cortex
Originates from Variations in Local Cortical Retinotopy
Sohrab Najafian, Jianzhong Jin, and X Jose-Manuel Alonso
Department of Biological and Visual Sciences, State University of New York College of Optometry, New York, New York 10036
The primary visual cortex contains a detailed map of retinal stimulus position (retinotopic map) and eye input (ocular dominance map)
that results from the precise arrangement of thalamic afferents during cortical development. For reasons that remain unclear, the
patterns of ocular dominance are very diverse across species and can take the shape of highly organized stripes, convoluted beads, or no
pattern at all. Here, we use a new image-processing algorithm to measure ocular dominance patterns more accurately than in the past. We
use these measurements to demonstrate that ocular dominance maps follow a common organizing principle that makes the cortical axis
with the slowest retinotopic gradient orthogonal to the ocular dominance stripes. We demonstrate this relation in multiple regions of the
primary visual cortex from individual animals, and different species. Moreover, consistent with the increase in the retinotopic gradient
with visual eccentricity, we demonstrate a strong correlation between eccentricity and ocular dominance stripe width. We also show that
an eye/polarity grid emerges within the visual cortical map when the representation of light and dark stimuli segregates along an axis
orthogonal to the ocular dominance stripes, as recently demonstrated in cats. Based on these results, we propose a developmental model
of visual cortical topography that sorts thalamic afferents by eye input and stimulus polarity, and then maximizes the binocular retinotopic match needed for depth perception and the light-dark retinotopic mismatch needed to process stimulus orientation. In this model,
the different ocular dominance patterns simply emerge from differences in local retinotopic cortical topography.
Key words: ocular dominance; plasticity; receptive field; thalamocortical; thalamus; visual cortex
Significance Statement
Thalamocortical afferents segregate in primary visual cortex by eye input and light-dark polarity. This afferent segregation forms
cortical patterns that vary greatly across species for reasons that remain unknown. Here we show that the formation of ocular
dominance patterns follows a common organizing principle across species that aligns the cortical axis of ocular dominance
segregation with the axis of slowest retinotopic gradient. Based on our results, we propose a model of visual cortical topography
that sorts thalamic afferents by eye input and stimulus polarity along orthogonal axes with the slowest and fastest retinotopic
gradients, respectively. This organization maximizes the binocular retinotopic match needed for depth perception and the lightdark retinotopic mismatch needed to process stimulus orientation in carnivores and primates.
Introduction
The human primary visual cortex needs to accommodate ⬃3
million afferents from the lateral geniculate nucleus of the thalamus (Selemon and Begovic, 2007) and sort them by their receptive field position within the retina. The afferent sorting along the
anteroposterior and mediolateral axes of the cortex creates a detailed map of retinal spatial position known as the retinotopic
Received May 17, 2019; revised Sept. 16, 2019; accepted Sept. 19, 2019.
Author contributions: S.N. and J.-M.A. designed research; S.N., J.J., and J.-M.A. performed research; S.N., J.J., and
J.-M.A. contributed unpublished reagents/analytic tools; S.N., J.J., and J.-M.A. analyzed data; S.N. and J.-M.A. wrote
the first draft of the paper; S.N., J.J., and J.-M.A. edited the paper; S.N. and J.-M.A. wrote the paper.
This work was supported by National Institutes of Health Grant EY027361.
The authors declare no competing financial interests.
Correspondence should be addressed to Jose-Manuel Alonso at jalonso@sunyopt.edu.
https://doi.org/10.1523/JNEUROSCI.1151-19.2019
Copyright © 2019 the authors
map (Daniel and Whitteridge, 1961). In species with high visual
acuity and large brains, such as humans and macaques, the retinotopic map splits into two intercalated copies for each eye forming an ocular dominance map with a zebra pattern (LeVay et al.,
1975; Hubel and Wiesel, 1977). In addition, evidence from cat
visual cortex (Kremkow et al., 2016) indicates that the retinotopic
map for each eye also splits into two copies for each contrast
polarity (light and dark), forming an eye/polarity grid (Kremkow
and Alonso, 2018).
The ocular dominance pattern of primary visual cortex varies
across species (LeVay et al., 1985; Anderson et al., 1988; Spatz,
1989; Adams and Horton, 2003; Adams et al., 2007; Takahata et
al., 2014), within the same species (Adams and Horton, 2003,
2006), and across local cortical regions of the same individual
animal (Adams et al., 2007). Although this pronounced variabil-
Najafian et al. • Diversity of Ocular Dominance Patterns
9146 • J. Neurosci., November 13, 2019 • 39(46):9145–9163
ity appears capricious (Adams and Horton, 2003), some ocular
dominance maps show consistent relations with cortical retinotopy, suggesting a common organizing principle. For example, in
humans and macaques, ocular dominance stripes run orthogonal
to the cortical border between areas V1 and V2, which represents
the retinotopy of the visual vertical meridian (Hubel and Wiesel,
1974a; Tootell et al., 1988; Blasdel and Campbell, 2001; Adams et
al., 2007). Because cortical retinotopy changes slower along than
across the V1/V2 border (Blasdel and Campbell, 2001), the ocular
dominance stripes also run orthogonal to the cortical axis with
the slowest retinotopy gradient.
Hubel and Wiesel interpreted the organization of ocular dominance stripes at the V1/V2 border as a cortical need to accommodate thalamic inputs sampling two copies of the same image
with two frontal eyes (LeVay et al., 1975; Hubel and Wiesel,
1977). In the ice-cube model of Hubel and Wiesel, each image
copy needs a square patch of cortex, and two image copies form a
rectangle with a length/width ratio of two. Under this interpretation, the longest axis of the ocular dominance stripes should run
orthogonal to the longest side of the retinotopic rectangle, which
is the V1/V2 border. Although this reasoning is compelling, the
relation between retinotopy and ocular dominance remained unclear. First, the length/width retinotopy ratio of the ocular dominance stripes is considerably ⬍2 and closer to 1.4 (Tootell et al.,
1982, 1988; Van Essen et al., 1984; Blasdel and Campbell, 2001).
Second, the relation between the ocular dominance domains and
the V1/V2 border is weak or absent in many species, including
cats and squirrel monkeys (Anderson et al., 1988; Horton and
Hocking, 1996). Third, ocular dominance maps are absent in
many animals with extensive binocular cortices, including some
squirrel monkeys (Adams and Horton, 2003, 2006), and adult
marmosets that did not experience monocular deprivation (DeBruyn and Casagrande, 1981; Spatz, 1989). Finally, the ocular
dominance map of an individual human can have diverse local
patterns with some regions resembling the organized stripes of
macaques and others the more disorganized stripes of squirrel
monkeys (Adams et al., 2007). There is currently no explanation
for this puzzling variability of ocular dominance patterns. Here,
we reveal a strong relationship between local cortical retinotopy
and ocular dominance in different species and different regions
of the same cortical map. Based on these results, we propose a
cortical model that generates variations in ocular dominance patterns from variations in local cortical retinotopy. This model has
implications for cortical sampling in binocular vision and predicts systematic changes in eye dominance across the horizontal
axis of the visual field.
Materials and Methods
We used an image-processing algorithm to measure the image parameters that distinguish different ocular dominance patterns in nature.
We then used a difference-of-Gaussians filter to simulate the patterns
published in the scientific literature and replicate the measurements
obtained with the image-processing algorithm. The software is publicly
available on Github (https://github.com/LabAlonsoSunyOptometry/
NajafianODSoftware; RRID:SCR_017381).
Measuring ocular dominance patterns with an image-processing algorithm. The image-processing algorithm can be applied to published ocular dominance maps from different species of either sex (the sex is not
mentioned if it is not reported in the original publication). It starts by
selecting and counting all individual stripes within the pattern, each
stripe being defined as a group of white pixels surrounded by black pixels
or vice versa. Then, the algorithm uses the MATLAB “bwmorph” function to shrink the stripe to a minimally connected central line and measures the stripe length by counting the pixels within this central line. The
algorithm measures the average stripe width as the average length of
multiple lines orthogonal to the central line and connecting the two
stripe borders. It measures the average angle of the stripe with two different methods: pixel-by-pixel method and Fourier method. In the pixelby-pixel method, it measures the angle of a line crossing two neighboring
pixels within the central line of the stripe. In the Fourier method, it
measures the dominant angle of a cortical patch with Fourier image
analysis. We use the pixel-by-pixel method to describe the angle for each
pixel and the Fourier method to describe the dominant angle of a cortical
patch.
Modeling the afferent-sorting filter (ASF). We model a family of ASFs to
accurate replicate the different ocular dominance patterns published in
the scientific literature. The ASFs are inspired by the work of Swindale
(1980) and simulate the sorting of thalamic afferents in visual cortex
during cortical development. We first measure the parameters of published ocular dominance patterns with the image-processing algorithm
and then use these parameters to search for the ASF that best describes
each published ocular dominance pattern. We model ASFs with a difference of two multivariate Gaussian functions, as illustrated in Equations
1–3 as follows:
1
f 共 x,y, X , Y 兲 ⫽
冉冉
⫺
⫽
2 X X 冑1 ⫺ 2
exp
冊冋
1
2共1 ⫺ 2 兲
册冊
共 x ⫺ X 兲2 共 y ⫺ Y 兲2 2共 x ⫺ X 兲共 y ⫺ Y 兲
⫹
⫺
X Y
2X
2Y
X䡠Y
兩X兩 兩Y兩
ASF 共 x,y, c , sx , sy 兲 ⫽ f center 共 x,y, c 兲 ⫺ f surround 共 x,y, sx , sy 兲
(1)
(2)
(3)
Where x and y (lowercase) are positions in cortical space along the mediolateral (x) and anteroposterior (y) axis. X and Y (uppercase) are vectors containing a subset of x (X vector) and y (Y vector) positions. x and
y are the standard deviations (SDs) of the X and Y vectors, and x and y
are the central positions. is the angle of the filter (Eq. 2), which is
calculated as a dot product of the X and Y vectors divided by their magnitudes (i.e., the square root of the sum of the squared elements of each
vector). The ASF results from the difference of two multivariate Gaussian
functions representing the filter center ( fcenter) and surround ( fsurround).
c is the SD of the filter center. Because the filter center is always circular,
the SDs in the x and y coordinates are equal and can be described with a
single variable that is c. sx and sy are the SDs of the filter surround for
the x and y coordinates. Unlike the filter center, the filter surround can be
elliptical and needs to be described with two different SDs for the x and y
coordinates, sx and sy. We experimented with different filter structures
to find the simplest one that could accurately replicate published ocular
dominance patters. Filters without surround allowed us to simulate ocular dominance segregation but could not simulate the different shapes
of ocular dominance bands found in nature. Filters with elongated centers generated similar ocular dominance patterns to filters with circular
centers; therefore, we used the simplest center shape, which is a circle.
Modeling ocular dominance segregation with ASFs. We simulate ocular
dominance patterns in visual cortex by convolving the ASF with a 2D
cortical patch of white and black pixels ( C), each pixel representing an
afferent from the contralateral eye (white pixel with a value of 1) or the
ipsilateral eye (black pixel with a value of 0). In the convolution between
the ASF and the cortical patch (Eq. 4), the multivariate Gaussian functions give a different weight to each afferent depending on its cortical
distance from the filter center. The center-surround difference of the
multivariate Gaussian functions simulates the attraction and repulsion
interactions between afferents mediated by molecular gradients and/or
neuronal activity. The convolution sum (CS) determines whether the
afferent at the center of the filter is replaced by an afferent of a different
type (r ⫽ 1) or not (r ⫽ 0), as shown in Equation 5. A replacement
simulates the retraction of the axon growth cone from one afferent type
and replacement by another afferent type. The central afferent remains
unchanged when surrounded by afferents of the same type (i.e., positive
CS when the central afferent is from the contralateral eye; negative CS
Najafian et al. • Diversity of Ocular Dominance Patterns
J. Neurosci., November 13, 2019 • 39(46):9145–9163 • 9147
when the central afferent is from the ipsilateral eye). The central afferent
is replaced by an afferent of different type when surrounded by afferents
of different type (e.g., negative CS when the central afferent is from the
contralateral eye; positive CS when the central afferent is from the ipsilateral eye) as follows:
CS 共 x,y,s 兲 ⫽
冘
ASF 共 x,y, c , sx , sy 兲 䊟 C 共 x,y,s 兲
(4)
x,y
C 共 x,y,s ⫹ 1 兲 ⫽ 共 C 共 x,y,s 兲 ⫺ r 兲 2
再
r ⫽ 1, if C 共 x,y,s 兲 ⫽ 1 and CS 共 x,y,s 兲 ⬍ 0
if C 共 x,y,s 兲 ⫽ 0 and CS 共 x,y,s 兲 ⬎ 0
r ⫽ 0, otherwise
冎
(5)
The convolution is performed across all cortical positions (x, y) and is
repeated for 10 developmental steps (s). In the first step (s ⫽ 1), the
cortical patch has randomly organized afferents from the contralateral
and ipsilateral eyes. In the following developmental steps (s ⬎ 1), the
cortical patch has afferents sorted at the previous developmental step
(Eq. 5). After 10 developmental steps, the arrangement of the afferents
becomes stable and does no longer change. We measure the amount of
change by calculating a pattern of similarity (PS) between cortical
patches from contiguous developmental steps (Eq. 6). PS equals 1 ⫺
MSE, where MSE is the mean squared error difference between cortical
patches from contiguous developmental steps, C (x, y, s ⫺ 1) and C (x, y,
s). Each patch has n pixels that simulate different x, y positions in cortical
space. If the two contiguous patches are identical, MSE equals 0 and PS
reaches its maximum value, which is 1.
PS ⫽ 1 ⫺ MSE
MSE 共 s 兲 ⫽
1
n
冘
n
x⫽1,
y⫽1
共 C 共 x,y,s 兲 ⫺ C 共 x,y,s ⫺ 1 兲兲 2
(6)
Generating an image database of diverse ocular dominance patterns. We
simulate different ocular dominance patterns by using ASFs that differ in
just four parameters: center diameter, ratio between surround and center
(surround/center ratio), ratio between longest and shortest surround
axes (x-surround/y-surround ratio), and angle of the longest surround
axis. The center diameter is defined as 2 c in Equation 3. The
surround/center ratio is defined as the ratio between the SDs of
the shortest surround axis (minimum [sx sy]) and the center (c).
The x-surround/y-surround ratio is defined as the ratio between sx
and sy. We calculate the angle of the longest surround axis as the
angle of the ASF (). These four parameters are the minimum number
needed to reproduce the diversity of ocular dominance patterns found in
nature and provide a reasonable compromise between accuracy and speed
when searching for a specific ocular dominance pattern.
By using multiple combinations of these parameters, we generated
3000 different ocular dominance patterns with an image resolution of
31 ⫻ 31 pixels. We used six center diameters (2 c: 6, 8, 10, 12, 14, 16), five
surround/center ratios (minimum [sx sy]/c: 1, 2, 3, 4, 5), 10
x-surround/y-surround ratios (fsx/fsy: 1, 2, 3, 4, 5, 6, 7, 8, 9, 10), and 35
angles (: 0 –175 in steps of 5 degrees). We first generated 300 different
patterns by using all possible combinations of center diameter, surround/center ratio, and x-surround/y-surround ratio (6 ⫻ 5 ⫻ 10 ⫽
300). We then repeated each of these 300 parameter combinations 10
times with different random noise seeds to generate 3000 different patterns (300 ⫻ 10 ⫽ 3000 ocular dominance patterns). We then rotated
each pattern to match the angle of the cortical patch measured with
Fourier analysis. Therefore, when including the 35 possible pattern angles, the number of possible patterns that we generate with the four filter
parameters is 105,000 (3000 ⫻ 35).
To simulate a specific ocular dominance pattern measured in primary
visual cortex, we first select an image of an ocular dominance map published in the scientific literature. We then convert the map to a binary
scale of black and white pixels and divide it in multiple patches, each
patch containing an average of three ocular dominance cycles (six stripes:
three from the contralateral eye and three from the ipsilateral eye). We
then resize each patch to 31 ⫻ 31 pixels to match the image resolution in
our database. This resizing provides a resolution for each cortical patch of
4.78 pixels/mm for human, 10.31 pixels/mm for macaque, and 8.85
pixels/mm for cat (patch size: 6.48 mm for human, 3.01 mm for macaque,
and 3.5 mm for cat). We then search in our database for the pattern that best
matches the original published image. The search comparison is guided by
measures of four different parameters, all from ipsilateral-eye stripes: number, average length, average width, and average angle.
Searching for the best pattern match in our image database. We use a cost
function to search for the best match of a published ocular dominance
pattern in our image database. The cost function measures the error ( E)
between the original image and each image from our database (Eq. 7).
The error is calculated as the square difference between the average width
( W), length ( L), and number ( N) of ipsilateral stripes from the original
(Wo, Lo, No) and model images (Wm, Lm, Nm). We chose stripes from
the ipsilateral eye because they are more numerous than stripes from the
contralateral eye (see Results).
E ⫽ 共 Wo ⫺ Wm 兲 2 ⫹ 共 Lo ⫺ Lm 兲 2 ⫹ 共 No ⫺ Nm 兲 2
(7)
The angle is not included in the cost function because patterns with different
angles can be simply replicated by rotating the ASF after finding the best
match for number, width, and length. We obtained 3000 different values of
E, one for each image in the database. After finding the best image match for
ipsilateral stripe width, length, and number (i.e., image with smallest E), we
rotated the angle of the image with 5 degree precision to match the original
and select the ASF that generated the best match. We repeated this search
process for each cortical patch to generate a full set of ASFs for each ocular
dominance map published in the scientific literature.
Identifying ASFs associated with cortical retinotopic sections. We
searched for the ASFs that best described specific retinotopic sections
within an ocular dominance map. For this search, we first redrew the
retinotopic maps from human (Horton and Hoyt, 1991), macaque (Van
Essen et al., 1984), and cat (Tusa et al., 1978) published in the scientific
literature. We then stretched and rotated the retinotopic maps to match
the global shapes of the published ocular dominance maps. This process
was relatively simple when using human or macaque brains because the
general shapes of published ocular dominance and retinotopic maps are
very similar in these species. The process was more difficult (and more
prone to error) in the cat because the global shapes of published ocular
dominance and retinotopic maps are quite different (Tusa et al., 1978;
Kaschube et al., 2003). Therefore, to reduce the error, we restricted the
cat map comparison to a central 30 ⫻ 30 degrees patch. After finishing
the map alignment, we divided the ocular dominance map in retinotopic
sections and measured the retinotopic gradient asymmetries in each section. We then searched in our image database for the best match of the
ocular dominance pattern from each retinotopic section and selected the
ASF that generated it.
Simulations of ocular dominance and ON-OFF afferent segregation in
visual cortex. Based on our experimental data in cats (Kremkow et al.,
2016; Kremkow and Alonso, 2018), we propose a developmental model
in which cortical retinotopy guides the segregation of thalamic afferents
by eye input and contrast polarity (ON and OFF) along orthogonal cortical axes. In the model, thalamic afferents from contralateral and ipsilateral eyes segregate along the cortical axis with slowest retinotopic
gradient to maximize the binocular retinotopic match between ocular
dominance stripes. In turn, ON and OFF afferents segregate along the
cortical axis with fastest retinotopic gradient to maximize the ON-OFF
retinotopic mismatch between ON and OFF domains. The model allows
us to simulate the map of ON-OFF cortical segregation if we know the
map of ocular dominance segregation. We first take an ocular dominance
map published in the scientific literature to obtain the full set of ASFs
associated with ocular dominance segregation. Then, we rotate all filters
by 90 degrees to obtain the ASFs for ON-OFF cortical segregation. Then,
we use the ON-OFF ASFs to segregate ON and OFF thalamic afferents
with an algorithm similar to that used for ocular dominance segregation.
To generate the ON-OFF cortical map, we first convolve the ASFs with
a 2D cortical patch of white and black pixels ( C), as shown in Equation 4,
Najafian et al. • Diversity of Ocular Dominance Patterns
9148 • J. Neurosci., November 13, 2019 • 39(46):9145–9163
Figure 1. Image-processing algorithm used to quantify ocular dominance patterns in visual cortex. a, Description of the image processing algorithm used to measure ocular dominance patterns.
The algorithm is illustrated for zebra skin (first row) and leopard skin (second row). Both images are from public domain (Skin of a Grant’s Zebra from Hans Hillewaert and Leopard Skin from Alex
Borland). The algorithm includes the following stages, described from left to right: selection of single feature, tracing the central line of the feature with 1 pixel width, measuring the feature length
(686 pixels long), angle of the central line (average across pixels: 38 degrees), and width of the central line (average width: 69 pixels). b, Binary versions of the images shown in a (top) and histograms
illustrating distributions of length, angle, width (normalized by total number of pixels), and total area of white and black image features (bottom). The numbers on top of the histograms show
average (length, width) and normalized area for white and black features.
each pixel representing an ON afferent (white pixel with a value of 1) or
OFF afferent (black pixel with a value of 0). As for ocular dominance
segregation, the CS determines whether the afferent at the filter center is
replaced by an afferent of a different type (r ⫽ 1) or not (r ⫽ 0). However,
because the segregation for ON-OFF afferents is likely to occur later in
time than the segregation for ocular dominance (Kremkow and Alonso,
2018), we assume that is also weaker. To simulate a weak ON-OFF afferent segregation, we add a randomization factor to the segregation algorithm described in Equation 5 (Eq. 8). An ON afferent at the center of the
filter is replaced by an OFF afferent if the CS plus a random factor ranging
from 0 to 1 (rnd) is ⬎0.5. An OFF afferent is replaced by an ON afferent
if the CS plus the random factor is ⬍0.5. The CS is multiplied by a
constant factor (␣) to simulate cortices with different ON/OFF afferent
balances. When ON and OFF afferents are balanced in number (50%
OFF, 50% ON), ␣ equals 1. When OFF afferents dominate over ON
afferents (60% OFF, 40% ON), ␣ equals 10, as follows.
C 共 x,y,s ⫹ 1 兲 ⫽ 共 C 共 x,y,s 兲 ⫺ r 兲 2
再
r ⫽ 1, if C 共 x,y,s 兲 ⫽ 1 and 共 rnd ⫹ ␣ CS 共 x,y,s 兲兲 ⬎ 0.5
if C 共 x,y,s 兲 ⫽ 0 and 共 rnd ⫹ ␣ CS 共 x,y,s 兲兲 ⬍ 0.5
r ⫽ 0, otherwise
冎
(8)
The strength of ON-OFF segregation was measured by generating ONOFF maps in 101 consecutive developmental steps. We measured the
Najafian et al. • Diversity of Ocular Dominance Patterns
J. Neurosci., November 13, 2019 • 39(46):9145–9163 • 9149
Figure 2. Quantification of ocular dominance patterns in human, macaque, and cat cortex. a, Human ocular dominance map shown in a binary image. Reproduced with permission from Adams
et al. (2007). Unless specified differently, all ocular dominance maps and plots within the paper show the contralateral eye in white and ipsilateral eye in black. b, Histograms illustrating distributions
of length, angle, width, and total area of white and black image features. Contralateral-eye stripes (white) tend to be longer than ipsilateral-eye stripes (black). c, Same as in a, but for a male
macaque. Reproduced with permission from Adams et al. (2007). d, Same as in b, but for a male macaque. e, Same as in a, but for a cat. Reproduced with permission from Kaschube et al. (2003).
f, Same as in b, but for a cat. g–l, Comparison of ocular dominance stripes from contralateral- (x axis) and ipsilateral-eye stripes (y axis), shown for human (n ⫽ 3), macaque (n ⫽ 7; 4 males and 3
females), and cat cortical hemispheres (n ⫽ 1).
segregation strength at each developmental step (Ss) as the MSE between the ON-OFF map at each step and developmental Step 100. We
measured the minimum possible segregation (Smin) as the MSE (Eq. 6)
between the ON-OFF maps at developmental Steps 0 and 100. The
strength of ON-OFF segregation was defined as Ss-Smin and ranged
from a value of 1 (strongest segregation) to 0 (no segregation).
Data analysis. Statistical significance was assessed with Wilcoxon tests
when comparing value averages and with linear correlation when comparing the relation between two variables. All average comparisons are
described as average ⫾ SD. Probability values are provided for all statistical comparisons, and probability values ⬍0.0001 are described as p ⬍
0.0001.
9150 • J. Neurosci., November 13, 2019 • 39(46):9145–9163
Najafian et al. • Diversity of Ocular Dominance Patterns
Results
Thalamic afferents segregate by eye input in the primary visual
cortex of humans, macaques, and carnivores. The cortical patterns formed by this afferent segregation are very diverse across
species and cortical regions for reasons that remain poorly understood. To investigate the origin of this pattern diversity and its
possible functional implications, we measured the statistics of
ocular dominance segregation with two new methods: an imageprocessing algorithm and a filter that simulates each local ocular
dominance pattern.
Diversity of ocular dominance maps measured with image
processing analysis
We used an image-processing algorithm to quantify the different
ocular dominance patterns found in nature. The algorithm works
with any image that can be converted to a binary pattern of black
and white pixels (Fig. 1a). It starts by isolating a feature within the
pattern: an area of pixels with the same contrast polarity (e.g.,
white pixels) surrounded by pixels of opposite polarity (e.g.,
black pixels). Then, it traces the central line of the feature to
measure its length. Then, it traces multiple lines orthogonal to the
central line to measure the width and angle of the feature at each
pixel of the central line (Fig. 1a). Finally, the algorithm returns
histogram distributions of length, angle, width, and total area of
white and black features (Fig. 1b). The histograms show one measure of length per feature (e.g., 8 length measures for the white
stripes of Fig. 1b) and one measure of angle or width per pixel
(e.g., 4363 angle measures for the white stripes of Fig. 1b).
We started applying this algorithm to three ocular dominance
maps from three different species published in the scientific literature: human (Adams et al., 2007), macaque (Adams et al.,
2007), and cat (Kaschube et al., 2003). Before taking any measurements, we removed all map regions that had weak or missing
ocular dominance patterns, such as the monocular crescent, optic disk, and regions not properly reconstructed (Fig. 2a,c,e). We
then divided each ocular dominance map into multiple square
patches, each patch covering three contralateral-ipsilateral pairs
of ocular dominance stripes. To facilitate comparisons across
species, we aligned the longest axis of the map with the horizontal
axis of the image (Fig. 2a,c,e).
Our algorithm demonstrates that the contralateral eye occupies a larger cortical territory than the ipsilateral eye in the ocular
dominance maps of the three species. We demonstrate this contralateral dominance with measurements of overall cortical area
(Fig. 2a– g), as shown in the past (Shatz and Stryker, 1978; Anderson et al., 1988; Adams et al., 2007). In addition, we demonstrate an even stronger contralateral bias with measurements of
stripe length and number (Fig. 2a–l ). The stripes from the contralateral eye connected more frequently with each other than the
stripes from the ipsilateral eye. Indeed, many cortical maps had
nearly all contralateral-eye stripes connected forming a labyrinth
that created multiple isolated ipsilateral-eye stripes (Movie 1).
The average stripe length in cortical patches with three
contralateral-ipsilateral stripe cycles was 1.6 –3.9 times greater for
the contralateral eye than ipsilateral eye (Fig. 2h; human: 6.4 ⫾
5.96 vs 4.13 ⫾ 3.73 mm, p ⬍ 0.0001, n ⫽ 3 ocular dominance
maps; macaque: 4.88 ⫾ 4.95 vs 1.75 ⫾ 1.14 mm, p ⬍ 0.0001, n ⫽
7 ocular dominance maps; cat: 6.73 ⫾ 4.85 vs 1.74 ⫾ 1.01 mm,
p ⬍ 0.0001, n ⫽ 1 ocular dominance map, Wilcoxon tests). The
average stripe length in the entire cortical map was also 2.3–10.5
times greater for the contralateral than the ipsilateral eye (human:
19.03 ⫾ 143.77 vs 8.14 ⫾ 54.17 mm, p ⫽ 0.0085, n ⫽ 3 ocular
dominance maps; macaque: 18.02 ⫾ 112.44 vs 2.55 ⫾ 9.73 mm,
Movie 1. Analysis of single stripes from contralateral (white) and
ipsilateral eyes (black) in human cortex. The red outline shows the selected stripe. The top label shows the stripe number and length. The
bottom panel shows histograms with the distribution of stripe width
(left) and angle (right) measured at each image pixel and the mean
stripe width and angle (top label).
p ⬍ 0.0001, n ⫽ 7 ocular dominance maps; cat: 33.5 ⫾ 168.5 vs
3.2 ⫾ 5.78 mm, p ⬍ 0.0001, n ⫽ 1 ocular dominance map, Wilcoxon tests). Because the contralateral stripes were longer and
more connected with each other than the ipsilateral stripes, they
were also 2–7 times fewer in number (Fig. 2j). In addition to the
pronounced contralateral bias for stripe length, the average stripe
width was also slightly greater for the contralateral than ipsilateral
eyes in most cortical maps (Fig. 2k; but see exception in Fig. 2b;
0.806 ⫾ 0.449 vs 0.817 ⫾ 0.353 mm, p ⬍ 0.0001, Wilcoxon test).
Finally, many ocular dominance stripes showed a strong tendency to
be orthogonal to the longest axis of area V1 (Fig. 2a–f,l).
Correlation between stripe width and eccentricity
We hypothesize that the orthogonal relationship between the
longest axes of the ocular dominance stripes and the longest axis
of area V1 is a consequence of asymmetries in cortical retinotopy.
Because the V1 area of humans and macaques is more elongated
along its horizontal than vertical dimension, the two eye copies of
each retinotopic position should be better accommodated along
the horizontal cortical dimension. Moreover, because cortical
retinotopy changes faster as visual eccentricity increases, the
amount of cortical space needed to accommodate an ipsilateralcontralateral stripe-cycle should decrease with eccentricity, affecting stripe width. Consistent with this prediction, we found
that the stripes from the ipsilateral eye become increasingly thinner as azimuth eccentricity increases, even within central vision
(Fig. 3). This eccentricity thinning could be demonstrated in both
macaques (Fig. 3a–f ) and humans (Fig. 3g– h), and was very robust, even within the central 16 degrees (Fig. 3a– h). Contrary to
the ipsilateral-eye stripes, the width of the contralateral-eye
stripes increased with azimuth eccentricity (Fig. 3a–f,h) or did
not change (Fig. 3g). The relation between eccentricity and
ipsilateral-eye stripe width could be accurately fit with a power function, and the power values across different hemispheres were tightly
constrained within 15% of the mean range (between ⫺0.12 and
⫺0.16 in macaque maps and ⫺0.06 and ⫺0.08 in human maps).
Najafian et al. • Diversity of Ocular Dominance Patterns
J. Neurosci., November 13, 2019 • 39(46):9145–9163 • 9151
Figure 3. Correlation between cortical stripe width and eccentricity. a, Cortical map of ocular dominance from a male macaque (left) and scatter plot showing a reduction in the width of
ipsilateral-eye stripes (ipsi, in black) with eccentricity. The contralateral-eye stripes (contra, in white) tend to increase with eccentricity but the width change is not as pronounced as for the
ipsilateral-eye stripes. Lines indicate fits to the power functions (equations shown in white for contralateral eye and black for ipsilateral eye). Scale bars: human, 10 mm; macaque, 5 mm. Ocular
dominance map reproduced with permission from Horton and Hocking (1996). b–f, Same as in a, but for other male (b– d) and female (e, f ) macaques. Reproduced with permission from Horton
and Hocking (1996). g, h, Same for other human ocular dominance maps. Reproduced with permission from Adams et al. (2007).
The increase with eccentricity of contralateral-ipsilateral stripe width
differences matches the increase of nasal-temporal cell-density differences within the retina (Curcio and Allen, 1990). The nasal and
temporal retina feed the contralateral- and ipsilateral-eye stripes,
respectively. Therefore, the relation between nasal-temporal retinal
differences and contralateral-ipsilateral cortical differences supports
our hypothesis that cortical retinotopy shapes ocular dominance
segregation.
our prediction, nearly all measures of ocular dominance patterns
were correlated between hemispheres, including the stripe length
and width (Fig. 4c,d), stripe number (Fig. 4e), and stripe angle
(Fig. 4f ). All correlations were highly significant, except the correlations for stripe width in Macaques 2 and 3 (Fig. 4d), which
were constrained to a very narrow range of values. The correlations for stripe angle were also weaker but significant for all 3
macaques.
Similar ocular dominance patterns between left and
right hemispheres
If V1 retinotopy shapes ocular dominance segregation, the overall pattern of ocular dominance should be similar between left
and right hemispheres. Because the cell density of the two nasal
retinas is similar (Curcio and Allen, 1990), the organization of the
contralateral-eye stripes from the two hemispheres should also be
similar (the same argument applies to the two temporal retinas
and ipsilateral-eye stripes). At the same time, because the number
of retinal ganglion cells can vary by more than twofold across
individuals (Curcio and Allen, 1990), the pattern of ocular dominance should also show great individual variability. To test this
hypothesis, we compared the ocular dominance patterns of the
two hemispheres from 3 macaques and 1 human (Fig. 4a) published in the scientific literature (Horton and Hocking, 1996;
Adams et al., 2007). To quantify the comparison, we divided each
hemisphere in multiple eccentricity sections, and calculated the
different parameters for each section (Fig. 4a,b). Consistent with
Similar ocular dominance patterns in upper and lower
V1 halves
If cortical retinotopy shapes ocular dominance segregation, the
overall pattern of ocular dominance should be also similar between the lower and upper halves of area V1 within each hemisphere. Similar ocular dominance patterns are expected because
the two V1 halves receive input from the same two eyes, and the
two eyes from the same individual animal have similar density of
retinal ganglion cells (Curcio and Allen, 1990). To test this hypothesis, we used the same approach described above. We divided each hemisphere in multiple eccentricity sections and
compared the different ocular dominance patterns from the
lower and upper V1 halves (upper and lower visual field fields)
across cortical sections with similar azimuth eccentricity (Fig. 5).
As for left and right hemispheres, we found strong correlations
between the ocular dominance patterns of upper and lower V1
halves. Unlike for left and right hemispheres, most correlations
for stripe length did not reach significance (Fig. 5c). This lack of
9152 • J. Neurosci., November 13, 2019 • 39(46):9145–9163
Najafian et al. • Diversity of Ocular Dominance Patterns
Figure 4. Correlation between ocular dominance patterns from left and right V1 hemispheres. a, Maps of ocular dominance from the right hemisphere (RH) and left hemisphere (LH) obtained
from the scientific literature for a human (Human 1; reproduced with permission from Adams et al., 2007) and 3 macaques (male macaques: 1–2, female macaques: 3, reproduced with permission
from Horton and Hocking, 1996). Red lines indicate the retinotopic sectors from human and macaque retinotopic maps used to calculate the different ocular dominance parameters. Scale bars:
human, 10 mm; macaque, 5 mm. b, Correlations between LH and RH area for human 1 (top) and macaques 1–3 (bottom). Each circle represents a measurement from a different retinotopic sector
for the contralateral eye (white) and ipsilateral eye (black). c–f, Same as in b, but for stripe length (c), stripe width (d), stripe number (e), and stripe angle (f ).
significance for stripe length is likely to reflect a sample size limitation (4 –5 data points for each eye in Fig. 5 vs 8 –10 in Fig. 4)
since most correlations for stripe number were strong (Fig. 5e).
Diversity of ocular dominance maps measured with ASFs
To study the diversity of ocular dominance patterns in further
detail, we simulated the pattern of each local cortical patch published in the scientific literature with a difference of multivariate
Gaussian functions. We call this function afferent sorting filter or
ASF because it simulates the sorting of thalamic afferents from
contralateral and ipsilateral eyes during cortical development.
The filter size simulates the cortical region receiving afferents
with the same retinotopy (i.e., overlapping receptive fields), and
the surround elongation simulates local retinotopic cortical
asymmetries (e.g., slower change in retinotopy along the longer
than shorter surround axis of the filter). The size of the cortical
region receiving afferents with overlapping receptive fields is approximately constant across different eccentricities, as demonstrated by measurements of cortical regions representing the
same point in visual space (Hubel and Wiesel, 1974b; Albus,
1975; Harvey and Dumoulin, 2011). Therefore, the filter size
should remain constant across the entire cortex and independent
of changes in receptive field size and spatial frequency preference
with visual eccentricity. The ASF has a center-surround structure
(Fig. 6a), which can be circular (Fig. 6b) or elliptical (Fig. 6c). The
center diameter simulates the spread of attraction between afferents of the same type and determines the width of the ocular
dominance bands. The surround simulates the spread of repulsion between afferents of different type and its elongation determines the band shape (e.g., stripes or beads).
We simulate a patch of afferents in the cortical plate (i.e.,
cortex at earliest developmental stage) as a binary-noise image.
White pixels represent afferents from the contralateral eye, and
black pixels represent afferents from the ipsilateral eye (Fig. 6b,c).
The sorting process changes the polarity of the pixel at the center
of the filter when the polarity of the surrounding afferents is
different (e.g., change from white to black when the average convolution is negative). This process aims to simulate the retraction/expansion of axonal growth cones from contralateral- and
ipsilateral-eye afferents in the cortical plate (Rakic, 1977; Shatz
and Stryker, 1978; Huberman et al., 2008). The sorting of thalamic afferents is simulated through sequential convolutions between the noise image representing the cortical patch and the ASF
(Fig. 7a). In the first convolution (developmental Step 1), we
convolve the filter with an image of random noise (i.e., randomly
arranged afferents from contralateral and ipsilateral eyes). In the
second convolution (developmental Step 2), we convolve again
the filter with the image resulting from the first convolution, and
we continue performing these sequential convolutions for 10 developmental steps. Ocular dominance patterns emerge at the first
Najafian et al. • Diversity of Ocular Dominance Patterns
J. Neurosci., November 13, 2019 • 39(46):9145–9163 • 9153
Figure 5. Correlation between ocular dominance patterns between the two V1 halves representing lower and upper visual fields. a, Maps of ocular dominance from the V1 halves representing
the lower visual field (LF) and upper visual fields (UF), obtained from the same brains shown in Figure 4. Scale bars: human, 10 mm; macaque, 5 mm. b, Correlations between UF and LF areas for
Human 1 and Macaques 1–3. Each circle represents a measurement from a different retinotopic sector for the contralateral eye (white) and ipsilateral eye (black). c–f, Same as in b, but for stripe
length (c), stripe width (d), stripe number (e), and stripe angle (f ).
9154 • J. Neurosci., November 13, 2019 • 39(46):9145–9163
Najafian et al. • Diversity of Ocular Dominance Patterns
Figure 6. The ASF. a, ASFs are modeled as a difference of two multivariate Gaussian functions. The filter has a positive center (orange) and a negative surround (purple). b, c, Top, View of circular
(b) and elliptical (c) filters superimposed on cortical space. The filter sorts contralateral (white pixels) and ipsilateral afferents (black pixels) over the mediolateral (ML) and anteroposterior (AP) axes
of cortical space.
Figure 7. ASFs generate realistic ocular dominance patterns. a, The ASF is convolved with randomly organized afferents (Step 1), the result is convolved again (Step 2), and this process repeated
for 10 sequential convolutions (Steps 1–10). The value of the convolution determines whether the polarity of the afferent changes or remains the same in each developmental step. Circular filters
generate ocular dominance stripes with random orientations, and elliptical filters generate ocular dominance stripes oriented orthogonal to the longest axis of the filter. b, The ocular dominance
pattern becomes increasingly similar as the number of sequential convolutions increases.
convolution between the filter and afferent image and become
increasingly similar at subsequent convolutions (Fig. 7b). Because the patterns become nearly identical at developmental
Step 10 (Fig. 7b; pattern similarity ⬃ 1), we used 10 convolutions for all simulations (for more details, see Materials and
Methods). As illustrated in Figure 7, the shape of the filter
determines the shape of the ocular dominance pattern. Circular filters generate ocular dominance patterns with random
orientations (Fig. 7a, top row), whereas elliptical filters gen-
erate ocular dominance stripes orthogonal to the filter orientation (Fig. 7a, bottom three rows).
A large diversity of ocular dominance patterns can be generated by manipulating just four filter parameters: the center diameter, the ratio between surround and center diameter (surround/
center ratio), the ratio between the longest and shortest axes of
the surround (x-surround/y-surround ratio), and the angle of the
filter (Fig. 8a). Circular surrounds generate bead patterns (Fig.
8a; x-surround/y-surround ⫽ 1) and elliptical surrounds gener-
Najafian et al. • Diversity of Ocular Dominance Patterns
J. Neurosci., November 13, 2019 • 39(46):9145–9163 • 9155
Figure 8. Different filter parameters generate different ocular dominance patterns. a, Ocular dominance patterns generated by systematically changing two parameters of the ASF: the
X-surround/Y-surround ratio and the surround/center ratio. ASFs with a X-surround/Y-surround ratio of 1 generate beaded patterns that resemble the ocular dominance patterns of cats. ASFs with
surround/center ratios ⬎1 generate ocular dominance stripes resembling ocular dominance patterns of macaques and humans. Increasing the surround/center ratio or the X-surround/Y-surround
ratio makes the ocular dominance stripes and beads wider and longer (i.e., more frequently connected with each other). b, Changes in the length of ocular dominance stripes as a function of changes
in X-surround/Y-surround and surround/center ratios.
ate stripe patterns (Fig. 8a; x-surround/y-surround ⱖ 2). As we
increase the center diameter, the surround/center ratio, or the
x-surround/y-surround ratios, stripes and beads become wider,
start intersecting with each other, and increase their total length
(Fig. 8b). By systematically manipulating these four filter parameters, we generated a database of multiple ocular dominance patterns resembling those found in nature. We then used this
database to identify the filters associated with different local patterns of ocular dominance segregation in ocular dominance maps
published in the scientific literature.
We divided each published ocular dominance map in multiple patches, each patch containing three cycles of contralateral
and ipsilateral stripes. We then searched our image database for
the best match for each ocular dominance patch (i.e., similar
number of stripes, average length, width, and angle) and selected
the ASF that generated the best match. The selected filters generated patterns very similar to the originals, as verified by visual
inspection (Fig. 9a) and parameter correlations (Fig. 9b; n ⫽ 6
ocular dominance maps). Elongated filters reproduced stripe
patterns, circular filters reproduced beaded patterns, filters with
large center diameters reproduced patterns with wide features,
and those with small diameters reproduced patterns with thin
features (Fig. 9a). At the end of the search, we obtained a set of
local filters (Fig. 10a– c) with diverse sizes, elongation ratios, and
angles for each ocular dominance map. The average elongation
ratio of the filters was constrained between 2 and 3 in the three
species (Fig. 10d–f; long/short average filter axis: 3 in human, 2.3
in macaque, and 2.5 in cat). The longest axes of the filters also
showed a strong tendency to be parallel to the longest axis of area
V1 in humans and macaques but not in cats (Fig. 10g–j).
Ocular dominance bands segregate along the cortical axis
with the slowest local retinotopy gradient
The cortical representation of visual space is much more distorted in humans and macaques than cats. For example, a patch
of primary visual cortex representing the 10 ⫻ 10 central degrees
is an order of magnitude more elongated in humans and macaques. We hypothesize that the more distorted cortical represen-
9156 • J. Neurosci., November 13, 2019 • 39(46):9145–9163
Najafian et al. • Diversity of Ocular Dominance Patterns
Figure 9. Reproducing the ocular dominance pattern of each local cortical patch published in the scientific literature with ASFs. a, Ocular dominance patterns of cortical patches published in the
scientific literature (top) and selected best matches from a database (middle) of patterns generated with ASFs (bottom). b, Ocular dominance comparison between mean parameters of the original
cortical patches (x axis) and best matches obtained from our database (y axis), shown for mean length (left), mean width (middle), and number of features (right).
tation of visual space makes ocular dominance bands to segregate
along the longest axis of area V1. The longest axis helps to maximize the retinotopic match between adjacent stripes from contralateral and ipsilateral eyes. To test this hypothesis, we divided
each ocular dominance map into multiple retinotopy patches
drawn from retinotopic maps published in the scientific literature (Tusa et al., 1978; Van Essen et al., 1984; Horton and Hoyt,
1991). We then searched in our database for the filters that best
replicated the ocular dominance patterns associated with each
retinotopic patch. To reduce the measurement errors as much as
possible, we restricted our analysis to large retinotopic patches,
which included most of the cortex in humans and macaques and
a central patch of 30 ⫻ 30 degrees in cats (Fig. 11a,b).
Consistent with our hypothesis, the angle of the ASF associated with each ocular dominance patch (Fig. 11a, purple ellipses)
tended to be parallel to the axis of slowest retinotopy gradient
within the patch (Fig. 11b, dotted purple lines). The angle difference between the filter and gradient axes was constrained within
⬍20 degrees in example hemispheres from the three species (Fig.
11c; number of retinotopic sectors: 14 for human, 10 for macaque, and 17 for cat). The median angle difference was constrained to ⬍25 degrees across all macaque and human
hemispheres that we measured (macaques: 12.93 ⫾ 14.06, n ⫽
116 retinotopic sectors from 7 macaque hemispheres; humans:
22.7 ⫾ 22.61, n ⫽ 123 retinotopic sectors from 3 human hemispheres; p ⬍ 0.0001 that the angle difference is due to chance,
Wilcoxon tests). The filter angle is measured with Fourier analysis and is independent of the other filter parameters. Therefore,
the results from Figure 11 would be identical if the angle of each
retinotopic patch was calculated directly with Fourier analysis
without using ASFs.
The tight relation between ocular dominance segregation and
retinotopy that we demonstrate is particularly remarkable given
the potential large sources of measurement error. The measurements require aligning published retinotopic and ocular dominance maps that have different shapes, are not from the same
individual animal, and are measured with different methods:
postmortem histology (Kaschube et al., 2003; Adams et al., 2007),
electrophysiology (Tusa et al., 1978; Van Essen et al., 1984), or
MRI (Horton and Hoyt, 1991). The large variability in V1 shape
and size across individuals of the same species also introduces
measurement errors (Tusa et al., 1978; Van Essen et al., 1984).
The errors should be even larger in cats because the cortical distortions of visual space are smaller and the published retinotopic
and ocular dominance maps have very different shapes (compare
Fig. 10a– c with Fig. 11b). Despite these measurement errors, the
relationship between ocular dominance segregation and local
cortical retinotopy was very robust. Therefore, we conclude that
ocular dominance segregation is closely associated with cortical
retinotopy and that this association is present in the entire cortical map of different species, including humans.
Simulations of ocular dominance segregation in visual cortex
Our ASFs can be used to simulate differences in ocular dominance segregation across species. The cortical area covered by the
ASF represents the cortical region receiving input from thalamic
afferents with overlapping receptive fields (i.e., cortical region
representing the same binocular point in visual space). Therefore, we can simulate an evolutionary increase in cortical area per
binocular visual point simply by increasing the size of the ASF
(Fig. 12a). For simplicity, we perform these simulations with a
circular filter that generates randomly oriented ocular domi-
Najafian et al. • Diversity of Ocular Dominance Patterns
J. Neurosci., November 13, 2019 • 39(46):9145–9163 • 9157
Figure 10. ASF set associated with each ocular dominance map. a, Map of ocular dominance (left, same as Fig. 2a) and the associated set of ASFs (right) obtained from human cortex. b, Same
as in a, but for male macaque (same ocular dominance map as Fig. 2c). c, Same as in a, but for cat (same ocular dominance map as Fig. 2e). d–f, Distribution of ASF lengths obtained from human
(d), macaque (e), and cat (f ) ocular dominance maps, shown separately for the longest (left) and shortest (right) axes of the filter. g–i, Distribution of filter angles and mean (purple number)
obtained from ocular dominance maps from human (g), macaque (h), and cat (i). j, Distribution of filter angles obtained from other ocular dominance maps from human (left) and macaque (right).
Most filter angles tend to be parallel to the longest axes of area V1 in macaques.
nance bands and has a center/surround ratio of 2.5, which is the
average ratio obtained with our measurements (Fig. 10). We apply ASFs of different sizes to images of binary white noise representing afferents in the cortical plate. We then assign a value of 1
to the smallest filter, which represents a species with a very small
cortical area per binocular visual point (e.g., mouse cortex), and
measure relative changes in ocular dominance segregation with
filter size.
This simple simulation demonstrates that increasing the cortical area per binocular visual point (i.e., size of ASF) strengthens
ocular dominance segregation through a power law function.
With the filter parameters that we used, doubling the size of the
filter increases the strength of ocular dominance segregation by
3.6 times (Fig. 12b) and the stripe width by 2– 4 times depending
on filter size (Fig. 12c). These simulations explain the weaker
ocular dominance segregation in New World monkeys and their
thin (0.4 – 0.2 mm) or absent ocular dominance stripes (Adams
and Horton, 2003; Takahata et al., 2014). The simulations also
reproduce the strong ocular dominance segregation of cortices
that represent each binocular visual point with a large cortical
area, such as humans and macaques (Fig. 12a– c). In this model,
the large cortical region per binocular visual point allows accommodating many afferents from the two eyes with the same retinotopy. Within this cortical region, afferents from each eye fire
together more often and cooperate to connect neurons that are
also close together. Consequently, the afferent segregation by eye
input helps reduce the total amount of axon needed to make their
connections (Fig. 12d). In contrast, ocular dominance segregation is weak and random in the visual cortex of rodents and
lagomorphs because they have a small cortical region per binocular visual point and a limited number of afferents from the two
eyes sharing the same retinotopy (Fig. 12e). The cortices of rodents and lagomorphs also have to represent large panoramic
visual fields that are mostly monocular, leaving limited cortical
resources to represent binocular visual points. For example, although ferrets and rabbits have cortices with very similar surface
area (⬃80 mm 2), rabbits have lateral eyes, sample a much larger
visual field and, unlike ferrets, do not have ocular dominance
segregation in cortex (Mazade and Alonso, 2017).
Simulations of ON-OFF afferent segregation in visual cortex
Thalamic afferents segregate in visual cortex not only by eye input
(contralateral or ipsilateral) but also by contrast polarity (ON or
OFF). This ON-OFF segregation is thought to be important in
the development of cortical maps (Miller, 1994; Paik and
Ringach, 2011; Jang and Paik, 2017), has been demonstrated in
cats, ferrets, minks, and tree shrews (McConnell and LeVay,
1984; Norton et al., 1985; Zahs and Stryker, 1988; Jin et al., 2008;
Kremkow et al., 2016), and is likely to be also present in primates
(Kremkow et al., 2016; Kremkow and Alonso, 2018). In cat visual
cortex, ON-OFF polarity and ocular dominance segregate along
orthogonal cortical axes (Kremkow et al., 2016). Because of this
Najafian et al. • Diversity of Ocular Dominance Patterns
9158 • J. Neurosci., November 13, 2019 • 39(46):9145–9163
Figure 11. Ocular dominance bands segregate along the cortical axis with slowest local retinotopy gradient. a, Set of ASFs obtained from the same ocular dominance maps illustrated in Figures
2 and 10, but for larger cortical patches. Each ASF was obtained from a patch of an ocular dominance map that matched the size and position of a cortical retinotopic sector. Retinotopic sectors were
estimated from published cortical retinotopic maps from human (Horton and Hoyt, 1991), macaque (Van Essen et al., 1984), and cat (Tusa et al., 1978). b, Axis of slowest retinotopy gradient (dotted
purple line) shown for each retinotopic sector. Gray lines indicate axes of slowest retinotopic gradient in sectors that could not be properly measured. Because the retinotopic sectors in cat are small
and the alignment with the ocular dominance map is prone to large error, only central large sectors are measured. c, Comparison of distributions for axis of slowest retinotopy gradient (dotted purple
line) and longest axis of ASFs (solid purple line).
orthogonal relationship, we can simulate an ON-OFF map with
our ASFs if we know the ocular dominance map and make two
main assumptions. First, the orthogonal relationship between
ocular dominance and ON-OFF polarity is similar in cats, macaques, and humans. Second, the afferent segregation by ONOFF polarity is weaker than the afferent segregation by ocular
dominance because ON-OFF afferents segregate later (Speer et
al., 2010; Kremkow and Alonso, 2018). To simulate the ON-OFF
cortical map, we first took an ocular dominance map published
in the scientific literature (Fig. 13a) and generated the set of ASFs
for ocular dominance (Fig. 10a– c). We then rotated all ASFs by
90 degrees to obtain the set of ASFs for ON-OFF dominance (Fig.
13b– d). The algorithm for ON-OFF segregation is similar to the
algorithm used for ocular dominance segregation but has a randomization factor that makes the ON-OFF afferent segregation
weaker (see Materials and Methods). In these simulations, the
afferent segregation for ocular dominance is already very pronounced at the first convolution (developmental Step 1) as previously shown in Figure 7. In contrast, the ON-OFF afferent
segregation is nearly absent at developmental Step 1, particularly
when there is an equal number of ON and OFF afferents (Fig.
13b,d,e). When OFF afferents dominate (e.g., 60% OFF and 40%
ON), the ON-OFF afferent segregation is stronger (Fig. 13c– e).
The eye/polarity grid: a developmental model of visual
cortical topography
Together, our results support a developmental model of visual
cortical topography that sorts thalamic afferents by eye input and
contrast polarity along orthogonal cortical axis with different
retinotopic gradients. The model assumes that afferents showing
the strongest correlated firing are more likely to make connection
with the same cortical targets and become closer together in the
cortex. In the model, images perceived by the two eyes are very
similar and positively correlated (Fig. 14a,b), making afferents
from the two eyes fire together when they have overlapping receptive fields (Fig. 14b, inset). Consequently, afferents from the
two eyes with overlapping receptive fields tend to be close together within the cortex, making the ocular dominance segregation align with the cortical axis of slowest retinotopy gradient.
Conversely, images of light and dark features are negatively correlated (Fig. 14c,d), making ON and OFF afferents from the same
eye fire together when they have partially overlapping receptive
fields (Fig. 14d, inset). Consequently, ON and OFF afferents from
the same eye with partially overlapping receptive fields tend to be
close together within the cortex, making the ON-OFF segregation
align with the cortical axis of fastest retinotopy gradient. This
orthogonal arrangement of afferent segregation by eye input and
contrast polarity (Kremkow et al., 2016) simply reflects the fact
that the same point in visual space can be seen by both eyes but
cannot be both dark and light at the same time.
Discussion
We have demonstrated that the diversity of ocular dominance
patterns in visual cortex is closely associated with asymmetries in
cortical retinotopy. We show that ocular dominance stripes run
across the cortical axis with slowest retinotopy gradient, an arrangement that maximizes the cortical retinotopic match across
the ocular dominance border. The close relationship between
retinotopy and ocular dominance is demonstrated in different
regions of an individual cortical map, different species, and is
Najafian et al. • Diversity of Ocular Dominance Patterns
J. Neurosci., November 13, 2019 • 39(46):9145–9163 • 9159
Figure 12. Simulations of ocular dominance segregation in visual cortex. a, Simulation of cortical patches processed with different ASFs, from the smallest filter on the left (size 1) to the largest
on the right (size 32). The size of the filter represents the cortical region receiving afferents with overlapping receptive fields. Afferents at the borders of the filter have nonoverlapping receptive fields
(RF), as illustrated at the top of each figure panel. b, Strength of ocular dominance segregation (measured as maximum power in Fourier space) as a function of ASF size. The segregation strength
increases with filter size following a power function with an exponent of 1.85. Top, Equation. c, Same as in b, but for stripe width. The model assumes a normalized stripe width of 1 for human visual
cortex, which is used as reference to estimate the relative values for macaques and squirrel monkeys in c and then b. d, The model assumes that animals showing ocular dominance segregation have
many afferents from the contralateral (C) and ipsilateral (I) eyes with overlapping receptive fields. Afferents with the same retinotopy from the same eye tend to fire together, connect to common
targets and become neighbors in cortex, which reduces the axon needed for their connections. e, The model assumes that animals lacking ocular dominance segregation have very few afferents from
the two eyes with overlapping receptive fields. Because the number of afferents from the two eyes with the same retinotopy is very limited, they remain randomly distributed in cortex.
particularly pronounced in macaques and humans. We also demonstrate that the ocular dominance stripes from the ipsilateral eye
become thinner as visual cortical eccentricity increases, even
within central vision. This thinning of ipsilateral-eye stripes further supports a close relationship between retinotopy and ocular
dominance because retinal ganglion cell density decreases more
with eccentricity in the temporal retina (that feeds ipsilateral-eye
stripes) than the nasal retina (that feeds contralateral-eye stripes)
(Curcio and Allen, 1990).
We use these results to propose a developmental model of
visual cortical topography that segregates thalamic afferents by
eye input and ON-OFF polarity along orthogonal cortical axes.
Eye input segregates along the axis of slowest retinotopic gradient
to maximize the binocular retinotopic match across the border of
ocular dominance stripes, which is needed for depth perception.
ON-OFF polarity segregates along the axis with fastest retinotopic gradient to maximize the retinotopic mismatch across the
border of ON and OFF domains, which is needed to process
stimulus orientation.
Relationship between cortical retinotopy and
ocular dominance
Hubel and Wiesel were the first to predict a close relation between
ocular dominance and cortical retinotopy. In their ice-cube
model, Hubel and Wiesel represented each portion of visual
space with a rectangular piece of cortex that has a length/width
ratio of 2, one cortical square for each eye (Hubel and Wiesel,
1977). To accommodate the cortical machinery for the two eyes,
retinotopy needs to change two times slower along the length
than the width of this cortical rectangle. This prediction is consistent with the finding that ocular dominance stripes run orthogonal to the border between areas V1 and V2 because
retinotopy changes slower along than across this border (Hubel
and Wiesel, 1974a, 1977; Tootell et al., 1982, 1988; Blasdel and
Campbell, 2001; Adams et al., 2007). However, the prediction is
inconsistent with careful measurements of cortical patches near
the V1/V2 border showing retinotopic ratios considerably lower
than two. The retinotopic ratio within an ocular dominance
stripe can range from 1.2 (Tootell et al., 1982) to 1.7 (Blasdel and
Campbell, 2001), and the average cortical patch representing a
square of visual space has a length/width ratio of 1.37 ⫾ 0.15
(Blasdel and Campbell, 2001) (ratios averaged from V/H column
in Table 1). The ratio of the cortical lines representing vertical/
horizontal meridians in visual cortex is also 1.34 ⫾ 0.09 (Tootell
et al., 1988) (ratios averaged from Figure 13 using grabit from
MATLAB to extract the data).
Comparative measurements of ocular dominance and retinotopy have been restricted in the past to the neighborhood of the
V1/V2 border (Hubel and Wiesel, 1977; Blasdel and Campbell,
2001). Therefore, it has been suggested that the limited retinotopic distortion demonstrated at the V1/V2 border is a special
case that cannot be generalized to the entire cortex (Tootell et al.,
1982). The retinotopic ratio at the V1/V2 border could be larger
than one simply because the cortical line representing the vertical
meridian (V1/V2 border) is elliptical and longer than the straight
line representing the horizontal meridian (Tootell et al., 1982,
1988). Contrary to this explanation, our results demonstrate a
tight relation between cortical retinotopy and ocular dominance
9160 • J. Neurosci., November 13, 2019 • 39(46):9145–9163
Najafian et al. • Diversity of Ocular Dominance Patterns
Figure 13. Simulations of ON-OFF segregation in visual cortex. a, V1 maps of ocular dominance used to obtain the ASFs for ON-OFF segregation (same maps of Fig. 2). b, Simulation of ON-OFF
segregation for the V1 ocular dominance maps shown in a. c, Same as in b, but using 40% of ON afferents and 60% of OFF afferents instead of 50% of each. d, Detail of cortical square patch shown
in b and c to illustrate better the more pronounced ON-OFF segregation when the cortex is OFF dominated (40% of ON and 60% of OFF afferents vs 50% of each). e, Strength of ON-OFF segregation
for consecutive developmental steps showing the stronger segregation for the OFF dominated cortex. Top, Images represent example square patches from the macaque maps used to obtain the
values in the plot below.
that is not restricted to the V1/V2 border but includes multiple
cortical regions and different species. The strong relationship between retinotopy and ocular dominance that we demonstrate is
also consistent with the asymmetric shapes of axonal arbors from
thalamic afferents, which tend to run orthogonal to the V1/V2
border, even in cats (Kremkow and Alonso, 2018).
The eye/polarity grid of visual cortical topography
Hubel and Wiesel (1977) predicted that cortical retinotopy
should abruptly change at the border between ocular dominance
stripes. However, we could not find evidence for such abrupt
retinotopic disruptions in multielectrode recordings from cat visual cortex (Kremkow et al., 2016). Also contrary to this predic-
Najafian et al. • Diversity of Ocular Dominance Patterns
J. Neurosci., November 13, 2019 • 39(46):9145–9163 • 9161
Figure 14. The eye/polarity grid. A model of visual cortical topography. a, Stereogram illustrating the image of a bicycle processed by the contralateral eye (left, gray frame) and ipsilateral eye
(right, black frame). Images from public domain (http://vision.middlebury.edu/stereo/). b, Correlation between the two images illustrated in a, measured when the images are aligned (0 at x axis)
and misaligned within a range between ⫺1500 and 1500 pixels. The correlation between images seen by two frontal eyes is positive. Therefore, neighboring afferents from the two eyes should be
most strongly correlated when they have overlapping receptive fields (inset). c, Stereogram illustrating the image of the bicycle processed by OFF (left, blue frame) and ON (right, red frame) cortical
pathways. d, Image correlation of the two images illustrated in c. The correlation between images processed by ON and OFF cortical pathways is negative. Therefore, ON and OFF neighboring
afferents should be most strongly correlated when they have partially overlapping receptive fields (inset). e, Simulation of the eye/polarity grid in human cortex shown for the entire map (top) and
a cortical patch (gray square). Bottom, The cortical patch is shown in more detail. Thalamic afferents from contralateral (contra) and ipsilateral eyes (ipsi) segregate in a cortical axis (eye axis)
orthogonal to the axis for ON and OFF thalamic segregation (polarity axis). The retinotopic gradient within the cortical patch is slowest along the eye axis to maximize the binocular retinotopic match
across the ocular dominance border, which is needed for depth perception. It is fastest along the polarity axis to maximize the retinotopic mismatch across the ON-OFF border, which is needed to
process stimulus orientation.
tion, the receptive fields from the left and right eyes of binocular
neurons are exquisitely overlapped in visual space (Ohzawa et al.,
1996; Tsao et al., 2003; Kara and Boyd, 2009; Wang et al., 2015;
Kremkow et al., 2016). Therefore, in our model, retinotopy does
not change rapidly at the border of ocular dominance stripes. It
does exactly the opposite. It changes at the slowest rate across the
ocular dominance border to minimize differences in retinotopy
between neighboring afferents from the two eyes.
Unlike ocular dominance stripes, ON and OFF cortical domains need to be mismatched in retinotopy to process stimulus
orientation and help build cortical receptive fields with spatially
separate ON and OFF subregions (Reid and Alonso, 1995; Alonso
et al., 1996, 2001; Lien and Scanziani, 2013; Sedigh-Sarvestani et
al., 2017). In cat visual cortex, ON and OFF domains are cortically segregated, retinotopically mismatched, and run orthogonal
to ocular dominance bands (Kremkow et al., 2016). Therefore,
our model adopts this orthogonal relationship in what we call the
eye/polarity grid (Kremkow and Alonso, 2018). In the eye/polarity grid, thalamic afferents segregate both by eye input (eye axis)
and ON-OFF contrast polarity (polarity axis) along orthogonal
cortical axes. The eye axis shows the slowest changes in retinotopy
and the polarity axis the fastest changes. In the eye/polarity grid,
ocular dominance stripes run parallel to iso-eccentricity lines
(Hubel and Freeman, 1977; Adams et al., 2007) simply because
cortical retinotopy changes faster along than across these lines.
Moreover, the average length/width ratio of the local cortical
retinotopy is 1.4 and not 2 because cortical retinotopy changes
faster along the polarity than the eye axes.
Our model predicts that ocular dominance stripes should be
present in any brain structure that can accommodate a large
number of afferents with overlapping receptive fields from both
eyes. Moreover, it predicts that variations in ocular dominance
patterns should correlate with variations in the density of retinal
ganglion cells that feed the cortex through the thalamus (Mazade
and Alonso, 2017). Because retinal ganglion cell density can be
300% higher in the peripheral nasal than temporal retina (Curcio
and Allen, 1990), the ocular dominance stripes from the ipsilateral eye should be thinner in the visual periphery. Moreover,
because the density differences between nasal and temporal retina can be already present within 5 degree eccentricities (Curcio
and Allen, 1990), the decrease in ipsilateral-eye stripe width
should be already present within central vision, as our results
demonstrate.
Our results indicate that, as retinal eccentricity increases, the
cortex compensates limitations in retinal sampling by enhancing
the dominance of the contralateral eye, a topographic adjustment
that should have direct consequences in human vision (e.g.,
contralateral-eye dominance should increase with azimuth eccentricity). In our model, the decrease in retinal sampling with
eccentricity makes cortical retinotopy change faster. In turn, the
faster retinotopy gradient limits the ability of the cortex to accommodate a full ipsilateral-contralateral cycle of afferents with
overlapping receptive fields. Therefore, in our model, ocular
dominance segregation vanishes in the visual periphery of human
and macaque cortex for the same reason that it vanishes in the
cortex of some squirrel monkeys (Adams and Horton, 2003);
because the number of afferents from the two eyes with overlapping receptive fields is reduced. In support of our model, adding
more afferents from the ipsilateral eye to the cortex causes afferent segregation by eye input, even in rodents (Merlin et al., 2013;
Laing et al., 2015). Moreover, directing retinal afferents from the
two eyes into the same optic tectum causes ocular dominance
segregation in frogs and fish (Constantine-Paton and Law, 1978;
Boss and Schmidt, 1984).
9162 • J. Neurosci., November 13, 2019 • 39(46):9145–9163
Our model predicts close relationships among the V1 maps
for retinotopy, ocular dominance, and ON-OFF polarity. Unfortunately, experimental measures of these relationships are still
rare. Accurate reconstructions of individual V1 maps are available for ocular dominance but not for retinotopy or ON-OFF
polarity. Maps with enough spatial resolution to measure retinotopy are only available as averages from multiple animals, and
ON-OFF polarity maps are restricted to small cortical patches
(Jin et al., 2008; Smith et al., 2015; Wang et al., 2015; Kremkow et
al., 2016; Lee et al., 2016). Therefore, future experiments and new
tools will be needed to reconstruct more precisely the organization of visual cortical maps and the relationship among the representation of different stimulus dimensions (Hübener et al.,
1997; Nauhaus et al., 2016). These maps are crucial to understand
how the cerebral cortex represents visual information and to
guide future implants of cortical prosthesis in the blind.
References
Adams DL, Horton JC (2003) Capricious expression of cortical columns in
the primate brain. Nat Neurosci 6:113–114.
Adams DL, Horton JC (2006) Ocular dominance columns in strabismus.
Vis Neurosci 23:795– 805.
Adams DL, Sincich LC, Horton JC (2007) Complete pattern of ocular dominance columns in human primary visual cortex. J Neurosci 27:10391–
10403.
Albus K (1975) A quantitative study of the projection area of the central and
the paracentral visual field in area 17 of the cat: I. The precision of the
topography. Exp Brain Res 24:159 –179.
Alonso JM, Usrey WM, Reid RC (1996) Precisely correlated firing in cells of
the lateral geniculate nucleus. Nature 383:815– 819.
Alonso JM, Usrey WM, Reid RC (2001) Rules of connectivity between
geniculate cells and simple cells in cat primary visual cortex. J Neurosci
21:4002– 4015.
Anderson PA, Olavarria J, Van Sluyters RC (1988) The overall pattern of
ocular dominance bands in cat visual cortex. J Neurosci 8:2183–2200.
Blasdel G, Campbell D (2001) Functional retinotopy of monkey visual cortex. J Neurosci 21:8286 – 8301.
Boss VC, Schmidt JT (1984) Activity and the formation of ocular dominance patches in dually innervated tectum of goldfish. J Neurosci 4:2891–
2905.
Constantine-Paton M, Law MI (1978) Eye-specific termination bands in
tecta of three-eyed frogs. Science 202:639 – 641.
Curcio CA, Allen KA (1990) Topography of ganglion cells in human retina.
J Comp Neurol 300:5–25.
Daniel PM, Whitteridge D (1961) The representation of the visual field on
the cerebral cortex in monkeys. J Physiol 159:203–221.
DeBruyn EJ, Casagrande VA (1981) Demonstration of ocular dominance
columns in a new world primate by means of monocular deprivation.
Brain Res 207:453– 458.
Harvey BM, Dumoulin SO (2011) The relationship between cortical magnification factor and population receptive field size in human visual
cortex: constancies in cortical architecture. J Neurosci 31:13604 –
13612.
Horton JC, Hocking DR (1996) Anatomical demonstration of ocular dominance columns in striate cortex of the squirrel monkey. J Neurosci 16:
5510 –5522.
Horton JC, Hoyt WF (1991) The representation of the visual field in human
striate cortex: a revision of the classic Holmes map. Arch Ophthalmol
109:816 – 824.
Hubel DH, Freeman DC (1977) Projection into the visual field of ocular
dominance columns in macaque monkey. Brain Res 122:336 –343.
Hubel DH, Wiesel TN (1974a) Sequence regularity and geometry of orientation columns in the monkey striate cortex. J Comp Neurol 158:
267–293.
Hubel DH, Wiesel TN (1974b) Uniformity of monkey striate cortex: a parallel relationship between field size, scatter, and magnification factor.
J Comp Neurol 158:295–305.
Hubel DH, Wiesel TN (1977) Ferrier lecture: functional architecture of macaque monkey visual cortex. Proc R Soc Lond B Biol Sci 198:1–59.
Hübener M, Shoham D, Grinvald A, Bonhoeffer T (1997) Spatial relation-
Najafian et al. • Diversity of Ocular Dominance Patterns
ships among three columnar systems in cat area 17. J Neurosci
17:9270 –9284.
Huberman AD, Feller MB, Chapman B (2008) Mechanisms underlying development of visual maps and receptive fields. Annu Rev Neurosci
31:479 –509.
Jang J, Paik SB (2017) Interlayer repulsion of retinal ganglion cell mosaics
regulates spatial organization of functional maps in the visual cortex.
J Neurosci 37:12141–12152.
Jin JZ, Weng C, Yeh CI, Gordon JA, Ruthazer ES, Stryker MP, Swadlow HA,
Alonso JM (2008) On and off domains of geniculate afferents in cat
primary visual cortex. Nat Neurosci 11:88 –94.
Kara P, Boyd JD (2009) A micro-architecture for binocular disparity and
ocular dominance in visual cortex. Nature 458:627– 631.
Kaschube M, Wolf F, Puhlmann M, Rathjen S, Schmidt KF, Geisel T, Löwel S
(2003) The pattern of ocular dominance columns in cat primary visual
cortex: intra- and interindividual variability of column spacing and its
dependence on genetic background. Eur J Neurosci 18:3251–3266.
Kremkow J, Alonso JM (2018) Thalamocortical circuits and functional architecture. Annu Rev Vis Sci 4:263–285.
Kremkow J, Jin J, Wang Y, Alonso JM (2016) Principles underlying sensory
map topography in primary visual cortex. Nature 533:52–57.
Laing RJ, Turecek J, Takahata T, Olavarria JF (2015) Identification of eyespecific domains and their relation to callosal connections in primary
visual cortex of Long–Evans rats. Cereb Cortex 25:3314 –3329.
Lee KS, Huang X, Fitzpatrick D (2016) Topology of ON and OFF inputs in
visual cortex enables an invariant columnar architecture. Nature 533:
90 –94.
LeVay S, Hubel DH, Wiesel TN (1975) The pattern of ocular dominance
columns in macaque visual cortex revealed by a reduced silver stain.
J Comp Neurol 159:559 –576.
LeVay S, Connolly M, Houde J, Van Essen DC (1985) The complete pattern
of ocular dominance stripes in the striate cortex and visual field of the
macaque monkey. J Neurosci 5:486 –501.
Lien AD, Scanziani M (2013) Tuned thalamic excitation is amplified by
visual cortical circuits. Nat Neurosci 16:1315–1323.
Mazade R, Alonso JM (2017) Thalamocortical processing in vision. Vis
Neurosci 34:E007.
McConnell SK, LeVay S (1984) Segregation of on- and off-center afferents
in mink visual cortex. Proc Natl Acad Sci U S A 81:1590 –1593.
Merlin S, Horng S, Marotte LR, Sur M, Sawatari A, Leamey CA (2013) Deletion of ten-m3 induces the formation of eye dominance domains in
mouse visual cortex. Cereb Cortex 23:763–774.
Miller KD (1994) A model for the development of simple cell receptive fields
and the ordered arrangement of orientation columns through activitydependent competition between ON- and OFF-center inputs. J Neurosci
14:409 – 441.
Nauhaus I, Nielsen KJ, Callaway EM (2016) Efficient receptive field tiling in
primate V1. Neuron 91:893–904.
Norton TT, Rager G, Kretz R (1985) ON and OFF regions in layer IV of
striate cortex. Brain Res 327:319 –323.
Ohzawa I, DeAngelis GC, Freeman RD (1996) Encoding of binocular disparity by simple cells in the cat’s visual cortex. J Neurophysiol 75:1779 –
1805.
Paik SB, Ringach DL (2011) Retinal origin of orientation maps in visual
cortex. Nat Neurosci 14:919 –925.
Rakic P (1977) Prenatal development of the visual system in rhesus monkey.
Philos Trans R Soc Lond B Biol Sci 278:245–260.
Reid RC, Alonso JM (1995) Specificity of monosynaptic connections from
thalamus to visual cortex. Nature 378:281–284.
Sedigh-Sarvestani M, Vigeland L, Fernandez-Lamo I, Taylor MM, Palmer LA,
Contreras D (2017) Intracellular, in vivo, dynamics of thalamocortical
synapses in visual cortex. J Neurosci 37:5250 –5262.
Selemon LD, Begovic A (2007) Stereologic analysis of the lateral geniculate
nucleus of the thalamus in normal and schizophrenic subjects. Psychiatry
Res 151:1–10.
Shatz CJ, Stryker MP (1978) Ocular dominance in layer IV of the cat’s visual
cortex and the effects of monocular deprivation. J Physiol 281:267–283.
Smith GB, Whitney DE, Fitzpatrick D (2015) Modular representation of
luminance polarity in the superficial layers of primary visual cortex. Neuron 88:805– 818.
Najafian et al. • Diversity of Ocular Dominance Patterns
Spatz WB (1989) Loss of ocular dominance columns with maturity in the
monkey, Callithrix jacchus. Brain Res 488:376 –380.
Speer CM, Mikula S, Huberman AD, Chapman B (2010) The developmental remodeling of eye-specific afferents in the ferret dorsal lateral geniculate nucleus. Anat Rec (Hoboken) 293:1–24.
Swindale NV (1980) A model for the formation of ocular dominance
stripes. Proc R Soc Lond B Biol Sci 208:243–264.
Takahata T, Miyashita M, Tanaka S, Kaas JH (2014) Identification of ocular
dominance domains in new world owl monkeys by immediate-early gene
expression. Proc Natl Acad Sci U S A 111:4297– 4302.
Tootell RB, Silverman MS, Switkes E, De Valois RL (1982) Deoxyglucose
analysis of retinotopic organization in primate striate cortex. Science
218:902–904.
Tootell RB, Switkes E, Silverman MS, Hamilton SL (1988) Functional anat-
J. Neurosci., November 13, 2019 • 39(46):9145–9163 • 9163
omy of macaque striate cortex: II. Retinotopic organization. J Neurosci
8:1531–1568.
Tsao DY, Conway BR, Livingstone MS (2003) Receptive fields of disparitytuned simple cells in macaque V1. Neuron 38:103–114.
Tusa RJ, Palmer LA, Rosenquist AC (1978) The retinotopic organization of
area 17 (striate cortex) in the cat. J Comp Neurol 177:213–235.
Van Essen DC, Newsome WT, Maunsell JH (1984) The visual field representation in striate cortex of the macaque monkey: asymmetries, anisotropies, and individual variability. Vision Res 24:429 – 448.
Wang Y, Jin J, Kremkow J, Lashgari R, Komban SJ, Alonso JM (2015)
Columnar organization of spatial phase in visual cortex. Nat Neurosci
18:97–103.
Zahs KR, Stryker MP (1988) Segregation of ON and OFF afferents to ferret
visual cortex. J Neurophysiol 59:1410 –1429.
|
|
https://openalex.org/W2790920547
|
https://dash.harvard.edu/bitstream/1/35982257/1/5860691.pdf
|
English
| null |
Primary prevention of cardiovascular disease: The past, present, and future of blood pressure- and cholesterol-lowering treatments
|
PLoS medicine
| 2,018
|
cc-by
| 2,891
|
Permanent link http://nrs.harvard.edu/urn-3:HUL.InstRepos:35982257 Terms of Use This article was downloaded from Harvard University’s DASH repository, and is made available
under the terms and conditions applicable to Other Posted Material, as set forth at http://
nrs.harvard.edu/urn-3:HUL.InstRepos:dash.current.terms-of-use#LAA Citation Leening, Maarten J. G., and M. Arfan Ikram. 2018. “Primary prevention of cardiovascular
disease: The past, present, and future of blood pressure- and cholesterol-lowering treatments.”
PLoS Medicine 15 (3): e1002539. doi:10.1371/journal.pmed.1002539. http://dx.doi.org/10.1371/
journal.pmed.1002539. Published Version Published Version
doi:10.1371/journal.pmed.1002539 Share Your Story The Harvard community has made this article openly available. Please share how this access benefits you. Submit a story . Accessibility PERSPECTIVE a1111111111
a1111111111
a1111111111
a1111111111
a1111111111 Published: March 20, 2018 Copyright: © 2018 Leening, Ikram. This is an open
access article distributed under the terms of the
Creative Commons Attribution License, which
permits unrestricted use, distribution, and
reproduction in any medium, provided the original
author and source are credited. OPEN ACCESS Citation: Leening MJG, Ikram MA (2018) Primary
prevention of cardiovascular disease: The past,
present, and future of blood pressure- and
cholesterol-lowering treatments. PLoS Med 15(3):
e1002539. https://doi.org/10.1371/journal. pmed.1002539 Primary prevention of cardiovascular disease:
The past, present, and future of blood
pressure- and cholesterol-lowering
treatments Maarten J. G. Leening1,2,3, M. Arfan Ikram1* 1 Department of Epidemiology, Erasmus MC – University Medical Center Rotterdam, Rotterdam, the
Netherlands, 2 Department of Cardiology, Erasmus MC – University Medical Center Rotterdam, Rotterdam,
the Netherlands, 3 Department of Epidemiology, Harvard T.H. Chan School of Public Health, Boston,
Massachusetts, United States of America 1 Department of Epidemiology, Erasmus MC – University Medical Center Rotterdam, Rotterdam, the
Netherlands, 2 Department of Cardiology, Erasmus MC – University Medical Center Rotterdam, Rotterdam,
the Netherlands, 3 Department of Epidemiology, Harvard T.H. Chan School of Public Health, Boston,
Massachusetts, United States of America * m.a.ikram@erasmusmc.nl * m.a.ikram@erasmusmc.nl Shortly after World War II, coronary heart disease was recognized as an epidemic. Cardiovas-
cular disease (CVD) had become the leading cause of death in Western societies. This led to
the formation of the National Heart, Lung, and Blood Institute in the United States and initia-
tion of the seminal Framingham Heart Study in 1948. This population-based cohort study on
the etiology and consequences of CVD has shed light on many of the well-known causes of
CVD. In one of their initial publications in 1961, the investigators from the Framingham
Heart Study introduced the concept of “factors of risk in the development of coronary heart
disease”, nowadays known as traditional cardiovascular risk factors [1]. The identification of high blood pressure and cholesterol levels as causes of CVD led to
the idea of screening and treatment thereof in otherwise healthy persons, in order to halt
atherosclerosis and forestall the occurrence of cardiovascular events. Initial treatment recom-
mendations in the 1970s and 1980s were based on the levels of the specific risk factors, with
antihypertensive treatment recommended for “virtually all persons with a diastolic blood pres-
sure exceeding 105 mmHg” [2]. However, a gradual diversification then took place in preven-
tive cardiology: hypertension guidelines remained focused on blood pressure levels, whereas
cholesterol treatment guidelines moved towards more sophisticated approaches by recom-
mending pharmacological interventions informed by the individual’s cardiovascular risk
based on the presence and levels of multiple cardiovascular risk factors. Current treatment guidelines With the latest edition of the American College of Cardiology/American Heart Association
hypertension treatment guidelines, treatment targets have also become more individualized
based on cardiovascular disease risk, with lower blood pressure targets for patients with estab-
lished CVD or diabetes, as well as in those at a high 10-year predicted cardiovascular risk [3]. Yet, the indications for initiating blood pressure-lowering treatment remain primarily driven
by blood pressure levels, whereas for statin treatment, cardiovascular risk instead of cholesterol
level is the main driver of recommendations for treatment initiation. In this issue of PLOS
Medicine, Kazem Rahimi and colleagues, on behalf of the Blood Pressure Lowering Treatment
Trialists’ Collaboration, present an analysis of individual patient data indicating that, at the
population level, a strategy based on 5-year predicted cardiovascular risk—rather than blood Funding: The authors received no specific funding
for this work. * m.a.ikram@erasmusmc.nl pressure levels alone—could be more effective, especially in a primary prevention setting [4]. For example, they report that compared with a strategy of treating everyone with a systolic
blood pressure of 140 mmHg or more, treatment based on overall CVD risk would prevent
3.1% more events (95% CI 1.5% to 5.0%) for the same number of people treated. These data
reemphasize that we should target overall cardiovascular risk and aim to modify risk through
holistic risk management, as opposed to targeting only individual risk factor levels in isolation. Traditionally, the focus in cardiovascular risk assessment has been on 10-year probabilities
of developing CVD. However, such 10-year absolute cardiovascular risk estimates are often
abstract numbers to patients, and risk communication is therefore challenging, especially in
younger individuals with multiple risk factors in whom 10-year absolute cardiovascular risk
remains low by virtue of their age [5]. For this reason, in the latest iterations of the US and
British CVD prevention guidelines [6–8], attention is shifting towards lifetime perspectives. This transition is driven by data demonstrating that cardiovascular risk is best expressed by
cumulative exposure to risk factors over time, rather than risk factor levels at a single time
point, and aggressive risk factor management should be considered earlier in life [5,6,9,10]. For example, duration of exposure to hypertension in early adulthood is associated with the
amount of coronary atherosclerosis in middle age [11]. Barriers towards a full transition from
10-year risk to lifetime risk in preventive cardiology include the lack of established treatment
thresholds and suboptimal performance of the available risk models [9]. PLOS Medicine | https://doi.org/10.1371/journal.pmed.1002539
March 20, 2018 Funding: The authors received no specific funding
for this work. Competing interests: The authors have declared
that no competing interests exist. Competing interests: The authors have declared
that no competing interests exist. Abbreviations: CVD, cardiovascular disease. Abbreviations: CVD, cardiovascular disease. Provenance: Commissioned; not externally peer
reviewed. Provenance: Commissioned; not externally peer
reviewed. Provenance: Commissioned; not externally peer
reviewed. 1 / 5 PLOS Medicine | https://doi.org/10.1371/journal.pmed.1002539
March 20, 2018 including autoimmune, pulmonary, or liver diseases. Consequently, interactions with drugs
used to treat such conditions or disease-specific adverse effects of preventive treatment will
remain understudied. Similarly, direct trial evidence on the efficacy of preventive cardiovascu-
lar medication in such vulnerable populations will remain absent. Patient involvement and shared decision-making between patients and physicians has
become an increasingly important aspect of CVD prevention. Anticipated cardiovascular risk
reduction with blood pressure- or cholesterol-lowering treatment should be weighed against
the burden and costs of taking medication, as well as potential adverse effects. In many indi-
viduals free of CVD, the risk of developing CVD is relatively low. Hence, patient preferences
and attitudes towards preventive medication are key to decisions to initiate or intensify treat-
ment. However, informed decisions rely on a good understanding of an individual’s cardiovas-
cular risk and risk reduction conferred by preventive treatment. Therefore, comprehensible
metrics are needed to summarize anticipated treatment benefit, for instance, by expressing
lifetime benefits of treatment as gains in CVD-free life expectancy [5,19]. Challenges and opportunities With accruing evidence supporting safe generic medication for blood pressure and cholesterol
lowering, indications for treatment have been widened through drastic lowering of treatment
thresholds and targets. This is a consistent phenomenon across clinical practice guidelines in
high-income countries, to such an extent that nearly all individuals aged 65 years and over
now qualify for statin treatment [5,12]. Nonetheless, drug treatment comes with potential side
effects and costs. Hence, clinicians and policy makers should retain a healthy level of skepti-
cism towards unrestricted population-wide treatment. On the other hand, broadened indica-
tions and widespread use of blood pressure- and cholesterol-lowering medication should not
be erroneously labeled as unnecessary medicalization of society [13]: the burden of cardiovas-
cular risk factors remains very high because of unhealthy contemporary lifestyles and justifies
medical therapy in a substantial proportion of the population. Therefore, population-wide
screening programs for cardiovascular risk assessment could facilitate early detection of those
at high risk and identify individuals who would benefit most from early sustained blood pres-
sure- or cholesterol-lowering treatment. However, the optimal age at which to screen for car-
diovascular risk factors and subsequent age-specific treatment thresholds and targets are as yet
unknown [5,14]. Treatment recommendations from current risk-based prevention guidelines [6–8, 15–17]
do not directly reflect the evidence derived from clinical trials [12,18] but rather reflect gener-
alizations of findings derived from clinical trials showing greater absolute risk reduction in
those at higher observed cardiovascular risk. Very low-risk individuals may not be recom-
mended preventive treatment although trial evidence of statin efficacy is available for such
people [18]. Conversely, findings from trials of antihypertensives and statins are extrapolated
to high-risk individuals in whom the efficacy of preventive treatment has not been studied
[12,18]. For many of these high-risk individuals, it is highly unlikely that direct trial evidence
will ever be accrued, because not every specific patient group can be studied; therefore, extrap-
olation will always be necessary to some extent. A particularly vulnerable and understudied
group consists of persons with comorbidities associated with increased cardiovascular risk, PLOS Medicine | https://doi.org/10.1371/journal.pmed.1002539
March 20, 2018 2 / 5 Future directions Recent decades have seen a transition from risk factor-based treatment to treatment based on
overall short-term and lifetime risk. The work by Rahimi and colleagues in PLOS Medicine
contributes evidence that risk-based allocation represents a more efficient approach towards
blood pressure-lowering treatment in primary prevention of CVD [4]. Now, we should move
further towards personalized treatment guidance based on anticipated benefit [5,19,20]. Antic-
ipated treatment benefit is related to absolute overall risk of CVD but is not simply a rescaling
of risk. First, data from clinical trials on pharmacological interventions need to be incorporated
in the construction of estimates of anticipated treatment benefit [20–22]. For instance,
blood pressure- and cholesterol-lowering treatment remains effective at older age, yet the
relative risk reduction generally diminishes with age. Similarly, treatment effects may vary
by ethnicity, by specific risk factor levels, and by comorbidity, such as renal disease. For
example, let us compare 2 healthy men of similar age with identical 10-year cardiovascular
risk estimates. One man’s risk is driven by moderate hypertension, and the other man’s risk
by moderate hypercholesterolemia: it seems most reasonable to lower cardiovascular risk
with antihypertensives in the former and with statins in the latter [20]. Second, since preven-
tive treatment to lower blood pressure and cholesterol is often initiated with the intention of
being used for decades or even lifelong, life expectancy is a key variable in estimating antici-
pated benefit. For instance, smoking roughly doubles the risk of CVD, yet anticipated gains
in CVD-free life expectancy with lifelong statin therapy are similar for smokers and non-
smokers. This is caused by competing risks of smoking-related cancers and other life-limit-
ing diseases [19]. Combining information on demographics, risk factor levels, and data on relative risk reduc-
tions obtained from clinical trials can inform models to provide anticipated benefits expressed
as gains in CVD-free life expectancy with specific treatment options and thereby facilitate
more informed treatment decisions. Similar estimates of anticipated harms can be produced
for the risks associated with drug treatment. Healthcare policy makers can use such data to
project costs and benefits of treatment and thereby provide clinicians with generally acceptable
treatment thresholds. CVD remains a leading cause of morbidity and mortality worldwide. Despite great progress
in treatment of acute cardiovascular conditions, first manifestations of CVD are still often
lethal or result in long-term disability. PLOS Medicine | https://doi.org/10.1371/journal.pmed.1002539
March 20, 2018 References 1. Kannel WB, Dawber TR, Kagan A, Revotskie N, Stokes J 3rd. Factors of risk in the development of cor-
onary heart disease—six year follow-up experience: the Framingham Study. Ann Intern Med. 1961 Jul;
55:33–50. PMID: 13751193 2. Report of the Joint National Committee on Detection, Evaluation, and Treatment of High Blood Pres-
sure. JAMA. 1977 Jan 17; 237(3):255–61. PMID: 576159 3. Whelton PK, Carey RM, Aronow WS, Casey DE Jr, Collins KJ, Dennison Himmelfarb C, et al. 2017
ACC/AHA/AAPA/ABC/ACPM/AGS/APhA/ASH/ASPC/NMA/PCNA guideline for the prevention, detec-
tion, evaluation, and management of high blood pressure in adults: a report of the American College of
Cardiology/American Heart Association Task Force on Clinical Practice Guidelines. J Am Coll Cardiol. 2017 Nov 7: Epub ahead of print. 4. Karmali KN, Lloyd-Jones DM, van der Leeuw J, Goff DC Jr, Rahimi K, Yusuf S, et al. Blood pressure-
lowering treatment strategies based on cardiovascular risk versus blood pressure: A meta-analysis of
individual participant data. PLoS Med. 2018; 15(3): e1002538. https://doi.org/10.1371/journal.pmed. 2002538 5. Leening MJG, Cook NR, Ridker PM. Should we reconsider the role of age in treatment allocation for pri-
mary prevention of cardiovascular disease? Eur Heart J. 2017 May 21; 38(20):1542–7. https://doi.org/
10.1093/eurheartj/ehw287 PMID: 27357357 6. Stone NJ, Robinson JG, Lichtenstein AH, Bairey Merz CN, Blum CB, Eckel RH, et al. 2013 ACC/AHA
guideline on the treatment of blood cholesterol to reduce atherosclerotic cardiovascular risk in adults: a
report of the American College of Cardiology/American Heart Association Task Force on Practice
Guidelines. Circulation. 2014 Jun 24; 129(25 Suppl 2):S1–45. https://doi.org/10.1161/01.cir. 0000437738.63853.7a PMID: 24222016 7. Goff DC Jr, Lloyd-Jones DM, Bennett G, Coady S, D’Agostino RB, Gibbons R, et al. 2013 ACC/AHA
guideline on the assessment of cardiovascular risk: a report of the American College of Cardiology/
American Heart Association Task Force on Practice Guidelines. Circulation. 2014 Jun 24; 129(25 Suppl
2):S49–73. https://doi.org/10.1161/01.cir.0000437741.48606.98 PMID: 24222018 8. JBS3 Board. Joint British Societies’ consensus recommendations for the prevention of cardiovascular
disease (JBS3). Heart. 2014 Apr; 100 Suppl 2:ii1–ii67. 9. Leening MJG, Berry JD, Allen NB. Lifetime perspectives on primary prevention of atherosclerotic car-
diovascular disease. JAMA. 2016 Apr 12; 315(14):1449–50. https://doi.org/10.1001/jama.2016.1654
PMID: 26999711 10. Sniderman AD, Toth PP, Thanassoulis G, Pencina MJ, Furberg CD. Taking a longer term view of car-
diovascular risk: the causal exposure paradigm. BMJ. 2014; 348:g3047. https://doi.org/10.1136/bmj. g3047 PMID: 24850828 11. Future directions Therefore, optimal primary prevention should remain a 3 / 5 PLOS Medicine | https://doi.org/10.1371/journal.pmed.1002539
March 20, 2018 priority in the future. More new strategies will emerge, and clinical practice guidelines are
likely to evolve from risk factor-based treatment to risk-based treatment and beyond to treat-
ment based on anticipated benefit. Throughout this process, cardiovascular specialists, general
practitioners, and researchers need to keep prioritizing time and resources to offer preventive
measures to healthy individuals free from CVD and focus on compliance and persistence with
treatment, irrespective of the prevention strategy chosen. References Allen NB, Siddique J, Wilkins JT, Shay C, Lewis CE, Goff DC, et al. Blood pressure trajectories in early
adulthood and subclinical atherosclerosis in middle age. JAMA. 2014 Feb 5; 311(5):490–7. https://doi. org/10.1001/jama.2013.285122 PMID: 24496536 12. Pavlović J, Greenland P, Deckers JW, Brugts JJ, Kavousi M, Dhana K, et al. Comparison of ACC/AHA
and ESC guideline recommendations following trial evidence for statin use in primary prevention of car-
diovascular disease: results from the population-based Rotterdam Study. JAMA Cardiol. 2016 Sep 1; 1
(6):708–13. https://doi.org/10.1001/jamacardio.2016.1577 PMID: 27439175 13. Greenland P. Cardiovascular guideline skepticism vs lifestyle realism? JAMA. 2018 Jan 9; 319(2):117–
8. https://doi.org/10.1001/jama.2017.19675 PMID: 29242938 14. Greenland P, Peterson ED. The new 2017 ACC/AHA guidelines "up the pressure" on diagnosis and
treatment of hypertension. JAMA. 2017 Dec 5; 318(21):2083–4. https://doi.org/10.1001/jama.2017. 18605 PMID: 29159417 15. Piepoli MF, Hoes AW, Agewall S, Albus C, Brotons C, Catapano AL, et al. 2016 European Guidelines
on cardiovascular disease prevention in clinical practice: The Sixth Joint Task Force of the European PLOS Medicine | https://doi.org/10.1371/journal.pmed.1002539
March 20, 2018 4 / 5 Society of Cardiology and other societies on cardiovascular disease prevention in clinical practice. Eur
Heart J. 2016 Aug 1; 37(29):2315–81. https://doi.org/10.1093/eurheartj/ehw106 PMID: 27222591 16. Anderson TJ, Gregoire J, Pearson GJ, Barry AR, Couture P, Dawes M, et al. 2016 Canadian Cardiovas-
cular Society guidelines for the management of dyslipidemia for the prevention of cardiovascular dis-
ease in the adult. Can J Cardiol. 2016 Nov; 32(11):1263–82. https://doi.org/10.1016/j.cjca.2016.07.510
PMID: 27712954 17. US Preventive Services Task Force. Statin use for the primary prevention of cardiovascular disease in
adults: US Preventive Services Task Force recommendation statement. JAMA. 2016 Nov 15; 316
(19):1997–2007. https://doi.org/10.1001/jama.2016.15450 PMID: 27838723 18. Pavlović J, Greenland P, Deckers JW, Kavousi M, Hofman A, Ikram MA, et al. Assessing gaps in cho-
lesterol treatment guidelines for primary prevention of cardiovascular disease based on available rando-
mised clinical trial evidence: the Rotterdam Study. Eur J Prev Cardiol. 2018 Mar 1; 25(4):420–31. https://doi.org/10.1177/2047487317743352 PMID: 29171772 19. Ferket BS, van Kempen BJH, Heeringa J, Spronk S, Fleischmann KE, Nijhuis RLG, et al. Personalized
prediction of lifetime benefits with statin therapy for asymptomatic individuals: a modeling study. PLoS
Med. 2012; 9(12):e1001361. https://doi.org/10.1371/journal.pmed.1001361 PMID: 23300388 20. Thanassoulis G, Williams K, Altobelli KK, Pencina MJ, Cannon CP, Sniderman AD. Individualized statin
benefit for determining statin eligibility in the primary prevention of cardiovascular disease. Circulation. 2016 Apr 19; 133(16):1574–81. https://doi.org/10.1161/CIRCULATIONAHA.115.018383 PMID:
26945047 21. PLOS Medicine | https://doi.org/10.1371/journal.pmed.1002539
March 20, 2018 References Dorresteijn JAN, Visseren FLJ, Ridker PM, Wassink AMJ, Paynter NP, Steyerberg EW, et al. Estimating
treatment effects for individual patients based on the results of randomised clinical trials. BMJ. 2011;
343:d5888. https://doi.org/10.1136/bmj.d5888 PMID: 21968126 22. van der Leeuw J, Ridker PM, van der Graaf Y, Visseren FLJ. Personalized cardiovascular disease pre-
vention by applying individualized prediction of treatment effects. Eur Heart J. 2014 Apr; 35(13):837–
43. https://doi.org/10.1093/eurheartj/ehu004 PMID: 24513790 5 / 5
|
https://openalex.org/W2625680825
|
https://hal.science/hal-01565904/document
|
English
| null |
A new global and comprehensive model for ICU ventilator performances evaluation
|
Annals of intensive care
| 2,017
|
cc-by
| 10,974
|
© The Author(s) 2017. This article is distributed under the terms of the Creative Commons Attribution 4.0 International License
(http://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium,
provided you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license,
and indicate if changes were made. Marjanovic et al. Ann. Intensive Care (2017) 7:68
DOI 10.1186/s13613-017-0285-2 Open Access Methods *Correspondence: erwan.lher@chu‑brest.fr
5 Médecine Intensive et Réanimation, CHRU de Brest, Boulevard Tanguy
Prigent, 29200 Brest Cedex, France
Full list of author information is available at the end of the article Nicolas S. Marjanovic1,2, Agathe De Simone3, Guillaume Jegou3 and Erwan L’Her4,5* Nicolas S. Marjanovic1,2, Agathe De Simone3, Guillaume Jegou3 and Erwan L’Her4,5* Background the extent to which a device can be used by specified users
to achieve specific goals effectively, efficiently and satis-
factorily, in a specified context of use. Usability is mainly
related to the quality of the human–machine interface. Improved interface seems mandatory to limit human
errors that could exacerbate morbidity and mortality
[10–12]. There are few studies dedicated to ventilator
ergonomics evaluation, and those that do exist are often
limited to timed task and/or easy user-friendliness assess-
ments [13–16].h g
Mechanical ventilation is a fundamental part of critical
care, and the accuracy of ventilatory settings is of utmost
importance. When dealing with unstable patients, a bad
technological performance may cause a patient harm, while
low tidal volume (VT) and high positive expiratory pres-
sure (PEEP) are key points for protective ventilation [1, 2]. Ineffective effort and asynchrony correction [3], along with
effective triggering [4], may decrease inspiratory work and
improve patients’ outcome. Bench-test studies are essen-
tial to assess the technical characteristics of ventilators and
determine their efficiency during critical care [5–9]. The aims of this study were to provide a new global and
comprehensive evaluation of recent ICU ventilators, tak-
ing into account both their technical performance and a
comprehensive ergonomics evaluation (Fig. 1). fi
Besides technical performance, another major aspect of
a device’s reliability is its usability. Usability is defined as Abstract Background: This study aimed to provide a new global and comprehensive evaluation of recent ICU ventilators tak-
ing into account both technical performances and ergonomics. Methods: Six recent ICU ventilators were evaluated. Technical performances were assessed under two FIO2 lev-
els (100%, 50%), three respiratory mechanics combinations (Normal: compliance [C] = 70 mL cmH2O−1/resistance
[R] = 5 cmH2O L−1 s−1; Restrictive: C = 30/R = 10; Obstructive: C = 120/R = 20), four exponential levels of leaks (from
0 to 12.5 L min−1) and three levels of inspiratory effort (P0.1 = 2, 4 and 8 cmH2O), using an automated test lung. Ergonomics were evaluated by 20 ICU physicians using a global and comprehensive model involving physiological
response to stress measurements (heart rate, respiratory rate, tidal volume variability and eye tracking), psycho-cogni-
tive scales (SUS and NASA-TLX) and objective tasks completion. Results: Few differences in terms of technical performance were observed between devices. Non-invasive ventila-
tion modes had a huge influence on asynchrony occurrence. Using our global model, either objective tasks com-
pletion, psycho-cognitive scales and/or physiological measurements were able to depict significant differences in
terms of devices’ usability. The level of failure that was observed with some devices depicted the lack of adaptation of
device’s development to end users’ requests. Conclusions: Despite similar technical performance, some ICU ventilators exhibit low ergonomics performance and
a high risk of misusage. *Correspondence: erwan.lher@chu‑brest.fr
5 Médecine Intensive et Réanimation, CHRU de Brest, Boulevard Tanguy
Prigent, 29200 Brest Cedex, France
Full list of author information is available at the end of the article Technical performance Measurements were taken using an ASL5000™ lung
simulator (Ingmar, Pittsburgh, PA, USA) under con-
stant room temperature (22 °C), simulator temperature
(37 °C), under dry ambient pressure (ATPD) conditions,
and converted into body temperature and pressure,
saturated (BTPS) as previously described [6, 17]. Tech-
nical performances were assessed under two FIO2
levels (100%, 50%), three respiratory mechanics com-
binations (Normal: resistance [R] = 5 cmH2O L−1 s−1;
compliance [C] = 70 mL cmH2O−1, Restrictive:
R = 10 cmH2O L−1 s−1; C = 30 mL cmH2O−1 and Obstruc‑
tive: R = 20 cmH2O L−1 s−1; C = 120 mL cmH2O−1),
three exponential levels of leaks (L1 = 3.5–4.0 L min−1;
L2 = 5.0–7.0 L min−1; L3 = 9.0–12.5 L min−1) and three
levels of inspiratory effort (P0.1 = 2, 4 and 8 cmH2O) in
volume-controlled continuous mandatory ventilation f
Error was evaluated as the average difference between
set and true dimension value (VT, PEEP, pressure sup-
port). Accuracy was a priori considered for an error value
below 10% for all parameters. Precision of the dimension
was defined as the range value of the dimension, con-
sidering that a narrow range was the more precise. An
asynchrony index greater than 10% was also considered
clinically significant [3, 19]. i
Details about the technical performance evaluation are
provided in the online repository. Tested devices In order to assess ‘efficacy’ of a device, we considered that a global ergonomics evaluation required
evaluating concomitantly ‘efficiency’, ‘engagement’, ‘ease of use’ and ‘tolerance to error’, as these four dimensions may be considered as interdependent. Tolerance to error was evaluated through the objective tasks completion scenarios. While considering that an easy-to-use device should be easily
managed by a skilled physician, but not familiar to that specific device, we took particular attention to included naive subjects in the evaluation. Bench testings do explore the most important technical determinants of the efficiency of a device (tidal volume accuracy, triggering, etc.). Efficiency
assessment might also include interfaces’ performance evaluation. For such sake, we used pupillary diameter variation which can be considered as
a determinant of stress. Other eye-tracking tools such as blinking measurements or heat mapping may also have been used. Engagement during
use of the device was evaluated through the use of psycho-cognitive scales, combined with physiological parameters measurements. Heart rate
variability (HRV) may also have been measured for such sake manufacturers: (1) Dräger V500 (Lubeck, Germany); (2)
Covidien PB980 (Mansfield, MA, USA); (3) Philips V680
(Murrysville, PA, USA); (4) Hamilton S1 (Bonaduz, Swit-
zerland); (5) General Electrics R860 (Fairfield, CT, USA);
and (6) Maquet Servo-U (Göteborg, Sweden). General
characteristics of the devices are provided in the online
repository (Additional file 1: Table S1). manufacturers: (1) Dräger V500 (Lubeck, Germany); (2)
Covidien PB980 (Mansfield, MA, USA); (3) Philips V680
(Murrysville, PA, USA); (4) Hamilton S1 (Bonaduz, Swit-
zerland); (5) General Electrics R860 (Fairfield, CT, USA);
and (6) Maquet Servo-U (Göteborg, Sweden). General
characteristics of the devices are provided in the online
repository (Additional file 1: Table S1). (VC-CMV) and pressure-controlled continuous spon-
taneous ventilation (PC-CSV) [4]. Triggering capabili-
ties, volume and pressurization accuracy were evaluated
under the different respiratory mechanics at standard-
ized respiratory settings. The asynchrony index [18] was
defined as the number of asynchrony events divided by
the total respiratory rate and expressed in percentage [3]. Asynchronies were evaluated under the different respira-
tory mechanics and the three exponential levels of leaks
and inspiratory efforts. Tested devices Six ICU ventilators were evaluated in a dedicated bench
test. All ventilators were provided free of charge by Marjanovic et al. Ann. Intensive Care (2017) 7:68 Page 2 of 13 Fig. 1 The concept of devices’ global evaluation. In order to assess ‘efficacy’ of a device, we considered that a global ergonomics evaluation required
evaluating concomitantly ‘efficiency’, ‘engagement’, ‘ease of use’ and ‘tolerance to error’, as these four dimensions may be considered as interdependent. Tolerance to error was evaluated through the objective tasks completion scenarios. While considering that an easy-to-use device should be easily
managed by a skilled physician, but not familiar to that specific device, we took particular attention to included naive subjects in the evaluation. Bench testings do explore the most important technical determinants of the efficiency of a device (tidal volume accuracy, triggering, etc.). Efficiency
assessment might also include interfaces’ performance evaluation. For such sake, we used pupillary diameter variation which can be considered as
a determinant of stress. Other eye-tracking tools such as blinking measurements or heat mapping may also have been used. Engagement during
use of the device was evaluated through the use of psycho-cognitive scales, combined with physiological parameters measurements. Heart rate
variability (HRV) may also have been measured for such sake Fig. 1 The concept of devices’ global evaluation. In order to assess ‘efficacy’ of a device, we considered that a global ergonomics evaluation required
evaluating concomitantly ‘efficiency’, ‘engagement’, ‘ease of use’ and ‘tolerance to error’, as these four dimensions may be considered as interdependent. Tolerance to error was evaluated through the objective tasks completion scenarios. While considering that an easy-to-use device should be easily
managed by a skilled physician, but not familiar to that specific device, we took particular attention to included naive subjects in the evaluation. Bench testings do explore the most important technical determinants of the efficiency of a device (tidal volume accuracy, triggering, etc.). Efficiency
assessment might also include interfaces’ performance evaluation. For such sake, we used pupillary diameter variation which can be considered as
a determinant of stress. Other eye-tracking tools such as blinking measurements or heat mapping may also have been used. Engagement during
use of the device was evaluated through the use of psycho-cognitive scales, combined with physiological parameters measurements. Heart rate
variability (HRV) may also have been measured for such sake Fig. 1 The concept of devices’ global evaluation. Objective task completionh The ICU physicians had to complete 11 specific tasks for
each ventilator, four mainly dedicated to monitoring and
seven to setting: (a) alarm control (users must shut down
alarms, identify the reason and modify setting to stop
alarms); (b) mode recognition (exact reading of the ven-
tilator mode set by investigator); (c) identify humidifi-
cation system on the screen and modify it; (d) ventilator
setting reading (VT, ventilation rate, PEEP and trigger
value); (e) power on the ventilator; (f) start ventilation; (g)
set inspiratory flow to a value defined by the investigator
(40–80 L min−1); (h) ventilator mode modification; (i) set
cycling to 60%; (j) non-invasive ventilation mode activa-
tion; and (k) ventilator extinction (complete ventilator
powering down). In each group of tests (i.e. monitoring or
setting), tasks were to be performed in a randomized order. The test was a priori considered as a failure if the correct
response was given after more than 120 s, or if the physi-
cians did not provide a correct response or abandoned the
task. Due to technical constraints, we chose not to use a
high-fidelity environment with a manikin, but to perform
measurements with the ventilators connected solely to the
test lung. Besides a task failure rate evaluation, these sce-
narios were also dedicated to enable usability and mental
workload scoring using psycho-cognitive scales. Ergonomics For ergonomics evaluation, we included, as a reference,
the use of a device that was familiar to all physicians
(Avea, Carefusion, San Diego, CA, USA). Page 3 of 13 Marjanovic et al. Ann. Intensive Care (2017) 7:68 ICU physicians involved in ergonomics evaluationif an indicator for interface development, assessment and
comparison. NASA-TLX is a multidimensional tool that
was developed by the National Aeronautics and Space
Administration’s Ames Research Center in 1986 for per-
ceptual mental workload evaluations using the Task Load
Index measurement through three dimensions, depend-
ent on the user’s perception of the task (mental workload,
temporal workload and physical workload) and three
dimensions dependent on the interaction between the
subject and the task itself, which may be mostly related to
the interface (effort, performance and frustration). Each
dimension is rated using a Likert-type scale ranging from
0 to 100. The second part of Task Load Index calculation
intends to create an individual weighting of these dimen-
sions by letting the subjects compare them pairwise, based
on their perceived importance. These 15 comparison pair-
ings thus enable the inter-/intra-individual variability of
the overall score to decrease. The overall workload score
for each subject is composed by multiplying each rating
by the weight given to that factor by that subject. The sum
of the weighted ratings for each task is divided by 15 (the
sum of the weights). The higher the Task Load Index, the
higher the mental workload and the more difficult it is to
use the device. p y
g
Twenty senior ICU physicians from five different ICUs
were included in the evaluation. Each physician tested
3–4 devices in a randomized order; each device was
tested 11 or 12 times. All physicians used the Avea in
their daily clinical practice; we paid particular attention
to the fact that none of them were familiar with the tested
devices (naive subjects), even though some of them were
in some cases familiar with other devices from the same
manufacturer (see Additional file 1: Table S2, Additional
file 1: Table S3). Physiological measurements Several physiological parameters were recorded dur-
ing the completion of the objective tasks. Pupil diameter
modifications were assessed using an eye-tracking system
(SMI ETG 1, SensoMotoric Instruments GmbH, Teltow,
Germany) (Additional file 1: Fig. S1); heart and respira-
tory rate and thoracic volume variations were measured
using a biometric belt (Hexoskin, Montréal, Canada). Analysis consisted of a data treatment by a systems’ acti-
vation count, which corresponded to highly different val-
ues, as compared to baseline. Each of these activations is
numerically integrated in order to evaluate the number of
physiological variations in response to tasks. These acti-
vations are considered to be adequate stress indicators. Psycho‑cognitive scales evaluation Psycho-cognitive scorings were performed immediately
after all objective tasks completion. Asynchrony management (Fig. 4) There was a significant difference in terms of tidal
volume delivery precision between devices (Fig. 2a;
p = 0.0498). All devices except S1 depicted a median
tidal volume value within the 10% error range
(VT = 449 ± 2 mL; ΔDVT = −10.2%). PB980 had the
lowest error in terms of tidal volume delivery, but Servo-
U had the higher precision in response to respiratory
mechanics modifications. V500 and V680 had relatively Mean asynchrony indexes were equal to 31 and 14.5%
under standard or non-invasive ventilatory modes,
respectively, for all devices. All ventilators presented an
asynchrony index of over 10% without using the non-
invasive ventilation mode. While using non-invasive ven-
tilation algorithms, the asynchrony index was lower for
R860 and Servo-U (9.6%) as compared to V500 (14.6%), Fig. 2 Box plot of tidal volume (VT) (a), positive end-expiratory pressure (PEEP) (b) and pressure support (PS) (c). Dotted lines represent the 10% error
range. Black line represents exact VT value delivery. Values are provided as median and interquartile. A p value equal or below 0.05 was considered
significant. *p < 0.05; §p < 0.05 as compared to S1, PB980 and Servo-U. †p < 0.005 as compared to R860, PB980 and Servo-U. There was a significant
difference in terms of VT delivery precision between devices (a; p = 0.0498). All devices except S1 depicted a median VT value within the 10% error
range. PB980 had the lowest error in terms of VT delivery, but Servo-U had the highest precision in response to respiratory mechanics modifica-
tions. V500 and V680 had relatively low error, but low precision in response to respiratory mechanics modifications. While median PEEP delivery was
within the reliability range for all devices (b), V500 was significantly different to the other devices in terms of precision. S1 had the lowest error, but
with rather low precision. Servo-U, PB980 and R860 had low error and high precision. When considering mean values, three devices delivered PS
values higher than the reliability range (Servo-U: 13 ± 6%; PB980: 11 ± 12%; S1: 17 ± 14.4%; p < 0.001). Pressure support delivery was higher than
the 10% error range for two devices (Servo-U and S1). Precision in response to respiratory mechanics modifications was low for PB980, S1 and R860 Fig. 2 Box plot of tidal volume (VT) (a), positive end-expiratory pressure (PEEP) (b) and pressure support (PS) (c). The concept of global ergonomics evaluation The concept of global ergonomics evaluation
A global and comprehensive model for a device’s effi-
cacy evaluation needs to either assess technical perfor-
mances and ergonomics or thus to explore four different
dimensions (Fig. 1). Each of these four dimensions can
be explored separately, but they are all related one to
the other. Tolerance to error was evaluated through the
objective tasks completion scenarios. While considering
that an easy-to-use device should be easily managed by a
skilled physician, but not familiar to that specific device,
we took particular attention to included naive subjects
in the evaluation. Bench testings do explore the most
important technical determinants of the efficiency of a System Usability Scale (SUS) The SUS is a reliability tool,
developed to measure a device’s usability [20]. It consists
of a ten-item questionnaire and assesses usability by dif-
ferent aspects: effectiveness (ability of users to complete
tasks); efficiency (level of resource used in performing
tasks); and satisfaction (subjective reactions to using the
system). The SUS score has a range of 0–100, the highest
score being the best value (‘simple to use’). Mental workload evaluation using the NASA‑TLX Men-
tal workload is a subjective ergonomic measurement and Marjanovic et al. Ann. Intensive Care (2017) 7:68 Page 4 of 13 low error, but low precision in response to respiratory
mechanics modifications. device (tidal volume accuracy, triggering, etc.). Efficiency
evaluation might also include interface’s performance
evaluation. For such sake, we used pupillary diameter
variation which can be considered as a determinant of
stress. Engagement during use of the device was evalu-
ated through the use of psycho-cognitive scales, com-
bined with physiological parameters measurements. Pressurization accuracy (Fig. 2)f Pressurization accuracy differed between devices. Mean
PEEP accuracy was similar between devices, except
for V500 in the obstructive condition (pressurization
error = 18 ± 5%). Three ventilators delivered a mean
pressure support over the 10% error range (Servo-U:
13 ± 6%; PB980: 11 ± 12%; S1: 17 ± 14.4%; p < 0.001). S1 had the lowest error, but with rather low precision. Servo-U, PB980 and R860 had low error and high preci-
sion. V500 and V680 had low precision. Triggering evaluation (Fig. 3)f No differences were observed between devices in terms
of inspiratory triggering. Triggering delay was below
150 ms among the different respiratory mechanics, but
exceeded 200 ms in obstructive conditions, except for
PB980. Triggering pressure presented a large difference
among the devices (p = 0.0004) and was higher for S1
(p = 0.0003) and R860 (p = 0.001). Statistical analysis Parameters were calculated over 10–20 cycles after signal
stabilization and are provided as mean ± SD to calculate
error and as median ± interquartile to evaluate precision
of the dimension, in response to respiratory mechanics
changes. Data were compared using analysis of variance
(ANOVA) and nonparametric Friedman and Wilcoxon
signed-rank test. A p value <0.05 was considered sta-
tistically significant. Statistical analysis was performed
using MedCalc 12.7.4 for Windows (MedCalc software,
Ostend, Belgium). Asynchrony management (Fig. 4) Dotted lines represent the 10% error
range. Black line represents exact VT value delivery. Values are provided as median and interquartile. A p value equal or below 0.05 was considered
significant. *p < 0.05; §p < 0.05 as compared to S1, PB980 and Servo-U. †p < 0.005 as compared to R860, PB980 and Servo-U. There was a significant
difference in terms of VT delivery precision between devices (a; p = 0.0498). All devices except S1 depicted a median VT value within the 10% error
range. PB980 had the lowest error in terms of VT delivery, but Servo-U had the highest precision in response to respiratory mechanics modifica-
tions. V500 and V680 had relatively low error, but low precision in response to respiratory mechanics modifications. While median PEEP delivery was
within the reliability range for all devices (b), V500 was significantly different to the other devices in terms of precision. S1 had the lowest error, but
with rather low precision. Servo-U, PB980 and R860 had low error and high precision. When considering mean values, three devices delivered PS
values higher than the reliability range (Servo-U: 13 ± 6%; PB980: 11 ± 12%; S1: 17 ± 14.4%; p < 0.001). Pressure support delivery was higher than
the 10% error range for two devices (Servo-U and S1). Precision in response to respiratory mechanics modifications was low for PB980, S1 and R860 Marjanovic et al. Ann. Intensive Care (2017) 7:68 Page 5 of 13 Fig. 3 Triggering evaluation according to respiratory mechanics combinations. R: resistance; C: compliance; ΔP: maximal pressure drop required
to trigger inspiration; DT: triggering delay, from the onset of the airway pressure decay (beginning of the patient’s effort) to flow delivery (begin-
ning of ventilator pressurization); DP: pressurization delay, from the airway pressure signal rise to a return to positive pressure; DI: overall inspiratory
delay (DT + DP). Triangle and black line: results for the ‘Normal’ respiratory mechanics; circle and dotted line: results for the ‘Restrictive’ respira-
tory mechanics; cross and grey line: results for the ‘Obstructive’ respiratory mechanics (Normal: resistance [R] = 5 cmH2O L−1 s−1; Compliance
[C] = 70 mL cmH2O−1, Restrictive: R = 10 cmH2O L−1 s−1; C = 30 mL cmH2O−1 and Obstructive: R = 20 cmH2O L−1 s−1; C = 120 mL cmH2O−1). The
figure presents individual results for each ventilator at the different respiratory mechanics combinations. Asynchrony management (Fig. 4) First point of each curve represents inspira-
tory effort initiation; second point represents maximal depressurization (ΔP) before inspiratory pressure increase. There was no significant difference
in terms of DT and DI between ventilators, nor in terms of maximal depressurization (ΔP). DI variability according to respiratory mechanics was
higher for V500 and R860 Fig. 3 Triggering evaluation according to respiratory mechanics combinations. R: resistance; C: compliance; ΔP: maximal pressure drop required
to trigger inspiration; DT: triggering delay, from the onset of the airway pressure decay (beginning of the patient’s effort) to flow delivery (begin-
ning of ventilator pressurization); DP: pressurization delay, from the airway pressure signal rise to a return to positive pressure; DI: overall inspiratory
delay (DT + DP). Triangle and black line: results for the ‘Normal’ respiratory mechanics; circle and dotted line: results for the ‘Restrictive’ respira-
tory mechanics; cross and grey line: results for the ‘Obstructive’ respiratory mechanics (Normal: resistance [R] = 5 cmH2O L−1 s−1; Compliance
[C] = 70 mL cmH2O−1, Restrictive: R = 10 cmH2O L−1 s−1; C = 30 mL cmH2O−1 and Obstructive: R = 20 cmH2O L−1 s−1; C = 120 mL cmH2O−1). The
figure presents individual results for each ventilator at the different respiratory mechanics combinations. First point of each curve represents inspira-
tory effort initiation; second point represents maximal depressurization (ΔP) before inspiratory pressure increase. There was no significant difference
in terms of DT and DI between ventilators, nor in terms of maximal depressurization (ΔP). DI variability according to respiratory mechanics was
higher for V500 and R860 Fig. 3 Triggering evaluation according to respiratory mechanics combinations. R: resistance; C: compliance; ΔP: maximal pressure drop required
to trigger inspiration; DT: triggering delay, from the onset of the airway pressure decay (beginning of the patient’s effort) to flow delivery (begin-
ning of ventilator pressurization); DP: pressurization delay, from the airway pressure signal rise to a return to positive pressure; DI: overall inspiratory
delay (DT + DP). Triangle and black line: results for the ‘Normal’ respiratory mechanics; circle and dotted line: results for the ‘Restrictive’ respira-
tory mechanics; cross and grey line: results for the ‘Obstructive’ respiratory mechanics (Normal: resistance [R] = 5 cmH2O L−1 s−1; Compliance
[C] = 70 mL cmH2O−1, Restrictive: R = 10 cmH2O L−1 s−1; C = 30 mL cmH2O−1 and Obstructive: R = 20 cmH2O L−1 s−1; C = 120 mL cmH2O−1). Asynchrony management (Fig. 4) The
figure presents individual results for each ventilator at the different respiratory mechanics combinations. First point of each curve represents inspira-
tory effort initiation; second point represents maximal depressurization (ΔP) before inspiratory pressure increase. There was no significant difference
in terms of DT and DI between ventilators, nor in terms of maximal depressurization (ΔP). DI variability according to respiratory mechanics was
higher for V500 and R860 V680 (17.5%) and PB980 (18.3%; p < 0.05) (Additional
file 1: Fig. S2). Most frequent asynchronies were pro-
longed cycles and ineffective efforts, ineffective efforts
being most of the time associated with prolonged cycles. always took longer than the predefined 120 s time range. The V500 was the fastest ventilator to power on. Only
36% of the ICU physicians were able to power on Servo-
U and always with over a 1-min delay. No users could set
the inspiratory flow on Servo-U, and only 18% of them
succeeded in the same task with S1. Difficulties in acti-
vating the non-invasive ventilation mode were frequent
with Servo-U, V500 and V680. Servo-U had the worst
global results (tasks failure rate = 42%) compared to our
reference (Avea; tasks failure rate = 13%) (p = 0.12). The
lack of sensitivity of the S1 touch screen proved to be a Objective task completion (Table 1) Of all ventilators, our reference device the Avea had the
best success rate. In our comparison of the six ventilators,
Covidien PB980 had the best results and the Servo-U the
worst. A minority of users could power on Servo-U, but Marjanovic et al. Ann. Intensive Care (2017) 7:68 Page 6 of 13 Fig. 4 Asynchrony index with or without the non-invasive ventila-
tion mode. The asynchrony index (AI) is presented as mean ± SD
and was measured under three different levels of exponential leaks
(L1 = 3.5–4.0 L min−1; L2 = 5.0–7.0 L min−1; L3 = 9.0–12.5 L min−1)
and three levels of inspiratory effort (P0.1 = 2, 4, 8 cmH2O). ‘Standard’
represents measurements performed under PC-CSV using unmodi-
fied manufacturers’ settings in terms of inspiratory and expiratory
triggering; ‘NIV’ represents the measurements performed under the
same conditions, while switching the ventilator to the NIV mode. Dotted line represents the 10% AI clinical level of significance. *p value
<0.001. All ventilators presented an AI over 10% without NIV mode
(‘standard’ invasive PC-CSV setting). Under the same leaks condi-
tions, switching the ventilator to the NIV mode enabled a decrease in
the AI to below a 10% value for the R860 and Servo-U. V500 and S1
measurements did not depict a significant impact of the NIV mode,
while ‘standard’ settings provided rather satisfactory results in terms
of leak management between devices (p < 0.05). V500 caused the highest
pupillary diameter activation and differed significantly
from the reference (p = 0.03) and R860 (p = 0.019). between devices (p < 0.05). V500 caused the highest
pupillary diameter activation and differed significantly
from the reference (p = 0.03) and R860 (p = 0.019). Ergonomics assessmentf While huge effort has been made by manufacturers
to improve technical issues, increasing complexity of
devices may in fact result in design errors. Not only are
the devices’ full capabilities underutilized, but also their
main functions may often be handled improperly. On the radar chart presentation of the NASA-TLX,
except for our reference value (Avea, TLX = 41.6), the
R860 had the lowest TLX value (TLX = 44.7) and V680
had the highest (TLX = 63.2; p = 0.049). The main
dimensions involved in the higher mental workload were
‘performance’ and ‘effort’. Human error has been demonstrated to be a leading
cause of morbidity and death during medical care [10,
24–26]. Many devices have interfaces that are so poorly
designed and difficult to use that they can increase the
risks associated with the medical equipment and device-
induced human error. Human error may be to some
extent inevitable and equally caused by human perfor-
mance and machine performance. In order to limit the Discussion Within this panel of recently available ICU ventilators,
no technical features could be considered as differen-
tiating between devices, while a contrario, ergonomics
and interface features were considered inadequate, thus
increasing the risks of misusage and adverse events. Technical performances 4 Asynchrony index with or without the non-invasive ventila-
tion mode. The asynchrony index (AI) is presented as mean ± SD
and was measured under three different levels of exponential leaks
(L1 = 3.5–4.0 L min−1; L2 = 5.0–7.0 L min−1; L3 = 9.0–12.5 L min−1)
and three levels of inspiratory effort (P0.1 = 2, 4, 8 cmH2O). ‘Standard’
represents measurements performed under PC-CSV using unmodi-
fied manufacturers’ settings in terms of inspiratory and expiratory
triggering; ‘NIV’ represents the measurements performed under the
same conditions, while switching the ventilator to the NIV mode. Dotted line represents the 10% AI clinical level of significance. *p value
<0.001. All ventilators presented an AI over 10% without NIV mode
(‘standard’ invasive PC-CSV setting). Under the same leaks condi-
tions, switching the ventilator to the NIV mode enabled a decrease in
the AI to below a 10% value for the R860 and Servo-U. V500 and S1
measurements did not depict a significant impact of the NIV mode,
while ‘standard’ settings provided rather satisfactory results in terms
of leak management Triggering performances depicted within our study are
similar to those observed in a previous study concern-
ing emergency transport ventilators [8] and tended to
be higher than previously observed [9, 17]. These results
may be explained by different respiratory mechanics and
BTPS conditions [8]. No device enabled a triggering delay
faster than 50 ms, and it exceeded 100 ms for two devices
in normal respiratory mechanics conditions. As already
described, flow or pressure triggering has not varied sig-
nificantly over the last decade [7]. i
During non-invasive ventilation, patient–ventilator
asynchronies are frequent [21, 22] and mainly related
to leaks around the interfaces and/or overassistance. Our mean asynchrony index was close to that observed
in other studies [8, 23]. Non-invasive ventilation algo-
rithms that are implemented in most devices were able
to decrease the asynchrony index significantly and might
thus be systematically turned on during non-invasive
ventilation, in an attempt to limit non-invasive ventila-
tion failure. barrier for some task completion and a significant source
of confusion. The settings reading represented one of the
most difficult tasks, whatever the device. Sensitivity anal-
ysis while deleting powering on and switching off tasks
did not significantly modify the overall results. Psycho‑cognitive scales measurements (Figs. 5, 6) NASA-TLX and SUS scorings are presented in Fig. 5. V680 for NASA-TLX, V680 and S1 for SUS were the only
devices to differ significantly from the reference (Avea). Except for the reference, not a single device reached an
SUS score equal or higher than 68. Technical performances While volume delivery and pressurization accuracy are
critical issues, few differences were observed between
devices. However, volume delivery and pressurization
errors and precisions were important for some ventila-
tors. In all cases, volume delivery was lower than that
expected, as already observed [18]. Fig. 4 Asynchrony index with or without the non-invasive ventila-
tion mode. The asynchrony index (AI) is presented as mean ± SD
and was measured under three different levels of exponential leaks
(L1 = 3.5–4.0 L min−1; L2 = 5.0–7.0 L min−1; L3 = 9.0–12.5 L min−1)
and three levels of inspiratory effort (P0.1 = 2, 4, 8 cmH2O). ‘Standard’
represents measurements performed under PC-CSV using unmodi-
fied manufacturers’ settings in terms of inspiratory and expiratory
triggering; ‘NIV’ represents the measurements performed under the
same conditions, while switching the ventilator to the NIV mode. Dotted line represents the 10% AI clinical level of significance. *p value
<0.001. All ventilators presented an AI over 10% without NIV mode
(‘standard’ invasive PC-CSV setting). Under the same leaks condi-
tions, switching the ventilator to the NIV mode enabled a decrease in
the AI to below a 10% value for the R860 and Servo-U. V500 and S1
measurements did not depict a significant impact of the NIV mode,
while ‘standard’ settings provided rather satisfactory results in terms
of leak management Fig. 4 Asynchrony index with or without the non-invasive ventila-
tion mode. The asynchrony index (AI) is presented as mean ± SD
and was measured under three different levels of exponential leaks
(L1 = 3.5–4.0 L min−1; L2 = 5.0–7.0 L min−1; L3 = 9.0–12.5 L min−1)
and three levels of inspiratory effort (P0.1 = 2, 4, 8 cmH2O). ‘Standard’
represents measurements performed under PC-CSV using unmodi-
fied manufacturers’ settings in terms of inspiratory and expiratory
triggering; ‘NIV’ represents the measurements performed under the
same conditions, while switching the ventilator to the NIV mode. Dotted line represents the 10% AI clinical level of significance. *p value
<0.001. All ventilators presented an AI over 10% without NIV mode
(‘standard’ invasive PC-CSV setting). Under the same leaks condi-
tions, switching the ventilator to the NIV mode enabled a decrease in
the AI to below a 10% value for the R860 and Servo-U. V500 and S1
measurements did not depict a significant impact of the NIV mode,
while ‘standard’ settings provided rather satisfactory results in terms
of leak management Fig. Physiological measurements (Fig. 7) To the best of our knowledge, this is first time that such
an innovative ergonomics evaluation of ICU mechani-
cal ventilators has been performed, globally integrating
the four main dimensions that enable a comprehensive
approach to the problem: 1—tolerance to error; 2—ease
of use; 3—efficiency; and 4—engagement. Tolerance to
error may be directly linked to efficiency and ease of use
to engagement. While all four dimensions may be consid-
ered independently, they are in fact related one to each
other (Fig. 1). Most previous ergonomics evaluations have mainly focused on tolerance to error, while the three
other dimensions were often missing data.h number of errors, computing technology and human–
machine interface development should be designed to
correspond to human characteristics of reasoning and
memory constraints [11]. It is also well known that the
working memory of humans is limited and that the num-
ber of variables depicted on screens is excessive. This
results in a large cognitive load (i.e. mental workload) on
the user, which is also a determinant of human error [27]. An interface with a human-centred design increases effi-
ciency and satisfaction and decreases the rate of medical
error. While these data are integrated in the ventilators’
interface development by manufacturers, and while ergo-
nomics are as essential as technical performances, very
few studies have assessed the ergonomics, and many were
limited to timed tasks and subjective evaluation [13–15]. To the best of our knowledge, this is first time that such number of errors, computing technology and human–
machine interface development should be designed to
correspond to human characteristics of reasoning and
memory constraints [11]. It is also well known that the
working memory of humans is limited and that the num-
ber of variables depicted on screens is excessive. This
results in a large cognitive load (i.e. mental workload) on
the user, which is also a determinant of human error [27]. An interface with a human-centred design increases effi-
ciency and satisfaction and decreases the rate of medical
error. While these data are integrated in the ventilators’
interface development by manufacturers, and while ergo-
nomics are as essential as technical performances, very
few studies have assessed the ergonomics, and many were
limited to timed tasks and subjective evaluation [13–15].i The integration of pupil diameter measurement, heart
and respiratory rate or tidal volume activation to assess
ergonomics are new data in the ICU field. Physiological measurements (Fig. 7) For all parameters, our reference value depicted sig-
nificantly less activation. Pupillary diameter, respiratory
rate and tidal volume activations significantly differed Marjanovic et al. Ann. Intensive Care (2017) 7:68 Page 7 of 13 Table 1 Objective tasks completion rate
Device
Power on (%)
Start ventilation (%) Inspiratory flow
setting (%)
Ventilatory mode
modification (%)
Cycling setting (%)
NIV mode activation
(%)
Carefusion Avea
92
100
100
100
75
75
Dräger V500
100
92
100
100
92
50
Covidien PB980
100
100
100
91
100
100
Philips V680
100
100
50§
91
91
40†
Hamilton S1
91
91
18§
82
91
91
GE R860
100
100
100
91
100
91
Maquet Servo-U
36*
100
0¤
100
45‡
27†
Device
Ventilator offset
(%)
Alarm shut
down (%)
Mode recognition
(%)
Humidification sys-
tem recognition (%)
Settings reading (%)
Overall
Carefusion Avea
100
100
100
75
42
87%/85%
Dräger V500
100
75
100
42
25
80%/75%
Covidien PB980
73*
64
100
73
45
86%/86%
Philips V680
100
82
100
55
18
75%/70%
Hamilton S1
100
91
100
36
27
74%/70%
GE R860
100
64
100
9‖
27
80%/76%
Maquet Servo-U
100
82
100
0‖
45
58%#/55%‖ Table 1 Objective tasks completion rate Results are presented as the different objective tasks success rate, expressed as the percentage of successful attempts. ICU’s physicians with mechanical ventilation’s
knowledge had to complete 11 specific tasks of variable clinical importance for each ventilator, 4 mainly dedicated to monitoring and 7 to setting. Overall provides
results of the entire bunch of test (first value), or after powering on/switching off tasks exclusion (second value) number of errors, computing technology and human–
machine interface development should be designed to
correspond to human characteristics of reasoning and
memory constraints [11]. It is also well known that the
working memory of humans is limited and that the num-
ber of variables depicted on screens is excessive. This
results in a large cognitive load (i.e. mental workload) on
the user, which is also a determinant of human error [27]. An interface with a human-centred design increases effi-
ciency and satisfaction and decreases the rate of medical
error. While these data are integrated in the ventilators’
interface development by manufacturers, and while ergo-
nomics are as essential as technical performances, very
few studies have assessed the ergonomics, and many were
limited to timed tasks and subjective evaluation [13–15]. Physiological measurements (Fig. 7) Compared to
subjective psychological measurements, these are objec-
tive data that allow the estimation of the physiologi-
cal stress induced by a device’s interface and an indirect
assessment of the interface’s usability.h The objective tasks results are often considered as the
most representative of the devices’ ergonomics’ differ-
ences. Even if we entirely agree with the fact that not all
scenarios may have the same importance, it is still sur-
prising that some ventilators could not be powered on
by a majority of physicians or that the NIV mode could
not be easily activated. Excluding powering on/off tests
from analysis, considering that these may be very differ-
ent tasks that have been voluntarily been made difficult
by the manufacturers for safety reasons, did not modify
the overall results. While it may be one of the main tasks
routinely performed, ventilator setting readings had the
worst results of all tasks, probably because of the absence
of a homogenized terminology among manufacturers. As
already observed in another recent study, the lack of sen-
sitivity of the S1 touch screen was specifically considered To the best of our knowledge, this is first time that such
an innovative ergonomics evaluation of ICU mechani-
cal ventilators has been performed, globally integrating
the four main dimensions that enable a comprehensive
approach to the problem: 1—tolerance to error; 2—ease
of use; 3—efficiency; and 4—engagement. Tolerance to
error may be directly linked to efficiency and ease of use
to engagement. While all four dimensions may be consid-
ered independently, they are in fact related one to each
other (Fig. 1). Most previous ergonomics evaluations Marjanovic et al. Ann. Intensive Care (2017) 7:68 Page 8 of 13 Fig. 5 Task Load Index and System Usability Scale scores. Dotted line represents the mean value across all scores (Avea excluded). SUS consists of
a ten-item questionnaire and assesses usability from different aspects: effectiveness (ability of users to complete tasks); efficiency (level of resource
used in performing tasks); and satisfaction (subjective reactions to using the system). SUS score has a range of 0–100, the highest score being
the best value (‘simple to use’). NASA-TLX is a multidimensional tool developed for mental workload evaluation. Physiological measurements (Fig. 7) It explores three dimensions
dependent on user perception of the task (mental workload, temporal workload and physical workload) and three dimensions dependent on the
interaction between the subject and the task itself, which may be mostly related to the interface (effort, performance and frustration). An individual
weighting of these dimensions by letting the subjects compare them pairwise enables a decrease in the inter-/intra-individual variability of the
overall score. The higher the TLX, the lower the ergonomics. Our reference device (Avea) had the best TLX and SUS scores, and V680 the worst
(p = 0.049). For usability (SUS), a difference between our reference device (Avea) and S1 was also observed. *p value <0.05 Fig. 5 Task Load Index and System Usability Scale scores. Dotted line represents the mean value across all scores (Avea excluded). SUS consists of
a ten-item questionnaire and assesses usability from different aspects: effectiveness (ability of users to complete tasks); efficiency (level of resource
used in performing tasks); and satisfaction (subjective reactions to using the system). SUS score has a range of 0–100, the highest score being
the best value (‘simple to use’). NASA-TLX is a multidimensional tool developed for mental workload evaluation. It explores three dimensions
dependent on user perception of the task (mental workload, temporal workload and physical workload) and three dimensions dependent on the
interaction between the subject and the task itself, which may be mostly related to the interface (effort, performance and frustration). An individual
weighting of these dimensions by letting the subjects compare them pairwise enables a decrease in the inter-/intra-individual variability of the
overall score. The higher the TLX, the lower the ergonomics. Our reference device (Avea) had the best TLX and SUS scores, and V680 the worst
(p = 0.049). For usability (SUS), a difference between our reference device (Avea) and S1 was also observed. *p value <0.05 of differences in terms of task completion perceptions
among users while using these parameters. Importantly,
while the evolution of physiological parameters may not
provide comparable results to those obtained with the
psycho-cognitive scores, they are consistent with the
objective task completion rate results. by the participants as responsible for an increased mental
workload and higher rates of task failures [28]. Fig. 6 Radar chart of National Aeronautics and Space Administration—Task Load Index for each ventilator. The radar chart of the NASA-TLX indi-
cates both the overall mental workload evaluation (TLX value) and the different dimensions that are evaluated. Three dimensions are dependent
on user perception of the task (mental workload, temporal workload and physical workload) and three dimensions dependent on the interaction
between the subject and the task itself, which may be mostly related to the interface (effort, performance and frustration). The larger the area of the
radar chart, the higher the TLX and thus the mental workload, and the lower the ergonomics. Values of the TLX score are indicated for each ventila-
tor, our reference value being depicted in the upper left. Our reference value (Avea, in orange) had the lowest mental workload value (TLX = 41.6),
thus depicting the potential influence of experience on mental workload. For this reason, it is strictly mandatory to compare measurements per-
formed on naïve subjects. R860 had the lowest TLX value, and V680 had the highest (*p = 0.049). Dimensions of the mental workload that seemed
to require the most important improvements were performance and efforts Fig. 6 Radar chart of National Aeronautics and Space Administration—Task Load Index for each ventilator. The radar chart of the NASA-TLX indi-
cates both the overall mental workload evaluation (TLX value) and the different dimensions that are evaluated. Three dimensions are dependent
on user perception of the task (mental workload, temporal workload and physical workload) and three dimensions dependent on the interaction
between the subject and the task itself, which may be mostly related to the interface (effort, performance and frustration). The larger the area of the
radar chart, the higher the TLX and thus the mental workload, and the lower the ergonomics. Values of the TLX score are indicated for each ventila-
tor, our reference value being depicted in the upper left. Our reference value (Avea, in orange) had the lowest mental workload value (TLX = 41.6),
thus depicting the potential influence of experience on mental workload. For this reason, it is strictly mandatory to compare measurements per-
formed on naïve subjects. R860 had the lowest TLX value, and V680 had the highest (*p = 0.049). Dimensions of the mental workload that seemed
to require the most important improvements were performance and efforts Fig. Physiological measurements (Fig. 7) The physi-
cians praised the Servo-U interface, but the interface also
tended to induce high mental workload during specific
tasks, thus generating frustration and higher task failure
rates.h The System Usability Scale [20] and NASA Task Load
Index [35, 36] are validated psycho-cognitive tools to
assess devices’ interface.h The pupillary diameter variation is linked to mental
workload and is used to assess cognitive skills [29, 30]. However, we must consider the variability related to the
light reflex induced by the laboratory environment and
the devices themselves [31]. To some extent, this could
explain results from the V500 that has a screen luminos-
ity that is higher than that of other devices. Heart and
respiratory rates and/or tidal volume variations are linked
to emotional behaviour [32–34]. The better results that
were observed with the Avea can be explained by the fact
that this device was well known to all participants. Our
results on the other devices clearly enable the depiction The SUS is a very easy scale to administer to partici-
pants. It can be used on small sample sizes with reliable
results, and it can effectively differentiate between usable
and unusable systems. A SUS score above 68 would be
considered above average and anything below 68 is con-
sidered below average. The NASA-TLX is a flexible, well-established and
widely used multidimensional assessment tool that
enables quick and easy workload estimation in order Marjanovic et al. Ann. Intensive Care (2017) 7:68 Page 9 of 13 6 Radar chart of National Aeronautics and Space Administration—Task Load Index for each ventilator. The radar chart of the NASA-TLX indi-
cates both the overall mental workload evaluation (TLX value) and the different dimensions that are evaluated. Three dimensions are dependent
on user perception of the task (mental workload, temporal workload and physical workload) and three dimensions dependent on the interaction
between the subject and the task itself, which may be mostly related to the interface (effort, performance and frustration). The larger the area of the
radar chart, the higher the TLX and thus the mental workload, and the lower the ergonomics. Values of the TLX score are indicated for each ventila-
tor, our reference value being depicted in the upper left. Our reference value (Avea, in orange) had the lowest mental workload value (TLX = 41.6),
thus depicting the potential influence of experience on mental workload. For this reason, it is strictly mandatory to compare measurements per-
formed on naïve subjects. R860 had the lowest TLX value, and V680 had the highest (*p = 0.049). Dimensions of the mental workload that seemed
to require the most important improvements were performance and efforts Marjanovic et al. Ann. Intensive Care (2017) 7:68 Page 10 of 13 Fig. 7 Box plot of physiological measurements and eye-tracking activations. Several physiological parameters were recorded during objective
tasks completion. These parameters were evaluated while detecting statistically different values, as compared to baseline. Each of these detections
(‘activations’) is numerically integrated in order to evaluate the number of physiological variations in response to tasks. These activations are consid-
ered to be adequate stress indicators. The number of activations is represented as median and interquartile. Dotted line is mean of activations for all
ventilators, during all tests. For all parameters, our reference value depicted significantly fewer activations, thus validating our experimental concept. Significant papillary diameter, respiratory rate and tidal volume activations were observed for several devices (at least V500 and Servo-U). *p value
<0.05; #p value <0.005; §p value <0.0005 Fig. 7 Box plot of physiological measurements and eye-tracking activations. Several physiological parameters were recorded during objective
tasks completion. These parameters were evaluated while detecting statistically different values, as compared to baseline. Each of these detections
(‘activations’) is numerically integrated in order to evaluate the number of physiological variations in response to tasks. These activations are consid-
ered to be adequate stress indicators. The number of activations is represented as median and interquartile. Dotted line is mean of activations for all
ventilators, during all tests. For all parameters, our reference value depicted significantly fewer activations, thus validating our experimental concept. Significant papillary diameter, respiratory rate and tidal volume activations were observed for several devices (at least V500 and Servo-U). *p value
<0.05; #p value <0.005; §p value <0.0005 describes how much physical activity is required (e.g. pushing, pulling, turning). Temporal demand describes
how much time pressure is perceived to fulfil the task
(was it slow and leisurely? Or rapid and frantic?). Effort
describes how hard the task is to be fulfilled (mentally
and physically) in order to accomplish the required level
of performance. Performance describes how satisfied the
subject feels or whether he/she thinks they were suc-
cessful in accomplishing the goals. Frustration describes
how insecure, discouraged or irritated the subject feels
after accomplishing the task. The subscale rating enables
inter-/intra-individual variability to be decreased, thus
enabling the number of subjects in the experiment to be
reduced. to assess a task or a system. It has been used in a great
variety of domains and is considered as one of the most
reliable questionnaires to measure workload in a health-
care setting. The higher the weighed TLX, the higher
the mental workload and the more ‘difficult to use’ is the
device. Each individual dimension can also be considered
on its own, either those dependent on users’ perception
of the task (mental workload, temporal workload and
physical workload) or those dependent on the interac-
tion between the subject and the task itself, which may be
mostly related to the interface (effort, performance and
frustration). Mental demand describes how much men-
tal and perceptual activity is required to perform the task
(e.g. thinking, deciding, calculating). Physical demand to assess a task or a system. It has been used in a great
variety of domains and is considered as one of the most
reliable questionnaires to measure workload in a health-
care setting. The higher the weighed TLX, the higher
the mental workload and the more ‘difficult to use’ is the
device. Each individual dimension can also be considered
on its own, either those dependent on users’ perception
of the task (mental workload, temporal workload and
physical workload) or those dependent on the interac-
tion between the subject and the task itself, which may be
mostly related to the interface (effort, performance and
frustration). Mental demand describes how much men-
tal and perceptual activity is required to perform the task
(e.g. thinking, deciding, calculating). Physical demand Page 11 of 13 Page 11 of 13 Marjanovic et al. Ann. Intensive Care (2017) 7:68 human behaviour under test may be significantly affected
by the context and set-up of the experiment. However,
while we only included experts, it would have been dif-
ficult to reach our experimental goals while also trying
to run after a more important degree of immersion that
may not be necessary with these types of physicians. A
simulated condition may never reproduce all the com-
plexities of the interactions between a patient, a clinician
and a ventilator, especially if the tester is an experienced
clinician [39]. There are many techniques available for
usability evaluation, such as cognitive walk-through,
expert reviews, focus groups, Delphi technique, heuris-
tic evaluation or objective timed tasks completion, all of
them providing different information [38]. To the best
of our knowledge, our study is the only one to provide a
global and complete ergonomics evaluation, taking into
account different techniques. Third, we may also consider
that the small number of senior ICU physicians that were
included in our study does not enable firm conclusions
to be drawn. Considering the design of the ergonomics
evaluation, it required a huge amount of dedicated time
from the physicians to undergo the different scenarios
and various measurements for the experimental team. Moreover, none of them were familiar with the six tested
devices, which exacerbated the difficulty in recruitment. It was therefore unrealistic to use more testers, and such
a drawback also tended to be limited by the use of a
device that was known to everyone as a comparison and
by the fact that we included physicians from five different
ICUs. The pairwise comparison that is performed while
using the NASA-TLX also limits inter-/intra-individual
variability. Finally, the use of the Avea as a ‘reference’ also
depicts a specific limitation about the use of subjective
psycho-cognitive scales. The better results of the Avea,
with both the SUS and the NASA-TLX, clearly indicate
that these values may be highly influenced by previous
experience. Such a bias was limited within our evaluation
by the fact that, in an attempt to assess the ease of use, we
only included naive subjects in order to limit the impact
of such experience on the evaluation. Precedent studies have shown the influence of expe-
rience on SUS scores [37], and the better results of the
Avea can clearly be related to the users’ knowledge of and
experience with this device and not specifically to a bet-
ter interface. Given the overall expertise of all the physi-
cians from the five ICUs with this device, it was used in
the comparison as a reference value. When considering both psycho-cognitive assess-
ment tools, two devices (V680 and S1) could be consid-
ered as below our reference device in terms of usability
and induced mental workload. In terms of usability, all
devices except the R860 and the Servo-U were equal to or
below a SUS value of 60, far below the acceptable average
value of 68, which may enable us to consider that from
an ergonomics point of view, a huge amount of work has
to be done to improve the device’s usability. With regard
to the other ventilators, the SUS and NASA-TLX val-
ues did not differ, which corresponds with physiological
analyses. If devices’ interfaces are globally equivalent, the
level of failure observed for some devices, combined with
the high induced mental workload and the low usability
score, clearly depicts a lack of adaptation of the device’s
development to end users. Considering our results and
the impact of tasks on dimensions like performance
and effort for some devices, manufacturers may primar-
ily focus on interface simplification and rationalization,
immediately providing the most important settings and
alarms on a first screen, leaving expert settings to a sec-
ond one. However, given individual physicians’ hetero-
geneity, the perfect ventilator may be a difficult goal to
achieve, and even with experience, some element of frus-
tration and/or temporal workload may still occur, as with
our reference device. Limitations As with other bench tests, the main limitations of our
study may concern the inability to extrapolate our results
to the real clinical situation. First, our technical evalua-
tion was performed on a model, which cannot mimic
the complexity of all interactions between a patient and
a ventilator. The ASL5000 is a simulator and it remains
different from patients, mainly because the spontane-
ous inspiratory profile is not modified by pressurization
during the inspiratory phase. However, the bench simu-
lates most other situations and combinations that can be
encountered in the clinical field. Second, the objective
and subjective ergonomics measurements were assessed
during standardized conditions that may be considered
as different from real-life conditions. In order to be able
to use various physiological sensors during the ergonom-
ics evaluation, we chose not to use a high-fidelity envi-
ronment with a manikin. We do agree with the fact that Availability of data and materials All data generated or analysed during this study are included in this published
article and its supplementary information files. 21. Vignaux L, Vargas F, Roeseler J, Tassaux D, Thille AW, Kossowsky MP,
et al. Patient-ventilator asynchrony during non-invasive ventilation
for acute respiratory failure: a multicenter study. Intensive Care Med. 2009;35:840–6. Received: 3 December 2016 Accepted: 29 May 2017 Received: 3 December 2016 Accepted: 29 May 2017 23. Carteaux G, Lyazidi A, Cordoba-Izquierdo A, Vignaux L, Jolliet P, Thille
AW, et al. Patient-ventilator asynchrony during noninvasive ventilation: a
bench and clinical study. Chest. 2012;142:367–76. y
24. Giraud T, Dhainaut JF, Vaxelaire JF, Joseph T, Journois D. Bleichner G
Iatrogenic complications in adult intensive care units: a prospective two-
center study. CCM. 1993;21:40–51. Conclusionsh The choice of an ‘ideal’ ventilatory device is a difficult
task that may concomitantly consider technical perfor-
mances and ergonomics. While technical bench tests are
essential to assess technical performances and a ventila-
tor’s accuracy, a global ergonomics evaluation, taking into
account different variables and dimensions, is crucial to
enable physicians to focus on their patients, rather than
on technological problems. Despite significant techno-
logical improvements, several ICU ventilators do exhibit
low ergonomics performance and a high risk of misusage. Page 12 of 13 Page 12 of 13 Marjanovic et al. Ann. Intensive Care (2017) 7:68 Competing interests Erwan L’Her has been a consultant for Air Liquide Medical Systems, Novartis
and Smiths Medical. He is the co-founder of Oxy’nov Inc., a spin-off company
from the Universite Laval, Quebec, dedicated to automation in critical care. Nicolas Marjanovic has received honorarium from Weinmann Emergency and
Air Liquide Medical Systems for lectures. 18. Chao DC, Scheinhorn DJ, Stearn-Hassenpflug M. Patient-ventilator trigger
asynchrony in prolonged mechanical ventilation. Chest. 1997;112:1592–9. 19. Vitacca M, Bianchi L, Zanotti E, Vianello A, Barbano L, Porta R, et al. Assess-
ment of physiologic variables and subjective comfort under different
levels of pressure support ventilation. Chest. 2004;126:851–9. 20. Brooke J. SUS: A quick and dirty usability scale. In: Jordan PW, Weerd-
meester B, Thomas A, McLelland IL, editors. Usability evaluation in
industry. London: Taylor and Francis; 1996. Abbreviations ICU: intensive care unit; NIV: non-invasive ventilation; C: compliance; R:
resistance; VT: tidal volume; PEEP: positive end-expiratory pressure; VC-CMV:
volume-controlled continuous mandatory ventilation; PC-CSV: pressure-con-
trolled continuous spontaneous ventilation; BTPS: body temperature, pressure
and saturated; SUS: System Usability Scale; NASA-TLX: National Aeronautics
and Space Administration—Task Load Index. 9. Boussen S, Gainnier M, Michelet P. Evaluation of ventilators used
during transport of critically Ill patients: a bench study. Respir Care. 2013;58:1911–22. 9. Boussen S, Gainnier M, Michelet P. Evaluation of ventilators used
during transport of critically Ill patients: a bench study. Respir Care. 2013;58:1911–22. 10. Institute of Medicine. To err is human: building a safer health system. Washington: The National Academies Press; 1999. g
11. Horsky J, Zhang J, Patel VL. To err is not entirely human: complex technol-
ogy and user cognition. J Biomed Inform. 2005;38:264–6. Author details
1 1 Urgences Adultes/SAMU 86, CHU de Poitiers, 86000 Poitiers Cedex, France. 2 ABS‑Lab, Laboratoire d’Anatomie, Biomécanique et Simulation, Université
de Poitiers, Rue de la Milétrie, 86000 Poitiers Cedex, France. 3 B-Com Techni-
cal Research Institute, 29200 Brest Cedex, France. 4 CeSim/LaTIM INSERM
UMR 1101, Université de Bretagne Occidentale, Rue Camille Desmoulins,
29200 Brest Cedex, France. 5 Médecine Intensive et Réanimation, CHRU de
Brest, Boulevard Tanguy Prigent, 29200 Brest Cedex, France. 14. Uzawa Y, Yamada Y, Suzukawa M. Evaluation of the user interface simplic-
ity in the modern generation of mechanical ventilators. Respir Care. 2008;53:329–37. 15. Vignaux L, Tassaux D, Jolliet P. Evaluation of the user-friendliness of
seven new generation intensive care ventilators. Intensive Care Med. 2009;35:1687–91. 29200 Brest Cedex, France. 5 Médecine Intensive et Réanimation, CHRU de
Brest, Boulevard Tanguy Prigent, 29200 Brest Cedex, France. 16. Gonzalez-Bermejo J, Laplanche V, Husseini FE, Duguet A, Derenne JP,
Similowski T. Evaluation of the user-friendliness of 11 home mechanical
ventilators. Eur Respir J. 2006;27:1236–43. Acknowledgements
h
h
ld l k The authors would like to acknowledge the help of all ICU Physicians from
Brest, Morlaix and Quimper who made this study possible. 17. Garnier M, Quesnel C, Fulgencio JP, Degrain M, Carteaux G, Bonnet F, et al. Multifaceted bench comparative evaluation of latest intensive care unit
ventilators. Br J Anaesth. 2015;115:89–98. Authors’ contributions 12. Richard JC, Kacmarek RM. ICU mechanical ventilators, technological
advances vs. user friendliness: the right picture is worth a thousand
numbers. Intensive Care Med. 2009;35:1662–3. ELH designed the study. ELH, NM, ADS and GJ acquired data. ELH and NM
wrote the manuscript. All authors read and approved the final manuscript. 13. Templier F, Miroux P, Dolveck F, Descatha A, Goddet NS, Jeleff C, et al. Evaluation of the ventilator-user interface of 2 new advanced compact
transport ventilators. Respir Care. 2007;52:1701–9. Additional file 4. Chatburn RL, El-Khatib M, Mireles-Cabodevila E. A taxonomy for mechani-
cal ventilation: 10 fundamental maxims. Respir Care. 2014;59:1747–63. Additional file
Additional file 1: Table S1. Ventilator’s general characteristics. Table
S2. Randomisation table for device’s testings. Table S3. Participants’
list and cumulated knowledge. Figure S1. Pupillar diameter variation
measurements. Figure S2. Overall inspiratory triggering delay. Figure S3. Asynchrony types for each device in the noninvasive ventilation mode. 5. Richard JC, Carlucci A, Breton L, Langlais N, Jaber S, Maggiore S, et al. Bench testing of pressure support ventilation with three different genera-
tions of ventilators. Intensive Care Med. 2002;28:1049–57. Additional file 1: Table S1. Ventilator’s general characteristics. Table
S2. Randomisation table for device’s testings. Table S3. Participants’
list and cumulated knowledge. Figure S1. Pupillar diameter variation
measurements. Figure S2. Overall inspiratory triggering delay. Figure S3. Asynchrony types for each device in the noninvasive ventilation mode. 6. Lyazidi A, Thille AW, Carteaux G, Galia F, Brochard L, Richard JC. Bench
test evaluation of volume delivered by modern ICU ventilators during
volume-controlled ventilation. Intensive Care Med. 2010;36:2074–80. 7. Thille AW, Lyazidi A, Richard JC, Galia F, Brochard L. A bench study of
intensive-care-unit ventilators: new versus old and turbine-based versus
compressed gas-based ventilators. Intensive Care Med. 2009;35:1368–76. 8. L’Her E, Roy A, Marjanovic N. Bench-test comparison of 26 emergency
and transport ventilators. Crit Care. 2014;18:506. References 1. The Acute Respiratory Distress Syndrome Network. Ventilation with
lower tidal volumes as compared with traditional tidal volumes for
acute lung injury and the acute respiratory distress syndrome. NEJM. 2000;342:1301–8. y
25. Bracco D, Favre JB, Bissonnette B, Wasserfallen JB, Revelly JP, Ravussin
P, et al. Human errors in a multidisciplinary intensive care unit: a 1-year
prospective study. Intensive Care Med. 2001;27:137–45. 26. Garrouste-Orgeas M, Philippart F, Bruel C, Max A, Lau N, Misset B. Over-
view of medical errors and adverse events. Ann Intensive Care. 2012;2:2. 2. Mercat A, Richard JC, Vielle B, Jaber S, Osman D, Diehl JL, et al. Positive
end-expiratory pressure setting in adults with acute lung injury and
acute respiratory distress syndrome: a randomized controlled trial. JAMA
2008;299:646–55. 2. Mercat A, Richard JC, Vielle B, Jaber S, Osman D, Diehl JL, et al. Positive
end-expiratory pressure setting in adults with acute lung injury and
acute respiratory distress syndrome: a randomized controlled trial. JAMA. 2008;299:646–55. view of medical errors and adverse events. Ann Intensive Care. 2012;2:2. 27. Cowan N. The magical number 4 in short-term memory: a reconsidera-
tion of mental storage capacity. Behav Brain Sci. 2001;24:87–114 (discus-
sion 114–185) 27. Cowan N. The magical number 4 in short-term memory: a reconsidera-
tion of mental storage capacity. Behav Brain Sci. 2001;24:87–114 (discus-
sion 114–185). Publisher’s Note Springer Nature remains neutral with regard to jurisdictional claims in pub-
lished maps and institutional affiliations. Springer Nature remains neutral with regard to jurisdictional claims in pub-
lished maps and institutional affiliations. 22. Vignaux L, Tassaux D, Jolliet P. Performance of noninvasive ventilation
modes on ICU ventilators during pressure support: a bench model study. Intensive Care Med. 2007;33:1444–51. Marjanovic et al. Ann. Intensive Care (2017) 7:68 sion 114–185). 3. Thille AW, Rodriguez P, Cabello B, Lellouche F, Brochard L. Patient-venti-
lator asynchrony during assisted mechanical ventilation. Intensive Care
Med. 2000;32:1515–22. 28. Morita PP, Weinstein PB, Flewwelling CJ, Banez CA, Chiu TA, Iannuzzi M,
et al. The usability of ventilators: a comparative evaluation of use safety
and user experience. Crit Care. 2016;20:263. 28. Morita PP, Weinstein PB, Flewwelling CJ, Banez CA, Chiu TA, Iannuzzi M,
et al. The usability of ventilators: a comparative evaluation of use safety
and user experience. Crit Care. 2016;20:263. Page 13 of 13 Marjanovic et al. Ann. Intensive Care (2017) 7:68 Page 13 of 13 Page 13 of 13 35. Hart SG, Staveland LE. Development of NASA-TLX (Task Load Index):
results of empirical and theoretical research. In: Peter AH, Najmedin M,
editors. Advances in psychology. Amsterdam: North-Holland; 1988. p. 139–83. 29. Hess EH, Polt JM. Pupil size in relation to mental activity during simple
problem-solving. Science. 1964;143:1190–2. 30. Beatty J, Kahneman D. Pupillary changes in two memory tasks. Psychon
Sci. 2013;5:371–2. 36. Hart S. Nasa-Task Load Index (Nasa-TLX); 20 years later. In: Hart S (ed) Book
Nasa-Task Load Index (Nasa-TLX); 20 years later. City;2006. 31. Chen S, Epps J, Chen F. An investigation of pupil-based cognitive load
measurement with low cost infrared webcam under light reflex interfer-
ence. In: Conference proceedings: annual international conference of the
IEEE engineering in medicine and biology society IEEE engineering in
medicine and biology society annual conference. 2013. p. 3202–05. 37. McLellan S, Muddimer A, Peres SC. The effect of experien 37. McLellan S, Muddimer A, Peres SC. The effect of experience on Sys
Usability Scale ratings. J Usability Stud. 2012;7:56–67. Usability Scale ratings. J Usability Stud. 2012;7:56–67. 38. Martin JL, Norris BJ, Murphy E, Crowe JA. Medical device development:
the challenge for ergonomics. Appl Ergon. 2008;39:271–83. 32. Boiten FA. The effects of emotional behaviour on components of the
respiratory cycle. Biol Psychol. 1998;49:29–51. 39. Alessi SM. Fidelity in the design of instructional simulations. J Comput
Based Instruct. 1988;15:40–7. 33. Boiten FA, Frijda NH, Wientjes CJ. Emotions and respiratory patterns:
review and critical analysis. Int J Psychophysiol. 1994;17:103–28. 34. Anttonen J, Surakka V. Emotions and heart rate while sitting on a chair. In:
Anttonen J, Surakka V, editors. Book emotions and heart rate while sitting
on a chair. New York: ACM; 2005. p. 491–9.
|
https://openalex.org/W2601266996
|
https://pure.au.dk/ws/files/142968639/1_s2.0_S1053811917302501_main.pdf
|
English
| null |
Expectation violation and attention to pain jointly modulate neural gain in somatosensory cortex
|
NeuroImage
| 2,017
|
cc-by
| 14,530
|
Francesca Fardoa,b,c,⁎, Ryszard Auksztulewiczd,e, Micah Allenc,e, Martin J. Dietzf,
Andreas Roepstorffb,f, Karl J. Fristone a Danish Pain Centre, Department of Clinical Medicine, Aarhus University, 8000 Aarhus, Denmark
b Interacting Minds Centre, Aarhus University, 8000 Aarhus, Denmark
c Institute of Cognitive Neuroscience, University College London, London WC1N 3AR, United Kingdom
d Oxford Centre for Human Brain Activity, University of Oxford, Oxford OX3 7JX, United Kingdom
e Wellcome Trust Centre for Neuroimaging, University College London, London WC1N 3BG, United Kingdom
f Center for Functionally Integrative Neuroscience, Aarhus University, 8000 Aarhus, Denmark A B S T R A C T The neural processing and experience of pain are influenced by both expectations and attention. For example,
the amplitude of event-related pain responses is enhanced by both novel and unexpected pain, and by moving
the focus of attention towards a painful stimulus. Under predictive coding, this congruence can be explained by
appeal to a precision-weighting mechanism, which mediates bottom-up and top-down attentional processes by
modulating the influence of feedforward and feedback signals throughout the cortical hierarchy. The influence
of expectation and attention on pain processing can be mapped onto changes in effective connectivity between
or within specific neuronal populations, using a canonical microcircuit (CMC) model of hierarchical processing. We thus implemented a CMC within dynamic causal modelling for magnetoencephalography in human subjects,
to investigate how expectation violation and attention to pain modulate intrinsic (within-source) and extrinsic
(between-source) connectivity in the somatosensory hierarchy. This enabled us to establish whether both
expectancy and attentional processes are mediated by a similar precision-encoding mechanism within a network
of somatosensory, frontal and parietal sources. We found that both unexpected and attended pain modulated
the gain of superficial pyramidal cells in primary and secondary somatosensory cortex. This modulation
occurred in the context of increased lateralized recurrent connectivity between somatosensory and fronto-
parietal sources, driven by unexpected painful occurrences. Finally, the strength of effective connectivity
parameters in S1, S2 and IFG predicted individual differences in subjective pain modulation ratings. Our
findings suggest that neuromodulatory gain control in the somatosensory hierarchy underlies the influence of
both expectation violation and attention on cortical processing and pain perception. NeuroImage 153 (2017) 109–121 NeuroImage 153 (2017) 109–121 Contents lists available at ScienceDirect ⁎ Correspondence to: Nørrebrogade 44, Building 1A, 1st floor, 8000 Aarhus, Denmark.
E-mail address: francesca@clin.au.dk (F. Fardo). http://dx.doi.org/10.1016/j.neuroimage.2017.03.041
Received 13 July 2016; Accepted 20 March 2017
⁎ Correspondence to: Nørrebrogade 44, Building 1A, 1st floor, 8000 Aarhus, Denmark.
E-mail address: francesca@clin.au.dk (F. Fardo).
Available online 21 March 2017
1053-8119/ © 2017 The Authors. Published by Elsevier Inc. This is an open access article under the CC BY http://dx.doi.org/10.1016/j.neuroimage.2017.03.041
Received 13 July 2016; Accepted 20 March 2017
⁎ Correspondence to: Nørrebrogade 44, Building 1A, 1st floor, 8000 Aarhus, Denmark.
E-mail address: francesca@clin.au.dk (F. Fardo).
Available online 21 March 2017
1053-8119/ © 2017 The Authors. Published by Elsevier Inc. This is an open access article under the CC BY license (http://creativecommons.org/licenses/BY/ Stimuli, task and procedure Painful stimuli were delivered using two intra-epidermal electrodes
(Inui et al., 2002; Inui and Kakigi, 2012), via two Digitimer DS7A
stimulators (Digitimer, Hertfordshire, UK). One concentric bipolar
needle electrode was placed on the dorsum of each hand, over the
radial nerve. Each stimulus consisted of two rapid square-wave pulses
of 50 μs duration; with an inter-pulse interval of 5 ms. Stimulus
intensity was calibrated for each participant (and for each hand) to
induce a painful percept of 5 on a visuo-analog scale from 0 to 10. The
calibration was obtained via a staircase procedure, based on the 3 up 1
down rule, 6 reversals and decreasing intensity steps of 1, 0.5, 0.2 and
0.1 mA. Participants rated the intensity of each stimulus on a hor-
izontal visual-analogue scale (VAS; range=0–10, where 0 equals ‘‘no
pain sensation’’, 1 “just noticeable pain” and 10 ‘‘worst imaginable
pain”). A single intensity, corresponding to the arithmetic mean of the
intensity levels identified as “5” in the two hands, was used in the
experimental task. ,
;
,
;
g
,
,
)
Similar to expectation violation, the magnitude of pain-related
potentials is also modulated by attention, irrespective of the somato-
sensory (Miltner et al., 1989; Yamasaki et al., 2000; Van der Lubbe
et al., 2012) or nociceptive modality (Siedenberg and Treede, 1996;
Legrain et al., 2002; Lorenz and Garcia-Larrea, 2003). Generally,
attention enhances the sensitivity of neuronal populations encoding
attended sensory inputs, while inhibiting neurons responding to
irrelevant inputs (Desimone and Duncan, 1995). As in the case of
mismatch responses, attention has been shown to enhance neuronal
responses for a variety of perceptual modalities (Woldorffet al., 1993;
Hillyard and Anllo-Vento, 1998; Eimer and Forster, 2003). This
common effect on event-related potentials might suggest a similar
neural mechanism underlying the facilitative influence of both expecta-
tion violation and attention on pain processing. Specifically, these
effects can be reconciled in terms of bottom-up and top-down atten-
tional selection of nociceptive inputs (Legrain et al., 2009a), mediated
by changes in connectivity in somatosensory and fronto-parietal net-
works (Corbetta and Shulman, 2002). Under predictive coding, top-
down (prior) and bottom-up (sensory) signals are integrated in
proportion to their precision (i.e., inverse variance or uncertainty). Computationally, this corresponds to a weighting of prediction errors
by their precision (Friston, 2009), and has been previously operatio-
nalized in terms of changes in post-synaptic neuromodulatory gain
(Feldman and Friston, 2010). Stimuli, task and procedure For instance, precision-weighting in the
context of top-down visuo-spatial attention has been linked to the gain
of superficial pyramidal cells encoding prediction errors, in a way that
is consistent with biased competition (Brown and Friston, 2013). The experimental task consisted of a within-block manipulation of
sensory expectation using a roving oddball sequence (Garrido et al.,
2008; Ostwald et al., 2012; Allen et al., 2016) and a between-block
manipulation of top-down attention using instructions to attend to – or
away from – pain before each block (Fig. 1A). Within blocks (n=8),
sensory expectations were implicitly established by the roving oddball
sequence, given the higher probability of stimulus repetitions in the
same location (i.e., pain on the same hand) with respect to a change in
spatial location (i.e., pain on the other hand). After a minimum of 3 and
a maximum of 7 painful stimulus repetitions on the same hand, the
location of the stimulation switched to the dorsum of the other hand
repeatedly, throughout the duration of each block (Fig. 1B). The switch (i.e., deviant stimulus) commonly elicits an automatic
shift of attention towards the novel spatial location, particularly when
the unexpected sensory event is inherently salient or behaviorally
relevant. To establish temporal predictability of the stimulus sequence,
the inter-stimulus interval was held constant at 1 s. To define the
event-related responses of interest, the first painful stimuli of each
roving oddball sequence (i.e., pain from the unexpected spatial
location) driving a bottom-up shift of attention were defined as
deviants. For example, left deviants corresponded to unexpected
painful stimuli delivered on the left hand, after several stimulus
repetitions on the right hand. Conversely, the last painful stimuli of
each repeated train (i.e., pain from the expected spatial location) were
referred to as standards. For example, left standards consisted of
expected painful stimuli on the left hand, after several stimulus
repetitions on the same location. To ensure a balanced analysis, we
modeled only one repetition in each oddball sequence as standard (i.e.,
the last repetition before a change in location). The roving oddball
sequence ensured that left and right deviant and standard events had
identical stimulus features, but differed only with respect to expec-
tancy. Similar roving sequences have been extensively used to study
mismatch responses in different sensory domains, including the
somatosensory (Ostwald et al., 2012; Allen et al., 2016) and nociceptive
systems (Hu et al., 2013a; Zhao et al., 2014). Stimuli, task and procedure To test the hypothesis that expectation violation and attention to
pain are both mediated by a precision-weighting mechanism, we used
dynamic causal modelling (DCM) of MEG responses to painful stimuli
in a location-based roving oddball paradigm. First, we established the
hierarchy of somatosensory and fronto-parietal regions underlying
location-specific pain mismatch responses. Then, we used DCM to
investigate how expectation violation and attention modulated the gain
of superficial pyramidal cells, modelled by intrinsic self-connections,
and the extrinsic connectivity between different neuronal populations
in the somatosensory hierarchy. Our results establish that enhanced
neural activity driven by expectation violation and attention to pain
was similarly explained by increased precision-weighting or gain in
superficial pyramidal cells in somatosensory cortex, while expectation
violation also increased recurrent connectivity with fronto-parietal
sources. Introduction processes, where the precision or confidence afforded to ascending
prediction error signals is encoded by postsynaptic gain (i.e., cortical
gain control or excitation-inhibition balance; Feldman and Friston,
2010). Here, we tested the hypothesis that both expectation violation
and attention effects on pain processing are mediated by a precision-
weighting mechanism, using a canonical microcircuit model (CMC;
Bastos et al., 2012) of cortical dynamics for magnetoencephalography
(MEG). Expectation and attention both exert a strong influence on pain
perception (e.g., Wiech et al., 2008) and the magnitude of pain-related
neural responses (e.g., Dowman, 2007; Legrain et al., 2009b). Specifically, expectation violation and top-down attention have a
similar effect on pain processing, as both unexpected and attended
pain typically increase the amplitude of event-related potentials. Although the underlying neural mechanism of this common effect is
unknown, one potential explanation lies in the theory of hierarchical
predictive coding (Friston, 2005, 2008). This theory suggests that top-
down and bottom-up signals are integrated by precision-dependent The violation of sensory expectations is a salient event, which
typically elicits increased neural activity, irrespective of the deviant
stimulus feature or sensory domain (Mouraux et al., 2011). In
electrophysiological studies, expectation violation has been extensively F. Fardo et al. NeuroImage 153 (2017) 109–121 investigated with respect to the mismatch negativity (e.g., Näätänen
et al., 2012), a well-characterized difference in neural response elicited
by a novel stimulus embedded within a structured stream of repeated,
identical stimuli. Although most frequently studied in the auditory
domain (for reviews, see Näätänen et al., 2011, 2012), cortical
mismatch responses have been observed for all sensory modalities,
including somatosensation (e.g., Kekoni et al., 1997; Akatsuka et al.,
2007b; Ostwald et al., 2012; Allen et al., 2016) and nociception (Hu
et al., 2013a; Zhao et al., 2014; also see Legrain et al., 2002, 2005). analysis on account of excessive MEG artifacts. The final sample
included 22 participants (12 females; mean age=23 years; range=20–
29 years). The study was approved by the Ethical Committee of the
Central Region Denmark and conducted in accordance with the
Declaration of Helsinki. Participants Instead, when instructed to pay attention to the painful
muli, participants had to silently count the number of times the stimulation switched from the left to the right hand or vice versa (Block B). Block order was counterbalanced across
rticipants. At the end of each block, participants were required to report the number of switches, as well as to rate the average pain intensity experienced for each hand. B) The
obability of repetitions between 3–7 times was 5%, 15%, 60%, 15%, 5%, respectively. C) Mean and standard error of pain ratings for left and right hand and attended (white) and
attended (grey) pain, separately. Participants reported less intense pain when the painful somatosensory stimuli were unattended. g Fig. 1. A) Experimental task. Top-down attentional set was manipulated at the block level by presenting a verbal instruction to attend away from pain and towards the visual
stimulation on the screen (i.e., cue=“CROSS”; unattended pain) or attend to the painful stimuli perceived on the dorsum of the hands (i.e., cue=“HAND”; attended pain). In each block, a
total of 25 trains of painful stimuli were delivered to the dorsum of one hand at the time, using an oddball roving sequence. Each train included 3–7 stimulus repetitions at a constant
inter-stimulus interval of 1 s. Each painful stimulus consisted of two rapidly square-wave pulses of 50 μs duration, with an inter-pulse interval of 5 ms. The same stimulus intensity was
used for both left and right stimuli. We considered a deviant (d) the first stimulus in each train (i.e., change in stimulus location). To match the number of trials of deviant and standard
stimuli, we only modeled the last repetition before a change as a standard (s). While painful stimuli were delivered, a fixation cross on the screen changed in color from black to white or
vice versa every 2–5 s. The visual change never occurred at the same time as a change in the painful stimulus location. When instructed to pay attention to the visual stimulation,
participants had to silently count the number of time the cross changed in color from white to black or vice versa (Block A). Instead, when instructed to pay attention to the painful
stimuli, participants had to silently count the number of times the stimulation switched from the left to the right hand or vice versa (Block B). Block order was counterbalanced across
participants. Participants Participants 26 healthy volunteers were recruited from Aarhus University and
the local community. All participants were right-handed (Edinburgh
Handedness Inventory; 93.81% ± 1.20) and had normal or corrected-
to-normal vision. No participants reported a history of pain disorders,
neurological or psychiatric illness, or use of analgesics. All participants
received a reimbursement of 500 DKK for participation and gave their
informed consent before participation. Data from two participants were
not included in any analyses due to technical failures during data
collection. Two further participants were excluded from statistical Between blocks, top-down attention was manipulated via verbal
cues informing the participants to either attend towards pain (i.e.,
attended pain) or away from pain (i.e., unattended pain). Each block
began with the visual presentation of a word, either “HAND” (n=4
blocks) or “CROSS” (n=4 blocks), informing the participant of the
upcoming block condition. When the cue “HAND” was presented, 110 NeuroImage 153 (2017) 109–121 F. Fardo et al. g. 1. A) Experimental task. Top-down attentional set was manipulated at the block level by presenting a verbal instruction to attend away from pain and towards the visual
mulation on the screen (i.e., cue=“CROSS”; unattended pain) or attend to the painful stimuli perceived on the dorsum of the hands (i.e., cue=“HAND”; attended pain). In each block, a
al of 25 trains of painful stimuli were delivered to the dorsum of one hand at the time, using an oddball roving sequence. Each train included 3–7 stimulus repetitions at a constant
er-stimulus interval of 1 s. Each painful stimulus consisted of two rapidly square-wave pulses of 50 μs duration, with an inter-pulse interval of 5 ms. The same stimulus intensity was
ed for both left and right stimuli. We considered a deviant (d) the first stimulus in each train (i.e., change in stimulus location). To match the number of trials of deviant and standard
muli, we only modeled the last repetition before a change as a standard (s). While painful stimuli were delivered, a fixation cross on the screen changed in color from black to white or
e versa every 2–5 s. The visual change never occurred at the same time as a change in the painful stimulus location. When instructed to pay attention to the visual stimulation,
rticipants had to silently count the number of time the cross changed in color from white to black or vice versa (Block A). Participants At the end of each block, participants were required to report the number of switches, as well as to rate the average pain intensity experienced for each hand. B) The
probability of repetitions between 3–7 times was 5%, 15%, 60%, 15%, 5%, respectively. C) Mean and standard error of pain ratings for left and right hand and attended (white) and
unattended (grey) pain, separately. Participants reported less intense pain when the painful somatosensory stimuli were unattended. Further, they were asked to rate the average pain felt on the left and
right hands on a visual analogue scale from 0 to 10 (where 0 equals “no
pain sensation,” 1 “just noticeable pain” and 10 ‘‘worst imaginable
pain”). After each rating session, 5 s rest intervals separated contiguous
blocks. The blocks were presented with two possible pseudo-rando-
mized sequences to counterbalance order effects: ABBABAAB or
BAABABBA. Before the beginning of the experiment, participants
completed a brief training session comprising two blocks, one for each
attention condition. All participants reported that the two blocks were
sufficient to understand the task. The PsychoPy software package
v1.76.00 (Peirce, 2007, 2009) was used for instructions, stimulation
and presentation of VASs. participants were asked to pay attention to the painful stimuli delivered
on the dorsum of the hands and silently count the number of times the
stimulation switched from one hand to the other (i.e., pain attentional
set), while ignoring visual stimuli (i.e., cross changing in color). Conversely, when the cue “CROSS” was presented, participants were
asked to ignore the painful stimuli, pay attention to the fixation cross at
the center of a screen and to silently count the number of times the
color of the cross changed from white to black or vice versa (i.e., visual
attentional set). The active counting task was chosen to control for
attentional effort throughout the roving sequences; the cross fluctuated
in all attention conditions. As deviants (i.e. stimulus changes) were
unpredictable, this manipulation requires participants to maintain
similar attentional effort to deviants and standards within the attended
sensory modality. Source localization To establish the optimal network architecture underlying the
evoked responses, we first performed source localization on the
observed signals using a minimum-norm procedure (Hämäläinen and
Ilmoniemi, 1994; Litvak et al., 2011). The time window considered was
between 20 and 400 ms. The time interval from 0 to 20 ms was
excluded as it contained artefactual activity due to the electrical
stimulation. We identified eight cortical sources (Fig. 2) consisting of
bilateral primary and secondary somatosensory cortices (S1 and S2),
inferior parietal cortex (IPC) and inferior frontal gyrus (IFG). The
specific locations of left and right S1 sources (MNI coordinates: left
[−26, −36, 58]; right [32, −40, 64]) were derived by comparing right vs. left and left vs. right stimulation (i.e., laterality main effect). We
established bilateral S2 coordinates (MNI coordinates: left [−62, 14,
20]; right [62, 24, 26]) by comparing attended vs. unattended pain,
regardless the stimulation side. Finally, we identified left inferior
frontal and right inferior parietal regions (MNI coordinates: left IFG
[−54, 8, 16]; right IPC [36, −66, 40]) in the expectation violation main
effect, as well as right inferior frontal and left inferior parietal regions
(MNI coordinates: right IFG [54, 0, 10]; left IPC [−32, −64, 46]) in the
attention by expectation violation interaction. Averaging was computed separately for each condition, leading to 8
average waveforms corresponding to unattended left deviant (uLD),
unattended left standard (uLS), unattended right deviant (uRD),
unattended right standard (uRS), attended left deviant (aLD), attended
left standard (aLS), attended right deviant (aRD), and attended right
standard (aRS). These eight averages correspond to the cells of our
2×2×2 factorial design with three factors (deviant vs. standard, right
vs. left, and attended vs. unattended) After combining the planar
gradiometers, we converted the ERF time-series for each condition into
three-dimensional scalp maps over two-dimensional sensor-space (x,
y) and time (z) (Kilner and Friston, 2010; Litvak et al., 2011). For each
participant, each time point of the averaged conditions was trans-
formed into a two dimensional 64×64 pixel scalp map using linear
interpolation and concatenated over the interval from 20 to 600 ms. The resulting 3D scalp map volumes (i.e., 8 images for each partici-
pant) were smoothed with a low-pass kernel (6 mm×6 mm×6 ms full-
width at half maximum, FWHM) and entered into a general linear
model (GLM) for statistical parametric mapping. MEG acquisition and preprocessing Using DCM and Bayesian model comparison, we first specified the
network architecture of a somatosensory processing hierarchy, in terms
of extrinsic connections between somatosensory and fronto-parietal
sources identified in an auxiliary source-localization analysis. Having
identified the optimal network architecture, we then asked how
expectation violation and attention modulated between-region (extrin-
sic) and within-region (intrinsic) effective connectivity within this
somatosensory hierarchy. We used Bayesian Model Averaging to
summarize the posterior probability of our grand-average model
parameters. Finally, we inverted the winning model separately for each
subject to test whether individual differences in subjective pain
modulation correlated with intrinsic and extrinsic connectivity para-
meters. MEG data was acquired using an Elekta Neuromag TRIUX MEG
system with 204 planar gradiometers and 102 magnetometers. Blinks
and eye movements were monitored using vertical and horizontal
bipolar surface electrodes. The data were digitized with a sampling
frequency of 1 kHz, with analog filtering of 0.1–330 Hz. A continuous
measure of the head position with respect to the sensors was obtained
using four head-position indicator coils attached to the scalp. Further,
three fiducial markers (i.e., nasion, left and right pre-auricular points)
and around 100 scalp points were digitized to define a MEG coordinate
frame. The raw MEG signal was maxfiltered (MaxFilter 2.2.15 software;
Elekta Neuromag) to (1) remove externally generated noise using the
temporal extension of the signal source separation (tSSS) algorithm
(Taulu and Simola, 2006); (2) detect bad channels automatically; (3)
correct for head movements within session; and (4) correct for head
positions across participants. Further, preprocessing and statistical
analysis of MEG data were implemented using SPM12 (Statistical
Parametric Mapping 12, http://www.fil.ion.ucl.ac.uk/spm). The raw
(maxfiltered) data were epoched into 700 ms stimulus time-locked
epochs (−100/+600 ms), baseline corrected using the average pre-
stimulus activity at −100/−15 ms and downsampled to 300 Hz. Robust
averaging was applied for artifact removal (Wager et al., 2005), as this
method down-weights the contribution of extreme values that do not
occur at the same time points across trials (i.e., outliers). Predictive coding, DCM, and CMC Predictive coding entails a neurobiological implementation of the
inferential processes supporting perception, based on recurrent message-
passing in cortical hierarchies (Friston, 2010). The core notion of
predictive coding lies in the integration of top-down predictions (i.e.,
descending signals) and bottom-up prediction errors (i.e., ascending
signals). Prediction errors – discrepancies based on current predictions
and the inputs from hierarchically lower regions or (at the peripheral
level) from the sensorium – are passed to higher-order regions in a
feedforward fashion, to update high level representations. Conversely,
predictions descend to lower-order regions to suppress, or explain away,
prediction errors. The influence of prediction errors at each level of the
hierarchy is weighted by their relative precision or reliability. For
example, precise prior beliefs at higher levels of the hierarchy can
override sensory impressions or conversely, reliable (precise) sensory
evidence can override prior beliefs regardless of how unexpected they
are. The constitutive elements of perceptual inference; namely, predic-
tion, prediction error, and precision weighting can be mapped onto
particular neurobiological mechanisms; in particular, various subpopula-
tions in canonical microcircuits and their postsynaptic gain on excit-
ability. Clearly, to interpret neuronal responses and connectivity in terms
of predictive coding one has to adopt a model of canonical microcircuity. Statistical parametric mapping – sensor space analysis Source localization The time interval of
interest for the statistical analysis did not include the baseline period,
which by definition cannot differ across conditions and subjects. Predictive coding, DCM, and CMC Behavioral and subjective statistical analysis Overall, participants received a total of 1000 painful stimuli,
including 200 deviants equally distributed across the four conditions
defined by laterality (i.e., left, right) and attention (i.e., attention to
pain vs. away from pain). The probability of repetitions between 3 and
7 times was 5%, 15%, 60%, 15%, 5%, respectively (Fig. 1B). Blocks
contained on average 24 painful deviants (min=17, max=31), and 37
color changes (min=28, max=49). Following each block, participants
were asked to report the exact number of events counted during the
previous block, either changes in spatial location or changes in the
color of the fixation cross, using a numerical rating scale from 10 to 60. Detection accuracy of changes in pain spatial location and in the
color of the fixation cross were compared using paired t-tests. Further,
the subjective ratings of the perceived pain intensity were analyzed
using two-way repeated-measures ANOVAs, with the within-subject
factor “attention” (2 levels: attended and unattended pain) and
“laterality” (2 levels: left and right hand). Statistical significance was
set at p < .05, effect sizes were calculated using the partial η2, and the
Tukey HSD test was applied for post-hoc comparisons. 111 F. Fardo et al. NeuroImage 153 (2017) 109–121 Dynamic causal modeling – source space analysis Dynamic causal modeling – source space analysis Dynamic causal modelling Dynamic causal modelling inhibitory interneurons are modelled as receiving descending or top-
down inputs. In terms of predictive coding, superficial pyramidal cells
signal prediction errors to higher-order regions, while deep pyramidal
cells signal predictions to lower-order regions. Crucially, the intrinsic
excitability (or gain) of superficial pyramidal cells can now be inter-
preted as encoding the precision of prediction errors1 (Feldman and
Friston, 2010). This interpretation appears to have a degree of validity
in relation to attentional gain and enjoys the support of several
empirical studies (e.g., Fogelson et al., 2014; Pinotsis et al., 2014;
Auksztulewicz and Friston, 2015; Vossel et al., 2015). The implemen-
tation of the CMC model in DCM therefore enabled us to test the
directionality and modulation of the message passing in terms of
extrinsic connectivity (i.e., between-region passing of prediction errors
and prediction), as well as intrinsic connectivity (i.e., self-inhibition or
gain modulation representing the precision of prediction errors). Note
that CMC/DCM does not provide direct evidence of how distinct
neurons respond to sensory inputs. Instead, it represents an estimate
based on a neuronal mass model (see below, Canonical Microcircuit
Model). The cortical regions identified in the source localization were
entered into a set of dynamic causal models that embodied alternative
connectivity architectures. The connectivity structure was initially
optimized by considering 24 alternative networks of somatosensory
and fronto-parietal regions (Step 1; Fig. 3A). We then optimized the
intrinsic and extrinsic connectivity changes associated with our experi-
mental manipulation of expectation violation and attention, by invert-
ing and averaging models from a factorial model space (Step 2;
Fig. 3B). In both DCM analyses, the peristimulus time window was
20–400 ms. Sources were modeled as equivalent current dipoles and
corresponded to cortical patches of 16 mm radius, centered on the
locations above. Both contralateral S1 and S2 were specified as cortical
targets of thalamic input (i.e., left-hand inputs to right S1 and S2;
right-hand inputs to left S1 and S2; Fig. 3C), in agreement with
anatomical mapping of spinothalamic tract projections in monkeys
(Dum et al., 2009) and the evidence of parallel S1 and S2 activity in
response to somatosensory (Klingner et al., 2015) and nociceptive
stimuli in humans (Ploner et al., 1999; Liang et al., 2011; Bastuji et al.,
2016). The inputs were modeled as a Gaussian function with a prior
mean latency of 36 ms post-stimulus and a prior standard deviation of
16 ms. 1 Technically, predictive coding is a particular form of Bayesian filtering; e.g., Kalman
filtering. In this context, the precision corresponds to the Kalman gain. The Kalman gain
is applied to prediction errors to provide a Bayes optimal update to the prediction based
upon previous estimates. In other words, the precision that we are talking about here is
the precision of sensory information (or ascending prediction errors at each level of the
hierarchy). For this reason, we associate the encoding of uncertainty in predictive coding
(in the brain) with the precision of prediction errors – supposedly encoded by superficial
pyramidal cells. This means that a precise prior belief is encoded by the gain of superficial
pyramidal cells high in the hierarchy that therefore exert more influence over the deep
pyramidal cells (encoding expectations) at that level, relative to ascending prediction
errors that convey relatively imprecise information. In short, we can interpret the
excitability or postsynaptic gain of superficial pyramidal cells as mediating the precision
at the level of the hierarchy occupied by those cells. Statistical parametric mapping – sensor space analysis We conducted a general linear model (GLM) mass-univariate SPM
analysis on the combined planar gradiometers. The maximal activity
observed at the scalp level is approximately located superior to the
source,
providing
an
easier
interpretation
of
the
scalp
maps
(Hämäläinen et al., 1993). In a group-level 2×2×2 repeated measures
ANOVA, we modeled the experimental conditions (8 levels; uLS, uLD,
uRS, uRD, aLS, aLD, aRS, aRD) and the factor subject (22 levels), in
order to assess the main effect of laterality, attention, and expectation
violation, as well as the attention by expectation violation and
laterality by expectation violation interactions. Inferences were
corrected for multiple comparisons, across sensors and time points,
using Gaussian random field theory to control the cluster-wise error
rate (Kilner and Friston, 2010; Litvak et al., 2011). The threshold for
significant results was set at an uncorrected peak-level selection
threshold at p < .005 and at p < .05, family-wise error (FWE) cor-
rected at the cluster level. Accordingly, the CMC model postulates four neural populations
associated with distinct ascending and descending connectivity streams
that are integrated within each cortical column (Bastos et al., 2012). Spiny stellate and deep pyramidal cells are modelled as receiving
ascending or bottom-up inputs, while superficial pyramidal cells and 112 F. Fardo et al. NeuroImage 153 (2017) 109–121 Fig. 2. ERF and source reconstruction results. Summary of laterality (separately for left and right), attention, expectation violation main effects, and attention by expectation
violation interactions. In each panel, the first row depicts the timing and topography of event-related field effects at the scalp level. The left figure represents the posterior-anterior
displacement of the effect as a function of time (y axis, from 20 to 400 ms). The red arrow indicates the ERF maximal peak; e.g., in central-anterior locations, at around 80 ms (left
stimulation) and 50 ms (right stimulation) and in a central location at around 100 ms (attention). The central figure represents the left-right displacement as a function of time (y axis,
from 20 to 400 ms). Again, the red arrow indicates the ERF maximal peak; e.g., in the right hemisphere (left stimulation), left hemisphere (right stimulation) or close to the midline
(attention). Finally, the right figure depicts the topography of the ERF effect at the peak time point; e.g., over anterior right sensors (left stimulation), anterior left sensors (right
stimulation), widespread across posterior and anterior sensors (attention). Statistical parametric mapping – sensor space analysis The second row illustrates the topography of reconstructed sources. The analysis identified four bilateral
sources (S1, S2, IFG, and IPC). The specific locations of left and right S1 sources (MNI coordinates: left [−26, −36, 58]; right [32, −40, 64]) were derived by comparing right vs. left and
left vs. right stimulation (i.e., laterality main effect). We established bilateral S2 coordinates (MNI coordinates: left [−62, 14, 20]; right [62, 24, 26]) by comparing attended vs. unattended pain, regardless the stimulation side. Finally, we identified left inferior frontal and right inferior parietal regions (MNI coordinates: left IFG [−54, 8, 16]; right IPC [36, −66,
40]) in the expectation violation main effect, as well as right inferior frontal and left inferior parietal regions (MNI coordinates: right IFG [54, 0, 10]; left IPC [−32, −64, 46]) in the
attention by expectation violation interaction. The identified sources were then entered into a dynamic causal modelling specifying alternative connectivity architectures. (For
interpretation of the references to color in this figure legend, the reader is referred to the web version of this article.) Fig. 2. ERF and source reconstruction results. Summary of laterality (separately for left and right), attention, expectation violation main effects, and attention by expectation
violation interactions. In each panel, the first row depicts the timing and topography of event-related field effects at the scalp level. The left figure represents the posterior-anterior
displacement of the effect as a function of time (y axis, from 20 to 400 ms). The red arrow indicates the ERF maximal peak; e.g., in central-anterior locations, at around 80 ms (left
stimulation) and 50 ms (right stimulation) and in a central location at around 100 ms (attention). The central figure represents the left-right displacement as a function of time (y axis,
from 20 to 400 ms). Again, the red arrow indicates the ERF maximal peak; e.g., in the right hemisphere (left stimulation), left hemisphere (right stimulation) or close to the midline
(attention). Finally, the right figure depicts the topography of the ERF effect at the peak time point; e.g., over anterior right sensors (left stimulation), anterior left sensors (right
stimulation), widespread across posterior and anterior sensors (attention). The second row illustrates the topography of reconstructed sources. The analysis identified four bilateral
sources (S1, S2, IFG, and IPC). Dynamic causal modelling This prior latency did not overlap with the artefact period (0–
20 ms). The models were furnished with a spatial forward model,
mapping from the modeled source dipoles to observed MEG data,
based on a single shell (Nolte, 2003). Statistical parametric mapping – sensor space analysis The specific locations of left and right S1 sources (MNI coordinates: left [−26, −36, 58]; right [32, −40, 64]) were derived by comparing right vs. left and
left vs. right stimulation (i.e., laterality main effect). We established bilateral S2 coordinates (MNI coordinates: left [−62, 14, 20]; right [62, 24, 26]) by comparing attended vs. unattended pain, regardless the stimulation side. Finally, we identified left inferior frontal and right inferior parietal regions (MNI coordinates: left IFG [−54, 8, 16]; right IPC [36, −66,
40]) in the expectation violation main effect, as well as right inferior frontal and left inferior parietal regions (MNI coordinates: right IFG [54, 0, 10]; left IPC [−32, −64, 46]) in the
attention by expectation violation interaction. The identified sources were then entered into a dynamic causal modelling specifying alternative connectivity architectures. (For
interpretation of the references to color in this figure legend, the reader is referred to the web version of this article.) Canonical microcircuit model In the DCM analysis, each region was modeled using neural masses
corresponding to different populations that comprise a canonical
microcircuit (Bastos et al., 2012; Pinotsis et al., 2012). The dynamics
at each region are prescribed by ordinary differential equations
coupling the changes in postsynaptic voltage (V) to the changes in 113 Fig. 3. A) Architectures of 24 alternative models of somatosensory and fronto-parietal regions, fitted to grand-averaged ERF data. The models differed with respect to the inclusion of a
frontal (IFG) and/or parietal (IPC) node, as well as the hierarchical architecture and connections between somatosensory and fronto-parietal areas. B) Contextual modulations of
intrinsic connectivity by attention and expectation violation were optimized with respect to 1 null model, 12 bilateral alternative models (attention and expectation violation) and 16
contralateral somatosensory models (expectation violation). In the figure, the contralateral models are shown for left expectation violation. Right expectation violation models were
identical but with left lateralization of somatosensory regions. C) Both contralateral S1 and S2 were specified as cortical targets of thalamic input. D) All DCMs were tested using a
canonical microcircuit model. Each source was thus modeled as compromising 4 neuronal populations (superficial and deep pyramidal cells, spiny stellate and inhibitory interneurons). E) The winning model structure (M10), identified using fixed-effect Bayesian model selection, included all bilateral regions, with IPC at the highest hierarchical level, as well as
connections between S1 to both fronto-parietal nodes and connections between S2 and the frontal node. E) The winning model of connectivity modulation, identified using fixed-effect
Bayesian model selection, included changes in gain in bilateral primary, secondary somatosensory cortex, as well as and IFG by attention (M10). Further, the winning model revealed
changes in contralateral primary and bilateral secondary somatosensory cortex by expectation violation (M13). F. Fardo et al. NeuroImage 153 (2017) 109–121 NeuroImage 153 (2017) 109–121 F. Fardo et al. Fig. 3. A) Architectures of 24 alternative models of somatosensory and fronto-parietal regions, fitted to grand-averaged ERF data. The models differed with respect to the inclusion of a
frontal (IFG) and/or parietal (IPC) node, as well as the hierarchical architecture and connections between somatosensory and fronto-parietal areas. B) Contextual modulations of
intrinsic connectivity by attention and expectation violation were optimized with respect to 1 null model, 12 bilateral alternative models (attention and expectation violation) and 16
contralateral somatosensory models (expectation violation). In the figure, the contralateral models are shown for left expectation violation. Canonical microcircuit model Right expectation violation models were
identical but with left lateralization of somatosensory regions. C) Both contralateral S1 and S2 were specified as cortical targets of thalamic input. D) All DCMs were tested using a
canonical microcircuit model. Each source was thus modeled as compromising 4 neuronal populations (superficial and deep pyramidal cells, spiny stellate and inhibitory interneurons). E) The winning model structure (M10), identified using fixed-effect Bayesian model selection, included all bilateral regions, with IPC at the highest hierarchical level, as well as
connections between S1 to both fronto-parietal nodes and connections between S2 and the frontal node. E) The winning model of connectivity modulation, identified using fixed-effect
Bayesian model selection, included changes in gain in bilateral primary, secondary somatosensory cortex, as well as and IFG by attention (M10). Further, the winning model revealed
changes in contralateral primary and bilateral secondary somatosensory cortex by expectation violation (M13). İ = κ (−A σ(V )+γ
σ(V )+γ
σ(V )−γ
σ(V ))−2κ V −κ I
II
II
B
DP
SS→II
SS
DP→II
DP
II→II
II
II
II
II
2
II
V̇
= I
SP
SP
İ
= κ
(−A σ(V )+γ
σ(V )−γ
σ(V ))−2κ
V −κ
I
SP
SP
B
DP
SS→SP
SS
SP→SP
SP
SP
SP
SP
2
SP
V̇
= I
DP
DP
İ
= κ
(A σ(V )−γ
σ(V )−γ
σ(V ))−2κ
V −κ
I
DP
DP
F
SP
DP→DP
DP
II→DP
II
DP
DP
DP
2
DP
Here, AF and AB denote the ascending (i.e., forward) and descend- current (I) of each of the four neuronal populations (subscript SS: spiny
stellate cells in layer 4; SP: superficial pyramidal cells in layers 2/3; DP:
deep pyramidal cells in layers 5/6; II: inhibitory interneurons;
Fig. 3D): current (I) of each of the four neuronal populations (subscript SS: spiny
stellate cells in layer 4; SP: superficial pyramidal cells in layers 2/3; DP:
deep pyramidal cells in layers 5/6; II: inhibitory interneurons;
Fig. 3D): V̇
= I
SS
SS V̇
= I
DP
DP V̇
= I
II
II 114 NeuroImage 153 (2017) 109–121 F. Fardo et al. ing (i.e., backward) connections between regions (i.e., extrinsic con-
nectivity), and γm→n the connections from population m to n within
regions (i.e., intrinsic connectivity). Importantly, ascending connec-
tions are excitatory and descending connections inhibitory, due to their
mediation by inhibitory interneurons (Bastos et al., 2012). Canonical microcircuit model Similarly, all
intrinsic self-connections are modeled as polysynaptically inhibiting
their target populations (Auksztulewicz and Friston, 2015). The symbol
σ denotes a sigmoid operator, transforming the post-synaptic potential
into pre-synaptic firing rate and C scales the thalamic input u, while K
symbolizes a synaptic rate constant. Given a considerable degree of
realism in modelling laminar asymmetry in terms of the origin and
target of ascending connections (i.e., from superficial pyramidal cells to
spiny stellate cells and deep pyramidal cells) and descending connec-
tions (i.e., from deep pyramidal cells to inhibitory interneurons and
superficial pyramidal cells), this model has been used in several
previous studies of synaptic gain (Brown and Friston, 2013; Moran
et al., 2013; Auksztulewicz and Friston, 2015). extrinsic connections (Fig. 3B). Three contextual effects were included:
(1) attended vs. unattended pain (top-down attention modulation), (2)
left deviant vs. left standard (left expectation violation modulation), (3)
right deviant vs. right standard (right expectation violation modula-
tion). A set of 13 alternative models of top-down attentional modula-
tion allowed for all combinations of changes in intrinsic gain of
bilateral S1, S2, IFG, and/or IPC. To limit the model space and in
agreement with previous literature (Brown and Friston, 2013), the
modulation of recurrent extrinsic connectivity by top-down attention
was not tested (Brown and Friston, 2013). Although endogenous
attentional manipulations are clearly top-down, the attentional gain
produced by these top-down effects is generally expressed at lower
(sensory or domain specific) levels of cortical hierarchies. Conversely, the models of expectation violation allowed a joint
modulation of intrinsic and extrinsic connectivity by expectation
violation (Garrido et al., 2007, 2008; Dietz et al., 2014). With respect
to intrinsic connectivity, the models allowed for all combinations of
changes in intrinsic gain of S1, S2, IFG, and/or IPC (Fig. 3B, Bilateral),
as well as a further distinction between contralateral or bilateral
modulation of S1 and S2 gain (Fig. 3B, Contralateral/Bilateral). These included 12 alternative models of bottom-up attention modula-
tion allowing bilateral combinations of changes in intrinsic gain of the
8 sources, as well as 16 alternative models of expectation violation
modulation – allowing combinations of changes in intrinsic gain of
contralateral somatosensory and bilateral fronto-parietal sources. The
latter models included subsets of the following cases: contralateral S1,
bilateral IFG and/or IPC (N=4); contralateral S2, bilateral IFG and/or
IPC (N=4); contralateral S1 and S2, bilateral IFG and/or IPC (N=4);
contralateral S1, bilateral S2, IFG and/or IPC (N=4). Step 1: model structure optimization In the first step, the basic structure of the model was optimized
using fixed-effects Bayesian model comparison using the grand-aver-
aged evoked responses to unattended deviants. We entertained a fairly
comprehensive set of models, because the architecture underlying our
somatosensory oddball paradigm has not been previously identified (in
contrast to the architecture underlying mismatch negativity responses
in the auditory domain). We considered a model space comprising 24
alternative network architectures. All models included 8 cortical
regions (bilateral S1, S2, IFG and IPC) and differed with respect to
different combinations of between-region connections (Fig. 3A). For
example, the simplest model included connections between bilateral S1
and S2, while the full model included connections between somato-
sensory and fronto-parietal regions. When present, intra-hemispheric
and inter-hemispheric connections were reciprocal; i.e., with both
forward and backward extrinsic connections. Specifically, the models
were partitioned based on the inclusion of IFG only (models 1–3), IPC
only (models 13–15) or both IFG and IPC (models 4–12 and 16–24). When IFG and/or IPC were included, we specified inter-hemispheric
connections between homotopic frontal and/or parietal areas. Different
patterns of extrinsic intra-hemispheric connections between the two
somatosensory and the fronto-parietal regions were tested. Based on reconstructed source activity estimates (see below), the
modulatory effect of expectation violation was modeled as an interac-
tion with lateralization in terms of the main effect of laterality plus an
ipsilateral effect of violation. In other words, expectation violation was
modeled as increasing one or more connections in the hemisphere
contralateral to stimulation, with an additional increase when the
stimulus was unpredicted or surprising. These modulatory effects were
assumed to operate on the homologous connections in each hemi-
sphere. In Fig. 3A, Models 1–3 included bilateral S1, S2 and IFG and
connections between the somatosensory regions and the frontal source. Models 13–15 comprised bilateral S1, S2 and IPC and connections
between the somatosensory regions and the parietal source. Models 4–
12 and 16–24 included bilateral S1, S2, IFG and IPC, as well as
different combinations of connections between the somatosensory
regions and the fronto-parietal sources. Crucially, the model space
was designed to test (1) whether the model evidence was significantly
improved when both IFG and IPC regions were included; (2) which
fronto-parietal region corresponded to the highest hierarchical level in
the network; (3) whether S1 and/or S2 were effectively connected to
IFG and/or IPC. Canonical microcircuit model Finally, we
specified a null model with no intrinsic modulation. With respect to
extrinsic connectivity, all 29 alternative models allowed changes in
recurrent (feedforward and feedback) connectivity. Step 3: between-subject variability on pain perception and DCM
connectivity Attention main effect average winning model to single-subject data and tested for random
(between-subject) effects on the connection strengths using a para-
metric empirical Bayes (PEB) procedure (Friston et al., 2016). This
procedure used a between-subject GLM, with a first regressor model-
ling the mean connection strength across subjects and a second (Z-
scored) regressor modelling attentional effects. With respect to con-
ventional summary statistic tests, this empirical Bayesian procedure
takes into account the estimate of each connectivity parameter and its
estimated uncertainty. We then used Bayesian model reduction
(Friston et al., 2016) to prune redundant connections that did not
show a significant departure from the prior mean (of zero) or pain
modulation. This enabled us to identify connections for which there
was strong evidence for an effect of attentional modulation (i.e., with a
posterior probability > 95%). Finally, to assess the predictive validity of
the surviving DCM parameters, we performed leave-one-out cross
validation (Friston et al., 2016). This allowed us to quantify the ‘out
of sample’ effect size in terms of the correlation between pain
modulation ratings and DCM parameters. This empirical Bayesian
analysis of (subject specific) connectivity estimates furnishes a pre-
dictive validity for the DCM by showing that it is possible to predict the
extent of subjective pain modulation by attention using connectivity
estimates based only on neurophysiological responses. Attention main effect Attended vs. unattended painful stimuli elicited greater ERF
amplitudes in both early and late time intervals, at 83–200 ms (peak-
level Tmax=4.79; cluster-level pFWE < .001) and at 323–393 ms (peak-
level Tmax=4.95; cluster-level pFWE < .001) over frontal, parietal, tem-
poral and occipital sensors. At the source level, the attention effect was
primarily associated with increased response of left S2 (MNI coordi-
nates: −64 −14 18; peak-level Tmax=5.89; peak-level pUNC < .001) and
right S2 (MNI coordinates: 62 −22 26; peak-level Tmax=4.28; peak-
level pUNC < .001). To a lesser extent, other regions were identified in
visual and primary somatosensory regions. Attention by expectation violation interaction Attention by expectation violation interaction Attended deviants and unattended standards elicited greater ERF
amplitudes than unattended deviants and attended standards at 177–
263 ms over a left fronto-temporal region (peak-level Tmax=3.98;
cluster-level pFWE=.001); at 260–363 ms over a central occipital region
(peak-level Tmax=4.56; cluster-level pFWE < .001). At the source level,
this interaction was associated with an increased response of left IPC
(MNI coordinates: −32 −64 46; peak-level Tmax=4.87; peak-level pUNC
< .001)
and
right
IFG
(MNI coordinates:
50
0
10;
peak-level
Tmax=3.15; peak-level pUNC=.001). Step I: model optimization Given the selected candidate sources and their prior locations, we
first optimized the model structure in a fixed-effects Bayesian model
selection among 24 alternative models of grand-average responses to
unattended deviants. We thus identified the model structure that best
explained our data, in terms of (1) inclusion of a frontal and/or a
parietal region, (2) hierarchical arrangement of the fronto-parietal
sources and (3) connectivity between the somatosensory and the
fronto-parietal regions. The winning model (M10) was associated with
a log-evidence greater than 4.04 with respect to the second best model
(M23; Fig. 3E). This corresponds to strong evidence in favor of the
winning model (Penny et al., 2004). The selected model structure
comprised all eight sources, with IPC modeled as hierarchically above
IFG, reciprocal connections between S1 and both fronto-parietal
regions and between S2 and IFG (Fig. 3A, M10). Pleasingly, Dietz
et al. (2014) identified the same hierarchical organization of IPC and
IFG with respect to early sensory regions during left and right stimuli
in the auditory modality. Results Behavioral and subjective results Behavioral and subjective results When pain was attended, participants accurately counted 95.82 ±
1.28% of switches in spatial location from one hand to another. When
pain was unattended, participants accurately counted 94.32 ± 3.00% of
switches in the color of the fixation cross. No difference was found in
the detection accuracy between the two tasks; suggesting that partici-
pants were similarly engaged in both attention conditions. Further, the
statistical analysis of the pain ratings indicated that participants
perceived the painful stimuli as less intense when pain was unattended
compared to attended, and in a similar fashion for the left and right
hand (Fig. 1C). We thus found a main effect of top-down attention,
F(1,21)=10.01, p < .005, partial η2=.32. There was no main effect of
laterality or attention by laterality interaction. Finally, the difference
between left and right pain ratings did not correlate with the
participants’ handedness, as measured by the Edinburgh Inventory
(r(20)=−.22, p=.33). In summary, these results replicated the classic
findings of attentional modulation of pain (i.e., and analgesic effect of
distraction), as attended painful events were consistently rated as more
intense than unattended ones. Laterality main effect Laterality main effect Left vs. right painful stimulation evoked increased ERF amplitudes
at
53–153 ms
over
a
right
fronto-temporal
region
(peak-level
Tmax=6.28; cluster-level pFWE < .001). At the source level, this effect
was associated with increased responses in contralateral S1 (MNI
coordinates: 32 −24 62; peak-level Tmax=4.62; peak-level pUNC < .001)
and contralateral S2. Conversely, right vs. left painful stimulation
evoked increased ERF amplitudes at 93–220 ms over a left fronto-
temporal region (peak-level Tmax=7.31; cluster-level pFWE < .001). At
the source level, this effect was associated with increased responses in
contralateral S1 (MNI coordinates: −26 −36 62; peak-level Tmax=4.50;
peak-level pUNC < .001) and contralateral S2. In summary, laterality
effects were most pronounced in early-mid latency time windows and
were associated with increased field strength in contralateral primary
and secondary somatosensory regions. Step 2: intrinsic/extrinsic connectivity optimization Expectation violation main effect Deviant vs. standard painful stimuli elicited increased ERF ampli-
tudes at 150–400 ms over most sensors (peak-level Tmax=6.25; cluster-
level pFWE < .001), with peak responses over left fronto-temporal
sensors. At the source level, the effect was associated with an extensive
network of cortical sources, including bilateral primary and secondary
somatosensory cortices, as well as right IPC (MNI coordinates: 36 −66
40; peak-level Tmax=4.16; peak-level pUNC < .001), and left IFG (MNI
coordinates: −54 8 16; peak-level Tmax=3.13; peak-level pUNC=.001). Laterality by expectation violation interaction Laterality by expectation violation interaction Right deviant and left standard stimuli evoked greater ERF
amplitudes than left deviant and right standard at 117–160 ms over
a left fronto-temporal region (peak-level Tmax=4.19; cluster-level pFWE
< .001). At the source level, this interaction was associated with
increased responses of left S2 (peak-level Tmax=3.30; peak-level
pUNC=.001). Dynamic causal modelling – connectivity results Step 1: model structure optimization Models 4–6 and 16–18 had connections between S1
and IFG, but lacked S1-IPC connectivity. Vice versa, models 7–9 and
19–21 had connections between S1 and IPC, but lacked S1-IFG
connectivity. Models 10–12 and 22–24 included connections from S1
to both IFG and IPC. We reiterated the connections between S2 and the
fronto-parietal regions, following the same logic. Models [4, 7, 10, 16,
19, and 22] had connections between S2 and IFG, but lacked S2-IPC
connectivity. Vice versa, models [5, 8, 11, 17, 20, and 23] had
connections between S2 and IPC, but lacked S2-IFG connectivity. Finally, models [6, 9, 12, 18, 21, and 24] included connections from S2
to both IFG and IPC. The ensuing model space comprised 337 models (i.e., 13 bilateral
models of top-down attention×29 bilateral/contralateral models of
expectation violation). Each model was fitted to the grand-average
ERF data, under the assumption that each participant had the same
functional architecture. Models were compared using fixed-effects
Bayesian model comparison based on the free-energy approximation
to the model log-evidence (Friston et al., 2007), which embodies a
trade-offbetween model accuracy and complexity (Penny, 2012). To
accommodate uncertainty about which was the best model, we used
Bayesian model averaging (Hoeting et al., 1999; Penny et al., 2010) to
produce quantitative posterior estimates of effective connectivity in
response to expectation violation and attention. BMA does not rely on
the parameter estimates of a particular model, but instead uses the
entire model space by assigning a weight to the parameters of each
model according to its model evidence. Step 3: between-subject variability on pain perception and DCM
connectivity As a last validation step, we asked whether inter-individual
variability in the attentional modulation of subjective ratings of pain
correlated with the strength of intrinsic and extrinsic connectivity. At
the perceptual level, pain modulation by attention was indexed by the
difference between pain ratings associated with attended and unat-
tended blocks of painful trials. The greater this difference, the greater
the attentional modulation. At the neural level, we fitted the grand- Step 2: intrinsic and extrinsic connectivity optimization The DCM with the highest model evidence was then optimized with
respect to contextual (condition specific) modulations of intrinsic and 115 F. Fardo et al. NeuroImage 153 (2017) 109–121 Step 2: intrinsic/extrinsic connectivity optimization The winning model
provided an excellent fit between predicted and observed data for all
sensors and time points used for model inversion. However, to account
for residual uncertainty about the best model, Bayesian model aver-
aging (BMA) was used to provide the quantitative estimates of effective
connectivity and their modulation by contextual factors (Fig. 4). factors (1) top-down attention, (2) left expectation violation (i.e.,
unexpected pain on the left hand), (3) right expectation violation
(i.e., unexpected pain on the right hand) modulating a different subset
of intrinsic and/or extrinsic connections. These models were fitted to
grand-averaged evoked responses and the modulatory effects served to
explain the observed differences in ERF amplitude. A fixed-effects
Bayesian model selection suggested that the winning model allowed for
(1) attentional modulation of intrinsic connectivity in bilateral SI, SII
and IFG (Fig. 3B, Bilateral, M10), (2) expectation modulation of
intrinsic connectivity in contralateral SI and bilateral SII (Fig. 3B,
Contralateral/Bilateral, M13), and (3) expectation modulation of
recurrent extrinsic connectivity throughout the hierarchy. The differ-
ence in log-evidence compared to the second-best model was 190.97
(Fig. 3E, B Matrix), corresponding to a very strong evidence for the
winning model (i.e., posterior probability > 99%). The winning model
provided an excellent fit between predicted and observed data for all
sensors and time points used for model inversion. However, to account
for residual uncertainty about the best model, Bayesian model aver-
aging (BMA) was used to provide the quantitative estimates of effective
connectivity and their modulation by contextual factors (Fig. 4). ipsilateral hemisphere, while right deviants lead to a nominal disin-
hibition in the contralateral hemisphere and a significant disinhibition
in the ipsilateral hemisphere. With respect to extrinsic forward
connectivity, our findings showed lateralized changes in forward
connectivity, independent of the expected location, with an increased
influence from rS1 to right higher-level regions, but decreased influ-
ence from lS1 to left higher-level regions. In a similarly lateralized
fashion, effective connectivity from S1 to S2 to IFG to IPC was
increased in the left hemisphere, but decreased in the right hemi-
sphere. The descending connectivity modulation showed an asymmetric
pattern, with left deviants associated with increased bilateral inhibition
of S1 by descending connections from S2, and right deviants associated
with decreased bilateral inhibition of S1 from S2. In addition, left
deviants were associated with increased bilateral downstream inhibi-
tion with the exception of disinhibition of rIPC influences on rIFG. Step 2: intrinsic/extrinsic connectivity optimization The winning model was further optimized with respect to con-
textual changes in extrinsic and intrinsic connectivity. To this aim, we
inverted models from a factorial model space, with the experimental 116 NeuroImage 153 (2017) 109–121 F. Fardo et al. del Averaging (BMA) of the contextual modulation of intrinsic connectivity by attention, as well as intrinsic and extrinsic connectivity by expectation violation. y is showed in red, while decreased connectivity in blue. Attention increased (disinhibition) somatosensory gain, while decreasing (increased inhibition) frontal
ation increased the gain of contralateral primary and bilateral secondary somatosensory cortex. Further, expectation violation increased forward connectivity, in a
on irrespective of violation location, while decreasing backward connectivity mostly between contralateral regions. (For interpretation of the references to color in
e reader is referred to the web version of this article.) Fig. 4. Bayesian Model Averaging (BMA) of the contextual modulation of intrinsic connectivity by attention, as well as intrinsic and extrinsic connectivity by expectation violation. Increased connectivity is showed in red, while decreased connectivity in blue. Attention increased (disinhibition) somatosensory gain, while decreasing (increased inhibition) frontal
gain. Expectation violation increased the gain of contralateral primary and bilateral secondary somatosensory cortex. Further, expectation violation increased forward connectivity, in a
right-lateralized fashion irrespective of violation location, while decreasing backward connectivity mostly between contralateral regions. (For interpretation of the references to color in
this figure legend, the reader is referred to the web version of this article.) factors (1) top-down attention, (2) left expectation violation (i.e.,
unexpected pain on the left hand), (3) right expectation violation
(i.e., unexpected pain on the right hand) modulating a different subset
of intrinsic and/or extrinsic connections. These models were fitted to
grand-averaged evoked responses and the modulatory effects served to
explain the observed differences in ERF amplitude. A fixed-effects
Bayesian model selection suggested that the winning model allowed for
(1) attentional modulation of intrinsic connectivity in bilateral SI, SII
and IFG (Fig. 3B, Bilateral, M10), (2) expectation modulation of
intrinsic connectivity in contralateral SI and bilateral SII (Fig. 3B,
Contralateral/Bilateral, M13), and (3) expectation modulation of
recurrent extrinsic connectivity throughout the hierarchy. The differ-
ence in log-evidence compared to the second-best model was 190.97
(Fig. 3E, B Matrix), corresponding to a very strong evidence for the
winning model (i.e., posterior probability > 99%). Step 2: intrinsic/extrinsic connectivity optimization The analysis revealed a subset of
intrinsic and extrinsic parameters that successfully predicted pain modulation ratings
(r(20)=.55, p < .01). The x-axis represents single participants (N=22), while the y-axis
depicts the actual vs. predicted z-scored pain modulation ratings. p
g
(
,
)
Specifically, at lower hierarchical levels, sensory inputs from the
skin are conveyed by mechanoreceptive and nociceptive afferents
projecting to spatially organized somatotopic maps at spinal, thalamic
and cortical levels of the somatosensory system. At higher order levels,
fine-grained somatotopic maps of tactile and nociceptive inputs are
encoded in contralateral primary S1 (Mancini et al., 2012) and, with
lessened precision, in S2 and insular cortex (Brooks et al., 2005;
Mazzola et al., 2009; Baumgärtner et al., 2010). Topographic organiza-
tion is commonly viewed as an integral part of a neuronal encoding that
underlies localization ability; however, compelling evidence supports
the notion that activity within these brain regions is not sufficient for
the perception of spatial attributes of somatosensory experience. In
particular, the detection and localization of painful events have been
associated with the recruitment of an extensive network of brain
regions beyond somatosensory areas and including fronto-parietal
regions (Oshiro et al., 2007, 2009). right S2 to right S1, as well as from left IFG to left S2. On the other
hand, following right expectation violation, pain modulation ratings
predicted the strength of forward connectivity from left S1 and S2 to
higher-order regions (Fig. 5). The laterality of DCM parameters
covarying with perceptual modulation ratings was consistent with the
contralateral projections of incoming sensory signals. However, the
asymmetry in right hemisphere backward connectivity vs. left hemi-
sphere forward connectivity complemented functional differences ob-
served in response to expectation violation at the grand-average level. Following the leave-one-out procedure, our analysis revealed that
the subset of identified parameters successfully predicted pain mod-
ulation ratings (r(20)=.55, p < .01). These results are consistent with
the notion that pain perception is not strictly related to any single area
or connection, but with patterns of intrinsic and extrinsic connectivity
modulation throughout the somatosensory hierarchy. Note that this
(non-trivial) effect size (corresponding to over 25% variance explained)
was based upon cross-validation and is therefore an out-of-sample
estimate. In other words, this is the sort of predictability one would
expect by performing DCM on a new subject to estimate their
propensity for pain modulation by attention. Step 2: intrinsic/extrinsic connectivity optimization Both expectation violation and attention to pain modulate intrinsic
connectivity in somatosensory cortex Both unexpected and attended pain influenced the gain of S1 and
S2. Specifically, location-based expectation violation elicited a rede-
ployment of precision between left and right S1, by disinhibiting
contralateral but not ipsilateral superficial pyramidal cells in primary
regions. Further, expectation violation also influenced gain control in
S2, but in a lateralized fashion. These findings suggest that spatial
expectation violation selectively enhanced cortical gain in contralateral
S1 and right S2. While top-down attention also influenced the gain or
precision of superficial pyramidal neurons in S1 and S2, its disin-
hibitory effects were symmetrical, in accordance with the non-spatial
top-down attention manipulation. Further, attention to pain was also
associated with increased inhibition of right IFG. Step 2: intrinsic/extrinsic connectivity optimization The analysis revealed a subset of
intrinsic and extrinsic parameters that successfully predicted pain modulation ratings
(r(20)=.55, p < .01). The x-axis represents single participants (N=22), while the y-axis
depicts the actual vs. predicted z-scored pain modulation ratings. errors in S1 and S2. Expectation violation also modulated forward and
backward connections between somatosensory and fronto-parietal
sources. These findings generalize the interpretation of mismatch
responses as prediction error minimization and adjustment of a
predictive model of sensory causes (Garrido et al., 2009; Lieder
et al., 2013) to the domain of painful stimuli. Interestingly, expectation
violation increased forward connectivity in a similar fashion for left and
right deviants, while inducing mostly contralateral changes in back-
ward connectivity, depending on where the unexpected stimulus
occurred. Finally, we revealed that intrinsic and extrinsic connectivity
DCM parameters jointly predicted subjective pain modulation ratings. The somatosensory hierarchy underlying the spatial localization of
pain We found that the network architecture best accounting for
responses during spatial location of painful events produced neural
responses in contralateral S1 and bilateral S2, as well as IFG and IPC. Specifically, the winning architecture implicated bidirectional connec-
tions from somatosensory regions to IFG and IPC, with the parietal
node at the highest hierarchical level. This hierarchical architecture is
in agreement with recent findings in the auditory system (Dietz et al.,
2014) and may be mediated anatomically by the third branch of the
superior longitudinal fasciculus (Thiebaut de Schotten et al., 2011). Fig. 5. A) Parametric empirical Bayes analysis revealed that inter-individual variability
in pain modulation ratings correlated with intrinsic and extrinsic connectivity changes
driven by attention and expectation violation. The neural effect of attention on self-
inhibitory connections in left S1 and right S2 correlated with the degree to which
participants experienced pain enhancement by attention. Further, following left expecta-
tion violation, pain modulation ratings were predicted by the strength of backward
connectivity to somatosensory regions primarily in the right hemisphere. Instead,
following right expectation violation, pain modulation ratings were predicted by the
strength of forwards connectivity from S1 and S2 to higher order regions in the left
hemisphere. This pattern of results complemented hemispheric functional asymmetries
observed at the group level. B) Prediction accuracy at the leave-one-out cross validation,
which allowed us to quantify the ‘out of sample’ effect size in terms of the correlation
between pain modulation ratings and DCM parameters. Step 2: intrinsic/extrinsic connectivity optimization In
contrast, right deviants were associated with increased left-hemisphere
disinhibition at all hierarchical levels, and an overall increased right-
hemisphere inhibition (Fig. 4). These results suggest that location-
based expectation violation is mediated by location-independent
changes in forward connectivity. However, it involves lateralized
changes in backward connectivity that specifically depend on the
unexpected location. Following BMA, significant changes in effective connectivity corre-
sponded to a posterior probability > 99%. The gain of superficial
pyramidal cells in bilateral S1 and S2 was increased under attention,
reflecting a significant decrease in self-inhibition (i.e., disinhibition or
increased gain). In contrast, the gain of superficial pyramidal cells in
rIFG was decreased under attention, while it did not change markedly
in lIFG. Furthermore, the gain of superficial pyramidal cells in
contralateral S1 increased following expectation violation, reflecting a
significant decrease in self-inhibition in S1 contralateral to stimulus
presentation. In S2, left deviants were associated with significant
disinhibition in the contralateral hemisphere and inhibition in the Step 3: between-subject variability on pain perception and DCM
parameters We found that subjective ratings of pain modulation by attention
can be predicted (over subjects) by the degree of self-inhibition in left
S1 and right S2 under attention. Further, following left expectation
violations, pain modulation ratings predicted the strength of self-
inhibition in right S1 and backward connectivity from right IFG to 117 NeuroImage 153 (2017) 109–121 F. Fardo et al. Fig. 5. A) Parametric empirical Bayes analysis revealed that inter-individual variability
in pain modulation ratings correlated with intrinsic and extrinsic connectivity changes
driven by attention and expectation violation. The neural effect of attention on self-
inhibitory connections in left S1 and right S2 correlated with the degree to which
participants experienced pain enhancement by attention. Further, following left expecta-
tion violation, pain modulation ratings were predicted by the strength of backward
connectivity to somatosensory regions primarily in the right hemisphere. Instead,
following right expectation violation, pain modulation ratings were predicted by the
strength of forwards connectivity from S1 and S2 to higher order regions in the left
hemisphere. This pattern of results complemented hemispheric functional asymmetries
observed at the group level. B) Prediction accuracy at the leave-one-out cross validation,
which allowed us to quantify the ‘out of sample’ effect size in terms of the correlation
between pain modulation ratings and DCM parameters. Pain perception as an inferential process In the present study, we replicated the well-known effect of
attentional modulation of pain perception, as participants consistently
reported less pain when noxious stimuli were unattended (e.g.,
Bushnell et al., 1985; Miron et al., 1989). Further, we demonstrated
that the influence of attention on pain was associated with the degree to
which pain-related neural signals changed intrinsic connectivity in
somatosensory regions, and their feedforward and feedback message
passing from and to higher-order regions. Predictive coding theories
describe the brain as using probabilistic internal models to infer the
causes of sensory inputs (Rao and Ballard, 1999; Friston, 2005; Clark,
2013). Within this framework, both expectation violation and attention
are integral part of perceptual inference and act via precision-weighting
mechanisms in order to produce context-sensitive responses (Friston,
2009). This theory fits comfortably with the evidence that perceived
pain intensity is strongly regulated by top-down processes; rather than
solely reflecting nociceptive bottom-up sensory inputs (for a review, see
Büchel et al., 2014). Our results are consistent with a previous study
showing that increased recurrent connectivity between S1 and S2 was
associated
with
awareness
of
peri-threshold
tactile
stimuli
(Auksztulewicz et al., 2012), thus indicating a link between conscious
perception and the instantiation and resolution of prediction errors in
somatosensory regions (also see, Auksztulewicz and Friston, 2016). Our finding seems at odds with previous studies reporting that the
effect of attention lies in the modulation of between-region backward
connectivity (Auksztulewicz and Friston, 2015; Chennu et al., 2016). However, changes in between-region backward connectivity and with-
in-region postsynaptic gain have aligned functional interpretations. Backward connectivity refers to the modulatory effect of predictions on
lower-level activity. On the other hand, postsynaptic gain reflects the
precision of prediction errors, which is dependent to the degree of
mismatch between prior expectations (i.e., backward connectivity) and
incoming signal. In summary, our findings showed a common precision-weighting
synaptic mechanism for bottom-up and top-down attentional modula-
tion in the somatosensory hierarchy. Importantly, these results extend
previous modelling work on visual spatial attention (Brown and
Friston, 2013) and auditory temporal attention (Auksztulewicz and
Friston, 2015), pointing to a general predictive coding gain control
mechanism, across expectancy and attentional manipulations, that may
be generic to all sensory modalities. Discussion We used magnetoencephalography (MEG) and dynamic causal
modelling (DCM) to investigate how expectancy and attention pro-
cesses influence effective connectivity in a somatosensory network, in
the context of pain perception. Overall, both expectation violation and
attention to pain influenced the promulgation of painful somatosensory
inputs throughout a hierarchy of brain sources; including primary and
secondary somatosensory cortex (S1 and S2), inferior frontal gyrus
(IFG) and inferior parietal cortex (IPC). Crucially, we showed that a
similar gain control mechanism governs the influence of expectancy
and attentional effects on pain-related neural processing. This mechan-
ism is consistent with an enhanced precision-weighting of prediction From a general perspective, a recent theory of attention – within the
predictive coding framework – suggests that local changes in cortical gain
are a key mechanism in both top-down attention and bottom-up salience
(Feldman and Friston, 2010). Neurophysiologically, gain corresponds to the
sensitivity of neurons responding to prediction errors (i.e., superficial
pyramidal cells), resulting from the competition between excitatory and
inhibitory neuronal populations. Functionally, increased somatosensory
gain would lead to the increased influence of low-level sensory responses on
higher-level processing in frontal and parietal regions observed here. Furthermore, the implicit decreased precision-weighting of right IFG 118 NeuroImage 153 (2017) 109–121 F. Fardo et al. signals by attention, is consistent with enhanced weighting of low-level
somatosensory signals. This interpretation is supported by a previous EEG
study demonstrating that tactile mismatch signals originating in S1 and S2
encode a perceptual mechanism described as Bayesian surprise, an
information theoretic index reflecting the amount of prediction error
(Ostwald et al., 2012). At later time points, mismatch signals from frontal
and cingulate cortex were instead associated with stimulus change or
salience (Ostwald et al., 2012), likely reflecting active inferential processes. Future studies could manipulate surprise and salience orthogonally to
investigate whether gain-dependent mechanisms explain not only mis-
match (perception-related) signals from lower sensory regions, but also
salience (action-related) signals from frontal regions. whether attention is re-oriented towards the left or the right hemi-field
(Shulman et al., 2010). Finally, an alternative explanation is offered by
functional asymmetries in the central representation of sympathetic
and parasympathetic afferent projections (Craig, 2005). The right
hemisphere is primarily associated with sympathetic activity, mediat-
ing arousal, negative affect, pain, and interoceptive processing. It is
thus possible that this lateralization reflects the operation of a system
generally involved in interoceptive active inference (Seth, 2013). Limitations and future directions One limitation of the present study entails the lack of a behavioral
measure of expectation violation, thus precluding an assessment of the
effect of expectation on pain perception and pain-related neural
measures. Furthermore, our results are limited to spatial expectation
violation. An important issue for future research will be to test the
association between expectations, prediction error signals and pain
perception; for example, by using a cue-based paradigm combined with
trial-by-trial pain ratings. Finally, it would be important to assess
whether these results can be generalized to temporal and magnitude
expectations (e.g., manipulation of the expected intensity, as for
example in Fardo et al., 2015). Conclusions Predicting coding accounts of brain function offer a unifying and
principled explanation for expectation violation and attention in
shaping neural responses across sensory modalities. Here, we extended
previous work on the functional anatomy of mismatch responses and
attention to the somatosensory domain. Our work suggests that both
expectation violation and attention to pain jointly influence the cortical
gain of superficial pyramidal cells in the somatosensory hierarchy. These findings shed new light on the neurobiological mechanisms
associated with expectancy and attention in cortical pain-related
processing. Expectation violation modulates extrinsic connectivity in the
somatosensory hierarchy Beyond altering intrinsic connectivity, expectation violation also
increased recurrent extrinsic connectivity between somatosensory,
frontal and parietal regions. This is in line with the role of gain in
regulating neuronal message passing across the cortical hierarchy
(Friston, 2005). We observed bilateral increases of effective connectiv-
ity driven by expectation violation, regardless the location of deviant
pain. However, the pattern of ascending connectivity was asymmetrical
in the two hemispheres. At each level of the neural hierarchy, incoming
inputs are compared with top-down predictions, with any resulting
discrepancy (i.e., error) being passed to higher regions of the hierarchy
(Friston, 2008). Prediction errors thus signal the degree to which
higher-level expectations about sensory causes must be revised to
reduce overall surprise or free energy (Friston and Kiebel, 2009;
Friston, 2010). For instance, recurrent connectivity amongst a fron-
to-temporo-parietal network is commonly modulated in response to a
novel and unexpected change in the environment, as in the case of
auditory mismatch negativity (MMN) responses (Dietz et al., 2014;
Phillips et al., 2015). Further, the MNN has been shown to reflect a
failure in suppressing prediction errors (Garrido et al., 2007; 2008,
2009) and the adjustment (learning) of a new probabilistic model of the
environment (Lieder et al., 2013a, 2013b). We thus interpreted the
increased bilateral (but asymmetrical) forward and backward connec-
tivity as reflecting location-unspecific detection of expectation violation
and location-specific update mechanism, corresponding to feedforward
propagation of highly-precise sensory prediction errors and backward
signaling of updated spatial predictions. Further, the asymmetrical
forward-connectivity results are consistent with the notion of a
supramodal right-lateralized fronto-parietal network that has been
implicated in the reorienting of spatial attention. For instance, the
bottom-up reorienting of visuo-spatial attention towards a novel
location is thought to be specifically mediated by a right-lateralized
ventral attention network (Corbetta and Shulman, 2002), regardless References Topological inference for EEG and MEG. Ann. Appl. Stat. 4, 1272–1290. Klingner, C.M., Brodoehl, S., Huonker, R., Baumann, L., Witte, O.W., 2015. Parallel
processing of somatosensory information: evidence from dynamic causal modeling of
MEG data. NeuroImage 118, 193–198. Baumgärtner, U., Iannetti, G.D., Zambreanu, L., Stoeter, P., Treede, R.-D., Tracey, I.,
2010. Multiple somatotopic representations of heat and mechanical pain in the
operculo-insular cortex: a high-resolution fMRI study. J. Neurophysiol. 104,
2863–2872. Legrain, V., Guérit, J.-M., Bruyer, R., Plaghki, L., 2002. Attentional modulation of the
nociceptive processing into the human brain: selective spatial attention, probability
of stimulus occurrence, and target detection effects on laser evoked potentials. Pain
99, 21–39. Brooks, J.C.W., Zambreanu, L., Godinez, A., Craig, A.D., Tracey, I., 2005. Somatotopic
organisation of the human insula to painful heat studied with high resolution
functional imaging. NeuroImage 27, 201–209. Legrain, V., Bruyer, R., Guérit, J.-M., Plaghki, L., 2005. Involuntary orientation of
attention to unattended deviant nociceptive stimuli is modulated by concomitant
visual task difficulty. Evidence from laser evoked potentials. Clin. Neurophysiol. 116,
2165–2174. Brown, H.R., Friston, K.J., 2013. The functional anatomy of attention: a DCM study. Front. Hum. Neurosci. 7, 784. Büchel, C., Geuter, S., Sprenger, C., Eippert, F., 2014. Placebo analgesia: a predictive
coding perspective. Neuron 81, 1223–1239. Legrain, V., Damme, S.V., Eccleston, C., Davis, K.D., Seminowicz, D.A., Crombez, G.,
2009a. A neurocognitive model of attention to pain: behavioral and neuroimaging
evidence. Pain 144, 230–232. Bushnell, M.C., Duncan, G.H., Dubner, R., Jones, R.L., Maixner, W., 1985. Attentional
influences on noxious and innocuous cutaneous heat detection in humans and
monkeys. J. Neurosci. 5, 1103–1110. Legrain, V., Perchet, C., García-Larrea, L., 2009b. Involuntary orienting of attention to
nociceptive events: neural and behavioral signatures. J. Neurophysiol. 102,
2423–2434. Chennu, S., Noreika, V., Gueorguiev, D., Shtyrov, Y., Bekinschtein, T.A., Henson, R.,
2016. Silent expectations: dynamic causal modeling of cortical prediction and
attention to sounds that weren’t. J. Neurosci. 36, 8305–8316. Liang, M., Mouraux, A., Iannetti, G.D., 2011. Parallel processing of nociceptive and non-
nociceptive somatosensory information in the human primary and secondary
somatosensory cortices: evidence from dynamic causal modeling of functional
magnetic resonance imaging data. J. Neurosci. 31, 8976–8985. Clark, A., 2013. Whatever next? Predictive brains, situated agents, and the future of
cognitive science. Behav. Brain Sci. 36, 181–204. Corbetta, M., Shulman, G.L., 2002. Control of goal-directed and stimulus-driven
attention in the brain. Nat. Rev. Neurosci. 3, 201–215. References Lieder, F., Daunizeau, J., Garrido, M.I., Friston, K.J., Stephan, K.E., 2013a. Modelling
trial-by-trial changes in the mismatch negativity. PLoS Comput. Biol. 9, e1002911. Li d
F
S
h
K E
D
i
J
G
id
M I
F i
K J
2013b A attention in the brain. Nat. Rev. Neurosci. 3, 201–215. Craig, A.D., 2005. Forebrain emotional asymmetry: a neuroanatomical basis? Trends
Cogn. Sci. 9, 566–571. Lieder, F., Stephan, K.E., Daunizeau, J., Garrido, M.I., Friston, K.J., 2013b. A
neurocomputational model of the mismatch negativity. PLOS Comput. Biol. 9,
e1003288. g
Desimone, R., Duncan, J., 1995. Neural mechanisms of selective visual attention. Annu. Rev. Neurosci. 18, 193–222. Dietz, M.J., Friston, K.J., Mattingley, J.B., Roepstorff, A., Garrido, M.I., 2014. Effective
connectivity reveals right-hemisphere dominance in audiospatial perception:
implications for models of spatial neglect. J. Neurosci. 34, 5003–5011. Litvak, V., Mattout, J., Kiebel, S., Phillips, C., Henson, R., Kilner, J., Barnes, G.,
Oostenveld, R., Daunizeau, J., Flandin, G., Penny, W., Friston, K., 2011. EEG and
MEG data analysis in SPM8. Comput. Intell. Neurosci. 2011, 1–32. Dowman, R., 2007. Neural mechanisms of detecting and orienting attention toward
unattended threatening somatosensory targets. I. Intermodal effects. Psychophysiology 44, 407–419. Lorenz, J., Garcia-Larrea, L., 2003. Contribution of attentional and cognitive factors to
laser evoked brain potentials. Neurophysiol. Clin. Clin. Neurophysiol. 33, 293–301. Mancini, F., Haggard, P., Iannetti, G.D., Longo, M.R., Sereno, M.I., 2012. Fine-grained
nociceptive maps in primary somatosensory cortex. J. Neurosci. 32, 17155–17162. l
d
G é
iè
2009 S
i Dum, R.P., Levinthal, D.J., Strick, P.L., 2009. The spinothalamic system targets motor
and sensory areas in the cerebral cortex of monkeys. J. Neurosci. 29, 14223–14235. Dum, R.P., Levinthal, D.J., Strick, P.L., 2009. The spinothalamic system targets motor
and sensory areas in the cerebral cortex of monkeys. J. Neurosci. 29, 14223–14235. Eimer, M., Forster, B., 2003. Modulations of early somatosensory ERP components by
transient and sustained spatial attention. Exp. Brain Res. 151, 24–31. Mazzola, L., Isnard, J., Peyron, R., Guénot, M., Mauguière, F., 2009. Somatotopic
organization of pain responses to direct electrical stimulation of the human insular
cortex. Pain 146, 99–104. Eimer, M., Forster, B., 2003. Modulations of early somatosensory ERP components by
transient and sustained spatial attention. Exp. Brain Res. 151, 24–31. d
ll
i d
illi
ff
i i Fardo, F., Allen, M., Jegindø, E.-M.E., Angrilli, A., Roepstorff, A., 2015. Neurocognitive
evidence for mental imagery-driven hypoalgesic and hyperalgesic pain regulation. NeuroImage 120, 350–361. References Hämäläinen, M.S., Hari, R., Ilmoniemi, R.J., Knuutila, J., Lounasmaa, O.V., 1993. Magnetoencephalography – theory, instrumentation, and applications to noninvasive studies of the working human brain. Rev. Mod. Phys. 65, 413–505. Hillyard, S.A., Anllo-Vento, L., 1998. Event-related brain potentials in the study of visual
selective attention. Proc. Natl. Acad. Sci. USA 95, 781–787. Akatsuka, K., Wasaka, T., Nakata, H., Kida, T., Kakigi, R., 2007. The effect of stimulus
probability on the somatosensory mismatch field. Exp. Brain Res. 181, 607–614. Allen, M., Fardo, F., Dietz, M.J., Hillebrandt, H., Friston, K.J., Rees, G., Roepstorff, A.,
2016. Anterior insula coordinates hierarchical processing of tactile mismatch Akatsuka, K., Wasaka, T., Nakata, H., Kida, T., Kakigi, R., 2007. The effect of stimulus
probability on the somatosensory mismatch field. Exp. Brain Res. 181, 607–614. Allen, M., Fardo, F., Dietz, M.J., Hillebrandt, H., Friston, K.J., Rees, G., Roepstorff, A.,
2016. Anterior insula coordinates hierarchical processing of tactile mismatch
responses. NeuroImage 127, 34–43. Hoeting, J. a, Madigan, D., Raftery, a E., Volinsky, C.T., 1999. Bayesian model averaging:
a tutorial. Stat. Sci. 14, 382–401. Hu, L., Zhao, C., Li, H., Valentini, E., 2013. Mismatch responses evoked by nociceptive
stimuli. Psychophysiology 50, 158–173. Auksztulewicz, R., Friston, K., 2015. Attentional enhancement of auditory mismatch
responses: a DCM/MEG study. Cereb. Cortex, 1–11. Inui, K., Kakigi, R., 2012. Pain perception in humans: use of intraepidermal electrical
stimulation. J. Neurol. Neurosurg. Psychiatry 83, 551–556. Auksztulewicz, R., Friston, K., 2016. Repetition suppression and its contextual
determinants in predictive coding. Cortex 80, 125–140. Inui, K., Tran, T.D., Hoshiyama, M., Kakigi, R., 2002. Preferential stimulation of A-delta
fibers by intra-epidermal needle electrode in humans. Pain 96, 247–252. K k
i J
Hä
äläi
H
S
i
M
G öh
J
R i ik i
K
L ht k
ki A Auksztulewicz, R., Spitzer, B., Blankenburg, F., 2012. Recurrent neural processing and
somatosensory awareness. J. Neurosci. 32, 799–805. Kekoni, J., Hämäläinen, H., Saarinen, M., Gröhn, J., Reinikainen, K., Lehtokoski, A.,
Näätänen, R., 1997. Rate effect and mismatch responses in the somatosensory
system: ERP-recordings in humans. Biol. Psychol. 46, 125–142. Bastos, A.M., Usrey, W.M., Adams, R.A., Mangun, G.R., Fries, P., Friston, K.J., 2012. Canonical microcircuits for predictive coding. Neuron 76, 695–711. Bastuji, H., Frot, M., Perchet, C., Magnin, M., Garcia-Larrea, L., 2016. Pain networks
from the inside: spatiotemporal analysis of brain responses leading from nociception
to conscious perception. Hum. Brain Mapp. 37, 4301–4315. Kilner, J.M., Friston, K.J., 2010. Author contributions Friston, K., 2010. The free-energy principle: a unified brain theory? Nat. Rev. Neurosci. 11, 127–138. F.F., M.A., M.J.D. and A.R. designed the experiment. F.F. and M.A. acquired the data. F.F., R.A. and K.J.F. analyzed the data. F.F., R.A.,
M.A., M.J.D., and K.J.F. wrote the paper. Friston, K., Kiebel, S., 2009. Predictive coding under the free-energy principle. Philos. Trans. R. Soc. Lond. B Biol. Sci. 364, 1211–1221. Friston, K., Mattout, J., Trujillo-Barreto, N., Ashburner, J., Penny, W., 2007. Variational
free energy and the Laplace approximation. NeuroImage 34, 220–234. Friston, K.J., Litvak, V., Oswal, A., Razi, A., Stephan, K.E., van Wijk, B.C.M., Ziegler, G.,
Zeidman, P., 2016. Bayesian model reduction and empirical Bayes for group (DCM)
studies. NeuroImage 128, 413–431. Conflict of interest The authors declare no competing financial interests. 119 F. Fardo et al. NeuroImage 153 (2017) 109–121 Acknowledgments Garrido, M.I., Kilner, J.M., Kiebel, S.J., Stephan, K.E., Friston, K.J., 2007. Dynamic
causal modelling of evoked potentials: a reproducibility study. NeuroImage 36,
571–580. This work was supported by the Danish Agency for Science,
Technology,
and
Innovation's
University
Investment
Grant
to
MINDlab,
Aarhus
University
(F.F.,
M.J.D. and
A.R.)
and
the
Wellcome Trust (100227, M.A.). The Wellcome Trust Centre for
Neuroimaging is supported by core funding from the Wellcome Trust
091593. Garrido, M.I., Friston, K.J., Kiebel, S.J., Stephan, K.E., Baldeweg, T., Kilner, J.M., 2008. The functional anatomy of the MMN: a DCM study of the roving paradigm. NeuroImage 42, 936–944. g
Garrido, M.I., Kilner, J.M., Stephan, K.E., Friston, K.J., 2009. The mismatch negativity: a
review of underlying mechanisms. Clin. Neurophysiol. 120, 453–463. Hämäläinen, M.S., Ilmoniemi, R.J., 1994. Interpreting magnetic fields of the brain:
minimum norm estimates. Med. Biol. Eng. Comput. 32, 35–42. References Miltner, W., Johnson, R., Braun, C., Larbig, W., 1989. Somatosensory event-related
potentials to painful and non-painful stimuli: effects of attention. Pain 38, 303–312. i
G
C
h
i
h
ll
1989
ff
f
i
h Miron, D., Duncan, G.H., Catherine Bushnell, M., 1989. Effects of attention on the
intensity and unpleasantness of thermal pain. Pain 39, 345–352. d
h
l
i Feldman, H., Friston, K., 2010. Attention, uncertainty, and free-energy. Front. Hum. Neurosci. 4, 215. Moran, R.J., Campo, P., Symmonds, M., Stephan, K.E., Dolan, R.J., Friston, K.J., 2013. Free energy, precision and learning: the role of cholinergic neuromodulation. J. Neurosci. 33, 8227–8236. Fogelson, N., Litvak, V., Peled, A., Fernandez-del-Olmo, M., Friston, K.J., 2014. The
functional anatomy of schizophrenia: a dynamic causal modeling study of predictive
coding. Schizophr. Res. 158, 204–212. Mouraux, A., Diukova, A., Lee, M.C., Wise, R.G., Iannetti, G.D., 2011. A multisensory
investigation of the functional significance of the “pain matrix”. NeuroImage 54,
2237–2249. Friston, K., 2005. A theory of cortical responses. Philos. Trans. R. Soc. Lond. B Biol. Sci. 360, 815–836. Friston, K., 2008. Hierarchical models in the brain. PLOS Comput. Biol. 4, e1000211. Friston, K., 2009. The free-energy principle: a rough guide to the brain? Trends Cogn. Näätänen, R., Kujala, T., Winkler, I., 2011. Auditory processing that leads to conscious
perception: a unique window to central auditory processing opened by the mismatch 120 F. Fardo et al. NeuroImage 153 (2017) 109–121 and secondary somatosensory cortices in human pain processing. J. Neurophysiol. 81, 3100–3104. negativity and related responses. Psychophysiology 48, 4–22. Näätänen, R., Kujala, T., Escera, C., Baldeweg, T., Kreegipuu, K., Carlson, S., Ponton, C.,
2012. The mismatch negativity (MMN) – a unique window to disturbed central
auditory processing in ageing and different clinical conditions. Clin. Neurophysiol. 123, 424–458. Rao, R.P.N., Ballard, D.H., 1999. Predictive coding in the visual cortex: a functional
interpretation of some extra-classical receptive-field effects. Nat. Neurosci. 2, 79–87. S
h A K
2013 I
i
i f
i
d h
b di d
lf T
d C Seth, A.K., 2013. Interoceptive inference, emotion, and the embodied self. Trends Cogn. Sci. 17, 565–573. Nolte, G., 2003. The magnetic lead field theorem in the quasi-static approximation and
its use for magnetoencephalography forward calculation in realistic volume
conductors. Phys. Med. Biol. 48, 3637–3652. Shulman, G.L., Pope, D.L.W., Astafiev, S.V., McAvoy, M.P., Snyder, A.Z., Corbetta, M.,
2010. References Right hemisphere dominance during spatial selective attention and target
detection occurs outside the dorsal frontoparietal network. J. Neurosci. 30,
3640–3651. Oshiro, Y., Quevedo, A.S., Mchaffie, J.G., Kraft, R.A., Coghffi, R.C., 2007. Brain
mechanisms supporting spatial discrimination of pain. J. Neurosci. 27, 3388–3394. O hi
Y
Q
d
A S
M H ffi
J G
K
f
R A
C
hill R C
2009 B
i Siedenberg, R., Treede, R.-D., 1996. Laser-evoked potentials: exogenous and endogenous
components. Electroencephalogr. Clin. Neurophysiol. Potentials Sect. 100, 240–249. Oshiro, Y., Quevedo, A.S., McHaffie, J.G., Kraft, R.A., Coghill, R.C., 2009. Brain
mechanisms supporting discrimination of sensory features of pain: a new model. J. Neurosci. 29, 14924–14931. Taulu, S., Simola, J., 2006. Spatiotemporal signal space separation method for rejecting
nearby interference in MEG measurements. Phys. Med. Biol. 51, 1759–1768. Thi b
d
S h
M
D ll’A
F
F
k l S J
Si
A
V
i F
DGM Ostwald, D., Spitzer, B., Guggenmos, M., Schmidt, T.T., Kiebel, S.J., Blankenburg, F.,
2012. Evidence for neural encoding of Bayesian surprise in human somatosensation. NeuroImage 62, 177–188. Thiebaut de Schotten, M., Dell’Acqua, F., Forkel, S.J., Simmons, A., Vergani, F., DGM,
Murphy, Catani, M., 2011. A lateralized brain network for visuospatial attention. Nat. Neurosci. 14, 1245–1246. Peirce, J.W., 2007. PsychoPy – Psychophysics software in Python. J. Neurosci. Methods
162, 8–13. Van der Lubbe, R.H.J., Buitenweg, J.R., Boschker, M., Gerdes, B., MLA, Jongsma, 2012. The influence of transient spatial attention on the processing of intracutaneous
electrical stimuli examined with ERPs. Clin. Neurophysiol. 123, 947–959. Peirce, J.W., 2009. Generating stimuli for neuroscience using PsychoPy. Front. Neuroinform. 2, 10. Penny, W.D., 2012. Comparing dynamic causal models using AIC, BIC and free energy. NeuroImage 59, 319–330. Vossel, S., Mathys, C., Stephan, K.E., Friston, K.J., 2015. Cortical coupling reflects
Bayesian belief updating in the deployment of spatial attention. J. Neurosci. 35,
11532–11542. Penny, W.D., Stephan, K.E., Mechelli, A., Friston, K.J., 2004. Comparing dynamic causal
models. NeuroImage 22, 1157–1172. Wager, T.D., Keller, M.C., Lacey, S.C., Jonides, J., 2005. Increased sensitivity in
neuroimaging analyses using robust regression. NeuroImage 26, 99–113. h
l
f
d Penny, W.D., Stephan, K.E., Daunizeau, J., Rosa, M.J., Friston, K.J., Schofield, T.M., Leff,
A.P., 2010. Comparing families of dynamic causal models. PLoS Comput. Biol., 6. h ll
l
k
h
k
h Wiech, K., Ploner, M., Tracey, I., 2008. Neurocognitive aspects of pain perception. Trends
Cogn. Sci. 12, 306–313. References Phillips, H.N., Blenkmann, A., Hughes, L.E., Bekinschtein, T.A., Rowe, J.B., 2015. Hierarchical organization of frontotemporal networks for the prediction of stimuli
across multiple dimensions. J. Neurosci. 35, 9255–9264. Woldorff, M.G., Gallen, C.C., Hampson, S.A., Hillyard, S.A., Pantev, C., Sobel, D., Bloom,
F.E., 1993. Modulation of early sensory processing in human auditory cortex during
auditory selective attention. Proc. Natl. Acad. Sci. USA 90, 8722–8726. Pinotsis, D. a, Schwarzkopf, D.S., Litvak, V., Rees, G., Barnes, G., Friston, K.J., 2012. Dynamic causal modelling of lateral interactions in the visual cortex. NeuroImage
66C, 563–576. Yamasaki, H., Kakigi, R., Watanabe, S., Hoshiyama, M., 2000. Effects of distraction on
pain-related somatosensory evoked magnetic fields and potentials following painful
electrical stimulation. Cogn. Brain Res. 9, 165–175. Pinotsis, D.A., Brunet, N., Bastos, A., Bosman, C.A., Litvak, V., Fries, P., Friston, K.J.,
2014. Contrast gain control and horizontal interactions in V1: a DCM study. NeuroImage 92, 143–155. Zhao, C., Valentini, E., Hu, L., 2014. Functional features of crossmodal mismatch
responses. Exp. Brain Res. 233, 617–629. Ploner, M., Schmitz, F., Freund, H.-J., Schnitzler, A., 1999. Parallel activation of primary 121
|
https://openalex.org/W2165924832
|
https://europepmc.org/articles/pmc2789086?pdf=render
|
English
| null |
Mating changes the subcellular distribution and the functionality of estrogen receptors in the rat oviduct
|
Reproductive biology and endocrinology
| 2,009
|
cc-by
| 9,513
|
Reproductive Biology and Endocrinology 2009, 7:139
doi:10.1186/1477-7827-7-139 Reproductive Biology and Endocrinology 2009, 7:139
doi:10.1186/1477-7827-7-139 This article is available from: http://www.rbej.com/content/7/1/139 © 2009 Orihuela et al; licensee BioMed Central Ltd. ;
This is an Open Access article distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/2.0),
which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited. Reproductive Biology and
Endocrinology BioMed Central Open Access Published: 30 November 2009 Received: 23 September 2009
Accepted: 30 November 2009 Received: 23 September 2009
Accepted: 30 November 2009 Mating changes the subcellular distribution and the functionality o
estrogen receptors in the rat oviduct
Pedro A Orihuela*1,2, Lidia M Zuñiga3, Mariana Rios4, Alexis Parada-
Bustamante3, Walter D Sierralta5, Luis A Velásquez1,2 and
Horacio B Croxatto1,2,3 Address: 1Laboratorio de Inmunología de la Reproducción, Facultad de Química y Biología, Universidad de Santiago de Chile, Santiago, Chile,
2Centro para el Desarrollo en Nanociencia y Nanotecnología-CEDENNA, Santiago, Chile, 3Millennium Institute for Fundamental and Applied
Biology, Santiago, Chile, 4Unidad de Reproducción y Desarrollo, Facultad de Ciencias Biológicas, Pontificia Universidad Católica de Chile,
Santiago, Chile and 5Instituto de Nutrición y Tecnología de los Alimentos, Universidad de Chile, Santiago, Chile Email: Pedro A Orihuela* - pedro.orihuela@usach.cl; Lidia M Zuñiga - lzunigac@bio.puc.cl; Mariana Rios - mrios@bio.puc.cl; Alexis Parada-
Bustamante - aparada@uc.cl; Walter D Sierralta - wsierra@inta.cl; Luis A Velásquez - luis.velasquez.c@usach.cl;
Horacio B Croxatto - horacio.croxatto@usach.cl * Corresponding author Background g
It is well recognized that in the female mammal ovarian
steroids, embryonic signals or mating-associated signals
regulate egg transport through the oviduct (for review see
[1]). In cyclic rats, oocytes reach the uterus approximately
72 h after ovulation, whereas in pregnant rats embryos
take 96 h. Oocytes take 96 h to traverse the oviduct in rats
made pseudopregnant by mechanical stimulation of the
cervix in the evening of proestrus, therefore the above dif-
ference is dependent on mating-associated signals rather
on whether eggs are fertilized or not [2]. Eggs cross the
ampullary-isthmic junction 9 h earlier and egg surrogates
move at higher speed in the isthmic segment with most of
them arriving earlier to the intramural segment in preg-
nant rats than in cycling rats [3]. Thus, a broad change
occurs in oviductal functioning elicited by mating-associ-
ated signals. A single injection of 17β-estradiol (E2) on day 1 of the
cycle or pregnancy shortens oviductal transport of eggs
from the normal 72-96 h to less 24 h [1]. Previously, we
demonstrated that inhibitors of RNA and protein synthe-
sis block E2-induced oviductal embryo transport accelera-
tion in pregnant rats, but fail to do so in cyclic rats [4,5]. Furthermore, in cyclic rats exogenous E2 activates protein
phosphorylation in the oviduct via a nongenomic path-
way, since such activation occurs when mRNA synthesis is
completely suppressed by α-Amanitin [6]. Estradiol-
induced phosphorylation is essential for its effect on
oocyte transport in cycling rats since local administration
of a broad-spectrum inhibitor of protein kinases totally
blocks the effect of E2 on egg transport [7,8]. Thus, E2
accelerates oviductal egg transport through nongenomic
pathways in cyclic rats, while it does it through genomic
pathways in pregnant rats. Recently, this change in path-
ways has been designated "intracellular path shifting"
(IPS) [9]. Further investigation has shown that activity of
the enzyme Catechol-O-Methyltransferase (COMT) is
higher in the oviduct of cyclic than pregnant rats while
OR486 a selective inhibitor of COMT blocked the effect of
E2 on oviductal egg transport only in cyclic rats suggesting
that decreased activity of oviductal COMT induced by
mating is one of the underlying mechanisms of IPS [9]. Abstract Background: Mating changes the mode of action of 17beta-estradiol (E2) to accelerate oviductal
egg transport from a nongenomic to a genomic mode, although in both pathways estrogen
receptors (ER) are required. This change was designated as intracellular path shifting (IPS). Methods: Herein, we examined the subcellular distribution of ESR1 and ESR2 (formerly known as
ER-alpha and ER-beta) in oviductal epithelial cells of rats on day 1 of cycle (C1) or pregnancy (P1)
using immunoelectron microscopy for ESR1 and ESR2. The effect of mating on intraoviductal ESR1
or ESR2 signaling was then explored comparing the expression of E2-target genes c-fos, brain
creatine kinase (Ckb) and calbindin 9 kDa (s100g) in rats on C1 or P1 treated with selective
agonists for ESR1 (PPT) or ESR2 (DPN). The effect of ER agonists on egg transport was also
evaluated on C1 or P1 rats. Results: Receptor immunoreactivity was associated with the nucleus, cytoplasm and plasma
membrane of the epithelial cells. Mating affected the subcellular distribution of both receptors as
well as the response to E2. In C1 and P1 rats, PPT increased Ckb while both agonists increased c-
fos. DPN increased Ckb and s100g only in C1 and P1 rats, respectively. PPT accelerated egg
transport in both groups and DPN accelerated egg transport only in C1 rats. Conclusion: Estrogen receptors present a subcellular distribution compatible with E2 genomic
and nongenomic signaling in the oviductal epithelial cells of C1 and P1 although IPS occurs
independently of changes in the distribution of ESR1 and ESR2 in the oviductal epithelial cells. Mating affected intraoviductal ER-signaling and induced loss of functional involvement of ESR2 on
E2-induced accelerated egg transport. These findings reveal a profound influence on the ER
signaling pathways exerted by mating in the oviduct. Page 1 of 11
(page number not for citation purposes) (page number not for citation purposes) http://www.rbej.com/content/7/1/139 Reproductive Biology and Endocrinology 2009, 7:139 rodents [11]. The antiestrogen ICI 182780 blocks E2-
induced egg transport acceleration in cyclic and pregnant
rats [7] indicating that ER participates in both the
genomic and the nongenomic pathways involved in the
kinetic action of E2 on the oviduct. However, we have
found that levels of ESR1 and ESR2 mRNA and protein in
oviducts of pregnant rats were similar to those oviducts of
cycling rats, suggesting that IPS is not explained by
changes in the expression of ER in the oviduct [12]. Background Although the physiological relevance of IPS has not been
clearly established it is probable that decrease in the
COMT activity induced by mating in the oviduct protects
the embryos from the deleterious effect that methox-
yestradiols exert during the first stages of development
[10]. Abstract Herein, we determined the effect of mating on subcellular
distribution and functionality of ESR1 and ESR2 in the rat
oviduct. First, we compared immunoreactivity of both ER
associated to cell membrane, cytoplasm and nucleus
between epithelial cells of the ampullary and isthmic seg-
ments of cyclic and pregnant rats following treatment
with E2. We also determined the effect of selective ago-
nists for ESR1 (PPT) or ESR2 (DPN) on mRNA levels of
three E2-inducible genes c-fos, brain creatine kinase (Ckb)
and calbindin 9 kDa (s100g) [13,14] in the oviduct of
pregnant and cyclic rats. Additionally, the role of ESR1
and ESR2 on oviductal egg transport was evaluated in
cyclic or pregnant rats treated with PPT or DPN. Methods
Animals Sprague-Dawley rats (bred in house) weighing 200-260 g
were used. Animals were kept under controlled tempera-
ture (21-24°C), and lights were on from 0700 to 2100 h. Water and pelleted food were supplied ad libitum. The
phases of the estrous cycle were determined by daily vagi-
nal smears [15] and all females were used after showing
two consecutive 4-day cycles. Females in pro-estrus were
kept either isolated or caged with fertile males. The fol-
lowing day (estrus) was designated as C1 in the first
instance and day P1 in the second, provided spermatozoa
were found in the vaginal smear. The care and manipula-
tion of the animals was done in accordance with the ethi-
cal guidelines of our institution. Systemic administration of E2 On C1 or P1 E2 5 μg was injected subcutaneously (s.c.) as
a single dose dissolved in 0.1 mL propylene glycol. Con-
trol rats received propylene glycol alone. Page 2 of 11
(page number not for citation purposes) Immunoelectron microscopy these doses to approximately 1000-fold for local
(intrabursal) injection. To our knowledge these doses of
PPT or DPN did not change the plasmatic E2 and Proges-
terone level in the rat. py
Oviducts from vehicle and E2-treated rats were separated
into ampulla and isthmus and the excess mucus was
removed in each segment by flushing with 50 μL saline. Both segments were fixed in 4% freshly depolymerised
paraformaldehyde, 0.5% glutaraldehyde in phosphate
buffer pH 7.4 0.1 M containing saccharose 0.1 M, DMSO
1% and CaCl2 1% for 2-4 h at room temperature. The
fixed samples were washed three times with phosphate
buffer, dehydrated in a graded ethanol series and infil-
trated with LR Gold (Plano, München, FRG). Subse-
quently, the samples were transferred to gelatine capsules
filled with 0.8% (w/v) benzoyl peroxide in LR Gold and
kept for polymerization at a pressure of 500 mmHg. The
blocks were cured for 1-2 days at room temperature before
sectioning with a Sorvall-2000 ultramicrotome using a
diamond knife. The sections (50-80 nm) were mounted
on formvar-coated nickel grids and incubated on droplets
of 0.1 M glycine in PBS pH 7.6, and subsequently blocked
with 1% bovine foetal serum for 2 h at room temperature. The grids were then incubated for 2 h with a rabbit anti-
ESR1 (MC-20, Santa Cruz Biotechnology, Santa Cruz, CA)
or anti-ESR2 (clone 68-4, Chemicon International, Biller-
ica, MA) at 1:50 dilution. After washing with PBS, the
preparations were incubated for 1 h with goat anti-rabbit
immunoglobulin conjugated to 10 nm gold particles
(Kirkegaard & Perry Laboratories Inc, Gaithersburg, MD)
diluted 1:30. Sections were washed and contrasted with
Reynolds stain [23]. All sections were examined using a
Phillips-TECNAI 12 BIOTWIN EM Microscope (FEI Com-
pany, Hillsboro, OR) at 80 kV. As negative controls the
primary antibody was replaced by rabbit preimmune
serum or oviductal samples without prior incubation with
anti-ESR1 antibody or anti-ESR2 antibody were also
included. For further validation we also performed immu-
noelectron microscopy of the isthmic segment from vehi-
cle and E2-treated rats using a mouse anti-ESR1 from
another company (H-151, Calbiochem, La Jolla, CA), as
primary antibody diluted 1:30. Furthermore, we used
gold-labeled particles of 40 nm (Kirkegaard & Perry Labo-
ratories Inc) to obtain photomicrographs at low magnifi-
cation for show unspecific background labeling in the
oviductal lumen. Animal surgery and assessment of egg transport Intrabursal administration of agonists, which minimizes
the dose needed to affect the oviduct without systemic
effects, was performed on C1 or P1 as previously
described [5]. At this time, ovulation has already taken
place, so this treatment cannot affect the number of
oocytes ovulated. Egg transport was evaluated as previ-
ously published [2,4,5]. Twenty-four hours after treat-
ment, animals were sacrificed and their oviducts were
flushed individually with saline. Flushing was examined
under low-power magnification (25×), and the number of
eggs found was recorded. Local administration of selective agonist of ESR1 (PPT) or ESR2
(DPN) Local administration of selective agonist of ESR1 (PPT) or ESR2
(DPN) PPT (1,3,5-tris(4-hydroxyphenyl)-4-propyl-1H-pyrazole,
Sigma Chem. CO, St. Louis, MO) [16] or DPN (Diarylpro-
pionitrile, Tocris Cookson Inc. Ellisville, MO) [17] were
injected into each bursa at a concentration of 7.5, 22.5 or
67.5 ng/μL in DMSO 1%. Control rats received the corre-
sponding vehicle alone. Since the range of effective doses
of PPT and DPN given systemically is between 2.5 μg/μL-
250 μg/μL [18,19] we considered appropriate diminished Estrogens induce cellular changes in their target organs
through several different mechanisms that involve activa-
tion of estrogen receptors (ER). The two main forms of ER,
ESR1 and ESR2 (formerly known as ER-α and ER-β), have
distinct tissue expression patterns in both humans and Page 2 of 11
(page number not for citation purposes) Page 2 of 11
(page number not for citation purposes) Reproductive Biology and Endocrinology 2009, 7:139 http://www.rbej.com/content/7/1/139 Immunoelectron microscopy At least ten areas of 63 μm2 from differ-
ent epithelial cells and different sections of an oviduct
were photographed and the photomicrographs were digi-
talized in an iBook computer (Apple Computer Inc,
Cupertino, CA), and gold particles present only in the
cells were counted using the image analysis software
Adobe Photoshop 7.0 (Adobe Systems Inc, San Jose, CA)
by an observer blinded to the treatment groups. The
results of the immunolabeling are presented as the quo-
tient of the number of gold particles present divided by
the area and cell number inspected [24] py
Oviducts from vehicle and E2-treated rats were separated
into ampulla and isthmus and the excess mucus was
removed in each segment by flushing with 50 μL saline. Both segments were fixed in 4% freshly depolymerised
paraformaldehyde, 0.5% glutaraldehyde in phosphate
buffer pH 7.4 0.1 M containing saccharose 0.1 M, DMSO
1% and CaCl2 1% for 2-4 h at room temperature. The
fixed samples were washed three times with phosphate
buffer, dehydrated in a graded ethanol series and infil-
trated with LR Gold (Plano, München, FRG). Subse-
quently, the samples were transferred to gelatine capsules
filled with 0.8% (w/v) benzoyl peroxide in LR Gold and
kept for polymerization at a pressure of 500 mmHg. The
blocks were cured for 1-2 days at room temperature before
sectioning with a Sorvall-2000 ultramicrotome using a
diamond knife. The sections (50-80 nm) were mounted
on formvar-coated nickel grids and incubated on droplets
of 0.1 M glycine in PBS pH 7.6, and subsequently blocked
with 1% bovine foetal serum for 2 h at room temperature. Real-time PCR Whole oviducts on C1 (N = 4) or P1 (N = 4) were dis-
sected and flushed to avoid contamination with oocytes
or embryos mRNA. Oviductal RNA was isolated using Tri-
zol Reagent (Invitrogen, Gaithersburg, MD) and 1 μg of
total RNA of each sample was treated with Dnase I Ampli-
fication grade (Invitrogen). The single-strand cDNA was
synthesized by reverse transcription using the Superscript
III Reverse Transcriptase First Strand System for RT-PCR
(Invitrogen), according to the manufacturer's protocol. The Light Cycler instrument (Roche Diagnostics, GmbH
Mannheim, Germany) was used to quantify the relative
gene expression of c-fos, Ckb or s100g in the oviduct of
cyclic and pregnant rats; Gapdh was chosen as the house-
keeping gene for load control because we have previously
demonstrated that E2 or pregnancy did not affect its
expression [20]. The SYBR® Green I double-strand DNA
binding dye (Roche Diagnostics) was the reagent of
choice for these assays. Primers for c-fos were 5' CCG AGA
TTG CCA ATC TAC TG 3' (sense) and 5' AGA AGG AAC
CAG ACA GGT CC 3' (antisense), Ckb 5' AAG CTG GCA
GTA GAA GCC CT 3' (sense) 5' TTG TCG AAG AGG AAG
TGG TC 3' (antisense), s100g 5' GGC AGC ACT CAC TGA
CAG C 3' (sense) 5' CAG TAG GTG GTG TCG GAG C
3'(antisense) and for Gapdh were 5' ACC ACA GTC CAT
GCC ATC AC 3' (sense) and 5' TCC ACC ACC CTG TTG
CTG TA 3' (antisense). The thermal cycling conditions
included an initial activation step at 95°C for 25 min, fol-
lowed by 40 cycles of denaturizing and annealing-ampli-
fication (95°C for 15 sec, 60°C for 15 sec and 72°C for 30
sec) and finally one cycle of melting (95° to 60°C). To
verify specificity of the product, amplified products were
subject to melting curve analysis as well as electrophore-
sis, and product sequencing was performed to confirm
identity as described by Muscillo et al [21]. The expression
of transcripts was determined using a method previously
reported [22,9]. Page 3 of 11
(page number not for citation purposes) Page 3 of 11
(page number not for citation purposes) http://www.rbej.com/content/7/1/139 Reproductive Biology and Endocrinology 2009, 7:139 http://www.rbej.com/content/7/1/139 Statistical analysis and ESR2-reacting gold particles in oviductal epithelial
cells of the rat are shown in Figures 1, 2 and 3. Receptor
immunoreactivity was found associated with the nucleus,
cytoplasm and plasma membrane, including cilia, of the
epithelial cells. Data for subcellular distribution of ESR1
in ampulla and isthmus are shown in Figure 4. Mating
increased the immunoreactivity of ESR1 in the plasma
membrane and cytoplasm of the ampullary segment
although it did not affect the ESR1 immunoreactivity in
the isthmus. In cyclic rats, E2 treatment increased the den-
sity of ESR1 labeling in all three compartments from both
segments, except the nucleus of the isthmic segment,
whereas in pregnant rats a major increase in labeling was
observed only in the cytoplasm of the isthmic segment. The results are presented as mean ± SE. Overall analysis
was done by Kruskal-Wallis test, followed by Mann-Whit-
ney test for pair-wise comparisons when overall signifi-
cance was detected. Page 4 of 11
(page number not for citation purposes) Distribution of ESR1 and ESR2 in epithelial cells of mated
and non-mated rat oviducts treated with E2 Distribution of ESR1 and ESR2 in epithelial cells of mated
and non-mated rat oviducts treated with E2 At 09:00 h of C1 or P1, 8 rats were injected with E2 5 μg or
vehicle and 3 h later they were sacrificed and their ovi-
ducts were separated into ampulla and isthmus and proc-
essed for immunoelectron microscopy. Representative
photomicrographs of the subcellular distribution of ESR1 Representative photomicrographs of oviductal epithelial cells processed by immunoelectron microscopy with gold labeled-anti-
bodies of 10 nm for ESR1 (a-c) or ESR2 (d-f)
Figure 1
Representative photomicrographs of oviductal epithelial cells processed by immunoelectron microscopy with
gold labeled-antibodies of 10 nm for ESR1 (a-c) or ESR2 (d-f). Ampullary and isthmic sections of cyclic (a-c) and preg-
nant (d-f) rats are shown in the upper and lower panels, respectively. Arrows denotes ESR1-or ESR2 reacting gold particles in
the epithelial cells. Arowheads emphasize the association of estradiol receptor immunoreactivity with cilia (cl, see insert in d). Asterisks indicate unspecific background laleling. Bar: 0.5 μm. PM = plasma membrane, C = cytoplasm, N = nucleus, L =
lumen, SV = secretory vesicle.
PM
cl
C
L
L
cl
L
L
PM
PM
PM
C
C
C
N
C
N
C
cl
cl
cl
cl
PM
C
*
*
sv
sv
PM
cl
C
L
L
cl
PM
C
* N
C
L
L
PM
C
cl
cl
cl
cl
PM
C
N
C
L
PM
C
*
sv
sv C L
p
p
g
p
p
p
y
py
g
(
)
(
)
g
Representative photomicrographs of oviductal epithelial cells processed by immunoelectron microscopy with
gold labeled-antibodies of 10 nm for ESR1 (a-c) or ESR2 (d-f). Ampullary and isthmic sections of cyclic (a-c) and preg-
nant (d-f) rats are shown in the upper and lower panels, respectively. Arrows denotes ESR1-or ESR2 reacting gold particles in
the epithelial cells. Arowheads emphasize the association of estradiol receptor immunoreactivity with cilia (cl, see insert in d). Asterisks indicate unspecific background laleling. Bar: 0.5 μm. PM = plasma membrane, C = cytoplasm, N = nucleus, L =
lumen, SV = secretory vesicle. Distribution of ESR1 and ESR2 in epithelial cells of mated
and non-mated rat oviducts treated with E2 Page 4 of 11
(page number not for citation purposes) Reproductive Biology and Endocrinology 2009, 7:139 http://www.rbej.com/content/7/1/139
C Representative photomicrographs of oviductal epithelial cells processed by immunoelectron microscopy without prior incuba-
tion with anti-ESR1 antibody (a) or anti-ESR2 antibody (b), or incubated with rabbit preimmune serum (c)
Figure 2
Representative photomicrographs of oviductal epithelial cells processed by immunoelectron microscopy with-
out prior incubation with anti-ESR1 antibody (a) or anti-ESR2 antibody (b), or incubated with rabbit preim-
mune serum (c). Bar: 0.5 μm. PM = plasma membrane, C = cytoplasm, L = lumen, cl = cilia. Bar: 0.5 μm.
L
cl
PM
C
C
C
cl
cl
C
cl
L
PM
C L Represent
tion with a
Figure 2 p
p
g
p
p
p
y
py
p
y ( )
y ( ),
p
( )
g
Representative photomicrographs of oviductal epithelial cells processed by immunoelectron microscopy with-
out prior incubation with anti-ESR1 antibody (a) or anti-ESR2 antibody (b), or incubated with rabbit preim-
mune serum (c). Bar: 0.5 μm. PM = plasma membrane, C = cytoplasm, L = lumen, cl = cilia. Bar: 0.5 μm.
L
PM
C Representative photomicrographs at low magnification of oviductal epithelial cells processed by immunoelectron microscopy
with gold labelled-antibodies of 40 nm for ESR1 (a) or ESR2 (b)
Figure 3
Representative photomicrographs at low magnification of oviductal epithelial cells processed by immunoelec-
tron microscopy with gold labelled-antibodies of 40 nm for ESR1 (a) or ESR2 (b). Arrows denotes ESR1-or ESR2
reacting gold particles in the epithelial cells. Note scarce unspecific background laleling. Bar: 0.5 μm. PM = plasma membrane,
C = cytoplasm, L = lumen. PM
L
C
L
PM
C PM
L
C L PM L Represent
with gold
Figure 3 Representative photomicrographs at low magnification of oviductal epithelial cells processed by immunoelectron microscopy
with gold labelled antibodies of 40 nm for ESR1 (a) or ESR2 (b)
Figure 3
Representative photomicrographs at low magnification of oviductal epithelial cells processed by immunoelec-
tron microscopy with gold labelled-antibodies of 40 nm for ESR1 (a) or ESR2 (b). Arrows denotes ESR1-or ESR2
reacting gold particles in the epithelial cells. Note scarce unspecific background laleling. Bar: 0.5 μm. PM = plasma membrane,
C = cytoplasm, L = lumen. Distribution of ESR1 and ESR2 in epithelial cells of mated
and non-mated rat oviducts treated with E2 p
p
g
p
g
p
p
y
py
g
( )
( )
g
Representative photomicrographs at low magnification of oviductal epithelial cells processed by immunoelec-
tron microscopy with gold labelled-antibodies of 40 nm for ESR1 (a) or ESR2 (b). Arrows denotes ESR1-or ESR2
reacting gold particles in the epithelial cells. Note scarce unspecific background laleling. Bar: 0.5 μm. PM = plasma membrane,
C = cytoplasm, L = lumen. Page 5 of 11
(page number not for citation purposes) Page 5 of 11
(page number not for citation purposes) Reproductive Biology and Endocrinology 2009, 7:139 http://www.rbej.com/content/7/1/139 Mean number of the density of gold particles (X ± SE) observed for ESR1 in the plasma membrane (PM), cytoplasm (C) and
nucleus (N) of epithelial cells of oviductal ampulla (A) and isthmus (I) from rats on day 1 of the cycle or pregnancy, 3 hours
after treatment with oestradiol
Figure 4
Mean number of the density of gold particles (X ± SE) observed for ESR1 in the plasma membrane (PM), cyto-
plasm (C) and nucleus (N) of epithelial cells of oviductal ampulla (A) and isthmus (I) from rats on day 1 of the
cycle or pregnancy, 3 hours after treatment with oestradiol. Means with different letters are significantly different
from each other within each graph (P < 0.05). a ≠ b ≠ c. Replicas of this experiment are stated in the figure. Distribution of ESR1 and ESR2 in epithelial cells of mated
and non-mated rat oviducts treated with E2 PM
C
N
a
a
a
b
b
b
b
b
a
a
a
a
a
a
a
a
c
a,b
b
a
b,c
a,b
c
a
Cyclic
Ampulla
Ampulla
Isthmus
Isthmus
V
E2
Pregnant
V
E2
V
E2
V
E2
ER-α
gold particles/μm2/cell
(X±SE)
N= 8 cells
N= 9 cells
N= 9 cells
N= 9 cells
N= 10 cells
N= 9 cells
N= 11 cells
N= 11 cells
0
5
10
15
20
25
30
0
5
10
15
20
25
30
PM
C
N
0
5
10
15
20
25
30
PM
C
N
0
5
10
15
20
25
30
PM
C
N
gold particles/μm2/cell
(X±SE)
gold particles/μm2/cell
(X±SE)
gold particles/μm2/cell
(X±SE) PM
C
N
a
a
a
b
b
b
b
b
a
a
a
a
Cyclic
Ampulla
Isthmus
V
E2
V
E2
ER-α
gold particles/μm2/cell
(X±SE)
N= 8 cells
N= 9 cells
N= 9 cells
N= 9 cells
0
5
10
15
20
25
30
0
5
10
15
20
25
30
PM
C
N
gold particles/μm2/cell
(X±SE) a
a
a
a
c
a,b
b
a
b,c
a,b
c
a
Ampulla
Isthmus
Pregnant
V
E2
V
E2
N= 10 cells
N= 9 cells
N= 11 cells
N= 11 cells
0
5
10
15
20
25
30
PM
C
N
0
5
10
15
20
25
30
PM
C
N
gold particles/μm2/cell
(X±SE)
gold particles/μm2/cell
(X±SE) ea
u be o t e e s ty o go
pa t c es (
S ) obse ve
o
S
t e p as a
e b a e (
), cytop as
(C) a
uc eus ( ) o ep t e a ce s o ov ucta a pu a ( ) a
st
us ( )
o
ats o
ay
o t e cyc e o p eg a cy, 3 ou s
a te t eat e t w t oest a o
gu e
Mean number of the density of gold particles (X ± SE) observed for ESR1 in the plasma membrane (PM), cyto-
plasm (C) and nucleus (N) of epithelial cells of oviductal ampulla (A) and isthmus (I) from rats on day 1 of the
cycle or pregnancy, 3 hours after treatment with oestradiol. Means with different letters are significantly different
from each other within each graph (P < 0.05). a ≠ b ≠ c. Page 6 of 11
(page number not for citation purposes) Distribution of ESR1 and ESR2 in epithelial cells of mated
and non-mated rat oviducts treated with E2 Replicas of this experiment are stated in the figure. Effect of selective agonist of ESR1 or ESR2 on the level of
c-fos, Ckb and s100g in the oviduct of pregnant and cycle
rats Mating decreased immunoreactivity of ESR2 in the cyto-
plasm of the ampullary and isthmic segments although it
increased labeling of ESR2 in the plasma membrane of the
isthmus. In cyclic rats, E2 decreased ESR2 density in the
cytoplasm in the ampulla and in the nucleus of the isth-
mus, whereas in pregnant rats, E2 increased the density of
ESR2 labeling in the cytoplasm in the ampulla and
decreased it in the plasma membrane and nucleus in the
isthmus (Figure 5). The results using the mouse anti-ESR1
were similar to those obtained with the rabbit anti-ESR1
(not shown). Furthermore, low unspecific background
labeling of gold particles was found in the lumen of epi-
thelial cells in control experiments without primary anti-
body or incubation with rabbit preimmune serum (see
figures 2 and 3). All this supports the specificity in the rec-
ognition of ER immunoreactivity. Rats on C1 (N = 4) or P1 (N = 4) were locally treated with
67.5 ng/μL of PPT or DPN and 3 h later oviducts were
excised and their total RNA were processed by RT-PCR
using specific primers for c-fos, Ckb, s100g or Gapdh as
described above. Figure 6 shows that in cyclic rats oviduc-
tal levels of c-fos, Ckb and s100g were similar while in preg-
nant rats levels of Ckb were major than c-fos and s100g. PPT increased 5-fold and 3-fold the levels of Ckb and c-fos
in cyclic and pregnant rats while s100g was not affected in
both conditions. In cyclic rats, DPN increased 4- and 5-
fold Ckb and c-fos respectively, while in pregnant rats
DPN increased 2.5 fold c-fos and 2-fold s100g. Distribution of ESR1 and ESR2 in epithelial cells of mated
and non-mated rat oviducts treated with E2 a
a
c
a
a
a
a
a
a
c
c
a
a
a
b
a, b
b
a
a
b
b
Cyclic
Ampulla
Ampulla
Isthmus
Isthmus
V
E2
Pregnant
V
E2
V
E2
V
E2
ER-β
N= 9 cells
N= 9 cells
N= 8 cells
N= 9 cells
N= 10 cells
N= 9 cells
N= 11 cells
N= 11 cells
0
5
10
15
20
25
30
PM
C
N
0
5
10
15
20
25
30
PM
C
N
0
5
10
15
20
25
30
PM
C
N
0
5
10
15
20
25
30
PM
C
N
gold particles/μm2/cell
(X±SE)
gold particles/μm2/cell
(X±SE)
gold particles/μm2/cell
(X±SE)
gold particles/μm2/cell
(X±SE) a
a
c
a
b
a
a
b
b
Cyclic
Ampulla
Isthmus
V
E2
V
E2
ER-β
N= 9 cells
N= 9 cells
N= 8 cells
N= 9 cells
0
5
10
15
20
25
30
PM
C
N
0
5
10
15
20
25
30
PM
C
N
gold particles/μm2/cell
(X±SE)
gold particles/μm2/cell
(X±SE) a
a
a
a
a
c
c
a
a
a
b
a, b
Ampulla
Isthmus
Pregnant
V
E2
V
E2
N= 10 cells
N= 9 cells
N= 11 cells
N= 11 cells
0
5
10
15
20
25
30
PM
C
N
0
5
10
15
20
25
30
PM
C
N
gold particles/μm2/cell
(X±SE)
gold particles/μm2/cell
(X±SE) Mean number of the density of gold particles (X ± SE) observed for ESR2 in the plasma membrane (PM), cytoplasm (C) and
nucleus (N) of epithelial cells of oviductal ampulla (A) and isthmus (I) from rats on day 1 of the cycle or pregnancy, 3 hours
after treatment with oestradiol
Figure 5
Mean number of the density of gold particles (X ± SE) observed for ESR2 in the plasma membrane (PM), cyto-
plasm (C) and nucleus (N) of epithelial cells of oviductal ampulla (A) and isthmus (I) from rats on day 1 of the
cycle or pregnancy, 3 hours after treatment with oestradiol. Means with different letters are significantly different
from each other within each graph (P < 0.05). a ≠ b ≠ c. Replicas of this experiment are stated in the figure. ullary and isthmic segments of the rat oviduct [25,26]. Distribution of ESR1 and ESR2 in epithelial cells of mated
and non-mated rat oviducts treated with E2 Page 6 of 11
(page number not for citation purposes) Reproductive Biology and Endocrinology 2009, 7:139 http://www.rbej.com/content/7/1/139 Mean number of the density of gold particles (X ± SE) observed for ESR2 in the plasma membrane (PM), cytoplasm (C) and
nucleus (N) of epithelial cells of oviductal ampulla (A) and isthmus (I) from rats on day 1 of the cycle or pregnancy, 3 hours
after treatment with oestradiol
Figure 5
Mean number of the density of gold particles (X ± SE) observed for ESR2 in the plasma membrane (PM), cyto-
plasm (C) and nucleus (N) of epithelial cells of oviductal ampulla (A) and isthmus (I) from rats on day 1 of the
cycle or pregnancy, 3 hours after treatment with oestradiol. Means with different letters are significantly different
from each other within each graph (P < 0.05). a ≠ b ≠ c. Replicas of this experiment are stated in the figure. Distribution of ESR1 and ESR2 in epithelial cells of mated
and non-mated rat oviducts treated with E2 Here, we show for the first time that ESR1 and ESR2 were
found associated to the plasma membrane, cytoplasm
and nucleus of the epithelial cells of both oviductal seg-
ments in the rat. This is in agreement with previous
immunoelectron microscopy studies reporting localiza-
tion of ESR1 in non-nuclear sites in other cell types
[27,28]. Other studies have also shown the presence of
ESR1 and ESR2 in non-nuclear sites of various cell types
using western blot or ligand-blot of subcellular enriched
fractions [29-31]. Ligand activation of ER associated to
cell membrane and cytoplasm can modulate downstream
pathways that induce discrete signaling responses, includ-
ing stimulation of adenylyl cyclase in breast and vascular
tissues [32,33], activation of Ca2+ flux in arterial smooth
muscle [34] or the cascade Src/Ras/ERK [35]. Our findings
provide evidence of the presence of multiple pools of ER
that could initiate genomic and nongenomic responses to Effect of selective agonist of ESR1 or ESR2 on oviductal
egg transport in mated and non-mated rats Rats on C1 or P1 were locally treated with PPT or DPN 7.5,
22.5 or 67.5 ng/μL and 24 h after treatment egg transport
was evaluated in all groups as described above. The mean
number (X ± SE) of eggs recovered from the oviducts of
control or treated groups are shown in figure 7. Intrabur-
sal administration of PPT decreased the number of eggs
recovered from the oviduct in cyclic and pregnant rats
although at lower doses in cyclic rats. Administration of
DPN decreased the number of eggs only in C1, but not in
P1 rats. Mean numb
s100g in th
lowing intr
ESR1, PPT
Figure 6 p
(
)
,
f
g
y
y
p
g
y
g
g
,
,
g
Mean number of the relative copies (X ± SE) of Ckb,
c-fos and s100g in the rat oviducts on day 1 of cycle or
pregnancy following intrabursal treatments with the
selective agonists of ESR1, PPT or ESR2, DPN. V:
vehicle of drugs, PPT: 67.5 ng/μL, DPN: 67.5 ng/μL. All treat-
ments were given 3 h before autopsy. Each experiment con-
sisted of 4 replicas. Means with different letters are
significantly different from each other (P < 0.05), a ≠ b ≠ c ≠
d. p
(
)
,
f
g
y
y
p
g
y
g
g
,
,
g
Mean number of the relative copies (X ± SE) of Ckb,
c-fos and s100g in the rat oviducts on day 1 of cycle or
pregnancy following intrabursal treatments with the
selective agonists of ESR1, PPT or ESR2, DPN. V:
vehicle of drugs, PPT: 67.5 ng/μL, DPN: 67.5 ng/μL. All treat-
ments were given 3 h before autopsy. Each experiment con-
sisted of 4 replicas. Means with different letters are
significantly different from each other (P < 0.05), a ≠ b ≠ c ≠
d. d. Interestingly, ESR1 and ESR2 were also observed associ-
ated with the cilia of epithelial cells (see insert in figure
1d). Estradiol regulates differentiation and dedifferentia-
tion of ciliated cells of the mammalian oviduct [38]. Fur-
thermore, follicular fluid of human pre-ovulatory follicles
containing high concentrations of estradiol and progester-
one increased the ciliary beat frequency of human oviduc-
tal ciliated cells [39]. Our findings suggest that E2 could
regulate ciliary activity directly through a nongenomic
mechanism probably involving phosphorylation/dephos-
phorylation of some proteins (e.g. tubulin or dynein)
present in this structure. Recently, it has been shown that
ESR2 is colocalized with β-tubulin at stem portion of the
cilia of the oviductal epithelial cells in immature rats [40]. Additionally, gold particles for ESR1 and ESR2 were found E2. Furthermore, the data reported here show a dynamic
behavior of these ER pools in response to mating-associ-
ated signals and to an E2 pulse. E2. Furthermore, the data reported here show a dynamic
behavior of these ER pools in response to mating-associ-
ated signals and to an E2 pulse. Since mating induces IPS we expected that pregnant and
cyclic rats would exhibit different ER subcellular distribu-
tion, e.g. Discussion An important variable that influences E2 actions on its tar-
get cells is the differential distribution of ESR1 and ESR2. Previous works have reported presence of ESR1 and ESR2
in the epithelium and smooth muscle layers of the amp- Page 7 of 11
(page number not for citation purposes) Page 7 of 11
(page number not for citation purposes) http://www.rbej.com/content/7/1/139 Reproductive Biology and Endocrinology 2009, 7:139 Mean number of the relative copies (X ± SE) of Ckb, c-fos and
s100g in the rat oviducts on day 1 of cycle or pregnancy fol-
lowing intrabursal treatments with the selective agonists of
ESR1, PPT or ESR2, DPN
Figure 6
Mean number of the relative copies (X ± SE) of Ckb,
c-fos and s100g in the rat oviducts on day 1 of cycle or
pregnancy following intrabursal treatments with the
selective agonists of ESR1, PPT or ESR2, DPN. V:
vehicle of drugs, PPT: 67.5 ng/μL, DPN: 67.5 ng/μL. All treat-
ments were given 3 h before autopsy. Each experiment con-
sisted of 4 replicas. Means with different letters are
significantly different from each other (P < 0.05), a ≠ b ≠ c ≠
d.
b
b
a,b
b
a
c
b
c
c
c
c
a,b a,b
a,b
d
d
d
d
Relative copies of c-fos
(X±SE)
Relative copies of Ckb
(X±SE)
Relative copies of s100g
(X±SE)
N= 4
N= 4
N= 4
b
c
c
d
d
d
Relative copies of Ckb
(X±SE)
N= 4 lial cells induced by mating. ESR1 and ESR2 are also
expressed in the mucosa and muscle layer of the rat ovi-
duct [25,26] so that it is possible that mating stimulates
accumulation of ER in the nuclear compartment of other
cell types. The changes described in gold particle density
may reflect either change in antibody accessibility to
immunoreactive epitopes or true changes in the expres-
sion level of ER. The current data does not allow to distin-
guish between these two possibilities but increases up to
ten fold 3 hours after E2 administration seem more plau-
sible as a result of ER dissociation from scaffolding pro-
teins than a result of de novo synthesis. Discussion In fact, previously
we have found that levels of ESR1 and ESR2 mRNA and
protein in whole oviducts of pregnant rats were similar to
those in oviducts of cycling rats, suggesting that mating
does not regulate global expression of ER in the oviduct
[12].
b
b
c
c
c
d
Relative copies of c-fos
(X±SE)
N= 4 We observed different responses in the subcellular distri-
bution of ER in the epithelial cells of the two oviductal
segments. It is known that the relative proportion
between ciliated and secretory epithelial cells varies con-
siderably from ampulla to isthmus [36] so that is proba-
ble that mating-associated signals may have acted
differentially on these two cell types. Further analysis that
segregates the responses of ciliated from secretory epithe-
lial needs to be done. It has been reported that in the rat,
E2 acts only in the isthmic segment to accelerate egg trans-
port [37] while that isthmus-specific apoptosis of epithe-
lial cells and activation of cilia-localized ESR2A induced
by clomiphene citrate act in parallel to block egg transport
[19]. Thus, it is probable that differences in the distribu-
tion of ESR1 and ESR2 between ampulla and isthmus
could reflect specific contribution of these segments to
signals provided by E2 to regulate egg transport. Further-
more, we did not discard the possibility that E2 acts
directly on the smooth muscle cells because it has been
found presence of ESR1 and ESR2 in the myosalpinx of
the rat [25,26].
a,b
a
b
a,b a,b
a,b
Relative copies of s100g
(X±SE)
N= 4 Mean numb
s100g in th
lowing intr
ESR1, PPT
Figure 6 higher ESR1, ESR2 or both in the nuclear com-
partment. However, quantitative analyses showed that
mating increased the number of both receptors only in the
non-nuclear compartments. Furthermore, when E2 was
administered to pregnant or cyclic rats receptor immuno-
reactivity also accumulated in the non-nuclear compart-
ments. Thus, IPS occurs independently of the changes in
the distribution of ESR1 and ESR2 in the oviductal epithe- Page 8 of 11
(page number not for citation purposes) Page 8 of 11
(page number not for citation purposes) Reproductive Biology and Endocrinology 2009, 7:139 http://www.rbej.com/content/7/1/139 Mean number of eggs (X ± SE) recovered from rat oviducts on day 2 of the cycle or pregnancy, 24 h after intrabursal treatment
with different doses of the selective agonists of ESR1, PPT or ESR2, DPN
Figure 7
Mean number of eggs (X ± SE) recovered from rat oviducts on day 2 of the cycle or pregnancy, 24 h after
intrabursal treatment with different doses of the selective agonists of ESR1, PPT or ESR2, DPN. Figures inside
the bars indicate the number of animals used. Means with different letters are significantly different from each other within
each graph, a ≠ b ≠ c (P < 0.05). Mean numb
s100g in th
lowing intr
ESR1, PPT
Figure 6 0
2
4
6
8
10
12
14
Number of oviductal oocytes
(X±SE)
5
3
4
4
a
b
b
b
V
7.5
22.5
67.5
Cyclic
PPT ng/μL
0
2
4
6
8
10
12
14
6
4
6
6
a
b
c
d
Pregnant
Number of oviductal embryos
(X±SE)
PPT ng/μL
0
2
4
6
8
10
12
14
5
3
4
4
a
a
a
b
Number of oviductal oocytes
(X±SE)
DPN ng/μL
0
2
4
6
8
10
12
14
6
4
6
6
Number of oviductal embryos
(X±SE)
DPN ng/μL
Cyclic
Pregnant
V
7.5
22.5
67.5
V
7.5
22.5
67.5
V
7.5
22.5
67.5 0
2
4
6
8
10
12
14
Number of oviductal oocytes
(X±SE)
5
3
4
4
a
b
b
b
V
7.5
22.5
67.5
Cyclic
PPT ng/μL
0
2
4
6
8
10
12
14
6
4
6
6
a
b
c
d
Pregnant
Number of oviductal embryos
(X±SE)
PPT ng/μL
V
7.5
22.5
67.5 Pregnant 0
2
4
6
8
10
12
14
5
3
4
4
a
a
a
b
Number of oviductal oocytes
(X±SE)
DPN ng/μL
0
2
4
6
8
10
12
14
6
4
6
6
Number of oviductal embryos
(X±SE)
DPN ng/μL
Cyclic
Pregnant
V
7.5
22.5
67.5
V
7.5
22.5
67.5 Mean number of eggs (X ± SE) recovered from rat oviducts on day 2 of the cycle or pregnancy, 24 h after intrabursal treatment
with different doses of the selective agonists of ESR1, PPT or ESR2, DPN
Figure 7
Mean number of eggs (X ± SE) recovered from rat oviducts on day 2 of the cycle or pregnancy, 24 h after
intrabursal treatment with different doses of the selective agonists of ESR1, PPT or ESR2, DPN. Figures inside
the bars indicate the number of animals used. Means with different letters are significantly different from each other within
each graph, a ≠ b ≠ c (P < 0.05). Mean number of eggs (X ± SE) recovered from rat oviducts on day 2 of the cycle or pregnancy, 24 h after intrabursal treatment
with different doses of the selective agonists of ESR1, PPT or ESR2, DPN
Figure 7
Mean number of eggs (X ± SE) recovered from rat oviducts on day 2 of the cycle or pregnancy, 24 h after
intrabursal treatment with different doses of the selective agonists of ESR1, PPT or ESR2, DPN. Mean numb
s100g in th
lowing intr
ESR1, PPT
Figure 6 Figures inside
the bars indicate the number of animals used. Means with different letters are significantly different from each other within
each graph, a ≠ b ≠ c (P < 0.05). affect expression of ESR2 [43]. In ovine endothelial cells,
short-term treatment with E2 down-regulates ESR1, but
not ESR2 while long-term treatment up-regulates ESR1
and down-regulates ESR2 expression [44]. Our findings
provide the first evidence that E2 is able to differentially
regulate not just the expression level, but also the subcel-
lular distribution of ESR1 and ESR2 in a target cell. We
also observed different responses in the expression of
three E2-associated signaling genes, c-fos, Ckb and s100g,
in the oviduct of pregnant and cyclic rats when ESR1 or
ESR2 was activated. Moreover, activation of ESR1 or ESR2
increased expression of c-fos although mating only
blunted the effect of ESR2. This indicates that mating-
associated signals modulate intraoviductal signalling of
both ER providing evidence that mating may change the
functional role of these receptors in the rat oviduct. On associated to secretory vesicles. This corroborates previous
works reporting localization of ESR1 and ER2 in the rough
endoplasmic reticulum and secretory vesicles of the
female rat pituitary cells [41]. Although, the biological sig-
nificance of the localization of ESR1 and ESR2 in secretory
vesicles remains to be determined it is probable that a
Golgi-dependent pathway could exist for translation of ER
that could be translocated into the plasma membrane and
mediate nongenomic responses [41]. In other estrogen-sensitive tissues ER subtype expression
is differentially regulated by E2. In the human vena cava,
E2 down-regulates ESR1 expression [42] while deprivation
of E2 in the cerebral microvessels of ovariectomized rats is
associated with a decrease in the expression of both iso-
forms and E2 replacement up-regulates ESR1 but does not Page 9 of 11
(page number not for citation purposes) Page 9 of 11
(page number not for citation purposes) Reproductive Biology and Endocrinology 2009, 7:139 http://www.rbej.com/content/7/1/139 the other hand, the role of c-fos, Ckb or s100g on IPS
induction or E2-induced egg transport acceleration needs
to be disclosed. WDS collaborate in the design of the studies of the immu-
nomicroscopy of ESR1 and ESR2 and quantification of
gold particles for ESR1 and ESR2. LV and HBC partici-
pated in planning experiments and contributed to draft-
ing the manuscript. All authors have read and approved
the final manuscript. Mean numb
s100g in th
lowing intr
ESR1, PPT
Figure 6 Given that PPT is 400-fold more selective and DPN is only
70-fold more selective for ESR1 an ESR2 respectively, it
was not surprisingly that PPT would be more effective to
accelerate oviductal egg transport than DPN in cyclic rats. However, mating decreased effectiveness of PPT and
blocked the effect of DPN. Probably, the nongenomic
pathway by which E2 accelerates egg transport operates
through activation of either ESR1 and ESR2 while the
genomic pathway only operates through ESR1. The fact
that IPS is associated with suppression of ESR2 involve-
ment in the kinetic effect of E2 in the oviduct indicate that
mating exerts a profound influence on the biology of ER
in a target organ of E2 that merits further investigation. Competing interests 13. Mathieu CL, Mills SE, Burnett SH, Cloney DL, Bruns DE, Bruns ME:
The presence and estrogen control of immunoreactive calbi-
ndin-D9k in the fallopian tube of the rat. Endocrinology 1989,
125:2745-2750. p
g
The authors declare that there is no conflict of interest that
could be perceived as prejudicing the impartiality of the
research reported. 14. Nilsson S, Mäkelä S, Treuter E, Tujague M, Thomsen J, Andersson G,
Enmark E, Petterson K, Warner M, Gustafsson JA: Mechanisms of
estrogen action. Physiol Rev 2001, 81:1535-1565. Conclusion Estrogen receptors ESR1 and ESR2 present a subcellular
distribution in oviductal epithelial cells that is compatible
with genomic and nongenomic actions of E2 in the rat ovi-
duct. Mating is associated with changes in the basal and
E2-induced subcellular distribution of ESR1 and ESR2 in
these cells although it did not clearly explain IPS. Further-
more, mating affected signaling of both ER in the oviduct
and induced loss of functional involvement of ESR2 on
E2-induced accelerated egg transport. These findings
reveal a profound influence on the intraoviductal ER sig-
nalling pathways exerted by mating. 9. p
y
g
J
9. Parada-Bustamante A, Orihuela PA, Ríos M, Navarrete-Gómez PA,
Cuevas CA, Velasquez LA, Villalón MJ, Croxatto HB: Catechol-o-
methyltransferase and methoxyestradiols participate in the
intraoviductal nongenomic pathway through which estradiol
accelerates egg transport in cycling rats. Biol Reprod 2007,
77:934-941. 10. Lattanzi ML, Santos CB, Mudry MD, Baranao JL: Exposure of bovine
oocytes to the endogenous metabolite 2-methoxyestradiol
during in vitro maturation inhibits early embryonic develop-
ment. Biol Reprod 2003, 69:1793-1800. 11. Hyder SM, Chiappetta C, Stancel GM: Synthetic estrogen
17alpha-ethinyl estradiol induces pattern of uterine gene
expression similar to endogenous estrogen 17beta-estradiol. J Pharmacol Exp Ther 1999, 290:740-747. J
p
,
12. Orihuela PA, Zuñiga LM, Rios M, Llados C, Sierralta W, Croxatto HB:
Estrogen receptors in the rat oviduct. Mol Biol Cell 2004, 15:. p
huela PA, Zuñiga LM, Rios M, Llados C, Sierralta W, Croxatto References 1. Croxatto HB: Physiology of gamete and embryo transport
through the Fallopian tube. Reprod Biomed Online 2002,
4:160-169. 1. Croxatto HB: Physiology of gamete and embryo transport
through the Fallopian tube. Reprod Biomed Online 2002,
4:160-169. 2. Villalón M, Ortiz ME, Aguayo C, Muñoz J, Croxatto HB: Differential
transport of fertilized and unfertilized ova in the rat. Biol
Reprod 1982, 26:337-341. p
3. Müller AL, Ortiz ME, Ambriz GD, Andrade YN, Croxatto HB: Mat-
ing associated signals have profound effects on ovum trans-
port parameters in rats. Placenta 2005, 27:A57. Shao et al [19] have reported that subcutaneous adminis-
tration of DPN retard egg transport in the rat. In this
study, immature animals were treated, prior to DPN
administration, with gonatrophins to mimic the endog-
enous luteneizing hormone surge. Probably, this treat-
ment could have affected the response of the oviduct to
DPN. Another factor is that we recorded the number and
distribution of eggs in the genital tract within the first 24
h after treatment. In order to detect whether PPT or DPN
delay egg transport autopsies should be performed on day
4 or 5 of cycle or pregnancy respectively, but this was not
done. p
p
4. Ríos M, Orihuela PA, Croxatto HB: Intraoviductal administra-
tion of ribonucleic acid from estrogen-treated rats mimics
the effect of estrogen on ovum transport. Biol Reprod 1997,
56:279-283. p
p
4. Ríos M, Orihuela PA, Croxatto HB: Intraoviductal administra-
tion of ribonucleic acid from estrogen-treated rats mimics
the effect of estrogen on ovum transport. Biol Reprod 1997,
56:279-283. 5. Orihuela PA, Croxatto HB: Acceleration of oviductal transport
of oocytes induced by estradiol in cycling rats is mediated by
nongenomic stimulation of protein phosphorylation in the
oviduct. Biol Reprod 2001, 65:238-1245. p
6. Orihuela PA, Ríos M, Croxatto HB: Disparate effects of estradiol
on egg transport and oviductal protein synthesis in mated
and cyclic rats. Biol Reprod 2001, 65:1232-1237. y
p
7. Orihuela PA, Parada-Bustamante A, Cortés PP, Gatica C, Croxatto
HB: Estrogen Receptor, Cyclic Adenosine Monophosphate,
and Protein Kinase A Are Involved in the Nongenomic Path-
way by Which Estradiol Accelerates Oviductal Oocyte
Transport in Cyclic Rats. Biol Reprod 2003, 68:1225-1231. p
y
p
8. Orihuela PA, Parada-Bustamante A, Zuñiga LM, Croxatto HB: Inosi-
tol triphosphate participates in an oestradiol nongenomic
signalling pathway involved in accelerated oviductal trans-
port in cycling rats. J Endocrinol 2006, 188:579-588. Acknowledgements This work was supported by grants received from FONDECYT 1030315,
1080523, PROGRESAR (PRE 004/2003) and Proyecto BASAL FB0807. Page 10 of 11
(page number not for citation purposes) Authors' contributions Gonzales M, Reyes M, Damas C, Alonso R, Bello AR: Oestrogen
receptor α and β in female rat pituitary cells: An immuno-
chemical study. Gen Comp Endocrinol 2008, 155:857-868. 21. Muscillo M, La Rosa G, Marianelli C, Zaniratti S, Capobianchi MR,
Cantiani L, Carducci A: A new RT-PCR method for the identifi-
cation of reoviruses in seawater samples. Water Res 2001,
35:548-556. 42. Knauthe R, Diel P, Hegele-Hartung C, Engelhaupt A, Fritzemeier KH:
Sexual dimorphism of steroid hormone receptor messenger
ribonucleic acid expression and hormonal regulation in rat
vascular tissue. Endocrinology 1996, 137:3220-3227. gy
43. Jesmin S, Hattori Y, Sakuma I, Liu MY, Mowa CN, Kitabatake A:
Estrogen deprivation and replacement modulate cerebral
capillary density with vascular expression of angiogenic mol-
ecules in middle-aged female rats. J Cereb Blood Flow Metab
2003, 23:181-189. 22. Livak KJ, Schmittgen TD: Analysis of relative gene expression
data using real-time quantitative PCR and the 2(T)(-Delta
Delta C) method. Methods 2001, 25:402-408. )
23. Reynolds ES: The use of lead citrate at high pH as an electron-
opaque stain in electron microscopy. J Cell Biol 1963,
17:208-212. 44. Ihionkhan CE, Chambliss KL, Gibson LL, Hahner LD, Mendelsohn ME,
Shaul PW: Estrogen causes dynamic alterations in endothelial
estrogen receptor expression. Circ Res 2002, 91:814-820. 24. Sierralta WD: Immunoelectron microscopy in embryos. Meth-
ods 2001, 24:61-69. 25. Mowa CN, Iwanaga T: Differential distribution of oestrogen
receptor-α and -β mRNAs in the female reproductive organ
of rats as revealed by in situ hybridization. J Endocrinol 2000,
165:59-66. 26. Wang H, Eriksson H, Sahlin L: Estrogen receptors alpha and beta
in the female reproductive tract of the rat during the estrous
cycle. Biol Reprod 2000, 63:1331-1340. 27. Kessels MM, Qualmann B, Thole HH, Sierralta WD: Subcellular
localization of estradiol receptor in MCF7 cells studied with
nanogold-labelled antibody fragments. Eur J Histochem 1998,
42:259-270. 28. Qualmann B, Kessels MM, Thole HH, Sierralta WD: A hormone
pulse induces transient changes in the subcellular distribu-
tion and leads to a lysosomal accumulation of the estradiol
receptor alpha in target tissues. Eur J Cell Biol 2000, 79:383-393. p
p
g
J
,
29. Welshons WV, Lieberman ME, Gorski J: Nuclear localization of
unoccupied oestrogen receptors. Nature 1984, 307:747-749. 30. Márquez DC, Pietras RJ: Membrane-associated binding sites for
estrogen contribute to growth regulation of human breast
cancer cells. Oncogen 2001, 20:5420-5430. g
31. Authors' contributions g
y
15. Turner CD: Endocrinology of the ovary. In General Endocrinology
3rd edition. WB Saunders Company. Philadelphia; 1961:365-409. PAO participated in the design of the study, in directing
and completing all experimental analysis and in writing
the manuscript. LMZ, MR, APB performed the sampling of
the animals, carried out the Real-Time PCR, intrabursal
injections of drugs and assessment of the egg transport. p
y
p
16. Harris HA, Katzenellenbogen JA, Katzenellenbogen BS: Characteri-
zation of the biological roles of the estrogen receptors, ERα
and ERβ, in estrogen target tissues in vivo through the use of
an ERα-selective ligand. Endocrinology 2002, 143:4172-4177. g
gy
17. Frasor J, Barnett DH, Danes JM, Hess R, Parlow AF, Katzenellenbogen
BS: Response-specific and ligand dose-dependent modulation g
gy
17. Frasor J, Barnett DH, Danes JM, Hess R, Parlow AF, Katzenellenbogen
BS: Response-specific and ligand dose-dependent modulation Page 10 of 11
(page number not for citation purposes) Reproductive Biology and Endocrinology 2009, 7:139 http://www.rbej.com/content/7/1/139 38. Reeder RL, Shirley B: Deciliation in the ampulla of the rat ovi-
duct and effects of estrogen on the process. J Exp Zool 1999,
283:71-80. of estrogen receptor (ER) alpha activity by ERβ in the uterus. Endocrinology 2003, 144:3159-3166. gy
18. Lund TD, Rovis T, Chung WC, Handa RJ: Novel actions of estro-
gen receptors-beta on anxiety-related behaviors. Endocrinol-
ogy 2005, 146:797-807. 39. Lyons RA, Saridogan E, Djahanbakhch O: The effect of ovarian fol-
licular fluid and peritoneal fluid on Fallopian tube ciliary beat
frequency. Hum Reprod 2006, 21:52-56. 19. Shao R, Nutu M, Weijdegård B, Egecioglu E, Fernandez-Rodriguez J,
Karlsson-Lindahl L, Gemzell-Danielsson K, Bergh C, Billig H: Clomi-
phene citrate causes aberrant tubal apoptosis and estrogen
receptor activation in rat fallopian tube: implications for
tubal ectopic pregnancy. Biol Reprod 2009, 80:1262-1271. 40. Shao R, Weijdegård B, Fernandez-Rodriguez J, Egecioglu E, Zhu C,
Andersson N, Thurin-Kjellberg A, Bergh C, Billig H: Ciliated epithe-
lial-specific and regional-specific expression and regulation
of the estrogen receptor-β2 in the fallopian tubes of imma-
ture rats: a possible mechanism for estrogen-mediated
transport process in vivo. Am J Physiol 2007, 293:E147-E158. p
p
g
p
20. Ríos M, Hermoso M, Sánchez TM, Croxatto HB, Villalón MJ: Effect
of oestradiol and progesterone on the instant and directional
velocity of microsphere movements in the rat oviduct: gap
junctions mediate the kinetic effect of oestradiol. Reprod Fertil
Dev 2007, 19:634-640. p
p
J
y
41. Authors' contributions Monje P, Boland R: Subcellular distribution of native estrogen
receptor α and β isoforms in rabbit uterus and ovary. J Cell
Biochem 2001, 82:467-479. 32. Aronica SM, Krauss WL, Katzenellenbogen BS: Estrogen action via
the cAMP signaling pathway: Stimulation of adenylate
cyclase and cAMP-regulated gene transcription. PNAS USA
1994, 91:8517-8521. 33. Farhat MY, Abiyounes S, Dingaan B, Vargas R, Ramwell PW: Estra-
diol increases cyclic adenosine monophosphate in rat pulmo-
nary vascular smooth muscle cells by a nongenomic
mechanism. J Pharmacol Exp Ther 1996, 276:652-657. 34. Prakash YS, Togaibayeva AA, Kannan MS, Miller VM, Fitzpatrick LA,
Sieck GC: Estrogen increases Ca2+ efflux from female porcine
coronary arterial smooth muscle. Am J Physiol 1999,
45:H926-H934. Publish with BioMed Central and every
scientist can read your work free of charge
"BioMed Central will be the most significant development for
disseminating the results of biomedical research in our lifetime."
Sir Paul Nurse, Cancer Research UK
Your research papers will be:
available free of charge to the entire biomedical community
peer reviewed and published immediately upon acceptance
cited in PubMed and archived on PubMed Central
yours — you keep the copyright
Submit your manuscript here:
http://www.biomedcentral.com/info/publishing_adv.asp
BioMedcentral
Page 11 of 11
(page number not for citation purposes) Publish with BioMed Central and every
scientist can read your work free of charge
"BioMed Central will be the most significant development for
disseminating the results of biomedical research in our lifetime."
Sir Paul Nurse, Cancer Research UK
Your research papers will be:
available free of charge to the entire biomedical community
peer reviewed and published immediately upon acceptance
cited in PubMed and archived on PubMed Central
yours — you keep the copyright
Submit your manuscript here:
http://www.biomedcentral.com/info/publishing_adv.asp
BioMedcentral Publish with BioMed Central and every
scientist can read your work free of charge 35. Migliaccio A, Castoria G, Di Domenico M, de Falco A, Bilancio A,
Lombardi M, Barone MV, Ametrano D, Zannini MS, Abbondanza C,
Auricchio F: Steroid-induced androgen receptor-oestradiol
receptor beta-Src complex triggers prostate cancer cell pro-
liferation. EMBO J 2000, 19:5406-5417. 36. Hafez ESE, Black DL: The mammalian uterotubal junction. In
The Mammalian Oviduct: Comparative Biology and Methodology Edited
by: Hafez, Blandau. The University of Chicago Press. Chicago;
1969:85-128. 37. Moore GD, Croxatto HB: Effects of delayed transfer and treat-
ment with oestrogen on the transport of microspheres by
the rat oviduct. J Reprod Fertil 1998, 83:795-802. Authors' contributions Page 11 of 11
(page number not for citation purposes)
|
https://openalex.org/W2535940026
|
https://biblio.ugent.be/publication/8568623/file/8568805.pdf
|
English
| null |
Measurement of differential cross sections for top quark pair production using the <mml:math xmlns:mml="http://www.w3.org/1998/Math/MathML" display="inline"><mml:mrow><mml:mtext>lepton</mml:mtext><mml:mo>+</mml:mo><mml:mrow><mml:mtext>jets</mml:mtext></mml:mrow></mml:mrow></mml:math> final state in proton-proton collisions at 13 TeV
|
Physical review. D/Physical review. D.
| 2,017
|
cc-by
| 39,159
|
Measurement of differential cross sections for top quark
pair production using the lepton+jets final state in
proton-proton collisions at 13 TeV iv:1610.04191v2 [hep-ex] 10 May 2017 The CMS Collaboration∗ arXiv:1610.04191v2 [hep-ex] 10 May Published in Physical Review D as doi:10.1103/PhysRevD.95.092001. UROPEAN ORGANIZATION FOR NUCLEAR RESEARCH (CERN CMS-TOP-16-008 Measurement of differential cross sections for top quark
pair production using the lepton+jets final state in
proton-proton collisions at 13 TeV Abstract arXiv:1610.04191v2 [hep-ex] Differential and double-differential cross sections for the production of top quark
pairs in proton-proton collisions at 13 TeV are measured as a function of jet multiplic-
ity and of kinematic variables of the top quarks and the top quark-antiquark system. This analysis is based on data collected by the CMS experiment at the LHC corre-
sponding to an integrated luminosity of 2.3 fb−1. The measurements are performed
in the lepton+jets decay channels with a single muon or electron in the final state. The
differential cross sections are presented at particle level, within a phase space close to
the experimental acceptance, and at parton level in the full phase space. The results
are compared to several standard model predictions. Published in Physical Review D as doi:10.1103/PhysRevD.95.092001. c⃝2017 CERN for the benefit of the CMS Collaboration. CC-BY-3.0 license 1
Introduction Studying the differential production cross sections of top quark pairs (tt) at high energies is
a crucial ingredient in testing the standard model and searching for sources of new physics,
which could alter the production rate. In particular, the differential tt cross sections probe pre-
dictions of quantum chromodynamics (QCD) and facilitate the comparisons of the data with
state-of-the-art calculations. In addition, some of the measured distributions, especially dis-
tributions of invariant mass and rapidity of the tt system, can be used to improve our under-
standing of parton distribution functions (PDFs). A measurement of the tt differential and double-differential production cross sections as a func-
tion of jet multiplicity and of kinematic variables of the top quarks and the tt system is pre-
sented. The measurement is based on proton-proton collision data at a center-of-mass energy
of 13 TeV corresponding to an integrated luminosity of 2.3 fb−1 [1]. The data were recorded by
the CMS experiment at the CERN LHC in 2015. This measurement makes use of the tt decay
into the ℓ+jets (ℓ= e, µ) final state, where, after the decay of each top quark into a bottom quark
and a W boson, one of the W bosons decays hadronically and the other one leptonically. The τ
lepton decay mode is not considered here as signal. The differential cross sections are presented
as a function of the transverse momentum pT and the absolute rapidity |y| of the hadronically
(th) and the leptonically (tℓ) decaying top quarks; as a function of pT, |y|, and mass M of the tt
system. The cross section is also measured as a function of the number of additional jets in the
event. In addition, the differential cross sections as a function of pT(th) and pT(tt) are measured
in bins of jet multiplicity and double-differential cross sections for the following combinations
of variables are determined: |y(th)| vs. pT(th), M(tt) vs. |y(tt)|, and pT(tt) vs. M(tt). This measurement continues a series of differential tt production cross section measurements
in proton-proton collisions at the LHC. Previous measurements at 7 TeV [2, 3] and 8 TeV [4–8]
have been performed in various tt decay channels. The differential cross sections are presented in two different ways, at particle level and at par-
ton level. c⃝2017 CERN for the benefit of the CMS Collaboration. CC-BY-3.0 license ∗See Appendix C for the list of collaboration members ∗See Appendix C for the list of collaboration members 1 2
Signal and background modeling The Monte Carlo programs POWHEG [9–12] (v2) and MADGRAPH5 aMC@NLO [13] (v2.2.2)
(MG5 aMC@NLO) are used to simulate tt events. They include NLO QCD matrix element cal-
culations that are combined with the parton shower simulation of PYTHIA [14, 15] (v8.205)
(PYTHIA8) using the tune CUETP8M1 [16]. In addition, MG5 aMC@NLO is used to produce
simulations of tt events with additional partons. In one simulation all processes of up to three
additional partons are calculated at leading order (LO) and combined with the PYTHIA8 parton
shower simulation using the MLM [17] algorithm. In another simulation all processes of up to
two additional partons are calculated at NLO and combined with the PYTHIA8 parton shower
simulation using the FxFx [18] algorithm. The default parametrization of the PDF used in all
simulations is NNPDF30 nlo as 0118 [19]. A top quark mass mt = 172.5 GeV is used. When
compared to the data, simulations are normalized to an inclusive tt production cross section
of 832+40
−46 pb [20]. This value is calculated with next-to-NLO (NNLO) precision including the
resummation of next-to-next-to-leading-logarithmic (NNLL) soft gluon terms. Its given uncer-
tainty is due to the choice of hadronization/factorization scales and PDF. In all simulations, event weights are calculated that represent the usage of the uncertainty
eigenvector sets of the PDF. There are also event weights available that represent the changes
of factorization and renormalization scales by a factor of two or one half. These additional
weights allow for the calculation of systematic uncertainties due to the PDF and the scale
choices. For additional uncertainty estimations we use POWHEG+PYTHIA8 simulations with
top quark masses of 171.5 and 173.5 GeV, with parton shower scales varied up and down by a
factor of two, and a simulation with POWHEG combined with HERWIG++ [21] (v2.7.1) using the
tune EE5C [22]. The main backgrounds are produced using the same techniques. The MG5 aMC@NLO generator
is used for the simulation of W boson production in association with jets, t-channel single
top quark production, and Drell–Yan (DY) production in association with jets. The POWHEG
generator is used for the simulation of single top quark associated production with a W boson
(tW) and PYTHIA8 is used for multijet production. In all cases the parton shower and the
hadronization are described by PYTHIA8. The W boson and DY backgrounds are normalized
to their NNLO cross sections [23]. 1
Introduction For the particle-level measurement a proxy of the top quark is defined based on ex-
perimentally accessible quantities like jets, which consist of quasi-stable particles with a mean
lifetime greater than 30 ps. These are described by theoretical calculations that, in contrast to
pure matrix-element calculations, involve parton shower and hadronization models. These ob-
jects are required to match closely the experimental acceptance. A detailed definition is given
in Section 3. Such an approach has the advantage that it reduces theoretical uncertainties in the
experimental results by avoiding theory-based extrapolations from the experimentally accessi-
ble portion of the phase space to the full range, and from jets to partons. However, such results
cannot be compared to parton-level calculations. For the measurement at parton level, the top quarks are defined directly before decaying into a
bottom quark and a W boson. For this analysis the parton-level tt system is calculated at next-
to-leading order (NLO) and combined with a simulation of the parton shower. No restriction
of the phase space is applied for parton-level top quarks. The experimental signature is the same for both measurements and consists of two jets coming
from the hadronization of b quarks (b jets), two jets from a hadronically decaying W boson, a
transverse momentum imbalance associated with the neutrino, and a single isolated muon or
electron. This paper is organized as follows: In Section 2 we provide a description of the signal and back-
ground simulations, followed by the definition of the particle-level top quarks in Section 3. After a short overview of the CMS detector and the particle reconstruction in Section 4, we 2
Signal and background modeling 2 describe the object and event selections in Sections 5 and 6, respectively. Section 7 contains
a detailed description of the reconstruction of the tt system. Details on the background esti-
mation and the unfolding are presented in Sections 8 and 9. After a discussion on systematic
uncertainties in Section 10, the results are finally presented in Section 11. 2
Signal and background modeling The single top quark processes are normalized to NLO
calculations [24, 25], and the multijet simulation is normalized to the LO calculation [15]. The detector response is simulated using GEANT4 [26]. Afterwards, the same reconstruction
algorithms that are applied to the data are used. Multiple proton-proton interactions per bunch
crossing (pileup) are included in the simulation. To correct the simulation to be in agreement
with the pileup conditions observed during the data taking, the average number of pileup
events per bunch crossing is calculated for the measured instantaneous luminosity. The simu-
lated events are weighted, depending on their number of pileup interactions, to reproduce the
measured pileup distribution. 3 3
Particle-level top quark definition We take the sum of the four-momenta of all
selected neutrinos as the neutrino momentum pν from the leptonically decaying top quark and
find the permutation of jets that minimizes the quantity K2 = [M(pν + pℓ+ pbℓ) −mt]2 + [M(pj1 + pj2) −mW]2 + [M(pj1 + pj2 + pbh) −mt]2,
( (1) e pj1/2 are the four-momenta of two light-flavor jet candidates, pbℓ/h a where pj1/2 are the four-momenta of two light-flavor jet candidates, pbℓ/h are the four-momenta
of two b-jet candidates, pℓis the four-momentum of the lepton, and mW = 80.4 GeV is the mass
of the W boson. All jets with pT > 25 GeV and |η| < 2.5 are considered. At least four jets are
required, of which at least two must be b jets. If there are more than two b jets, we allow b
jets as decay products of the proxy for the hadronically decaying W boson. Due to a limited
efficiency of the b jet identification at detector level this improves the agreement between the
reconstructed top quarks and the particle-level top quarks. The remaining jets with the same
kinematic selection are considered as additional jets at particle level. where pj1/2 are the four-momenta of two light-flavor jet candidates, pbℓ/h are the four-momenta
of two b-jet candidates, pℓis the four-momentum of the lepton, and mW = 80.4 GeV is the mass
of the W boson. All jets with pT > 25 GeV and |η| < 2.5 are considered. At least four jets are
required, of which at least two must be b jets. If there are more than two b jets, we allow b
jets as decay products of the proxy for the hadronically decaying W boson. Due to a limited
efficiency of the b jet identification at detector level this improves the agreement between the
reconstructed top quarks and the particle-level top quarks. The remaining jets with the same
kinematic selection are considered as additional jets at particle level. It should be remarked that events with a hadronic and a leptonic particle-level top quark are
not required to be ℓ+jets events at the parton level. As an example, in Fig. 1 the relation between
the pT(th) values at particle and parton level is shown. 3
Particle-level top quark definition The following list describes the definitions of objects constructed from quasi-stable particles,
obtained from the predictions of tt event generators before any detector simulation. These
objects are further used to define the particle-level top quarks. • Muons and electrons that do not have their origin in a decay of a hadron are selected
and their momenta are corrected for the final-state radiation effects. The anti-kT jet
algorithm [27, 28] with a distance parameter of 0.1 is used to cluster the leptons
and photons not originating from hadron decays. Those photons that are clustered
together with a selected lepton are assumed to have been radiated by the lepton and
their momenta are added to the lepton momentum. However, the lepton is only
selected if the original pT is at least half of their corrected pT. • All neutrinos that do not have their origin in a decay of a hadron are selected. • Jets are clustered by the anti-kT jet algorithm with a distance parameter of 0.4. All
quasi-stable particles are considered, excluding the selected neutrinos and leptons
together with their radiated photons. • b jets at particle level are defined as those jets that contain a b hadron. As a result of
the short lifetime of b hadrons, only their decay products should be considered for
the jet clustering. However, to allow their association to a jet, the b hadrons are also
included with their momenta scaled down to a negligible value. This preserves the
information of their directions, but they have no impact on the jet clustering itself. Based on the invariant masses M of these objects, we construct a pair of particle-level top
quarks in the ℓ+jets final state. Events with exactly one muon or electron with pT > 30 GeV and
an absolute pseudorapidity |η| < 2.5 are selected. 4
The CMS detector Each column is normalized to the number of events per column at parton level in
the full phase space. are measured in gas-ionization detectors embedded in the steel flux-return yoke outside the
solenoid. A more detailed description of the CMS detector, together with a definition of the
coordinate system and relevant kinematic variables, can be found in Ref. [29]. are measured in gas-ionization detectors embedded in the steel flux-return yoke outside the
solenoid. A more detailed description of the CMS detector, together with a definition of the
coordinate system and relevant kinematic variables, can be found in Ref. [29]. The particle-flow (PF) event algorithm [30, 31] reconstructs and identifies each individual par-
ticle with an optimized combination of information from the various elements of the CMS
detector. The energy of photons is directly obtained from the ECAL measurement, corrected
for zero-suppression effects. The energy of electrons is determined from a combination of the
electron momentum at the primary interaction vertex as determined by the tracker, the energy
of the corresponding ECAL cluster, and the energy sum of all bremsstrahlung photons spa-
tially compatible with originating from the electron track. The energy of muons is obtained
from the curvature of the corresponding track. The energy of charged hadrons is determined
from a combination of their momentum measured in the tracker and the matching ECAL and
HCAL energy deposits, corrected for zero-suppression effects and for the response function of
the calorimeters to hadronic showers. Finally, the energy of neutral hadrons is obtained from
the corresponding corrected ECAL and HCAL energy. 4
The CMS detector Normalized at parton level per bin
0.05
0.1
0.15
0.2
0.25
0.3
0.35
0.4
0.45
0.13
0.03
0.01
0.01
0.01
0.01
0.00
0.00
0.01
0.06
0.16
0.06
0.03
0.02
0.01
0.01
0.01
0.01
0.02
0.04
0.17
0.07
0.03
0.02
0.01
0.01
0.02
0.01
0.01
0.03
0.18
0.07
0.03
0.02
0.01
0.01
0.00
0.00
0.01
0.03
0.21
0.08
0.02
0.01
0.01
0.00
0.00
0.00
0.01
0.03
0.24
0.07
0.02
0.01
0.00
0.01
0.03
0.28
0.05
0.01
0.00
0.00
0.01
0.04
0.36
0.05
0.00
0.00
0.03
0.42
)
h
(t
T
Bin at parton level p
1
2
3
4
5
6
7
8
9
)
h
(t
T
Bin at particle level p
1
2
3
4
5
6
7
8
9
CMS
Simulation ) [GeV]
h
(t
T
Particle level p
0
100
200
300
400
500
600
700
800
Fraction
0
0.2
0.4
0.6
0.8
1
1.2
Purity
Stability
Bin efficiency
Non parton-level signal
CMS
Simulation ) [GeV]
h
(t
T
Particle level p
400
500
600
700
800 Figure 1: Comparison between the pT(th) at particle and parton level, extracted fr Figure 1: Comparison between the pT(th) at particle and parton level, extracted from the
POWHEG+PYTHIA8 simulation. Left: fraction of parton-level top quarks in the same bin at
particle level (purity), fraction of particle-level top quarks in the same bin at parton level (sta-
bility), ratio of the number of particle- to parton-level top quarks, and fraction of events with a
particle-level top quark pair that are not considered as signal events at parton level. Right: bin
migrations between particle and parton level. The pT range of the bins can be taken from the
left panel. Each column is normalized to the number of events per column at parton level in
the full phase space. Figure 1: Comparison between the pT(th) at particle and parton level, extracted from the
POWHEG+PYTHIA8 simulation. Left: fraction of parton-level top quarks in the same bin at
particle level (purity), fraction of particle-level top quarks in the same bin at parton level (sta-
bility), ratio of the number of particle- to parton-level top quarks, and fraction of events with a
particle-level top quark pair that are not considered as signal events at parton level. Right: bin
migrations between particle and parton level. The pT range of the bins can be taken from the
left panel. 4
The CMS detector The central feature of the CMS detector is a superconducting solenoid of 6 m internal diameter,
providing a magnetic field of 3.8 T. Within the solenoid volume are a silicon pixel and strip
tracker, a lead tungstate crystal electromagnetic calorimeter (ECAL), and a brass and scintil-
lator hadron calorimeter (HCAL), each composed of a barrel and two endcap sections. For-
ward calorimeters extend the η coverage provided by the barrel and endcap detectors. Muons 5
Physics object reconstruction 4 ) [GeV]
h
(t
T
Particle level p
0
100
200
300
400
500
600
700
800
Fraction
0
0.2
0.4
0.6
0.8
1
1.2
Purity
Stability
Bin efficiency
Non parton-level signal
CMS
Simulation
Normalized at parton level per bin
0.05
0.1
0.15
0.2
0.25
0.3
0.35
0.4
0.45
0.13
0.03
0.01
0.01
0.01
0.01
0.00
0.00
0.01
0.06
0.16
0.06
0.03
0.02
0.01
0.01
0.01
0.01
0.02
0.04
0.17
0.07
0.03
0.02
0.01
0.01
0.02
0.01
0.01
0.03
0.18
0.07
0.03
0.02
0.01
0.01
0.00
0.00
0.01
0.03
0.21
0.08
0.02
0.01
0.01
0.00
0.00
0.00
0.01
0.03
0.24
0.07
0.02
0.01
0.00
0.01
0.03
0.28
0.05
0.01
0.00
0.00
0.01
0.04
0.36
0.05
0.00
0.00
0.03
0.42
)
h
(t
T
Bin at parton level p
1
2
3
4
5
6
7
8
9
)
h
(t
T
Bin at particle level p
1
2
3
4
5
6
7
8
9
CMS
Simulation
Figure 1: Comparison between the pT(th) at particle and parton level, extracted from the
POWHEG+PYTHIA8 simulation. Left: fraction of parton-level top quarks in the same bin at
particle level (purity), fraction of particle-level top quarks in the same bin at parton level (sta-
bility), ratio of the number of particle- to parton-level top quarks, and fraction of events with a
particle-level top quark pair that are not considered as signal events at parton level. Right: bin
migrations between particle and parton level. The pT range of the bins can be taken from the
left panel. Each column is normalized to the number of events per column at parton level in
the full phase space. 5
Physics object reconstruction This analysis depends on the reconstruction and identification of muons, electrons, jets, and
missing transverse momentum associated with a neutrino. Only leptons are selected that are
compatible with originating from the primary vertex, defined as the vertex at the beam posi-
tion with the highest sum of p2
T of the associated tracks. Leptons from tt decays are typically
isolated, i.e., separated in ∆R =
p
(∆φ)2 + (∆η)2 from other particles. A requirement on the
lepton isolation is used to reject leptons produced in decays of hadrons. The muon isolation variable is defined as the sum of the pT of all tracks, except for the muon
track, originating from the tt interaction vertex within a cone of ∆R = 0.3. It is required to be
less than 5% of the muon pT. The muon reconstruction and selection [32] efficiency is measured 5 in the data using tag-and-probe techniques [33]. Depending on the pT and η of the muon it is
90–95%. in the data using tag-and-probe techniques [33]. Depending on the pT and η of the muon it is
90–95%. For electrons the isolation variable is the sum of the pT of neutral hadrons, charged hadrons,
and photon PF candidates in a cone of ∆R = 0.3 around the electron. Contributions of the
electron to the isolation variable are suppressed excluding a small region around the electron. This isolation variable is required to be smaller than 7% of the electron pT. An event-by-event
correction is applied that maintains a constant electron isolation efficiency with respect to the
number of pileup interactions [34]. The measured reconstruction and identification [35] effi-
ciency for electrons is 70–85% with a pT and η dependence. Jets are reconstructed from PF objects clustered using the anti-kT jet algorithm with a distance
parameter of 0.4 using the FASTJET package [28]. Charged particles originating from a vertex of
a pileup interaction are excluded. The total energy of the jets is corrected for energy depositions
from pileup. In addition, pT- and η-dependent corrections are applied to correct for detector
response effects [36]. Those jets identified as isolated muons or electrons are removed from
consideration. For the identification of b jets the combined secondary vertex algorithm [37] is used. It provides
a discriminant between light-flavor and b jets based on the combined information of secondary
vertices and the impact parameter of tracks at the primary vertex. 5
Physics object reconstruction A jet is identified as b jet if the
associated value of the discriminant exceeds a threshold criterion. Two threshold criteria are
used: a tight threshold with an efficiency of about 70% and a light-flavor jet rejection probability
of 95%, and a loose one with an efficiency of about 80% and a rejection probability of 85%. The missing transverse momentum ⃗pmiss
T
is calculated as the negative of the vectorial sum of
transverse momenta of all PF candidates in the event. Jet energy corrections are also propa-
gated to improve the measurement of ⃗pmiss
T
. 7
Reconstruction of the top quark-antiquark system The goal of the tt reconstruction is the correct identification of reconstructed objects as parton-
or particle-level top quark decay products. To test the performance of the reconstruction algo-
rithm an assignment between detector level and particle- (parton-) level objects is needed. For
the particle-level measurement this relationship is straightforward. Reconstructed leptons and
jets can be matched spatially to corresponding objects at the particle level. For the parton-level
measurement we need to define how to match the four initial quarks from a tt decay with re-
constructed jets. This is not free of ambiguities since a quark does generally not lead to a single
jet. One quark might shower into several jets or multiple quarks might be clustered into one jet
if they are not well separated. We introduce an unambiguous matching criterion that matches
the reconstructed jet with the highest pT within ∆R = 0.4 to a quark from the tt decay. If two
quarks are matched with the same jet, the event has a merged topology and is considered as
“not reconstructible” in the context of this analysis. The same matching criterion is also used to assign particle-level jets to the tt decay products at
parton level. Those particle-level jets with pT > 25 GeV and |η| < 2.5, which are not assigned
to one of the initial quarks, are considered as additional jets at parton level. For the reconstruction of the top quark-antiquark system all possible permutations of jets that
assign reconstructed jets to the decay products of the tt system are tested and a likelihood that
a certain permutation is correct is evaluated. Permutations are considered only if the two jets
with the highest b tagging probabilities are the two b-jet candidates. In addition, the pT of at
least one b-jet candidate and at least one jet candidate from the W boson decay have to be above
35 GeV. In each event the permutation with the highest probability is selected. The likelihoods
are evaluated separately for the particle- and the parton-level measurements. The first reconstruction step involves the determination of the neutrino four-momentum pν. This is performed using the algorithm described in Ref. [38]. The idea is to find all possible
solutions for the three components of the neutrino momentum using the two mass constraints
(pν + pℓ)2 = m2
W and (pν + pℓ+ pbℓ)2 = m2
t . 6
Event selection Events are selected if they pass single-lepton triggers. These require pT > 22 GeV for electrons
and pT > 20 GeV for muons, as well as various quality and isolation criteria. To reduce the background contributions and optimize the tt reconstruction additional, more
stringent, requirements on the events are imposed. Events with exactly one muon or electron
with pT > 30 GeV and |η| < 2.1 are selected. No additional muons or electrons with pT >
15 GeV and |η| < 2.4 are allowed. In addition to the lepton, at least four jets with pT > 30 GeV
and |η| < 2.4 are required. At least two of these jets must be tagged as b jets. At least one jet has
to fulfill the tight b-jet identification criterion while for the second b jet only the loose criterion
is required. At least one of the two jets with the highest value of the b tagging discriminant and
at least one of the remaining jets are required to have pT > 35 GeV. We compare several kinematic distributions of the muon and electron channels to the simu-
lation to verify that there are no unexpected differences. The ratios of the measured to the
expected event yields in the two channels agree within the uncertainty in the lepton recon-
struction and selection efficiencies. In the remaining steps of the analysis the two channels are
combined by adding their distributions. 7
Reconstruction of the top quark-antiquark system 6 7
Reconstruction of the top quark-antiquark system Each equation describes an ellipsoid in the
three-dimensional momentum space of the neutrino. The intersection of these two ellipsoids is
usually an ellipse. We select pν as the point on the ellipse for which the distance Dν,min between
the ellipse projection onto the transverse plane and ⃗pmiss
T
is minimal. This algorithm leads to a
unique solution for the longitudinal neutrino momentum and an improved resolution for the
transverse component. The minimum distance Dν,min can also be used to identify the correct
bℓ. In the cases with an invariant mass of the lepton and the bℓcandidate above mt no solution
can be found and we continue with the next permutation. The likelihood λ is maximized to select the best permutation of jets. It uses constraints of the
top quark and W boson masses on the hadronic side and the Dν,min value from the neutrino
reconstruction, and is defined through (2) −log(λ) = −log(Pm(m2, m3)) −log(Pν(Dν,min)),
(2) where Pm is the two-dimensional probability distribution of the invariant masses of correctly
reconstructed W bosons and top quarks. This probability is calculated for the invariant mass
of the two jets m2 tested as the W boson decay products, and the invariant mass of the three
jets m3 tested as the decay products of the hadronically decaying top quark. The distributions
for the correct jet assignments, taken from the POWHEG+PYTHIA8 simulation and normalized
to unity, are shown in Fig. 2 for the particle- and parton-level measurements. Permutations
with probabilities of less than 0.1% of the highest value are rejected. This part of the likelihood
is sensitive to the correct reconstruction of the hadronically decaying top quark, modulo a 7 permutation of the two jets from the W boson, but none of the measured kinematic variables
will be affected by this ambiguity. The probability Pν describes the distribution of Dν,min for a correctly selected bℓ. In Fig. 2 the
normalized distributions of Dν,min for bℓand for other jets are shown. On average, the distance
Dν,min for correctly selected bℓis smaller and has a lower tail compared to the distance obtained
for other jets. Permutations with values of Dν,min > 150 GeV are rejected since they are very
unlikely to originate from a correct bℓassociation. This part of the likelihood is sensitive to the
correct reconstruction of the leptonically decaying top quark. 7
Reconstruction of the top quark-antiquark system ) [GeV]
h
M(t
0
50 100 150 200 250 300 350 400 450 500
M(W) [GeV]
0
50
100
150
200
250
300
350
400
450
500
7
−
10
6
−
10
5
−
10
4
−
10
CMS
Simulation
parton level
+jets
µ
e/ ) [GeV]
h
M(t
0
50 100 150 200 250 300 350 400 450 500
M(W) [GeV]
0
50
100
150
200
250
300
350
400
450
500
7
−
10
6
−
10
5
−
10
4
−
10
CMS
Simulation
particle level
+jets
µ
e/ ) [GeV]
h
M(t
50
00
50 500 [GeV]
,min
ν
D
0
20
40
60
80
100
120
140
0
0.005
0.01
0.015
0.02
0.025
0.03
0.035
0.04
l
Correct b
l
Wrong b
CMS
Simulation
parton level
+jets
µ
e/ [GeV]
,min
ν
D
0
20
40
60
80
100
120
140
0
0.005
0.01
0.015
0.02
0.025
0.03
0.035
0.04
l
Correct b
l
Wrong b
CMS
Simulation
particle level
+jets
µ
e/ Figure 2: Top: normalized two-dimensional mass distribution of the correct reconstructed
hadronically decaying W bosons M(W) and the correct reconstructed top quarks M(th) for the
parton- (left) and the particle- (right) level measurements. Bottom: normalized distributions of
the distance Dν,min for correctly and wrongly selected b jets from the leptonically decaying top
quarks. The distributions are taken from the POWHEG+PYTHIA8 tt simulation. Figure 2: Top: normalized two-dimensional mass distribution of the correct reconstructed
hadronically decaying W bosons M(W) and the correct reconstructed top quarks M(th) for the
parton- (left) and the particle- (right) level measurements. Bottom: normalized distributions of
the distance Dν,min for correctly and wrongly selected b jets from the leptonically decaying top
quarks. The distributions are taken from the POWHEG+PYTHIA8 tt simulation. The likelihood λ combines the probabilities from the reconstruction of the hadronically and lep-
tonically decaying top quarks and provides information on reconstructing the whole tt system. The performance of the reconstruction algorithm is tested using the three tt simulations gen-
erated with POWHEG combined with PYTHIA8 or HERWIG++, and MG5 aMC@NLO+PYTHIA8
where we use the input distributions Pm and Pν from POWHEG+PYTHIA8. The efficiency of
the reconstruction algorithm is defined as the probability that the most likely permutation, as
identified through the maximization of the likelihood λ, is the correct one, given that all decay
products from the tt decay are reconstructed and selected. 7
Reconstruction of the top quark-antiquark system ) [GeV]
h
M(t
0
50 100 150 200 250 300 350 400 450 500
M(W) [GeV]
0
50
100
150
200
250
300
350
400
450
500
7
−
10
6
−
10
5
−
10
4
−
10
CMS
Simulation
parton level
+jets
µ
e/
) [GeV]
h
M(t
0
50 100 150 200 250 300 350 400 450 500
M(W) [GeV]
0
50
100
150
200
250
300
350
400
450
500
7
−
10
6
−
10
5
−
10
4
−
10
CMS
Simulation
particle level
+jets
µ
e/
[GeV]
,min
ν
D
0
20
40
60
80
100
120
140
0
0.005
0.01
0.015
0.02
0.025
0.03
0.035
0.04
l
Correct b
l
Wrong b
CMS
Simulation
parton level
+jets
µ
e/
[GeV]
,min
ν
D
0
20
40
60
80
100
120
140
0
0.005
0.01
0.015
0.02
0.025
0.03
0.035
0.04
l
Correct b
l
Wrong b
CMS
Simulation
particle level
+jets
µ
e/
Figure 2: Top: normalized two-dimensional mass distribution of the correct reconstructed
hadronically decaying W bosons M(W) and the correct reconstructed top quarks M(th) for the
parton- (left) and the particle- (right) level measurements. Bottom: normalized distributions of
the distance Dν,min for correctly and wrongly selected b jets from the leptonically decaying top
quarks. The distributions are taken from the POWHEG+PYTHIA8 tt simulation. 7
Reconstruction of the top quark-antiquark system These efficiencies as a function of
the jet multiplicity are shown in Fig. 3. Since the number of permutations increases drastically
with the number of jets, it is more likely to select a wrong permutation if there are additional
jets. The small differences observed in different simulations are taken into account for the un- 7
Reconstruction of the top quark-antiquark system 8 certainty estimations. We observe a lower reconstruction efficiency for the particle-level mea-
surement. This is caused by the weaker mass constraints for a particle-level top quark, where,
in contrast to the parton-level top quark, exact matches to the top quark and W boson masses
are not required. This can be seen in the mass distributions of Fig. 2 and the likelihood distri-
butions in Fig. 4. Here the signal simulation is divided into the following categories: correctly
reconstructed tt systems (tt right reco), events where all decay products are available, but the al-
gorithm failed to identify the correct permutation (tt wrong reco), ℓ+jets tt events where at least
one decay product is missing (tt not reconstructible), and nonsignal tt events (tt background). However, the lower reconstruction efficiency of the particle-level top quark is compensated by
the higher number of reconstructible events. Additional jets
0
1
2
3
4
≥
reconstruction eff. tt
0
0.1
0.2
0.3
0.4
0.5
0.6
0.7
0.8
0.9
1
POWHEG P8
POWHEG H++
MG5_aMC@NLO P8
CMS
Simulation
parton level
+jets
µ
e/
Additional jets
0
1
2
3
4
≥
reconstruction eff. tt
0
0.1
0.2
0.3
0.4
0.5
0.6
0.7
0.8
0.9
1
POWHEG P8
POWHEG H++
MG5_aMC@NLO P8
CMS
Simulation
particle level
+jets
µ
e/
Figure 3: Reconstruction efficiency of the tt system as a function of the number of additional
jets for the parton- (left) and particle- (right) level measurements calculated based on the
simulations with POWHEG+PYTHIA8 (P8), POWHEG+HERWIG++ (H++), and MG5 aMC@NLO
+PYTHIA8. Additional jets
0
1
2
3
4
≥
reconstruction eff. tt
0
0.1
0.2
0.3
0.4
0.5
0.6
0.7
0.8
0.9
1
POWHEG P8
POWHEG H++
MG5_aMC@NLO P8
CMS
Simulation
parton level
+jets
µ
e/ Additional jets
0
1
2
3
4
≥
reconstruction eff. 7
Reconstruction of the top quark-antiquark system tt
0
0.1
0.2
0.3
0.4
0.5
0.6
0.7
0.8
0.9
1
POWHEG P8
POWHEG H++
MG5_aMC@NLO P8
CMS
Simulation
particle level
+jets
µ
e/ Figure 3: Reconstruction efficiency of the tt system as a function of the number of additional
jets for the parton- (left) and particle- (right) level measurements calculated based on the
simulations with POWHEG+PYTHIA8 (P8), POWHEG+HERWIG++ (H++), and MG5 aMC@NLO
+PYTHIA8. )
λ
-log(
Events / 0.4
0
1000
2000
3000
4000
5000
6000
7000
Data
right reco
tt
wrong reco
tt
not reconstructible
tt
background
tt
Single t
W/DY+jets
Multijet
(13 TeV)
-1
2.3 fb
parton level
+jets
µ
e/
CMS
)
λ
-log(
10
12
14
16
18
20
22
Theory
Data
0.6
0.8
1
1.2
1.4 )
λ
-log(
Events / 0.4
0
1000
2000
3000
4000
5000
6000
7000
Data
right reco
tt
wrong reco
tt
not reconstructible
tt
background
tt
Single t
W/DY+jets
Multijet
(13 TeV)
-1
2.3 fb
parton level
+jets
µ
e/
CMS
)
λ
-log(
10
12
14
16
18
20
22
Theory
Data
0.6
0.8
1
1.2
1.4
)
λ
-log(
Events / 0.4
0
1000
2000
3000
4000
5000
6000
7000
Data
right reco
tt
wrong reco
tt
not reconstructible
tt
background
tt
Single t
W/DY+jets
Multijet
(13 TeV)
-1
2.3 fb
particle level
+jets
µ
e/
CMS
)
λ
-log(
10
12
14
16
18
20
22
Theory
Data
0.6
0.8
1
1.2
1.4
Figure 4: Distribution of the negative log-likelihood for the selected best permutation in the
parton- (left) and the particle- (right) level measurements in data and simulations. The simula-
tion of POWHEG+PYTHIA8 is used to describe the tt production. Experimental (cf. Section 10)
and statistical uncertainties (hatched area) are shown for the total simulated yield, which is nor-
malized to the measured integrated luminosity. The ratios of data to the sum of the expected
yields are provided at the bottom of each panel. In Fig. 5 the distributions of pT and |y| of the reconstructed hadronically decaying top quarks
for the parton- and particle-level measurements are compared to the simulation. In Fig. 6 the
distributions of pT(tt), |y(tt)|, M(tt), and the number of additional jets are shown. 7
Reconstruction of the top quark-antiquark system In general, 9 good agreement is observed between the data and the simulation though the overall yield in the
data is slightly lower, but within the experimental uncertainties. The observed jet multiplicities
are lower than predicted. ) [GeV]
h
(t
T
p
Events / 40 GeV
2
10
3
10
4
10
5
10
Data
signal
tt
background
tt
Single t
W/DY+jets
Multijet
(13 TeV)
-1
2.3 fb
parton level
+jets
µ
e/
CMS
) [GeV]
h
(t
T
p
0
100
200
300
400
500
600
700
800
Theory
Data
0.6
0.8
1
1.2
1.4 ) [GeV]
h
(t
T
p
Events / 40 GeV
2
10
3
10
4
10
5
10
Data
signal
tt
background
tt
Single t
W/DY+jets
Multijet
(13 TeV)
-1
2.3 fb
parton level
+jets
µ
e/
CMS
) [GeV]
h
(t
T
p
0
100
200
300
400
500
600
700
800
Theory
Data
0.6
0.8
1
1.2
1.4
) [GeV]
h
(t
T
p
Events / 40 GeV
2
10
3
10
4
10
5
10
Data
signal
tt
background
tt
Single t
W/DY+jets
Multijet
(13 TeV)
-1
2.3 fb
particle level
+jets
µ
e/
CMS
) [GeV]
h
(t
T
p
0
100
200
300
400
500
600
700
800
Theory
Data
0.6
0.8
1
1.2
1.4
)|
h
|y(t
Events / 0.25
0
2000
4000
6000
8000
10000
12000
14000
Data
signal
tt
background
tt
Single t
W/DY+jets
Multijet
(13 TeV)
-1
2.3 fb
parton level
+jets
µ
e/
CMS
)|
h
|y(t
0
0.5
1
1.5
2
2.5
Theory
Data
0.6
0.8
1
1.2
1.4
)|
h
|y(t
Events / 0.25
0
2000
4000
6000
8000
10000
12000
14000
16000
18000
Data
signal
tt
background
tt
Single t
W/DY+jets
Multijet
(13 TeV)
-1
2.3 fb
particle level
+jets
µ
e/
CMS
)|
h
|y(t
0
0.5
1
1.5
2
2.5
Theory
Data
0.6
0.8
1
1.2
1.4
Figure 5: Comparisons of the reconstructed pT(th) (top) and |y(th)| (bottom) in data and simu-
lations for the parton (left) and the particle (right) level. The simulation of POWHEG+PYTHIA8
is used to describe the tt production. Experimental (cf. Section 10) and statistical uncertain-
ties (hatched area) are shown for the total simulated yield, which is normalized according to
the measured integrated luminosity. The ratios of data to the expected yields are given at the
bottom of each panel. 7
Reconstruction of the top quark-antiquark system ) [GeV]
h
(t
T
p
Events / 40 GeV
2
10
3
10
4
10
5
10
Data
signal
tt
background
tt
Single t
W/DY+jets
Multijet
(13 TeV)
-1
2.3 fb
particle level
+jets
µ
e/
CMS
) [GeV]
h
(t
T
p
0
100
200
300
400
500
600
700
800
Theory
Data
0.6
0.8
1
1.2
1.4 ) [
]
h
(
T
p
h
T
)|
h
|y(t
Events / 0.25
0
2000
4000
6000
8000
10000
12000
14000
16000
18000
Data
signal
tt
background
tt
Single t
W/DY+jets
Multijet
(13 TeV)
-1
2.3 fb
particle level
+jets
µ
e/
CMS
)|
h
|y(t
0
0.5
1
1.5
2
2.5
Theory
Data
0.6
0.8
1
1.2
1.4 ) [
]
h
(
T
p
)|
h
|y(t
Events / 0.25
0
2000
4000
6000
8000
10000
12000
14000
Data
signal
tt
background
tt
Single t
W/DY+jets
Multijet
(13 TeV)
-1
2.3 fb
parton level
+jets
µ
e/
CMS
)|
h
|y(t
0
0.5
1
1.5
2
2.5
Theory
Data
0.6
0.8
1
1.2
1.4 Figure 5: Comparisons of the reconstructed pT(th) (top) and |y(th)| (bottom) in data and simu-
lations for the parton (left) and the particle (right) level. The simulation of POWHEG+PYTHIA8
is used to describe the tt production. Experimental (cf. Section 10) and statistical uncertain-
ties (hatched area) are shown for the total simulated yield, which is normalized according to
the measured integrated luminosity. The ratios of data to the expected yields are given at the
bottom of each panel. 7
Reconstruction of the top quark-antiquark system In general, )
λ
-log(
Events / 0.4
0
1000
2000
3000
4000
5000
6000
7000
Data
right reco
tt
wrong reco
tt
not reconstructible
tt
background
tt
Single t
W/DY+jets
Multijet
(13 TeV)
-1
2.3 fb
particle level
+jets
µ
e/
CMS
)
λ
-log(
10
12
14
16
18
20
22
Theory
Data
0.6
0.8
1
1.2
1.4 Figure 4: Distribution of the negative log-likelihood for the selected best permuta Figure 4: Distribution of the negative log-likelihood for the selected best permutation in the
parton- (left) and the particle- (right) level measurements in data and simulations. The simula-
tion of POWHEG+PYTHIA8 is used to describe the tt production. Experimental (cf. Section 10)
and statistical uncertainties (hatched area) are shown for the total simulated yield, which is nor-
malized to the measured integrated luminosity. The ratios of data to the sum of the expected
yields are provided at the bottom of each panel. Figure 4: Distribution of the negative log-likelihood for the selected best permutation in the
parton- (left) and the particle- (right) level measurements in data and simulations. The simula-
tion of POWHEG+PYTHIA8 is used to describe the tt production. Experimental (cf. Section 10)
and statistical uncertainties (hatched area) are shown for the total simulated yield, which is nor-
malized to the measured integrated luminosity. The ratios of data to the sum of the expected
yields are provided at the bottom of each panel. parton (left) and the particle (right) level measurements in data and simulations. The
tion of POWHEG+PYTHIA8 is used to describe the tt production. Experimental (cf. Sec
and statistical uncertainties (hatched area) are shown for the total simulated yield, whic
malized to the measured integrated luminosity. The ratios of data to the sum of the e
yields are provided at the bottom of each panel. In Fig. 5 the distributions of pT and |y| of the reconstructed hadronically decaying top quarks
for the parton- and particle-level measurements are compared to the simulation. In Fig. 6 the
distributions of pT(tt), |y(tt)|, M(tt), and the number of additional jets are shown. In general, In Fig. 5 the distributions of pT and |y| of the reconstructed hadronically decaying top quarks
for the parton- and particle-level measurements are compared to the simulation. In Fig. 6 the
distributions of pT(tt), |y(tt)|, M(tt), and the number of additional jets are shown. 8
Background subtraction After the event selection and tt reconstruction about 65 000 (53 000) events are observed in the
particle- (parton-) level measurements. A small contribution of about 9% of single top quark,
DY, W boson, and multijet events is expected. These have to be estimated and subtracted from
the selected data. The background from single top quark production is subtracted based on its simulation. Its
overall contribution corresponds to about 4% of the selected data. Single top quark production
cross sections are calculated with precisions of a few percent [24, 25]. Since the calculations
have a limited reliability after tt selection we assume an overall uncertainty of 50%. However,
this conservative estimate has negligible impact on the final results and their accuracy. The simulations of multijet, DY, and W boson production contain limited numbers of events
after the full selection. 8
Background subtraction We extract the shapes of the distributions of these backgrounds from 8
Background subtraction 10 ) [GeV]
t
(t
T
p
Events / 20 GeV
2
10
3
10
4
10
5
10
Data
signal
tt
background
tt
Single t
W/DY+jets
Multijet
(13 TeV)
-1
2.3 fb
parton level
+jets
µ
e/
CMS
) [GeV]
t
(t
T
p
0
50
100
150
200
250
300
350
400
450
500
Theory
Data
0.6
0.8
1
1.2
1.4
) [GeV]
t
(t
T
p
Events / 20 GeV
2
10
3
10
4
10
5
10
Data
signal
tt
background
tt
Single t
W/DY+jets
Multijet
(13 TeV)
-1
2.3 fb
particle level
+jets
µ
e/
CMS
) [GeV]
t
(t
T
p
0
50
100
150
200
250
300
350
400
450
500
Theory
Data
0.6
0.8
1
1.2
1.4
) [GeV]
t
M(t
Events / 80 GeV
2
10
3
10
4
10
5
10
Data
signal
tt
background
tt
Single t
W/DY+jets
Multijet
(13 TeV)
-1
2.3 fb
parton level
+jets
µ
e/
CMS
) [GeV]
t
M(t
200
400
600
800
1000
1200
1400
Theory
Data
0.6
0.8
1
1.2
1.4
) [GeV]
t
M(t
Events / 80 GeV
2
10
3
10
4
10
5
10
Data
signal
tt
background
tt
Single t
W/DY+jets
Multijet
(13 TeV)
-1
2.3 fb
particle level
+jets
µ
e/
CMS
) [GeV]
t
M(t
200
400
600
800
1000
1200
1400
Theory
Data
0.6
0.8
1
1.2
1.4
)|
t
|y(t
Events / 0.25
0
2000
4000
6000
8000
10000
12000
14000
16000
18000
20000
Data
signal
tt
background
tt
Single t
W/DY+jets
Multijet
(13 TeV)
-1
2.3 fb
parton level
+jets
µ
e/
CMS
)|
t
|y(t
0
0.5
1
1.5
2
2.5
Theory
Data
0.6
0.8
1
1.2
1.4
additional jets
Events
0
10000
20000
30000
40000
50000
Data
signal
tt
background
tt
Single t
W/DY+jets
Multijet
(13 TeV)
-1
2.3 fb
particle level
+jets
µ
e/
CMS
Additional jets
0
1
2
3
4
5
6
≥
Theory
Data
0.6
0.8
1
1.2
1.4
Figure 6: Comparisons of the reconstructed distributions of pT(tt) (top) and M(tt) (middle)
for the parton- (left) and the particle- (right) level measurements in data and simulations. Bot-
tom: distributions of |y(tt)| (left) and the number of additional jets (right). The simulation
of POWHEG+PYTHIA8 is used to describe the tt production. Experimental (cf. 8
Background subtraction ) [GeV]
t
(t
T
p
Events / 20 GeV
2
10
3
10
4
10
5
10
Data
signal
tt
background
tt
Single t
W/DY+jets
Multijet
(13 TeV)
-1
2.3 fb
particle level
+jets
µ
e/
CMS
) [GeV]
t
(t
T
p
0
50
100
150
200
250
300
350
400
450
500
Theory
Data
0.6
0.8
1
1.2
1.4 ) [GeV]
t
(t
p
Events / 20 GeV
2
10
3
10
4
10
5
10
Data
signal
tt
background
tt
Single t
W/DY+jets
Multijet
(13 TeV)
-1
2.3 fb
parton level
+jets
µ
e/
CMS
) [GeV]
t
(t
T
p
0
50
100
150
200
250
300
350
400
450
500
Theory
Data
0.6
0.8
1
1.2
1.4 ) [GeV]
t
(t
T
pT
) [GeV]
t
M(t
Events / 80 GeV
2
10
3
10
4
10
5
10
Data
signal
tt
background
tt
Single t
W/DY+jets
Multijet
(13 TeV)
-1
2.3 fb
parton level
+jets
µ
e/
CMS
) [GeV]
t
M(t
200
400
600
800
1000
1200
1400
Theory
Data
0.6
0.8
1
1.2
1.4 ) [GeV]
t
(t
T
pT
) [GeV]
t
M(t
Events / 80 GeV
2
10
3
10
4
10
5
10
Data
signal
tt
background
tt
Single t
W/DY+jets
Multijet
(13 TeV)
-1
2.3 fb
particle level
+jets
µ
e/
CMS
) [GeV]
t
M(t
200
400
600
800
1000
1200
1400
Theory
Data
0.6
0.8
1
1.2
1.4 Events / 80 GeV ) [GeV]
t
M(t ) [GeV]
t
M(t
) [GeV]
t
M(t ) [GeV]
t
M(t
)|
t
|y(t
Events / 0.25
0
2000
4000
6000
8000
10000
12000
14000
16000
18000
20000
Data
signal
tt
background
tt
Single t
W/DY+jets
Multijet
(13 TeV)
-1
2.3 fb
parton level
+jets
µ
e/
CMS
)|
t
|y(t
0
0.5
1
1.5
2
2.5
Theory
Data
0.6
0.8
1
1.2
1.4
additional jets
Events
0
10000
20000
30000
40000
50000
Data
signal
tt
background
tt
Single t
W/DY+jets
Multijet
(13 TeV)
-1
2.3 fb
particle level
+jets
µ
e/
CMS
Additional jets
0
1
2
3
4
5
6
≥
Theory
Data
0.6
0.8
1
1.2
1.4 ) [GeV]
t
M(t ) [GeV]
t
M(t
additional jets
Events
0
10000
20000
30000
40000
50000
Data
signal
tt
background
tt
Single t
W/DY+jets
Multijet
(13 TeV)
-1
2.3 fb
particle level
+jets
µ
e/
CMS
Additional jets
0
1
2
3
4
5
6
≥
Theory
Data
0.6
0.8
1
1.2
1.4 Events Figure 6: Comparisons of the reconstructed distributions of pT(tt) (top) and M(tt) (middle)
for the parton- (left) and the particle- (right) level measurements in data and simulations. 8
Background subtraction Section 10) and
statistical uncertainties (hatched area) are shown for the total simulated yield, which is normal-
ized according to the measured integrated luminosity. The ratios of data to the expected yields
are given at the bottom of each panel. 8
Background subtraction Bot-
tom: distributions of |y(tt)| (left) and the number of additional jets (right). The simulation
of POWHEG+PYTHIA8 is used to describe the tt production. Experimental (cf. Section 10) and
statistical uncertainties (hatched area) are shown for the total simulated yield, which is normal-
ized according to the measured integrated luminosity. The ratios of data to the expected yields
are given at the bottom of each panel. 11 a control region in the data, similar to the signal region, but requiring no b-tagged jet in the
event. In this selection the contribution of tt events is estimated to be about 15%. The remaining
fraction consists of multijet, DY, and W boson events. The reconstruction algorithm is exactly
the same as for the signal selection. To estimate the shape dependency in the control region
on the selection we vary the selection threshold of the b tagging discriminant. This changes
the top quark contribution and the flavor composition, however, we find the observed shape
variation to be negligible. For the background subtraction, the distributions extracted from
the control region are normalized to the yield of multijet, DY, and W boson events predicted
by the simulation in the signal region. In the control region the expected and measured event
yields agree within their statistical uncertainties. Taking into account the statistical uncertainty
of the normalization factor and the shape differences between the signal and control regions in
the simulation, we estimate an overall uncertainty of 20% in this background estimation. The
overall contribution to the selected data is about 5%. For the parton-level measurement, special care has to be taken with the contribution of nonsignal
tt events, i.e., dilepton, all-jets, and τ+jets events. For the particle-level measurement care is
needed with all tt events for which no pair of particle-level top quarks exists. The behavior of
this background depends on the tt cross section and a subtraction according to the expected
value can result in a bias of the measurement, especially if large differences between the simu-
lation and the data are observed. However, the shapes of the distributions show an agreement
within uncertainties between data and simulation and we subtract the predicted relative frac-
tions from the remaining event yields. 10
Systematic uncertainties We study several sources of experimental and theoretical uncertainty. Uncertainties in the jet
and ⃗pmiss
T
calibrations, in the pileup modeling, in the b tagging and lepton selection efficiencies,
and in the integrated luminosity measurement fall into the first category. Uncertainties in the jet energy calibration are estimated by shifting the energies of jets in the
simulation up and down by their pT- and η-dependent uncertainties of 3–7% [36]. At the same
time ⃗pmiss
T
is recalculated according to the rescaled jet energies. The recomputed backgrounds,
response matrices, and acceptances are used to unfold the data. The observed differences be-
tween these and the original results are taken as an uncertainty in the unfolded event yields. The same technique is used to calculate the impact of the uncertainties in the jet energy reso-
lution, the uncertainty in ⃗pmiss
T
not related to the jet energy calibration, in the b tagging, and in
the pileup modeling. The b tagging efficiency in the simulation is corrected using scale factors determined from the
data [37]. These have an uncertainty of about 2–5% depending on the pT of the b jet. The effect on the measurement due to the uncertainty in the modeling of pileup in the simu-
lation is estimated by varying the average number of pileup events per bunch crossing by 5%
and reweighting the simulated events accordingly. The trigger, reconstruction, and identification efficiencies of leptons are evaluated with tag-
and-probe techniques using Z boson dilepton decays [33]. The uncertainties in the scale factors,
which are used to correct the simulation to match the data, take into account the different lepton
selection efficiencies in events with high jet multiplicities. The overall uncertainty in the lepton
reconstruction and selection efficiencies is 3%. The relative uncertainty in the integrated luminosity measurement is 2.3% [1]. Uncertainties in the PDFs, the choice of factorization and renormalization scales, the modeling
of the parton shower and hadronization, the effect of different NLO event generation methods,
and the top quark mass fall into the second category of theoretical uncertainties. The effects of these uncertainties are estimated either by using the various event weights intro-
duced in Section 2, e.g., in the case of PDFs, factorization scale, and renormalization scale, or
by using a different tt signal simulation. The POWHEG simulation combined with HERWIG++ is
used to estimate the effect of different parton shower and hadronization models. ization. For the two-dimensional measurements with n bins in one and m bins in the other quan-
tity the D’Agostini unfolding can be generalized using a vector of n · m entries of the form:
b1,1, b2,1 . . . bn,1, . . . b1,m, b2,m . . . bn,m with a corresponding (n · m) × (n · m) migration matrix. The
number of iterations is optimized in the same way. 9
Unfolding For the unfolding, the iterative D’Agostini method [39] is used. The migration matrices and
the acceptances are needed as input. The migration matrix relates the quantities at particle
(parton) level and at detector level. It accounts for the effects from the parton shower and
hadronization as well as the detector response, where the former has a large impact on the
parton-level measurement. For the central results the migration matrices and the acceptances
are taken from the POWHEG+PYTHIA8 simulation and other simulations are used to estimate
the uncertainties. The binning in the unfolding is optimized based on the resolution in the
simulation. We utilize for the minimal bin widths that, according to the resolution, at least 50%
of the events are reconstructed in the correct bin. The iterative D’Agostini method takes the number of iterations as an input parameter to control
the level of regularization. A small number of iterations corresponds to a large regularization,
which may bias the unfolded results. The level of regularization and hence the bias decreases
with the number of iterations – but with the drawback of increasing variances in the unfolded
spectra. To optimize the number of iterations, we chose the criterion that the compatibility be-
tween a model and the unfolded data at particle (parton) level is the same as the compatibility
between the folded model and the data at detector level. The compatibilities are determined by
χ2 tests at both levels based on all available simulations and several modified spectra obtained
by reweighting the pT(t), |y(t)|, or pT(tt) distributions in the POWHEG+PYTHIA8 simulation. The reweighted spectra are chosen in such a way that they cover the observed differences be-
tween the data and the unmodified simulation. We find the above criterion fulfilled for the number of iterations such that a second χ2 test
between the detector-level spectrum with its statistical uncertainty and the refolded spectrum
exceeds a probability of 99.9%. The refolded spectrum is obtained by inverting the unfolding
step. This consists of a multiplication with the response matrix and does not need any regular- 12 10
Systematic uncertainties 10
Systematic uncertainties In addition,
POWHEG+PYTHIA8 samples with a parton shower scale varied by a factor of two are used to
study the parton shower modeling uncertainties. The result obtained with MG5 aMC@NLO is
used to estimate the effect of different NLO event generation methods. The effect due to un-
certainties in the top quark mass is estimated using simulations with altered top quark masses. We quote as uncertainty the cross section differences observed for a top quark mass variation
of 1 GeV around the central value of 172.5 GeV used in the central simulation. ckground predictions, response matrices, and acceptances obtained from these simula- 13 tions are used to unfold the data. The observed deviations with respect to the original result
are quoted as an uncertainty in the unfolded event yield. tions are used to unfold the data. The observed deviations with respect to the original result
are quoted as an uncertainty in the unfolded event yield. For the PDF uncertainty only the variation in the acceptance is taken into account while varia-
tions due to migrations between bins are neglected. It is calculated according to the uncertain-
ties in the NNPDF30 nlo as 0118 [19] parametrization. In addition, the uncertainties obtained
using the PDF sets derived with varied values of the strong coupling constant αs = 0.117 and
0.119 are considered. An overview of the uncertainties in the differential cross sections is provided in Table 1, where
the typical ranges of uncertainties in the bins are shown. In the double-differential measure-
ments the jet energy scale uncertainty is about 15% in bins of high jet multiplicities and the
dominant uncertainties due to hadronization modeling and NLO calculation reach up to 30%
for the parton-level measurements. Table 1: Overview of the uncertainties in the differential cross section measurements at particle
and at parton level. Typical ranges of uncertainties in the bins are shown. Overview of the uncertainties in the differential cross section measurements at particle
parton level. Typical ranges of uncertainties in the bins are shown. Source
Particle
Parton
level [%]
level [%]
Statistical uncertainty
1–5
1–5
Jet energy scale
5–8
6–8
Jet energy resolution
<1
<1
⃗pmiss
T
(non jet)
<1
<1
b tagging
2–3
2–3
Pileup
<1
<1
Lepton selection
3
3
Luminosity
2.3
2.3
Background
1–3
1–3
PDF
<1
<1
Fact./ren. scale
<1
<1
Parton shower scale
2–5
2–9
POWHEG+PYTHIA8 vs. 10
Systematic uncertainties HERWIG++
1–5
1–12
NLO event generation
1–5
1–10
mt
1–2
1–3 11
Cross section results The cross section σ in each bin is calculated as the ratio of the unfolded signal yield and the
integrated luminosity. These are further divided by the bin width (the product of the two bin
widths) to obtain the single- (double-) differential results. The measured differential cross sections are compared to the predictions of POWHEG and
MG5 aMC@NLO, each combined with the parton shower simulations of PYTHIA8 and HER-
WIG++. In addition, the tt multiparton simulations of MG5 aMC@NLO at LO and NLO with
a PYTHIA8 parton shower are shown in Fig. 7 (8) as a function of the top quark pT and |y|
at parton (particle) level. In Figs. 9 and 10 the cross sections as a function of kinematic vari-
ables of the tt system and the number of additional jets are compared to the same theoretical
predictions. In Fig. 11 the parton-level results are compared to theoretical predictions of various accuracies. The first is an approximate NNLO [40] QCD calculation using the CT14 NNLO [41] PDF and 11
Cross section results 14 mt = 172.5 GeV. The factorization and renormalization scales are fixed at mt. The second is an
approximate next-to-NNLO (NNNLO) [42, 43] QCD calculation using the MSTW2008nnlo [44]
PDF, mt = 172.5 GeV and factorization and renormalization scales fixed at mt. The third com-
bines the NLO QCD calculation with an improved NNLL QCD calculation (NLO+NNLL’) [45]
using the MSTW2008nnlo PDF, mt = 173.2 GeV, and the renormalization and factorization
scales of MT =
q
m2
t + p2
T(t) for the pT(t) calculation and M(tt)/2 for the M(tt) calculation. The fourth is a full NNLO [46] QCD calculation using the NNPDF3.0 PDF, mt = 173.3 GeV,
and the renormalization and factorization scales of MT/2 for the pT(t) calculation and one-
fourth of the sum of the pT of all partons for the other distributions.The displayed uncertainties
come from varying the scales up and down by a factor of two. Only the uncertainties in the
approximate NNLO calculation include PDF uncertainties and a mt variation of 1 GeV. 11
Cross section results ) [GeV]
h
(t
T
p
]
-1
[pb GeV
)
h
(t
T
dp
σ
d
2
−
10
1
−
10
1
(13 TeV)
-1
2.3 fb
parton level
+jets
µ
e/
CMS
Data
stat
⊕
Sys
Stat
POWHEG P8
POWHEG H++
MG5 P8
MG5 H++
MG5 P8 [FxFx]
MG5 P8 [MLM]
) [GeV]
h
(t
T
p
0
100
200
300
400
500
600
700
800
Data
Theory
0.8
1
1.2
) [GeV]
l(t
T
p
]
-1
[pb GeV
)
l(t
T
dp
σ
d
2
−
10
1
−
10
1
(13 TeV)
-1
2.3 fb
parton level
+jets
µ
e/
CMS
Data
stat
⊕
Sys
Stat
POWHEG P8
POWHEG H++
MG5 P8
MG5 H++
MG5 P8 [FxFx]
MG5 P8 [MLM]
) [GeV]
l(t
T
p
0
100
200
300
400
500
600
700
800
Data
Theory
0.8
1
1.2
1.4
)|
h
|y(t
[pb]
)|
h
d|y(t
σ
d
40
60
80
100
120
140
160
180
200
(13 TeV)
-1
2.3 fb
parton level
+jets
µ
e/
CMS
Data
stat
⊕
Sys
Stat
POWHEG P8
POWHEG H++
MG5 P8
MG5 H++
MG5 P8 [FxFx]
MG5 P8 [MLM]
)|
h
|y(t
0
0.5
1
1.5
2
2.5
Data
Theory
0.8
0.9
1
1.1
1.2
)|
l
|y(t
[pb]
)|
l
d|y(t
σ
d
40
60
80
100
120
140
160
180
(13 TeV)
-1
2.3 fb
parton level
+jets
µ
e/
CMS
Data
stat
⊕
Sys
Stat
POWHEG P8
POWHEG H++
MG5 P8
MG5 H++
MG5 P8 [FxFx]
MG5 P8 [MLM]
)|
l
|y(t
0
0.5
1
1.5
2
2.5
Data
Theory
0.8
0.9
1
1.1
1.2
Figure 7: Differential cross sections at parton level as a function of pT(t) (top) and |y(t)|
(bottom) measured separately for the hadronically (left) and leptonically (right) decaying top
quarks. The cross sections are compared to the predictions of POWHEG and MG5 aMC@NLO
(MG5) combined with PYTHIA8 (P8) or HERWIG++ (H++) and the multiparton simulations
MG5 aMC@NLO+PYTHIA8 MLM and MG5 aMC@NLO+PYTHIA8 FxFx. The ratios of the vari-
ous predictions to the measured cross sections are shown at the bottom of each panel together
with the statistical and systematic uncertainties of the measurement. 11
Cross section results ) [GeV]
l(t
T
p
]
-1
[pb GeV
)
l(t
T
dp
σ
d
2
−
10
1
−
10
1
(13 TeV)
-1
2.3 fb
parton level
+jets
µ
e/
CMS
Data
stat
⊕
Sys
Stat
POWHEG P8
POWHEG H++
MG5 P8
MG5 H++
MG5 P8 [FxFx]
MG5 P8 [MLM]
) [GeV]
l(t
T
p
0
100
200
300
400
500
600
700
800
Data
Theory
0.8
1
1.2
1.4 ) [GeV]
h
(t
T
p
]
-1
[pb GeV
)
h
(t
T
dp
σ
d
2
−
10
1
−
10
1
(13 TeV)
-1
2.3 fb
parton level
+jets
µ
e/
CMS
Data
stat
⊕
Sys
Stat
POWHEG P8
POWHEG H++
MG5 P8
MG5 H++
MG5 P8 [FxFx]
MG5 P8 [MLM]
) [GeV]
h
(t
T
p
0
100
200
300
400
500
600
700
800
Data
Theory
0.8
1
1.2 l
T
)|
l
|y(t
[pb]
)|
l
d|y(t
σ
d
40
60
80
100
120
140
160
180
(13 TeV)
-1
2.3 fb
parton level
+jets
µ
e/
CMS
Data
stat
⊕
Sys
Stat
POWHEG P8
POWHEG H++
MG5 P8
MG5 H++
MG5 P8 [FxFx]
MG5 P8 [MLM]
)|
l
|y(t
0
0.5
1
1.5
2
2.5
Data
Theory
0.8
0.9
1
1.1
1.2 Figure 7: Differential cross sections at parton level as a function of pT(t) (top) and |y(t)|
(bottom) measured separately for the hadronically (left) and leptonically (right) decaying top
quarks. The cross sections are compared to the predictions of POWHEG and MG5 aMC@NLO
(MG5) combined with PYTHIA8 (P8) or HERWIG++ (H++) and the multiparton simulations
MG5 aMC@NLO+PYTHIA8 MLM and MG5 aMC@NLO+PYTHIA8 FxFx. The ratios of the vari-
ous predictions to the measured cross sections are shown at the bottom of each panel together
with the statistical and systematic uncertainties of the measurement. The differential cross sections as a function of pT(th) and pT(tt) in bins of the number of ad-
ditional jets are shown in Fig. 12 (13) at parton (particle) level. The double-differential cross
sections as a function of |y(th)| vs. pT(th), M(tt) vs. |y(tt)|, and pT(tt) vs. M(tt) are shown at
parton level in Figs. 14–15 and at particle level in Figs. 16–17. The results are compared to the
predictions of the event generators. 11
Cross section results ) [GeV]
h
(t
T
p
]
-1
[pb GeV
)
h
(t
T
dp
σ
d
2
−
10
1
−
10
1
(13 TeV)
-1
2.3 fb
particle level
+jets
µ
e/
CMS
Data
stat
⊕
Sys
Stat
POWHEG P8
POWHEG H++
MG5 P8
MG5 H++
MG5 P8 [FxFx]
MG5 P8 [MLM]
) [GeV]
h
(t
T
p
0
100
200
300
400
500
600
700
800
Data
Theory
0.8
1
1.2
1.4 ) [GeV]
l(t
T
p
]
-1
[pb GeV
)
l(t
T
dp
σ
d
2
−
10
1
−
10
1
(13 TeV)
-1
2.3 fb
particle level
+jets
µ
e/
CMS
Data
stat
⊕
Sys
Stat
POWHEG P8
POWHEG H++
MG5 P8
MG5 H++
MG5 P8 [FxFx]
MG5 P8 [MLM]
) [GeV]
l(t
T
p
0
100
200
300
400
500
600
700
800
Data
Theory
1
1.2
1.4 ) [GeV]
h
(t
T
p
]
-1
[pb GeV
)
h
(t
T
dp
σ
d
2
−
10
1
−
10
1
(13 TeV)
-1
2.3 fb
particle level
+jets
µ
e/
CMS
Data
stat
⊕
Sys
Stat
POWHEG P8
POWHEG H++
MG5 P8
MG5 H++
MG5 P8 [FxFx]
MG5 P8 [MLM]
) [GeV]
h
(t
T
p
0
100
200
300
400
500
600
700
800
Data
Theory
0.8
1
1.2
1.4
) [GeV]
l(t
T
p
]
-1
[pb GeV
)
l(t
T
dp
σ
d
2
−
10
1
−
10
1
(13 TeV)
-1
2.3 fb
particle level
+jets
µ
e/
CMS
Data
stat
⊕
Sys
Stat
POWHEG P8
POWHEG H++
MG5 P8
MG5 H++
MG5 P8 [FxFx]
MG5 P8 [MLM]
) [GeV]
l(t
T
p
0
100
200
300
400
500
600
700
800
Data
Theory
1
1.2
1.4
)|
h
|y(t
[pb]
)|
h
d|y(t
σ
d
10
20
30
40
50
60
70
80
90
(13 TeV)
-1
2.3 fb
particle level
+jets
µ
e/
CMS
Data
stat
⊕
Sys
Stat
POWHEG P8
POWHEG H++
MG5 P8
MG5 H++
MG5 P8 [FxFx]
MG5 P8 [MLM]
)|
h
|y(t
0
0.5
1
1.5
2
2.5
Data
Theory
0.8
0.9
1
1.1
1.2
)|
l
|y(t
[pb]
)|
l
d|y(t
σ
d
10
20
30
40
50
60
70
80
(13 TeV)
-1
2.3 fb
particle level
+jets
µ
e/
CMS
Data
stat
⊕
Sys
Stat
POWHEG P8
POWHEG H++
MG5 P8
MG5 H++
MG5 P8 [FxFx]
MG5 P8 [MLM]
)|
l
|y(t
0
0.5
1
1.5
2
2.5
Data
Theory
0.9
1
1.1
1.2 ) [GeV]
h
(t
T
p
) [GeV]
h
(t
T
p
0
100
200
300
400
500
600
700
800
T
0.8
1
) [GeV]
l(t
T
p
) [GeV]
l(t
T
p
0
100
200
300
400
500
600
700
800
T
1
)|
h
|y(t
[pb]
)|
h
d|y(t
σ
d
10
20
30
40
50
60
70
80
90
(13 TeV)
-1
2.3 fb
particle level
+jets
µ
e/
CMS
Data
stat
⊕
Sys
Stat
POWHEG P8
POWHEG H++
MG5 P8
MG5 H++
MG5 P8 [FxFx]
MG5 P8 [MLM]
)|
h
|y(t
0
0.5
1
1.5
2
2.5
Data
Theory
0.8
0.9
1
1.1
1.2
)|
l
|y(t
[pb]
)|
l
d|y(t
σ
d
10
20
30
40
50
60
70
80
(13 TeV)
-1
2.3 fb
particle level
+jets
µ
e/
CMS
Data
stat
⊕
Sys
Stat
POWHEG P8
POWHEG H++
MG5 P8
MG5 H++
MG5 P8 [FxFx]
MG5 P8 [MLM]
)|
l
|y(t
0
0.5
1
1.5
2
2.5
Data
Theory
0.9
1
1.1
1.2 l
T
)|
l
|y(t
[pb]
)|
l
d|y(t
σ
d
10
20
30
40
50
60
70
80
(13 TeV)
-1
2.3 fb
particle level
+jets
µ
e/
CMS
Data
stat
⊕
Sys
Stat
POWHEG P8
POWHEG H++
MG5 P8
MG5 H++
MG5 P8 [FxFx]
MG5 P8 [MLM]
)|
l
|y(t
0
0.5
1
1.5
2
2.5
Data
Theory
0.9
1
1.1
1.2 Figure 8: Differential cross sections at particle level as a function of pT(t) (top) and |y(t)|
(bottom) measured separately for the hadronically (left) and leptonically (right) decaying
particle-level top quarks. 11
Cross section results All cross section values together with their statistical and
systematic uncertainties are listed in Appendices A and B for the parton- and particle-level The differential cross sections as a function of pT(th) and pT(tt) in bins of the number of ad-
ditional jets are shown in Fig. 12 (13) at parton (particle) level. The double-differential cross
sections as a function of |y(th)| vs. pT(th), M(tt) vs. |y(tt)|, and pT(tt) vs. M(tt) are shown at
parton level in Figs. 14–15 and at particle level in Figs. 16–17. The results are compared to the
predictions of the event generators. 11
Cross section results All cross section values together with their statistical and
systematic uncertainties are listed in Appendices A and B for the parton- and particle-level 15 ) [GeV]
h
(t
T
p
]
-1
[pb GeV
)
h
(t
T
dp
σ
d
2
−
10
1
−
10
1
(13 TeV)
-1
2.3 fb
particle level
+jets
µ
e/
CMS
Data
stat
⊕
Sys
Stat
POWHEG P8
POWHEG H++
MG5 P8
MG5 H++
MG5 P8 [FxFx]
MG5 P8 [MLM]
) [GeV]
h
(t
T
p
0
100
200
300
400
500
600
700
800
Data
Theory
0.8
1
1.2
1.4
) [GeV]
l(t
T
p
]
-1
[pb GeV
)
l(t
T
dp
σ
d
2
−
10
1
−
10
1
(13 TeV)
-1
2.3 fb
particle level
+jets
µ
e/
CMS
Data
stat
⊕
Sys
Stat
POWHEG P8
POWHEG H++
MG5 P8
MG5 H++
MG5 P8 [FxFx]
MG5 P8 [MLM]
) [GeV]
l(t
T
p
0
100
200
300
400
500
600
700
800
Data
Theory
1
1.2
1.4
)|
h
|y(t
[pb]
)|
h
d|y(t
σ
d
10
20
30
40
50
60
70
80
90
(13 TeV)
-1
2.3 fb
particle level
+jets
µ
e/
CMS
Data
stat
⊕
Sys
Stat
POWHEG P8
POWHEG H++
MG5 P8
MG5 H++
MG5 P8 [FxFx]
MG5 P8 [MLM]
)|
h
|y(t
0
0.5
1
1.5
2
2.5
Data
Theory
0.8
0.9
1
1.1
1.2
)|
l
|y(t
[pb]
)|
l
d|y(t
σ
d
10
20
30
40
50
60
70
80
(13 TeV)
-1
2.3 fb
particle level
+jets
µ
e/
CMS
Data
stat
⊕
Sys
Stat
POWHEG P8
POWHEG H++
MG5 P8
MG5 H++
MG5 P8 [FxFx]
MG5 P8 [MLM]
)|
l
|y(t
0
0.5
1
1.5
2
2.5
Data
Theory
0.9
1
1.1
1.2
Figure 8: Differential cross sections at particle level as a function of pT(t) (top) and |y(t)|
(bottom) measured separately for the hadronically (left) and leptonically (right) decaying
particle-level top quarks. The cross sections are compared to the predictions of POWHEG and
MG5 aMC@NLO (MG5) combined with PYTHIA8 (P8) or HERWIG++ (H++) and the multiparton
simulations MG5 aMC@NLO+PYTHIA8 MLM and MG5 aMC@NLO+PYTHIA8 FxFx. The ratios of
the various predictions to the measured cross sections are shown at the bottom of each panel
together with the statistical and systematic uncertainties of the measurement. 11
Cross section results The cross sections are compared to the predictions of POWHEG and
MG5 aMC@NLO (MG5) combined with PYTHIA8 (P8) or HERWIG++ (H++) and the multiparton
simulations MG5 aMC@NLO+PYTHIA8 MLM and MG5 aMC@NLO+PYTHIA8 FxFx. The ratios of
the various predictions to the measured cross sections are shown at the bottom of each panel
together with the statistical and systematic uncertainties of the measurement. 11
Cross section results 16 ) [GeV]
t
(t
T
p
]
-1
[pb GeV
)t
(t
T
dp
σ
d
1
−
10
1
10
(13 TeV)
-1
2.3 fb
parton level
+jets
µ
e/
CMS
Data
stat
⊕
Sys
Stat
POWHEG P8
POWHEG H++
MG5 P8
MG5 H++
MG5 P8 [FxFx]
MG5 P8 [MLM]
) [GeV]
t
(t
T
p
0
50
100
150
200
250
300
350
400
450
500
Data
Theory
0.8
1
1.2
1.4
)|
t
|y(t
[pb]
|)
t
d|y(t
σ
d
40
60
80
100
120
140
160
180
200
220
(13 TeV)
-1
2.3 fb
parton level
+jets
µ
e/
CMS
Data
stat
⊕
Sys
Stat
POWHEG P8
POWHEG H++
MG5 P8
MG5 H++
MG5 P8 [FxFx]
MG5 P8 [MLM]
)|
t
|y(t
0
0.2
0.4
0.6
0.8
1
1.2
1.4
1.6
1.8
2
2.2
Data
Theory
0.8
0.9
1
1.1
1.2
) [GeV]
t
m(t
]
-1
[pb GeV
)t
dM(t
σ
d
2
−
10
1
−
10
1
(13 TeV)
-1
2.3 fb
parton level
+jets
µ
e/
CMS
Data
stat
⊕
Sys
Stat
POWHEG P8
POWHEG H++
MG5 P8
MG5 H++
MG5 P8 [FxFx]
MG5 P8 [MLM]
) [GeV]
t
M(t
400
600
800
1000
1200
1400
1600
1800
2000
Data
Theory
0.5
1
1.5
n-jets
[pb]
σ
20
40
60
80
100
120
140
(13 TeV)
-1
2.3 fb
parton level
+jets
µ
e/
CMS
Data
stat
⊕
Sys
Stat
POWHEG P8
POWHEG H++
MG5 P8
MG5 H++
MG5 P8 [FxFx]
MG5 P8 [MLM]
Additional jets
0
1
2
3
4
≥
Data
Theory
1
1.5
Figure 9: Differential cross sections at parton level as a function of pT(tt), |y(tt)|, M(tt), and
cross sections as a function of the number of additional jets compared to the predictions of
POWHEG and MG5 aMC@NLO (MG5) combined with PYTHIA8 (P8) or HERWIG++ (H++) and the
multiparton simulations MG5 aMC@NLO+PYTHIA8 MLM and MG5 aMC@NLO+PYTHIA8 FxFx. 11
Cross section results The ratios of the various predictions to the measured cross sections are shown at the bottom of
each panel together with the statistical and systematic uncertainties of the measurement. ) [GeV]
t
(t
T
p
]
-1
[pb GeV
)t
(t
T
dp
σ
d
1
−
10
1
10
(13 TeV)
-1
2.3 fb
parton level
+jets
µ
e/
CMS
Data
stat
⊕
Sys
Stat
POWHEG P8
POWHEG H++
MG5 P8
MG5 H++
MG5 P8 [FxFx]
MG5 P8 [MLM]
) [GeV]
t
(t
T
p
0
50
100
150
200
250
300
350
400
450
500
Data
Theory
0.8
1
1.2
1.4 )|
t
|y(t
[pb]
|)
t
d|y(t
σ
d
40
60
80
100
120
140
160
180
200
220
(13 TeV)
-1
2.3 fb
parton level
+jets
µ
e/
CMS
Data
stat
⊕
Sys
Stat
POWHEG P8
POWHEG H++
MG5 P8
MG5 H++
MG5 P8 [FxFx]
MG5 P8 [MLM]
)|
t
|y(t
0
0.2
0.4
0.6
0.8
1
1.2
1.4
1.6
1.8
2
2.2
Data
Theory
0.8
0.9
1
1.1
1.2 )|
|y(
n-jets
[pb]
σ
20
40
60
80
100
120
140
(13 TeV)
-1
2.3 fb
parton level
+jets
µ
e/
CMS
Data
stat
⊕
Sys
Stat
POWHEG P8
POWHEG H++
MG5 P8
MG5 H++
MG5 P8 [FxFx]
MG5 P8 [MLM]
Additional jets
0
1
2
3
4
≥
Data
Theory
1
1.5 ) [GeV]
t
m(t
]
-1
[pb GeV
)t
dM(t
σ
d
2
−
10
1
−
10
1
(13 TeV)
-1
2.3 fb
parton level
+jets
µ
e/
CMS
Data
stat
⊕
Sys
Stat
POWHEG P8
POWHEG H++
MG5 P8
MG5 H++
MG5 P8 [FxFx]
MG5 P8 [MLM]
) [GeV]
t
M(t
400
600
800
1000
1200
1400
1600
1800
2000
Data
Theory
0.5
1
1.5 Additional jets Figure 9: Differential cross sections at parton level as a function of pT(tt), |y(tt)|, Figure 9: Differential cross sections at parton level as a function of pT(tt), |y(tt)|, M(tt), and
cross sections as a function of the number of additional jets compared to the predictions of
POWHEG and MG5 aMC@NLO (MG5) combined with PYTHIA8 (P8) or HERWIG++ (H++) and the
multiparton simulations MG5 aMC@NLO+PYTHIA8 MLM and MG5 aMC@NLO+PYTHIA8 FxFx. The ratios of the various predictions to the measured cross sections are shown at the bottom of
each panel together with the statistical and systematic uncertainties of the measurement. 11
Cross section results 17 ) [GeV]
t
(t
T
p
]
-1
[pb GeV
)t
(t
T
dp
σ
d
1
−
10
1
(13 TeV)
-1
2.3 fb
particle level
+jets
µ
e/
CMS
Data
stat
⊕
Sys
Stat
POWHEG P8
POWHEG H++
MG5 P8
MG5 H++
MG5 P8 [FxFx]
MG5 P8 [MLM]
) [GeV]
t
(t
T
p
0
50
100
150
200
250
300
350
400
450
500
Data
Theory
0.8
1
1.2
1.4
)|
t
|y(t
[pb]
|)
t
d|y(t
σ
d
20
40
60
80
100
(13 TeV)
-1
2.3 fb
particle level
+jets
µ
e/
CMS
Data
stat
⊕
Sys
Stat
POWHEG P8
POWHEG H++
MG5 P8
MG5 H++
MG5 P8 [FxFx]
MG5 P8 [MLM]
)|
t
|y(t
0
0.2
0.4
0.6
0.8
1
1.2
1.4
1.6
1.8
2
2.2
Data
Theory
1
1.2
) [GeV]
t
m(t
]
-1
[pb GeV
)t
dM(t
σ
d
2
−
10
1
−
10
1
(13 TeV)
-1
2.3 fb
particle level
+jets
µ
e/
CMS
Data
stat
⊕
Sys
Stat
POWHEG P8
POWHEG H++
MG5 P8
MG5 H++
MG5 P8 [FxFx]
MG5 P8 [MLM]
) [GeV]
t
M(t
400
600
800
1000
1200
1400
1600
1800
2000
Data
Theory
0.8
1
1.2
1.4
n-jets
[pb]
σ
10
20
30
40
50
(13 TeV)
-1
2.3 fb
particle level
+jets
µ
e/
CMS
Data
stat
⊕
Sys
Stat
POWHEG P8
POWHEG H++
MG5 P8
MG5 H++
MG5 P8 [FxFx]
MG5 P8 [MLM]
Additional jets
0
1
2
3
4
≥
Data
Theory
1
1.5
Figure 10: Differential cross sections at particle level as a function of pT(tt), |y(tt)|, M(tt), and
cross sections as a function of the number of additional jets compared to the predictions of
POWHEG and MG5 aMC@NLO (MG5) combined with PYTHIA8 (P8) or HERWIG++ (H++) and the
multiparton simulations MG5 aMC@NLO+PYTHIA8 MLM and MG5 aMC@NLO+PYTHIA8 FxFx. The ratios of the various predictions to the measured cross sections are shown at the bottom of
each panel together with the statistical and systematic uncertainties of the measurement. 11
Cross section results ) [GeV]
t
(t
T
p
]
-1
[pb GeV
)t
(t
T
dp
σ
d
1
−
10
1
(13 TeV)
-1
2.3 fb
particle level
+jets
µ
e/
CMS
Data
stat
⊕
Sys
Stat
POWHEG P8
POWHEG H++
MG5 P8
MG5 H++
MG5 P8 [FxFx]
MG5 P8 [MLM]
) [GeV]
t
(t
T
p
0
50
100
150
200
250
300
350
400
450
500
Data
Theory
0.8
1
1.2
1.4 )|
t
|y(t
[pb]
|)
t
d|y(t
σ
d
20
40
60
80
100
(13 TeV)
-1
2.3 fb
particle level
+jets
µ
e/
CMS
Data
stat
⊕
Sys
Stat
POWHEG P8
POWHEG H++
MG5 P8
MG5 H++
MG5 P8 [FxFx]
MG5 P8 [MLM]
)|
t
|y(t
0
0.2
0.4
0.6
0.8
1
1.2
1.4
1.6
1.8
2
2.2
Data
Theory
1
1.2 )|
|y(
n-jets
[pb]
σ
10
20
30
40
50
(13 TeV)
-1
2.3 fb
particle level
+jets
µ
e/
CMS
Data
stat
⊕
Sys
Stat
POWHEG P8
POWHEG H++
MG5 P8
MG5 H++
MG5 P8 [FxFx]
MG5 P8 [MLM]
Additional jets
0
1
2
3
4
≥
Data
Theory
1
1.5 Figure 10: Differential cross sections at particle level as a function of pT(tt), |y(tt)|, M(tt), and
cross sections as a function of the number of additional jets compared to the predictions of
POWHEG and MG5 aMC@NLO (MG5) combined with PYTHIA8 (P8) or HERWIG++ (H++) and the
multiparton simulations MG5 aMC@NLO+PYTHIA8 MLM and MG5 aMC@NLO+PYTHIA8 FxFx. The ratios of the various predictions to the measured cross sections are shown at the bottom of
each panel together with the statistical and systematic uncertainties of the measurement. 11
Cross section results 18 ) [GeV]
h
(t
T
p
]
-1
[pb GeV
)
h
(t
T
dp
σ
d
2
−
10
1
−
10
1
(13 TeV)
-1
2.3 fb
parton level
+jets
µ
e/
CMS
Data
stat
⊕
Sys
Stat
Powheg P8
MG5 P8
appr. NNLO
appr. NNNLO
NLO+NNLL'
NNLO
) [GeV]
h
(t
T
p
0
100
200
300
400
500
600
700
800
Data
Theory
1
1.5
2
)|
h
|y(t
[pb]
)|
h
d|y(t
σ
d
40
60
80
100
120
140
160
180
200
(13 TeV)
-1
2.3 fb
parton level
+jets
µ
e/
CMS
Data
stat
⊕
Sys
Stat
Powheg P8
MG5 P8
appr. NNLO
appr. 11
Cross section results NNNLO
NNLO
)|
h
|y(t
0
0.5
1
1.5
2
2.5
Data
Theory
0.8
1
1.2
) [GeV]
t
(t
T
p
]
-1
[pb GeV
)t
(t
T
dp
σ
d
1
−
10
1
10
(13 TeV)
-1
2.3 fb
parton level
+jets
µ
e/
CMS
Data
stat
⊕
Sys
Stat
Powheg P8
MG5 P8
NNLO
) [GeV]
t
(t
T
p
0
50
100
150
200
250
300
350
400
450
500
Data
Theory
0.8
1
1.2
1.4
)|
t
|y(t
[pb]
|)
t
d|y(t
σ
d
40
60
80
100
120
140
160
180
200
220
(13 TeV)
-1
2.3 fb
parton level
+jets
µ
e/
CMS
Data
stat
⊕
Sys
Stat
Powheg P8
MG5 P8
NNLO
)|
t
|y(t
0
0.2
0.4
0.6
0.8
1
1.2
1.4
1.6
1.8
2
2.2
Data
Theory
0.8
0.9
1
1.1
1.2
) [GeV]
t
m(t
]
-1
[pb GeV
)t
dM(t
σ
d
2
−
10
1
−
10
1
(13 TeV)
-1
2.3 fb
parton level
+jets
µ
e/
CMS
Data
stat
⊕
Sys
Stat
Powheg P8
MG5 P8
NLO+NNLL'
NNLO
) [GeV]
t
M(t
400
600
800
1000
1200
1400
1600
1800
2000
Data
Theory
0.5
1
1.5
Figure 11: Differential cross sections at parton level as a function of pT(t), |y(t)|, pT(tt), |y(tt)|,
and M(tt) compared to the available predictions of an approximate NNLO calculation [40], an
approximate NNNLO calculation [42, 43], a NLO+NNLL’ calculation [45], and a full NNLO
calculation [46]. For these models uncertainties due to the choices of scales are shown. To im-
prove the visibility the theoretical predictions are horizontally shifted. The ratios of the various
predictions to the measured cross sections are shown at the bottom of each panel together with
the statistical and systematic uncertainties of the measurement. ) [GeV]
h
(t
T
p
]
-1
[pb GeV
)
h
(t
T
dp
σ
d
2
−
10
1
−
10
1
(13 TeV)
-1
2.3 fb
parton level
+jets
µ
e/
CMS
Data
stat
⊕
Sys
Stat
Powheg P8
MG5 P8
appr. NNLO
appr. 11
Cross section results NNNLO
NLO+NNLL'
NNLO
) [GeV]
h
(t
T
p
0
100
200
300
400
500
600
700
800
Data
Theory
1
1.5
2
)|
h
|y(t
[pb]
)|
h
d|y(t
σ
d
40
60
80
100
120
140
160
180
200
(13 TeV)
-1
2.3 fb
parton level
+jets
µ
e/
CMS
Data
stat
⊕
Sys
Stat
Powheg P8
MG5 P8
appr. NNLO
appr. NNNLO
NNLO
)|
h
|y(t
0
0.5
1
1.5
2
2.5
Data
Theory
0.8
1
1.2
) [GeV]
t
(t
T
p
]
-1
[pb GeV
)t
(t
T
dp
σ
d
1
−
10
1
10
(13 TeV)
-1
2.3 fb
parton level
+jets
µ
e/
CMS
Data
stat
⊕
Sys
Stat
Powheg P8
MG5 P8
NNLO
) [GeV]
t
(t
T
p
0
50
100
150
200
250
300
350
400
450
500
Data
Theory
0.8
1
1.2
1.4
)|
t
|y(t
[pb]
|)
t
d|y(t
σ
d
40
60
80
100
120
140
160
180
200
220
(13 TeV)
-1
2.3 fb
parton level
+jets
µ
e/
CMS
Data
stat
⊕
Sys
Stat
Powheg P8
MG5 P8
NNLO
)|
t
|y(t
0
0.2
0.4
0.6
0.8
1
1.2
1.4
1.6
1.8
2
2.2
Data
Theory
0.8
0.9
1
1.1
1.2 )|
h
|y(t
[pb]
)|
h
d|y(t
σ
d
40
60
80
100
120
140
160
180
200
(13 TeV)
-1
2.3 fb
parton level
+jets
µ
e/
CMS
Data
stat
⊕
Sys
Stat
Powheg P8
MG5 P8
appr. NNLO
appr. NNNLO
NNLO
)|
h
|y(t
0
0.5
1
1.5
2
2.5
Data
Theory
0.8
1
1.2 ) [GeV]
h
(t
T
p
]
-1
[pb GeV
)
h
(t
T
dp
σ
d
2
−
10
1
−
10
1
(13 TeV)
-1
2.3 fb
parton level
+jets
µ
e/
CMS
Data
stat
⊕
Sys
Stat
Powheg P8
MG5 P8
appr. NNLO
appr. measurements, respectively. measurements, respectively. measurements, respectively. The precision of the measurement is limited by systematic uncertainties, dominated by jet en-
ergy scale uncertainties on the experimental side and parton shower and hadronization model-
ing uncertainties on the theoretical side. As expected, the theoretical uncertainties are reduced
in the particle-level measurements since these are less dependent on theory-based extrapola-
tions. We evaluate the level of agreement between the measured differential cross sections and the
various theoretical predictions using χ2 tests. In these tests we take into account the full co-
variance matrix obtained from the unfolding procedure for the statistical uncertainty. For each
of the studied systematic uncertainties we assume a full correlation among all bins. No uncer-
tainties in the theoretical predictions are considered for this comparison. However, these un-
certainties are known to be large. Typically, differences between the various models are used
to assess their uncertainties. From the χ2 values and the numbers of degrees of freedom, which
corresponds to the number of bins in the distributions, the p-values are calculated. The results
are shown in Table 2 for the parton-level and in Table 3 for the particle-level measurements. The observed cross sections are slightly lower than expected. However, taking into account the
systematic uncertainties, that are highly correlated among the bins, there is no significant devi-
ation. In general, the measured distributions are in agreement with the predictions of the event
generators with some exceptions in the pT(tt) and M(tt) distributions. The jet multiplicities are
lower than predicted by almost all simulations. The measured pT of the top quarks is slightly
softer than predicted. Such an effect has already been observed in previous measurements [2–
5]. However, the comparison between the HERWIG++ and PYTHIA8 simulations together with
the same matrix-element calculations show the large impact of the parton shower and hadron-
ization modeling. The parton-level results are well described by the matrix-element calcula-
tions. Especially, the soft pT of the top quarks is predicted by the NNLO and NLO+NNLL’
QCD calculation. 11
Cross section results NNNLO
NLO+NNLL'
NNLO
) [GeV]
h
(t
T
p
0
100
200
300
400
500
600
700
800
Data
Theory
1
1.5
2 h
)|
t
|y(t
[pb]
|)
t
d|y(t
σ
d
40
60
80
100
120
140
160
180
200
220
(13 TeV)
-1
2.3 fb
parton level
+jets
µ
e/
CMS
Data
stat
⊕
Sys
Stat
Powheg P8
MG5 P8
NNLO
)|
t
|y(t
0
0.2
0.4
0.6
0.8
1
1.2
1.4
1.6
1.8
2
2.2
Data
Theory
0.8
0.9
1
1.1
1.2 ) [GeV]
t
(t
T
p
]
-1
[pb GeV
)t
(t
T
dp
σ
d
1
−
10
1
10
(13 TeV)
-1
2.3 fb
parton level
+jets
µ
e/
CMS
Data
stat
⊕
Sys
Stat
Powheg P8
MG5 P8
NNLO
) [GeV]
t
(t
T
p
0
50
100
150
200
250
300
350
400
450
500
Data
Theory
0.8
1
1.2
1.4 Data
Theory T
) [GeV]
t
m(t
]
-1
[pb GeV
)t
dM(t
σ
d
2
−
10
1
−
10
1
(13 TeV)
-1
2.3 fb
parton level
+jets
µ
e/
CMS
Data
stat
⊕
Sys
Stat
Powheg P8
MG5 P8
NLO+NNLL'
NNLO
) [GeV]
t
M(t
400
600
800
1000
1200
1400
1600
1800
2000
Data
Theory
0.5
1
1.5 ) [GeV]
t
m(t
]
-1
[pb GeV
)t
dM(t
σ
d
2
−
10
1
−
10
1
(13 TeV)
-1
2.3 fb
parton level
+jets
µ
e/
CMS
Data
stat
⊕
Sys
Stat
Powheg P8
MG5 P8
NLO+NNLL'
NNLO
) [GeV]
t
M(t
400
600
800
1000
1200
1400
1600
1800
2000
Data
Theory
0.5
1
1.5 ) [GeV]
t
M(t Figure 11: Differential cross sections at parton level as a function of pT(t), |y(t)|, pT(tt), |y(tt)|,
and M(tt) compared to the available predictions of an approximate NNLO calculation [40], an
approximate NNNLO calculation [42, 43], a NLO+NNLL’ calculation [45], and a full NNLO
calculation [46]. For these models uncertainties due to the choices of scales are shown. To im-
prove the visibility the theoretical predictions are horizontally shifted. The ratios of the various
predictions to the measured cross sections are shown at the bottom of each panel together with
the statistical and systematic uncertainties of the measurement. 19 12
Summary gen
]
-1
[pb GeV
)
h
(t
T
dp
σ
d
3
−
10
2
−
10
1
−
10
1
(13 TeV)
-1
2.3 fb
parton level
+jets
µ
e/
CMS
0 additional jets
Data
stat
⊕
Sys
Stat
POWHEG P8
POWHEG H++
MG5 P8
MG5 H++
MG5 P8 [FxFx]
MG5 P8 [MLM]
) [GeV]
h
(t
T
p
0
100
200
300
400
500
600
700
800
Data
Theory
0.8
1
1.2
1.4 gen
]
-1
[pb GeV
)
h
(t
T
dp
σ
d
2
−
10
1
−
10
1
(13 TeV)
-1
2.3 fb
parton level
+jets
µ
e/
CMS
1 additional jet
Data
stat
⊕
Sys
Stat
POWHEG P8
POWHEG H++
MG5 P8
MG5 H++
MG5 P8 [FxFx]
MG5 P8 [MLM]
) [GeV]
h
(t
T
p
0
100
200
300
400
500
600
700
800
Data
Theory
0.8
1
1.2
1.4 ) [GeV]
h
(t
T
p
00
800 ) [GeV]
h
(t
T
p
gen
]
-1
[pb GeV
)
h
(t
T
dp
σ
d
3
−
10
2
−
10
1
−
10
(13 TeV)
-1
2.3 fb
parton level
+jets
µ
e/
CMS
3 additional jets
≥
Data
stat
⊕
Sys
Stat
POWHEG P8
POWHEG H++
MG5 P8
MG5 H++
MG5 P8 [FxFx]
MG5 P8 [MLM]
) [GeV]
h
(t
T
p
0
100
200
300
400
500
600
700
800
Data
Theory
1
1.5 gen
]
-1
[pb GeV
)
h
(t
T
dp
σ
d
3
−
10
2
−
10
1
−
10
1
(13 TeV)
-1
2.3 fb
parton level
+jets
µ
e/
CMS
2 additional jets
Data
stat
⊕
Sys
Stat
POWHEG P8
POWHEG H++
MG5 P8
MG5 H++
MG5 P8 [FxFx]
MG5 P8 [MLM]
) [GeV]
h
(t
T
p
0
100
200
300
400
500
600
700
800
Data
Theory
0.8
1
1.2
1.4 Data
Theory h
T
gen
]
-1
[pb GeV
)t
(t
T
dp
σ
d
2
−
10
1
−
10
1
(13 TeV)
-1
2.3 fb
parton level
+jets
µ
e/
CMS
1 additional jet
Data
stat
⊕
Sys
Stat
POWHEG P8
POWHEG H++
MG5 P8
MG5 H++
MG5 P8 [FxFx]
MG5 P8 [MLM]
) [GeV]
t
(t
T
p
0
50
100
150
200
250
300
350
400
450
500
Data
Theory
1
1.5 gen
]
-1
[pb GeV
)t
(t
T
dp
σ
d
4
−
10
3
−
10
2
−
10
1
−
10
1
10
(13 TeV)
-1
2.3 fb
parton level
+jets
µ
e/
CMS
0 additional jets
Data
stat
⊕
Sys
Stat
POWHEG P8
POWHEG H++
MG5 P8
MG5 H++
MG5 P8 [FxFx]
MG5 P8 [MLM]
) [GeV]
t
(t
T
p
0
50
100
150
200
250
300
350
400
450
500
Data
Theory
0
10
20 T
gen
]
-1
[pb GeV
)t
(t
T
dp
σ
d
2
−
10
1
−
10
(13 TeV)
-1
2.3 fb
parton level
+jets
µ
e/
CMS
3 additional jets
≥
Data
stat
⊕
Sys
Stat
POWHEG P8
POWHEG H++
MG5 P8
MG5 H++
MG5 P8 [FxFx]
MG5 P8 [MLM]
) [GeV]
t
(t
T
p
0
50
100
150
200
250
300
350
400
450
500
Data
Theory
0.5
1
1.5 Figure 12: Differential cross sections at parton level as a function of pT(th) (upper two rows)
and pT(tt) (lower two rows) in bins of the number of additional jets. 12
Summary Measurements of the differential and double-differential cross sections for tt production in
proton-proton collisions at 13 TeV have been presented. The data correspond to an integrated
luminosity of 2.3 fb−1 recorded by the CMS experiment. The tt production cross section is
measured in the lepton+jets channel as a function of transverse momentum pT and rapidity |y|
of the top quarks; pT, |y|, and invariant mass of the tt system; and the number of additional
jets. The measurement at parton level is dominated by the uncertainties in the parton shower
and hadronization modeling. The dependence on these theoretical models is reduced for the
particle-level measurement, for which the experimental uncertainties of jet energy calibration
and b tagging efficiency are dominant. The results are compared to several standard model predictions that use different methods and
approximations for their calculations. In general, the measured cross sections are slightly lower
than predicted, but within the uncertainty compatible with the expectation. The measured dis-
tributions are in agreement with the predictions of the event generators with some exceptions
in the pT(tt) and M(tt) distributions. The number of additional jets is lower and the measured
pT of the top quarks is slightly softer than predicted by most of the event generators. A softer
pT of the top quarks has already been observed in previous measurements and is predicted by
the NNLO and the NLO+NNLL’ QCD calculation. 12
Summary 20 12
Summary gen
]
-1
[pb GeV
)
h
(t
T
dp
σ
d
3
−
10
2
−
10
1
−
10
1
(13 TeV)
-1
2.3 fb
parton level
+jets
µ
e/
CMS
0 additional jets
Data
stat
⊕
Sys
Stat
POWHEG P8
POWHEG H++
MG5 P8
MG5 H++
MG5 P8 [FxFx]
MG5 P8 [MLM]
) [GeV]
h
(t
T
p
0
100
200
300
400
500
600
700
800
Data
Theory
0.8
1
1.2
1.4
gen
]
-1
[pb GeV
)
h
(t
T
dp
σ
d
2
−
10
1
−
10
1
(13 TeV)
-1
2.3 fb
parton level
+jets
µ
e/
CMS
1 additional jet
Data
stat
⊕
Sys
Stat
POWHEG P8
POWHEG H++
MG5 P8
MG5 H++
MG5 P8 [FxFx]
MG5 P8 [MLM]
) [GeV]
h
(t
T
p
0
100
200
300
400
500
600
700
800
Data
Theory
0.8
1
1.2
1.4
gen
]
-1
[pb GeV
)
h
(t
T
dp
σ
d
3
−
10
2
−
10
1
−
10
1
(13 TeV)
-1
2.3 fb
parton level
+jets
µ
e/
CMS
2 additional jets
Data
stat
⊕
Sys
Stat
POWHEG P8
POWHEG H++
MG5 P8
MG5 H++
MG5 P8 [FxFx]
MG5 P8 [MLM]
) [GeV]
h
(t
T
p
0
100
200
300
400
500
600
700
800
Data
Theory
0.8
1
1.2
1.4
gen
]
-1
[pb GeV
)
h
(t
T
dp
σ
d
3
−
10
2
−
10
1
−
10
(13 TeV)
-1
2.3 fb
parton level
+jets
µ
e/
CMS
3 additional jets
≥
Data
stat
⊕
Sys
Stat
POWHEG P8
POWHEG H++
MG5 P8
MG5 H++
MG5 P8 [FxFx]
MG5 P8 [MLM]
) [GeV]
h
(t
T
p
0
100
200
300
400
500
600
700
800
Data
Theory
1
1.5
gen
]
-1
[pb GeV
)t
(t
T
dp
σ
d
4
−
10
3
−
10
2
−
10
1
−
10
1
10
(13 TeV)
-1
2.3 fb
parton level
+jets
µ
e/
CMS
0 additional jets
Data
stat
⊕
Sys
Stat
POWHEG P8
POWHEG H++
MG5 P8
MG5 H++
MG5 P8 [FxFx]
MG5 P8 [MLM]
) [GeV]
t
(t
T
p
0
50
100
150
200
250
300
350
400
450
500
Data
Theory
0
10
20
gen
]
-1
[pb GeV
)t
(t
T
dp
σ
d
2
−
10
1
−
10
1
(13 TeV)
-1
2.3 fb
parton level
+jets
µ
e/
CMS
1 additional jet
Data
stat
⊕
Sys
Stat
POWHEG P8
POWHEG H++
MG5 P8
MG5 H++
MG5 P8 [FxFx]
MG5 P8 [MLM]
) [GeV]
t
(t
T
p
0
50
100
150
200
250
300
350
400
450
500
Data
Theory
1
1.5
gen
]
-1
[pb GeV
)t
(t
T
dp
σ
d
1
−
10
1
(13 TeV)
-1
2.3 fb
parton level
+jets
µ
e/
CMS
2 additional jets
Data
stat
⊕
Sys
Stat
POWHEG P8
POWHEG H++
MG5 P8
MG5 H++
MG5 P8 [FxFx]
MG5 P8 [MLM]
) [GeV]
t
(t
T
p
0
50
100
150
200
250
300
350
400
450
500
Data
Theory
1
1.5
gen
]
-1
[pb GeV
)t
(t
T
dp
σ
d
2
−
10
1
−
10
(13 TeV)
-1
2.3 fb
parton level
+jets
µ
e/
CMS
3 additional jets
≥
Data
stat
⊕
Sys
Stat
POWHEG P8
POWHEG H++
MG5 P8
MG5 H++
MG5 P8 [FxFx]
MG5 P8 [MLM]
) [GeV]
t
(t
T
p
0
50
100
150
200
250
300
350
400
450
500
Data
Theory
0.5
1
1.5
gure 12: Differential cross sections at parton level as a function of pT(th) (upper two r
nd pT(tt) (lower two rows) in bins of the number of additional jets. 12
Summary The measuremen
mpared to the predictions of POWHEG and MG5 aMC@NLO (MG5) combined with PYT
8) or HERWIG++ (H++) and the multiparton simulations MG5 aMC@NLO+PYTHIA8 MLM
G5 aMC@NLO+PYTHIA8 FxFx. The ratios of the predictions to the measured cross section
hown at the bottom of each panel together with the statistical and systematic uncertaint
e measurement. 12
Summary gen
]
-1
[pb GeV
)
h
(t
T
dp
σ
d
3
−
10
2
−
10
1
−
10
(13 TeV)
-1
2.3 fb
particle level
+jets
µ
e/
CMS
1 additional jet
Data
stat
⊕
Sys
Stat
POWHEG P8
POWHEG H++
MG5 P8
MG5 H++
MG5 P8 [FxFx]
MG5 P8 [MLM]
) [GeV]
h
(t
T
p
0
100
200
300
400
500
600
700
800
Data
Theory
0.8
1
1.2 gen
]
-1
[pb GeV
)
h
(t
T
dp
σ
d
3
−
10
2
−
10
1
−
10
1
(13 TeV)
-1
2.3 fb
particle level
+jets
µ
e/
CMS
0 additional jets
Data
stat
⊕
Sys
Stat
POWHEG P8
POWHEG H++
MG5 P8
MG5 H++
MG5 P8 [FxFx]
MG5 P8 [MLM]
) [GeV]
h
(t
T
p
0
100
200
300
400
500
600
700
800
Data
Theory
0.8
1
1.2
1.4 ) [GeV]
h
(t
T
p
700
800 ) [GeV]
h
(t
T
p
gen
]
-1
[pb GeV
)
h
(t
T
dp
σ
d
3
−
10
2
−
10
1
−
10
(13 TeV)
-1
2.3 fb
particle level
+jets
µ
e/
CMS
3 additional jets
≥
Data
stat
⊕
Sys
Stat
POWHEG P8
POWHEG H++
MG5 P8
MG5 H++
MG5 P8 [FxFx]
MG5 P8 [MLM]
) [GeV]
h
(t
T
p
0
100
200
300
400
500
600
700
800
Data
Theory
1
1.5 gen
]
-1
[pb GeV
)
h
(t
T
dp
σ
d
3
−
10
2
−
10
1
−
10
(13 TeV)
-1
2.3 fb
particle level
+jets
µ
e/
CMS
2 additional jets
Data
stat
⊕
Sys
Stat
POWHEG P8
POWHEG H++
MG5 P8
MG5 H++
MG5 P8 [FxFx]
MG5 P8 [MLM]
) [GeV]
h
(t
T
p
0
100
200
300
400
500
600
700
800
Data
Theory
0.8
1
1.2
1.4 ) [GeV]
h
(t
T
p
700
800 h
T
gen
]
-1
[pb GeV
)t
(t
T
dp
σ
d
2
−
10
1
−
10
1
(13 TeV)
-1
2.3 fb
particle level
+jets
µ
e/
CMS
1 additional jet
Data
stat
⊕
Sys
Stat
POWHEG P8
POWHEG H++
MG5 P8
MG5 H++
MG5 P8 [FxFx]
MG5 P8 [MLM]
) [GeV]
t
(t
T
p
0
50
100
150
200
250
300
350
400
450
500
Data
Theory
0.8
1
1.2 gen
]
-1
[pb GeV
)t
(t
T
dp
σ
d
4
−
10
3
−
10
2
−
10
1
−
10
1
(13 TeV)
-1
2.3 fb
particle level
+jets
µ
e/
CMS
0 additional jets
Data
stat
⊕
Sys
Stat
POWHEG P8
POWHEG H++
MG5 P8
MG5 H++
MG5 P8 [FxFx]
MG5 P8 [MLM]
) [GeV]
t
(t
T
p
0
50
100
150
200
250
300
350
400
450
500
Data
Theory
2
4
6 Figure 13: Differential cross sections at particle level as a function of pT(th) (upper two rows)
and pT(tt) (lower two rows) in bins of the number of additional jets. 12
Summary The measurements are
compared to the predictions of POWHEG and MG5 aMC@NLO (MG5) combined with PYTHIA8
(P8) or HERWIG++ (H++) and the multiparton simulations MG5 aMC@NLO+PYTHIA8 MLM and
MG5 aMC@NLO+PYTHIA8 FxFx. The ratios of the predictions to the measured cross sections are
shown at the bottom of each panel together with the statistical and systematic uncertainties of
the measurement. 12
Summary 21 gen
]
-1
[pb GeV
)
h
(t
T
dp
σ
d
3
−
10
2
−
10
1
−
10
1
(13 TeV)
-1
2.3 fb
particle level
+jets
µ
e/
CMS
0 additional jets
Data
stat
⊕
Sys
Stat
POWHEG P8
POWHEG H++
MG5 P8
MG5 H++
MG5 P8 [FxFx]
MG5 P8 [MLM]
) [GeV]
h
(t
T
p
0
100
200
300
400
500
600
700
800
Data
Theory
0.8
1
1.2
1.4
gen
]
-1
[pb GeV
)
h
(t
T
dp
σ
d
3
−
10
2
−
10
1
−
10
(13 TeV)
-1
2.3 fb
particle level
+jets
µ
e/
CMS
1 additional jet
Data
stat
⊕
Sys
Stat
POWHEG P8
POWHEG H++
MG5 P8
MG5 H++
MG5 P8 [FxFx]
MG5 P8 [MLM]
) [GeV]
h
(t
T
p
0
100
200
300
400
500
600
700
800
Data
Theory
0.8
1
1.2
gen
]
-1
[pb GeV
)
h
(t
T
dp
σ
d
3
−
10
2
−
10
1
−
10
(13 TeV)
-1
2.3 fb
particle level
+jets
µ
e/
CMS
2 additional jets
Data
stat
⊕
Sys
Stat
POWHEG P8
POWHEG H++
MG5 P8
MG5 H++
MG5 P8 [FxFx]
MG5 P8 [MLM]
) [GeV]
h
(t
T
p
0
100
200
300
400
500
600
700
800
Data
Theory
0.8
1
1.2
1.4
gen
]
-1
[pb GeV
)
h
(t
T
dp
σ
d
3
−
10
2
−
10
1
−
10
(13 TeV)
-1
2.3 fb
particle level
+jets
µ
e/
CMS
3 additional jets
≥
Data
stat
⊕
Sys
Stat
POWHEG P8
POWHEG H++
MG5 P8
MG5 H++
MG5 P8 [FxFx]
MG5 P8 [MLM]
) [GeV]
h
(t
T
p
0
100
200
300
400
500
600
700
800
Data
Theory
1
1.5
gen
]
-1
[pb GeV
)t
(t
T
dp
σ
d
4
−
10
3
−
10
2
−
10
1
−
10
1
(13 TeV)
-1
2.3 fb
particle level
+jets
µ
e/
CMS
0 additional jets
Data
stat
⊕
Sys
Stat
POWHEG P8
POWHEG H++
MG5 P8
MG5 H++
MG5 P8 [FxFx]
MG5 P8 [MLM]
) [GeV]
t
(t
T
p
0
50
100
150
200
250
300
350
400
450
500
Data
Theory
2
4
6
gen
]
-1
[pb GeV
)t
(t
T
dp
σ
d
2
−
10
1
−
10
1
(13 TeV)
-1
2.3 fb
particle level
+jets
µ
e/
CMS
1 additional jet
Data
stat
⊕
Sys
Stat
POWHEG P8
POWHEG H++
MG5 P8
MG5 H++
MG5 P8 [FxFx]
MG5 P8 [MLM]
) [GeV]
t
(t
T
p
0
50
100
150
200
250
300
350
400
450
500
Data
Theory
0.8
1
1.2
gen
]
-1
[pb GeV
)t
(t
T
dp
σ
d
2
−
10
1
−
10
(13 TeV)
-1
2.3 fb
particle level
+jets
µ
e/
CMS
2 additional jets
Data
stat
⊕
Sys
Stat
POWHEG P8
POWHEG H++
MG5 P8
MG5 H++
MG5 P8 [FxFx]
MG5 P8 [MLM]
) [GeV]
t
(t
T
p
0
50
100
150
200
250
300
350
400
450
500
Data
Theory
0.8
1
1.2
1.4
1.6
gen
]
-1
[pb GeV
)t
(t
T
dp
σ
d
2
−
10
1
−
10
(13 TeV)
-1
2.3 fb
particle level
+jets
µ
e/
CMS
3 additional jets
≥
Data
stat
⊕
Sys
Stat
POWHEG P8
POWHEG H++
MG5 P8
MG5 H++
MG5 P8 [FxFx]
MG5 P8 [MLM]
) [GeV]
t
(t
T
p
0
50
100
150
200
250
300
350
400
450
500
Data
Theory
1
1.5
gure 13: Differential cross sections at particle level as a function of pT(th) (upper two r
nd pT(tt) (lower two rows) in bins of the number of additional jets. 12
Summary The measuremen
mpared to the predictions of POWHEG and MG5 aMC@NLO (MG5) combined with PYT
8) or HERWIG++ (H++) and the multiparton simulations MG5 aMC@NLO+PYTHIA8 MLM
G5 aMC@NLO+PYTHIA8 FxFx. The ratios of the predictions to the measured cross section
own at the bottom of each panel together with the statistical and systematic uncertaint
e measurement. 12
Summary gen
]
-1
[pb GeV
)
h
(t
T
)| dp
h
d|y(t
σ
2
d
2
−
10
1
−
10
1
(13 TeV)
-1
2.3 fb
parton level
+jets
µ
e/
CMS
)| < 0.5
h
0 < |y(t
Data
stat
⊕
Sys
Stat
POWHEG P8
POWHEG H++
MG5 P8
MG5 H++
MG5 P8 [FxFx]
MG5 P8 [MLM]
) [GeV]
h
(t
T
p
0
100
200
300
400
500
600
700
800
Data
Theory
0.8
1
1.2 gen
]
-1
[pb GeV
)
h
(t
T
)| dp
h
d|y(t
σ
2
d
2
−
10
1
−
10
1
(13 TeV)
-1
2.3 fb
parton level
+jets
µ
e/
CMS
)| < 1
h
0.5 < |y(t
Data
stat
⊕
Sys
Stat
POWHEG P8
POWHEG H++
MG5 P8
MG5 H++
MG5 P8 [FxFx]
MG5 P8 [MLM]
) [GeV]
h
(t
T
p
0
100
200
300
400
500
600
700
800
Data
Theory
0.8
1
1.2
1.4 ) [GeV]
h
(t
T
p
gen
]
-1
[pb GeV
)
h
(t
T
)| dp
h
d|y(t
σ
2
d
3
−
10
2
−
10
1
−
10
1
(13 TeV)
-1
2.3 fb
parton level
+jets
µ
e/
CMS
)| < 2.5
h
1.5 < |y(t
Data
stat
⊕
Sys
Stat
POWHEG P8
POWHEG H++
MG5 P8
MG5 H++
MG5 P8 [FxFx]
MG5 P8 [MLM]
) [GeV]
h
(t
T
p
0
100
200
300
400
500
600
700
800
Data
Theory
1
1.5 gen
]
-1
[pb GeV
)
h
(t
T
)| dp
h
d|y(t
σ
2
d
2
−
10
1
−
10
1
(13 TeV)
-1
2.3 fb
parton level
+jets
µ
e/
CMS
)| < 1.5
h
1 < |y(t
Data
stat
⊕
Sys
Stat
POWHEG P8
POWHEG H++
MG5 P8
MG5 H++
MG5 P8 [FxFx]
MG5 P8 [MLM]
) [GeV]
h
(t
T
p
0
100
200
300
400
500
600
700
800
Data
Theory
1
1.5 gen
]
-1
[pb GeV
)|
t
) d|y(t
t
dM(t
σ
2
d
0.1
0.15
0.2
0.25
0.3
0.35
0.4
0.45
0.5
(13 TeV)
-1
2.3 fb
parton level
+jets
µ
e/
CMS
) < 625 GeV
t
450 < M(t
Data
stat
⊕
Sys
Stat
POWHEG P8
POWHEG H++
MG5 P8
MG5 H++
MG5 P8 [FxFx]
MG5 P8 [MLM]
)|
t
|y(t
0
0.2
0.4
0.6
0.8
1
1.2
1.4
1.6
1.8
2
2.2
Data
Theory
0.8
1
1.2 gen
]
-1
[pb GeV
)|
t
) d|y(t
t
dM(t
σ
2
d
0.2
0.3
0.4
0.5
0.6
(13 TeV)
-1
2.3 fb
parton level
+jets
µ
e/
CMS
) < 450 GeV
t
300 < M(t
Data
stat
⊕
Sys
Stat
POWHEG P8
POWHEG H++
MG5 P8
MG5 H++
MG5 P8 [FxFx]
MG5 P8 [MLM]
)|
t
|y(t
0
0.2
0.4
0.6
0.8
1
1.2
1.4
1.6
1.8
2
2.2
Data
Theory
0.8
1
1.2 gen
]
-1
[pb GeV
)|
t
) d|y(t
t
dM(t
σ
2
d
0.2
0.3
0.4
0.5
0.6
(13 TeV)
-1
2.3 fb
parton level
+jets
µ
e/
CMS
) < 450 GeV
t
300 < M(t
Data
stat
⊕
Sys
Stat
POWHEG P8
POWHEG H++
MG5 P8
MG5 H++
MG5 P8 [FxFx]
MG5 P8 [MLM]
)|
t
|y(t
0
0.2
0.4
0.6
0.8
1
1.2
1.4
1.6
1.8
2
2.2
Data
Theory
0.8
1
1.2
gen
]
-1
[pb GeV
)|
t
) d|y(t
t
dM(t
σ
2
d
0.02
0.04
0.06
0.08
0.1
0.12
0.14
0.16
0.18
(13 TeV)
-1
2.3 fb
parton level
+jets
µ
e/
CMS
) < 850 GeV
t
625 < M(t
Data
stat
⊕
Sys
Stat
POWHEG P8
POWHEG H++
MG5 P8
MG5 H++
MG5 P8 [FxFx]
MG5 P8 [MLM]
)|
t
|y(t
0
0.2
0.4
0.6
0.8
1
1.2
1.4
1.6
1.8
2
2.2
Data
Theory
0.8
1
1.2
bl
d ff
l ]
-1
[pb GeV
)|
t
) d|y(t
t
dM(t
σ
2
d )|
t
|y(t
gen
]
-1
[pb GeV
)|
t
) d|y(t
t
dM(t
σ
2
d
0.002
0.004
0.006
0.008
0.01
0.012
0.014
(13 TeV)
-1
2.3 fb
parton level
+jets
µ
e/
CMS
) < 2000 GeV
t
850 < M(t
Data
stat
⊕
Sys
Stat
POWHEG P8
POWHEG H++
MG5 P8
MG5 H++
MG5 P8 [FxFx]
MG5 P8 [MLM]
)|
t
|y(t
0
0.2
0.4
0.6
0.8
1
1.2
1.4
1.6
1.8
2
2.2
Data
Theory
0.8
1
1.2
1.4 Figure 14: Double-differential cross sections at parton level as a function of |y(th)| vs. 12
Summary The measurements are
compared to the predictions of POWHEG and MG5 aMC@NLO (MG5) combined with PYTHIA8
(P8) or HERWIG++ (H++) and the multiparton simulations MG5 aMC@NLO+PYTHIA8 MLM and
MG5 aMC@NLO+PYTHIA8 FxFx. The ratios of the predictions to the measured cross sections are
shown at the bottom of each panel together with the statistical and systematic uncertainties of
the measurement. 12
Summary 22 12
Summary gen
]
-1
[pb GeV
)
h
(t
T
)| dp
h
d|y(t
σ
2
d
2
−
10
1
−
10
1
(13 TeV)
-1
2.3 fb
parton level
+jets
µ
e/
CMS
)| < 0.5
h
0 < |y(t
Data
stat
⊕
Sys
Stat
POWHEG P8
POWHEG H++
MG5 P8
MG5 H++
MG5 P8 [FxFx]
MG5 P8 [MLM]
) [GeV]
h
(t
T
p
0
100
200
300
400
500
600
700
800
Data
Theory
0.8
1
1.2
gen
]
-1
[pb GeV
)
h
(t
T
)| dp
h
d|y(t
σ
2
d
2
−
10
1
−
10
1
(13 TeV)
-1
2.3 fb
parton level
+jets
µ
e/
CMS
)| < 1
h
0.5 < |y(t
Data
stat
⊕
Sys
Stat
POWHEG P8
POWHEG H++
MG5 P8
MG5 H++
MG5 P8 [FxFx]
MG5 P8 [MLM]
) [GeV]
h
(t
T
p
0
100
200
300
400
500
600
700
800
Data
Theory
0.8
1
1.2
1.4
gen
]
-1
[pb GeV
)
h
(t
T
)| dp
h
d|y(t
σ
2
d
2
−
10
1
−
10
1
(13 TeV)
-1
2.3 fb
parton level
+jets
µ
e/
CMS
)| < 1.5
h
1 < |y(t
Data
stat
⊕
Sys
Stat
POWHEG P8
POWHEG H++
MG5 P8
MG5 H++
MG5 P8 [FxFx]
MG5 P8 [MLM]
) [GeV]
h
(t
T
p
0
100
200
300
400
500
600
700
800
Data
Theory
1
1.5
gen
]
-1
[pb GeV
)
h
(t
T
)| dp
h
d|y(t
σ
2
d
3
−
10
2
−
10
1
−
10
1
(13 TeV)
-1
2.3 fb
parton level
+jets
µ
e/
CMS
)| < 2.5
h
1.5 < |y(t
Data
stat
⊕
Sys
Stat
POWHEG P8
POWHEG H++
MG5 P8
MG5 H++
MG5 P8 [FxFx]
MG5 P8 [MLM]
) [GeV]
h
(t
T
p
0
100
200
300
400
500
600
700
800
Data
Theory
1
1.5
gen
]
-1
[pb GeV
)|
t
) d|y(t
t
dM(t
σ
2
d
0.2
0.3
0.4
0.5
0.6
(13 TeV)
-1
2.3 fb
parton level
+jets
µ
e/
CMS
) < 450 GeV
t
300 < M(t
Data
stat
⊕
Sys
Stat
POWHEG P8
POWHEG H++
MG5 P8
MG5 H++
MG5 P8 [FxFx]
MG5 P8 [MLM]
)|
t
|y(t
0
0.2
0.4
0.6
0.8
1
1.2
1.4
1.6
1.8
2
2.2
Data
Theory
0.8
1
1.2
gen
]
-1
[pb GeV
)|
t
) d|y(t
t
dM(t
σ
2
d
0.1
0.15
0.2
0.25
0.3
0.35
0.4
0.45
0.5
(13 TeV)
-1
2.3 fb
parton level
+jets
µ
e/
CMS
) < 625 GeV
t
450 < M(t
Data
stat
⊕
Sys
Stat
POWHEG P8
POWHEG H++
MG5 P8
MG5 H++
MG5 P8 [FxFx]
MG5 P8 [MLM]
)|
t
|y(t
0
0.2
0.4
0.6
0.8
1
1.2
1.4
1.6
1.8
2
2.2
Data
Theory
0.8
1
1.2
gen
]
-1
[pb GeV
)|
t
) d|y(t
t
dM(t
σ
2
d
0.02
0.04
0.06
0.08
0.1
0.12
0.14
0.16
0.18
(13 TeV)
-1
2.3 fb
parton level
+jets
µ
e/
CMS
) < 850 GeV
t
625 < M(t
Data
stat
⊕
Sys
Stat
POWHEG P8
POWHEG H++
MG5 P8
MG5 H++
MG5 P8 [FxFx]
MG5 P8 [MLM]
)|
t
|y(t
0
0.2
0.4
0.6
0.8
1
1.2
1.4
1.6
1.8
2
2.2
Data
Theory
0.8
1
1.2
gen
]
-1
[pb GeV
)|
t
) d|y(t
t
dM(t
σ
2
d
0.002
0.004
0.006
0.008
0.01
0.012
0.014
(13 TeV)
-1
2.3 fb
parton level
+jets
µ
e/
CMS
) < 2000 GeV
t
850 < M(t
Data
stat
⊕
Sys
Stat
POWHEG P8
POWHEG H++
MG5 P8
MG5 H++
MG5 P8 [FxFx]
MG5 P8 [MLM]
)|
t
|y(t
0
0.2
0.4
0.6
0.8
1
1.2
1.4
1.6
1.8
2
2.2
Data
Theory
0.8
1
1.2
1.4
gure 14: Double-differential cross sections at parton level as a function of |y(th)| vs. 12
Summary p
upper two rows) and M(tt) vs. |y(tt)| (lower two rows). The measurements are
ared to the predictions of POWHEG and MG5 aMC@NLO (MG5) combined with PYTHIA
r HERWIG++ (H++) and the multiparton simulations MG5 aMC@NLO+PYTHIA8 MLM
G5 aMC@NLO+PYTHIA8 FxFx. The ratios of the predictions to the measured cross sec
re shown at the bottom of each panel together with the statistical and systematic uncerta
f the measurement. 12
Summary 23 gen
]
-2
[pb GeV
)t
(t
T
) dp
t
dM(t
σ
2
d
4
−
10
3
−
10
2
−
10
(13 TeV)
-1
2.3 fb
parton level
+jets
µ
e/
CMS
) < 35 GeV
t
(t
T
0 < p
Data
stat
⊕
Sys
Stat
POWHEG P8
POWHEG H++
MG5 P8
MG5 H++
MG5 P8 [FxFx]
MG5 P8 [MLM]
) [GeV]
t
M(t
400
600
800
1000
1200
1400
1600
1800
2000
Data
Theory
0.5
1
1.5
gen
]
-2
[pb GeV
)t
(t
T
) dp
t
dM(t
σ
2
d
4
−
10
3
−
10
2
−
10
(13 TeV)
-1
2.3 fb
parton level
+jets
µ
e/
CMS
) < 80 GeV
t
(t
T
35 < p
Data
stat
⊕
Sys
Stat
POWHEG P8
POWHEG H++
MG5 P8
MG5 H++
MG5 P8 [FxFx]
MG5 P8 [MLM]
) [GeV]
t
M(t
400
600
800
1000
1200
1400
1600
1800
2000
Data
Theory
0.5
1
1.5
gen
]
-2
[pb GeV
)t
(t
T
) dp
t
dM(t
σ
2
d
5
−
10
4
−
10
3
−
10
2
−
10
(13 TeV)
-1
2.3 fb
parton level
+jets
µ
e/
CMS
) < 140 GeV
t
(t
T
80 < p
Data
stat
⊕
Sys
Stat
POWHEG P8
POWHEG H++
MG5 P8
MG5 H++
MG5 P8 [FxFx]
MG5 P8 [MLM]
) [GeV]
t
M(t
400
600
800
1000
1200
1400
1600
1800
2000
Data
Theory
1
1.5
gen
]
-2
[pb GeV
)t
(t
T
) dp
t
dM(t
σ
2
d
5
−
10
4
−
10
3
−
10
(13 TeV)
-1
2.3 fb
parton level
+jets
µ
e/
CMS
) < 500 GeV
t
(t
T
140 < p
Data
stat
⊕
Sys
Stat
POWHEG P8
POWHEG H++
MG5 P8
MG5 H++
MG5 P8 [FxFx]
MG5 P8 [MLM]
) [GeV]
t
M(t
400
600
800
1000
1200
1400
1600
1800
2000
Data
Theory
1
1.5
2 gen
]
-2
[pb GeV
)t
(t
T
) dp
t
dM(t
σ
2
d
4
−
10
3
−
10
2
−
10
(13 TeV)
-1
2.3 fb
parton level
+jets
µ
e/
CMS
) < 35 GeV
t
(t
T
0 < p
Data
stat
⊕
Sys
Stat
POWHEG P8
POWHEG H++
MG5 P8
MG5 H++
MG5 P8 [FxFx]
MG5 P8 [MLM]
) [GeV]
t
M(t
400
600
800
1000
1200
1400
1600
1800
2000
Data
Theory
0.5
1
1.5
gen
]
-2
[pb GeV
)t
(t
T
) dp
t
dM(t
σ
2
d
4
−
10
3
−
10
2
−
10
(13 TeV)
-1
2.3 fb
parton level
+jets
µ
e/
CMS
) < 80 GeV
t
(t
T
35 < p
Data
stat
⊕
Sys
Stat
POWHEG P8
POWHEG H++
MG5 P8
MG5 H++
MG5 P8 [FxFx]
MG5 P8 [MLM]
) [GeV]
t
M(t
400
600
800
1000
1200
1400
1600
1800
2000
Data
Theory
0.5
1
1.5
gen
]
-2
[pb GeV
)t
(t
T
) dp
t
dM(t
σ
2
d
5
−
10
4
−
10
3
−
10
2
−
10
(13 TeV)
-1
2.3 fb
parton level
+jets
µ
e/
CMS
) < 140 GeV
t
(t
T
80 < p
Data
stat
⊕
Sys
Stat
POWHEG P8
POWHEG H++
MG5 P8
MG5 H++
MG5 P8 [FxFx]
MG5 P8 [MLM]
) [GeV]
t
M(t
400
600
800
1000
1200
1400
1600
1800
2000
Data
Theory
1
1.5
gen
]
-2
[pb GeV
)t
(t
T
) dp
t
dM(t
σ
2
d
5
−
10
4
−
10
3
−
10
(13 TeV)
-1
2.3 fb
parton level
+jets
µ
e/
CMS
) < 500 GeV
t
(t
T
140 < p
Data
stat
⊕
Sys
Stat
POWHEG P8
POWHEG H++
MG5 P8
MG5 H++
MG5 P8 [FxFx]
MG5 P8 [MLM]
) [GeV]
t
M(t
400
600
800
1000
1200
1400
1600
1800
2000
Data
Theory
1
1.5
2
Figure 15:
Double-differential cross section at parton level as a function of pT(tt) vs. 12
Summary pT(th)
(upper two rows) and M(tt) vs. |y(tt)| (lower two rows). The measurements are com-
pared to the predictions of POWHEG and MG5 aMC@NLO (MG5) combined with PYTHIA8 (P8)
or HERWIG++ (H++) and the multiparton simulations MG5 aMC@NLO+PYTHIA8 MLM and
MG5 aMC@NLO+PYTHIA8 FxFx. The ratios of the predictions to the measured cross sections
are shown at the bottom of each panel together with the statistical and systematic uncertainties
of the measurement. 12
Summary gen
]
-1
[pb GeV
)
h
(t
T
)| dp
h
d|y(t
σ
2
d
2
−
10
1
−
10
1
(13 TeV)
-1
2.3 fb
particle level
+jets
µ
e/
CMS
)| < 0.5
h
0 < |y(t
Data
stat
⊕
Sys
Stat
POWHEG P8
POWHEG H++
MG5 P8
MG5 H++
MG5 P8 [FxFx]
MG5 P8 [MLM]
) [GeV]
h
(t
T
p
0
100
200
300
400
500
600
700
800
Data
Theory
0.8
1
1.2 gen
]
-1
[pb GeV
)
h
(t
T
)| dp
h
d|y(t
σ
2
d
2
−
10
1
−
10
1
(13 TeV)
-1
2.3 fb
particle level
+jets
µ
e/
CMS
)| < 1
h
0.5 < |y(t
Data
stat
⊕
Sys
Stat
POWHEG P8
POWHEG H++
MG5 P8
MG5 H++
MG5 P8 [FxFx]
MG5 P8 [MLM]
) [GeV]
h
(t
T
p
0
100
200
300
400
500
600
700
800
Data
Theory
0.8
1
1.2
1.4 ) [
]
h
(
T
p
gen
]
-1
[pb GeV
)
h
(t
T
)| dp
h
d|y(t
σ
2
d
3
−
10
2
−
10
1
−
10
(13 TeV)
-1
2.3 fb
particle level
+jets
µ
e/
CMS
)| < 2.5
h
1.5 < |y(t
Data
stat
⊕
Sys
Stat
POWHEG P8
POWHEG H++
MG5 P8
MG5 H++
MG5 P8 [FxFx]
MG5 P8 [MLM]
) [GeV]
h
(t
T
p
0
100
200
300
400
500
600
700
800
Data
Theory
0.8
1
1.2
1.4
1.6 gen
]
-1
[pb GeV
)
h
(t
T
)| dp
h
d|y(t
σ
2
d
3
−
10
2
−
10
1
−
10
1
(13 TeV)
-1
2.3 fb
particle level
+jets
µ
e/
CMS
)| < 1.5
h
1 < |y(t
Data
stat
⊕
Sys
Stat
POWHEG P8
POWHEG H++
MG5 P8
MG5 H++
MG5 P8 [FxFx]
MG5 P8 [MLM]
) [GeV]
h
(t
T
p
0
100
200
300
400
500
600
700
800
Data
Theory
0.8
1
1.2
1.4
1.6 gen
]
-1
[pb GeV
)|
t
) d|y(t
t
dM(t
σ
2
d
0.02
0.04
0.06
0.08
0.1
0.12
0.14
0.16
0.18
0.2
0.22
(13 TeV)
-1
2.3 fb
particle level
+jets
µ
e/
CMS
) < 625 GeV
t
450 < M(t
Data
stat
⊕
Sys
Stat
POWHEG P8
POWHEG H++
MG5 P8
MG5 H++
MG5 P8 [FxFx]
MG5 P8 [MLM]
)|
t
|y(t
0
0.2
0.4
0.6
0.8
1
1.2
1.4
1.6
1.8
2
2.2
Data
Theory
0.8
1
1.2
1.4 ]
-1
[pb GeV
)|
t
) d|y(t
t
dM(t
σ
2
d )|
t
|y(t
gen
]
-1
[pb GeV
)|
t
) d|y(t
t
dM(t
σ
2
d
0.002
0.004
0.006
0.008
0.01
0.012
(13 TeV)
-1
2.3 fb
particle level
+jets
µ
e/
CMS
) < 2000 GeV
t
850 < M(t
Data
stat
⊕
Sys
Stat
POWHEG P8
POWHEG H++
MG5 P8
MG5 H++
MG5 P8 [FxFx]
MG5 P8 [MLM]
)|
t
|y(t
0
0.2
0.4
0.6
0.8
1
1.2
1.4
1.6
1.8
2
2.2
Data
Theory
0.8
1
1.2
1.4 Figure 16: Double-differential cross sections at particle level as a function of |y(th)| vs. 12
Summary M(tt). The measurements are compared to the predictions of POWHEG and MG5 aMC@NLO
(MG5) combined with PYTHIA8 (P8) or HERWIG++ (H++) and the multiparton simulations
MG5 aMC@NLO+PYTHIA8 MLM and MG5 aMC@NLO+PYTHIA8 FxFx. The ratios of the pre-
dictions to the measured cross sections are shown at the bottom of each panel together with the
statistical and systematic uncertainties of the measurement. ) [GeV]
t
M(t ) [GeV]
t
M(t ) [GeV]
t
M(t Figure 15:
Double-differential cross section at parton level as a function of pT(tt) vs. M(tt). The measurements are compared to the predictions of POWHEG and MG5 aMC@NLO
(MG5) combined with PYTHIA8 (P8) or HERWIG++ (H++) and the multiparton simulations
MG5 aMC@NLO+PYTHIA8 MLM and MG5 aMC@NLO+PYTHIA8 FxFx. The ratios of the pre-
dictions to the measured cross sections are shown at the bottom of each panel together with the
statistical and systematic uncertainties of the measurement. 12
Summary 24 12
Summary gen
]
-1
[pb GeV
)
h
(t
T
)| dp
h
d|y(t
σ
2
d
2
−
10
1
−
10
1
(13 TeV)
-1
2.3 fb
particle level
+jets
µ
e/
CMS
)| < 0.5
h
0 < |y(t
Data
stat
⊕
Sys
Stat
POWHEG P8
POWHEG H++
MG5 P8
MG5 H++
MG5 P8 [FxFx]
MG5 P8 [MLM]
) [GeV]
h
(t
T
p
0
100
200
300
400
500
600
700
800
Data
Theory
0.8
1
1.2
gen
]
-1
[pb GeV
)
h
(t
T
)| dp
h
d|y(t
σ
2
d
2
−
10
1
−
10
1
(13 TeV)
-1
2.3 fb
particle level
+jets
µ
e/
CMS
)| < 1
h
0.5 < |y(t
Data
stat
⊕
Sys
Stat
POWHEG P8
POWHEG H++
MG5 P8
MG5 H++
MG5 P8 [FxFx]
MG5 P8 [MLM]
) [GeV]
h
(t
T
p
0
100
200
300
400
500
600
700
800
Data
Theory
0.8
1
1.2
1.4
gen
]
-1
[pb GeV
)
h
(t
T
)| dp
h
d|y(t
σ
2
d
3
−
10
2
−
10
1
−
10
1
(13 TeV)
-1
2.3 fb
particle level
+jets
µ
e/
CMS
)| < 1.5
h
1 < |y(t
Data
stat
⊕
Sys
Stat
POWHEG P8
POWHEG H++
MG5 P8
MG5 H++
MG5 P8 [FxFx]
MG5 P8 [MLM]
) [GeV]
h
(t
T
p
0
100
200
300
400
500
600
700
800
Data
Theory
0.8
1
1.2
1.4
1.6
gen
]
-1
[pb GeV
)
h
(t
T
)| dp
h
d|y(t
σ
2
d
3
−
10
2
−
10
1
−
10
(13 TeV)
-1
2.3 fb
particle level
+jets
µ
e/
CMS
)| < 2.5
h
1.5 < |y(t
Data
stat
⊕
Sys
Stat
POWHEG P8
POWHEG H++
MG5 P8
MG5 H++
MG5 P8 [FxFx]
MG5 P8 [MLM]
) [GeV]
h
(t
T
p
0
100
200
300
400
500
600
700
800
Data
Theory
0.8
1
1.2
1.4
1.6
gen
]
-1
[pb GeV
)|
t
) d|y(t
t
dM(t
σ
2
d
0.02
0.04
0.06
0.08
0.1
0.12
0.14
0.16
0.18
0.2
0.22
(13 TeV)
-1
2.3 fb
particle level
+jets
µ
e/
CMS
) < 450 GeV
t
300 < M(t
Data
stat
⊕
Sys
Stat
POWHEG P8
POWHEG H++
MG5 P8
MG5 H++
MG5 P8 [FxFx]
MG5 P8 [MLM]
)|
t
|y(t
0
0.2
0.4
0.6
0.8
1
1.2
1.4
1.6
1.8
2
2.2
Data
Theory
1
1.2
gen
]
-1
[pb GeV
)|
t
) d|y(t
t
dM(t
σ
2
d
0.02
0.04
0.06
0.08
0.1
0.12
0.14
0.16
0.18
0.2
0.22
(13 TeV)
-1
2.3 fb
particle level
+jets
µ
e/
CMS
) < 625 GeV
t
450 < M(t
Data
stat
⊕
Sys
Stat
POWHEG P8
POWHEG H++
MG5 P8
MG5 H++
MG5 P8 [FxFx]
MG5 P8 [MLM]
)|
t
|y(t
0
0.2
0.4
0.6
0.8
1
1.2
1.4
1.6
1.8
2
2.2
Data
Theory
0.8
1
1.2
1.4
gen
]
-1
[pb GeV
)|
t
) d|y(t
t
dM(t
σ
2
d
0.02
0.04
0.06
0.08
0.1
(13 TeV)
-1
2.3 fb
particle level
+jets
µ
e/
CMS
) < 850 GeV
t
625 < M(t
Data
stat
⊕
Sys
Stat
POWHEG P8
POWHEG H++
MG5 P8
MG5 H++
MG5 P8 [FxFx]
MG5 P8 [MLM]
)|
t
|y(t
0
0.2
0.4
0.6
0.8
1
1.2
1.4
1.6
1.8
2
2.2
Data
Theory
0.8
1
1.2
gen
]
-1
[pb GeV
)|
t
) d|y(t
t
dM(t
σ
2
d
0.002
0.004
0.006
0.008
0.01
0.012
(13 TeV)
-1
2.3 fb
particle level
+jets
µ
e/
CMS
) < 2000 GeV
t
850 < M(t
Data
stat
⊕
Sys
Stat
POWHEG P8
POWHEG H++
MG5 P8
MG5 H++
MG5 P8 [FxFx]
MG5 P8 [MLM]
)|
t
|y(t
0
0.2
0.4
0.6
0.8
1
1.2
1.4
1.6
1.8
2
2.2
Data
Theory
0.8
1
1.2
1.4
gure 16: Double-differential cross sections at particle level as a function of |y(th)| vs. 12
Summary p
upper two rows) and M(tt) vs. |y(tt)| (lower two rows). The measurements are c
ared to the predictions of POWHEG and MG5 aMC@NLO (MG5) combined with PYTHIA8
r HERWIG++ (H++) and the multiparton simulations MG5 aMC@NLO+PYTHIA8 MLM
G5 aMC@NLO+PYTHIA8 FxFx. The ratios of the predictions to the measured cross sec
re shown at the bottom of each panel together with the statistical and systematic uncertai
f the measurement. 12
Summary pT(th)
(upper two rows) and M(tt) vs. |y(tt)| (lower two rows). The measurements are com-
pared to the predictions of POWHEG and MG5 aMC@NLO (MG5) combined with PYTHIA8 (P8)
or HERWIG++ (H++) and the multiparton simulations MG5 aMC@NLO+PYTHIA8 MLM and
MG5 aMC@NLO+PYTHIA8 FxFx. The ratios of the predictions to the measured cross sections
are shown at the bottom of each panel together with the statistical and systematic uncertainties
of the measurement. 12
Summary 25 gen
]
-2
[pb GeV
)t
(t
T
) dp
t
dM(t
σ
2
d
4
−
10
3
−
10
2
−
10
(13 TeV)
-1
2.3 fb
particle level
+jets
µ
e/
CMS
) < 35 GeV
t
(t
T
0 < p
Data
stat
⊕
Sys
Stat
POWHEG P8
POWHEG H++
MG5 P8
MG5 H++
MG5 P8 [FxFx]
MG5 P8 [MLM]
) [GeV]
t
M(t
400
600
800
1000
1200
1400
1600
1800
2000
Data
Theory
1
1.5
gen
]
-2
[pb GeV
)t
(t
T
) dp
t
dM(t
σ
2
d
4
−
10
3
−
10
2
−
10
(13 TeV)
-1
2.3 fb
particle level
+jets
µ
e/
CMS
) < 80 GeV
t
(t
T
35 < p
Data
stat
⊕
Sys
Stat
POWHEG P8
POWHEG H++
MG5 P8
MG5 H++
MG5 P8 [FxFx]
MG5 P8 [MLM]
) [GeV]
t
M(t
400
600
800
1000
1200
1400
1600
1800
2000
Data
Theory
0.8
1
1.2
1.4
1.6
gen
]
-2
[pb GeV
)t
(t
T
) dp
t
dM(t
σ
2
d
5
−
10
4
−
10
3
−
10
(13 TeV)
-1
2.3 fb
particle level
+jets
µ
e/
CMS
) < 140 GeV
t
(t
T
80 < p
Data
stat
⊕
Sys
Stat
POWHEG P8
POWHEG H++
MG5 P8
MG5 H++
MG5 P8 [FxFx]
MG5 P8 [MLM]
) [GeV]
t
M(t
400
600
800
1000
1200
1400
1600
1800
2000
Data
Theory
0.8
1
1.2
1.4
gen
]
-2
[pb GeV
)t
(t
T
) dp
t
dM(t
σ
2
d
5
−
10
4
−
10
(13 TeV)
-1
2.3 fb
particle level
+jets
µ
e/
CMS
) < 500 GeV
t
(t
T
140 < p
Data
stat
⊕
Sys
Stat
POWHEG P8
POWHEG H++
MG5 P8
MG5 H++
MG5 P8 [FxFx]
MG5 P8 [MLM]
) [GeV]
t
M(t
400
600
800
1000
1200
1400
1600
1800
2000
Data
Theory
0.8
1
1.2
1.4
Figure 17:
Double-differential cross section at particle level as a function of pT(tt) vs. M(tt). The measurements are compared to the predictions of POWHEG and MG5 aMC@NLO
(MG5) combined with PYTHIA8 (P8) or HERWIG++ (H++) and the multiparton simulations
MG5 aMC@NLO+PYTHIA8 MLM and MG5 aMC@NLO+PYTHIA8 FxFx. 12
Summary The ratios of the pre-
dictions to the measured cross sections are shown at the bottom of each panel together with the
statistical and systematic uncertainties of the measurement. ) [GeV]
t
M(t ) [GeV]
t
M(t ) [GeV]
t
M(t Figure 17:
Double-differential cross section at particle level as a function of pT(tt) vs. M(tt). The measurements are compared to the predictions of POWHEG and MG5 aMC@NLO
(MG5) combined with PYTHIA8 (P8) or HERWIG++ (H++) and the multiparton simulations
MG5 aMC@NLO+PYTHIA8 MLM and MG5 aMC@NLO+PYTHIA8 FxFx. The ratios of the pre-
dictions to the measured cross sections are shown at the bottom of each panel together with the
statistical and systematic uncertainties of the measurement. 26 12
Summary 12
Summary Table 2: Comparison between the measured distributions at parton level and the predictions
of POWHEG and MG5 aMC@NLO combined with PYTHIA8 (P8) or HERWIG++ (H++) and the
multiparton simulations MG5 aMC@NLO MLM and MG5 aMC@NLO FxFx, as well as the pre-
dictions of an approximate NNNLO calculation [42, 43], a NLO+NNLL’ calculation [45], and
a full NNLO calculation [46]. We list the results of the χ2 tests together with the numbers of
degrees of freedom (dof) and the corresponding p-values. For the comparison no uncertainties
in the theoretical predictions are taken into account. g
(
)
p
g p
p
in the theoretical predictions are taken into account. Distribution
χ2/dof
p-value
χ2/dof
p-value
χ2/dof
p-value
POWHEG+P8
POWHEG+H++
MG5 aMC@NLO+P8 MLM
Order: NLO
Order: NLO
Order: LO, up to 3 add. partons
pT(th)
10.7/9
0.295
8.01/9
0.533
19.0/9
0.025
|y(th)|
3.91/7
0.790
4.33/7
0.741
4.49/7
0.721
pT(tℓ)
14.9/9
0.093
9.03/9
0.435
41.8/9
< 0.01
|y(tℓ)|
11.4/7
0.121
13.1/7
0.070
12.0/7
0.100
M(tt)
5.61/8
0.691
10.9/8
0.206
45.0/8
< 0.01
pT(tt)
0.941/5
0.967
4.34/5
0.501
16.8/5
< 0.01
|y(tt)|
1.95/6
0.924
2.04/6
0.916
5.55/6
0.476
Additional jets
8.22/5
0.145
6.88/5
0.230
5.82/5
0.324
Additional jets vs. pT(tt)
85.3/20
< 0.01
132/20
< 0.01
135/20
< 0.01
Additional jets vs. pT(th)
89.0/36
< 0.01
43.1/36
0.193
71.7/36
< 0.01
|y(th)| vs. pT(th)
55.3/36
0.021
52.4/36
0.038
60.7/36
< 0.01
M(tt) vs. |y(tt)|
19.3/24
0.734
18.3/24
0.788
49.4/24
< 0.01
pT(tt) vs. M(tt)
14.5/32
0.997
26.2/32
0.755
100/32
< 0.01
MG5 aMC@NLO+P8
MG5 aMC@NLO+H++
MG5 aMC@NLO+P8 FxFx
Order: NLO
Order: NLO
Order: NLO, up to 2 add. 12
Summary partons
pT(th)
8.68/9
0.467
15.3/9
0.084
9.35/9
0.406
|y(th)|
4.11/7
0.767
5.42/7
0.608
3.91/7
0.790
pT(tℓ)
13.0/9
0.162
26.8/9
< 0.01
11.7/9
0.228
|y(tℓ)|
14.3/7
0.046
10.7/7
0.151
16.4/7
0.022
M(tt)
9.91/8
0.271
5.93/8
0.655
28.0/8
< 0.01
pT(tt)
31.1/5
< 0.01
24.6/5
< 0.01
18.4/5
< 0.01
|y(tt)|
1.97/6
0.923
2.04/6
0.916
2.49/6
0.870
Additional jets
21.5/5
< 0.01
4.21/5
0.520
7.98/5
0.158
Additional jets vs. pT(tt)
319/20
< 0.01
259/20
< 0.01
121/20
< 0.01
Additional jets vs. pT(th)
90.9/36
< 0.01
45.0/36
0.145
52.5/36
0.037
|y(th)| vs. pT(th)
73.1/36
< 0.01
111/36
< 0.01
48.1/36
0.086
M(tt) vs. |y(tt)|
26.1/24
0.347
17.8/24
0.811
36.7/24
0.047
pT(tt) vs. M(tt)
229/32
< 0.01
71.5/32
< 0.01
97.6/32
< 0.01
appr. NNLO
appr. NNNLO
NLO+NNLL’
pT(th)
14.3/9
0.111
36.7/9
< 0.01
6.29/9
0.710
|y(th)|
5.30/7
0.623
2.59/7
0.920
—
—
pT(tℓ)
12.1/9
0.209
92.1/9
< 0.01
3.06/9
0.962
|y(tℓ)|
3.77/7
0.805
4.34/7
0.739
—
—
M(tt)
—
—
—
—
6.70/8
0.569
NNLO
pT(th)
5.78/9
0.762
|y(th)|
2.20/7
0.948
pT(tℓ)
5.54/9
0.785
|y(tℓ)|
6.48/7
0.485
M(tt)
5.88/8
0.660
pT(tt)
3.50/5
0.623
|y(tt)|
1.42/6
0.965 27 Table 3: Comparison between the measured distributions at particle level and the predictions
of POWHEG and MG5 aMC@NLO combined with PYTHIA8 (P8) or HERWIG++ (H++) and the
multiparton simulations MG5 aMC@NLO MLM and MG5 aMC@NLO FxFx. We list the results
of the χ2 tests together with the numbers of degrees of freedom (dof) and the corresponding
p-values. For the comparison no uncertainties in the theoretical predictions are taken into
account. Distribution
χ2/dof
p-value
χ2/dof
p-value
χ2/dof
p-value
POWHEG+P8
POWHEG+H++
MG5 aMC@NLO+P8 MLM
Order: NLO
Order: NLO
Order: LO, up to 3 add. partons
pT(th)
14.2/9
0.115
24.0/9
< 0.01
32.8/9
< 0.01
|y(th)|
3.47/7
0.838
5.66/7
0.579
6.64/7
0.468
pT(tℓ)
20.8/9
0.013
38.2/9
< 0.01
49.7/9
< 0.01
|y(tℓ)|
6.37/7
0.497
9.69/7
0.207
16.1/7
0.025
M(tt)
9.03/8
0.340
148/8
< 0.01
12.0/8
0.151
pT(tt)
2.15/5
0.829
29.4/5
< 0.01
49.2/5
< 0.01
|y(tt)|
0.869/6
0.990
2.06/6
0.914
13.2/6
0.040
Additional jets
28.2/5
< 0.01
17.2/5
< 0.01
36.8/5
< 0.01
Additional jets vs. pT(tt)
70.7/20
< 0.01
86.1/20
< 0.01
161/20
< 0.01
Additional jets vs. pT(th)
91.6/36
< 0.01
200/36
< 0.01
162/36
< 0.01
|y(th)| vs. pT(th)
56.2/36
0.017
197/36
< 0.01
114/36
< 0.01
M(tt) vs. |y(tt)|
26.6/24
0.324
263/24
< 0.01
38.1/24
0.034
pT(tt) vs. Acknowledgments We congratulate our colleagues in the CERN accelerator departments for the excellent perfor-
mance of the LHC and thank the technical and administrative staffs at CERN and at other
CMS institutes for their contributions to the success of the CMS effort. In addition, we grate-
fully acknowledge the computing centers and personnel of the Worldwide LHC Computing
Grid for delivering so effectively the computing infrastructure essential to our analyses. Acknowledgments Fi-
nally, we acknowledge the enduring support for the construction and operation of the LHC
and the CMS detector provided by the following funding agencies: the Austrian Federal Min-
istry of Science, Research and Economy and the Austrian Science Fund; the Belgian Fonds de
la Recherche Scientifique, and Fonds voor Wetenschappelijk Onderzoek; the Brazilian Fund-
ing Agencies (CNPq, CAPES, FAPERJ, and FAPESP); the Bulgarian Ministry of Education and
Science; CERN; the Chinese Academy of Sciences, Ministry of Science and Technology, and Na-
tional Natural Science Foundation of China; the Colombian Funding Agency (COLCIENCIAS);
the Croatian Ministry of Science, Education and Sport, and the Croatian Science Foundation;
the Research Promotion Foundation, Cyprus; the Secretariat for Higher Education, Science,
Technology and Innovation, Ecuador; the Ministry of Education and Research, Estonian Re-
search Council via IUT23-4 and IUT23-6 and European Regional Development Fund, Estonia;
the Academy of Finland, Finnish Ministry of Education and Culture, and Helsinki Institute of
Physics; the Institut National de Physique Nucl´eaire et de Physique des Particules / CNRS, and
Commissariat `a l’´Energie Atomique et aux ´Energies Alternatives / CEA, France; the Bundes-
ministerium f¨ur Bildung und Forschung, Deutsche Forschungsgemeinschaft, and Helmholtz-
Gemeinschaft Deutscher Forschungszentren, Germany; the General Secretariat for Research
and Technology, Greece; the National Scientific Research Foundation, and National Innova-
tion Office, Hungary; the Department of Atomic Energy and the Department of Science and
Technology, India; the Institute for Studies in Theoretical Physics and Mathematics, Iran; the
Science Foundation, Ireland; the Istituto Nazionale di Fisica Nucleare, Italy; the Ministry of
Science, ICT and Future Planning, and National Research Foundation (NRF), Republic of Ko-
rea; the Lithuanian Academy of Sciences; the Ministry of Education, and University of Malaya
(Malaysia); the Mexican Funding Agencies (BUAP, CINVESTAV, CONACYT, LNS, SEP, and
UASLP-FAI); the Ministry of Business, Innovation and Employment, New Zealand; the Pak-
istan Atomic Energy Commission; the Ministry of Science and Higher Education and the Na-
tional Science Centre, Poland; the Fundac¸˜ao para a Ciˆencia e a Tecnologia, Portugal; JINR,
Dubna; the Ministry of Education and Science of the Russian Federation, the Federal Agency
of Atomic Energy of the Russian Federation, Russian Academy of Sciences, and the Russian
Foundation for Basic Research; the Ministry of Education, Science and Technological Devel-
opment of Serbia; the Secretar´ıa de Estado de Investigaci´on, Desarrollo e Innovaci´on and Pro-
grama Consolider-Ingenio 2010, Spain; the Swiss Funding Agencies (ETH Board, ETH Zurich,
PSI, SNF, UniZH, Canton Zurich, and SER); the Ministry of Science and Technology, Taipei; the
Thailand Center of Excellence in Physics, the Institute for the Promotion of Teaching Science
and Technology of Thailand, Special Task Force for Activating Research and the National Sci-
ence and Technology Development Agency of Thailand; the Scientific and Technical Research
Council of Turkey, and Turkish Atomic Energy Authority; the National Academy of Sciences
of Ukraine, and State Fund for Fundamental Researches, Ukraine; the Science and Technology
Facilities Council, UK; the US Department of Energy, and the US National Science Foundation. 12
Summary M(tt)
13.4/32
0.998
459/32
< 0.01
89.0/32
< 0.01
MG5 aMC@NLO+P8
MG5 aMC@NLO+H++
MG5 aMC@NLO+P8 FxFx
Order: NLO
Order: NLO
Order: NLO, up to 2 add. partons
pT(th)
11.9/9
0.221
5.51/9
0.788
4.17/9
0.900
|y(th)|
7.34/7
0.394
10.6/7
0.156
5.93/7
0.547
pT(tℓ)
11.0/9
0.274
6.37/9
0.702
6.51/9
0.688
|y(tℓ)|
12.3/7
0.092
6.04/7
0.535
14.3/7
0.047
M(tt)
9.57/8
0.296
28.7/8
< 0.01
28.5/8
< 0.01
pT(tt)
37.1/5
< 0.01
7.92/5
0.161
29.6/5
< 0.01
|y(tt)|
1.75/6
0.942
1.98/6
0.922
2.87/6
0.825
Additional jets
29.6/5
< 0.01
12.2/5
0.032
11.6/5
0.041
Additional jets vs. pT(tt)
197/20
< 0.01
163/20
< 0.01
85.3/20
< 0.01
Additional jets vs. pT(th)
151/36
< 0.01
57.7/36
0.012
40.4/36
0.282
|y(th)| vs. pT(th)
36.6/36
0.441
82.5/36
< 0.01
42.2/36
0.222
M(tt) vs. |y(tt)|
21.4/24
0.612
47.9/24
< 0.01
52.3/24
< 0.01
pT(tt) vs. M(tt)
119/32
< 0.01
164/32
< 0.01
107/32
< 0.01 28 12
Summary 12
Summary the Agentschap voor Innovatie door Wetenschap en Technologie (IWT-Belgium); the Ministry
of Education, Youth and Sports (MEYS) of the Czech Republic; the Council of Science and In-
dustrial Research, India; the HOMING PLUS program of the Foundation for Polish Science,
cofinanced from European Union, Regional Development Fund, the Mobility Plus program of
the Ministry of Science and Higher Education, the National Science Center (Poland), contracts
Harmonia 2014/14/M/ST2/00428, Opus 2013/11/B/ST2/04202, 2014/13/B/ST2/02543 and
2014/15/B/ST2/03998, Sonata-bis 2012/07/E/ST2/01406; the Thalis and Aristeia programs
cofinanced by EU-ESF and the Greek NSRF; the National Priorities Research Program by Qatar
National Research Fund; the Programa Clar´ın-COFUND del Principado de Asturias; the Rachada-
pisek Sompot Fund for Postdoctoral Fellowship, Chulalongkorn University and the Chula-
longkorn Academic into Its 2nd Century Project Advancement Project (Thailand); and the
Welch Foundation, contract C-1845. References [1] CMS Collaboration, “CMS Luminosity Measurement for the 2015 Data Taking Period”,
CMS Physics Analysis Summary CMS-PAS-LUM-15-001, 2017. [2] CMS C ll b
i
“M
f diff
i l
k
i
d
i [1] CMS Collaboration, “CMS Luminosity Measurement for the 2015 Data Taking Period”,
CMS Physics Analysis Summary CMS-PAS-LUM-15-001, 2017. [2] CMS Collaboration, “Measurement of differential top-quark pair production cross
sections in pp colisions at √s = 7 TeV”, Eur. Phys. J. C 73 (2013) 2339,
doi:10.1140/epjc/s10052-013-2339-4, arXiv:1211.2220. [3] ATLAS Collaboration, “Differential top-antitop cross-section measurements as a function
of observables constructed from final-state particles using pp collisions at √s = 7 TeV in
the ATLAS detector”, JHEP 06 (2015) 100, doi:10.1007/JHEP06(2015)100,
arXiv:1502.05923. [4] CMS Collaboration, “Measurement of the differential cross section for top quark pair
production in pp collisions at √s = 8 TeV”, Eur. Phys. J. C 75 (2015) 542,
doi:10.1140/epjc/s10052-015-3709-x, arXiv:1505.04480. [5] ATLAS Collaboration, “Measurements of top-quark pair differential cross-sections in the
lepton+jets channel in pp collisions at √s = 8 TeV using the ATLAS detector”, Eur. Phys. J. C 76 (2016) 538, doi:10.1140/epjc/s10052-016-4366-4, arXiv:1511.04716. [6] ATLAS Collaboration, “Measurement of the differential cross-section of highly boosted
top quarks as a function of their transverse momentum in √s = 8 TeV proton-proton
collisions using the ATLAS detector”, Phys. Rev. D 93 (2016) 032009,
doi:10.1103/PhysRevD.93.032009, arXiv:1510.03818. [7] CMS Collaboration, “Measurement of the tt production cross section in the all-jets final
state in pp collisions at √s = 8 TeV”, Eur. Phys. J. C 76 (2016) 128,
doi:10.1140/epjc/s10052-016-3956-5, arXiv:1509.06076. [8] CMS Collaboration, “Measurement of the integrated and differential t¯t production cross
sections for high-pT top quarks in pp collisions at √s = 8 TeV”, Phys. Rev. D 94 (2016)
072002, doi:10.1103/PhysRevD.94.072002, arXiv:1605.00116. [9] P. Nason, “A new method for combining NLO QCD with shower Monte Carlo
algorithms”, JHEP 11 (2004) 040, doi:10.1088/1126-6708/2004/11/040,
arXiv:hep-ph/0409146. [10] S. Frixione, P. Nason, and C. Oleari, “Matching NLO QCD computations with Parton
Shower simulations: the POWHEG method”, JHEP 11 (2007) 070,
doi:10.1088/1126-6708/2007/11/070, arXiv:0709.2092. [11] S. Alioli, P. Nason, C. Oleari, and E. Re, “A general framework for implementing NLO
calculations in shower Monte Carlo programs: the POWHEG BOX”, JHEP 06 (2010) 043,
doi:10.1007/JHEP06(2010)043, arXiv:1002.2581. [12] J. M. Campbell, R. K. Ellis, P. Nason, and E. Re, “Top-pair production and decay at NLO
matched with parton showers”, JHEP 04 (2015) 114,
doi:10.1007/JHEP04(2015)114, arXiv:1412.1828. [13] J. Acknowledgments Individuals have received support from the Marie-Curie program and the European Research
Council and EPLANET (European Union); the Leventis Foundation; the A. P. Sloan Founda- Individuals have received support from the Marie-Curie program and the European Research
Council and EPLANET (European Union); the Leventis Foundation; the A. P. Sloan Founda-
tion; the Alexander von Humboldt Foundation; the Belgian Federal Science Policy Office; the
Fonds pour la Formation `a la Recherche dans l’Industrie et dans l’Agriculture (FRIA-Belgium); 29 the Agentschap voor Innovatie door Wetenschap en Technologie (IWT-Belgium); the Ministry
of Education, Youth and Sports (MEYS) of the Czech Republic; the Council of Science and In-
dustrial Research, India; the HOMING PLUS program of the Foundation for Polish Science,
cofinanced from European Union, Regional Development Fund, the Mobility Plus program of
the Ministry of Science and Higher Education, the National Science Center (Poland), contracts
Harmonia 2014/14/M/ST2/00428, Opus 2013/11/B/ST2/04202, 2014/13/B/ST2/02543 and
2014/15/B/ST2/03998, Sonata-bis 2012/07/E/ST2/01406; the Thalis and Aristeia programs
cofinanced by EU-ESF and the Greek NSRF; the National Priorities Research Program by Qatar
National Research Fund; the Programa Clar´ın-COFUND del Principado de Asturias; the Rachada-
pisek Sompot Fund for Postdoctoral Fellowship, Chulalongkorn University and the Chula-
longkorn Academic into Its 2nd Century Project Advancement Project (Thailand); and the
Welch Foundation, contract C-1845. the Agentschap voor Innovatie door Wetenschap en Technologie (IWT-Belgium); the Ministry
of Education, Youth and Sports (MEYS) of the Czech Republic; the Council of Science and In-
dustrial Research, India; the HOMING PLUS program of the Foundation for Polish Science,
cofinanced from European Union, Regional Development Fund, the Mobility Plus program of
the Ministry of Science and Higher Education, the National Science Center (Poland), contracts
Harmonia 2014/14/M/ST2/00428, Opus 2013/11/B/ST2/04202, 2014/13/B/ST2/02543 and
2014/15/B/ST2/03998, Sonata-bis 2012/07/E/ST2/01406; the Thalis and Aristeia programs
cofinanced by EU-ESF and the Greek NSRF; the National Priorities Research Program by Qatar
National Research Fund; the Programa Clar´ın-COFUND del Principado de Asturias; the Rachada-
pisek Sompot Fund for Postdoctoral Fellowship, Chulalongkorn University and the Chula-
longkorn Academic into Its 2nd Century Project Advancement Project (Thailand); and the
Welch Foundation, contract C-1845. 30 References References Alwall et al., “The automated computation of tree-level and next-to-leading order
differential cross sections, and their matching to parton shower simulations”, JHEP 07
(2014) 079, doi:10.1007/JHEP07(2014)079, arXiv:1405.0301. 31 References [14] T. Sj¨ostrand, S. Mrenna, and P. Skands, “PYTHIA 6.4 physics and manual”, JHEP 05
(2006) 026, doi:10.1088/1126-6708/2006/05/026, arXiv:hep-ph/0603175. [15] T. Sj¨ostrand, S. Mrenna, and P. Z. Skands, “A brief introduction to PYTHIA 8.1”, Comput. Phys. Commun. 178 (2008) 852, doi:10.1016/j.cpc.2008.01.036,
arXiv:0710.3820. [16] P. Skands, S. Carrazza, and J. Rojo, “Tuning PYTHIA 8.1: the Monash 2013 Tune”, Eur. Phys. J. C 74 (2014) 3024, doi:10.1140/epjc/s10052-014-3024-y,
arXiv:1404.5630. [17] J. Alwall et al., “Comparative study of various algorithms for the merging of parton
showers and matrix elements in hadronic collisions”, Eur. Phys. J. C 53 (2008) 473,
doi:10.1140/epjc/s10052-007-0490-5, arXiv:0706.2569. [18] R. Frederix and S. Frixione, “Merging meets matching in MC@NLO”, JHEP 12 (2012)
061, doi:10.1007/JHEP12(2012)061, arXiv:1209.6215. [19] NNPDF Collaboration, “Parton distributions for the LHC Run II”, JHEP 04 (2015) 040,
doi:10.1007/JHEP04(2015)040, arXiv:1410.8849. [20] M. Czakon and A. Mitov, “Top++: A Program for the Calculation of the Top-Pair
Cross-Section at Hadron Colliders”, Comput. Phys. Commun. 185 (2014) 2930,
doi:10.1016/j.cpc.2014.06.021, arXiv:1112.5675. [21] M. B¨ahr et al., “Herwig++ physics and manual”, Eur. Phys. J. C 58 (2008) 639,
doi:10.1140/epjc/s10052-008-0798-9, arXiv:0803.0883. [22] M. H. Seymour and A. Siodmok, “Constraining MPI models using σe f f and recent
Tevatron and LHC Underlying Event data”, JHEP 10 (2013) 113,
doi:10.1007/JHEP10(2013)113, arXiv:1307.5015. [23] Y. Li and F. Petriello, “Combining QCD and electroweak corrections to dilepton
production in FEWZ”, Phys. Rev. D 86 (2012) 094034,
doi:10.1103/PhysRevD.86.094034, arXiv:1208.5967. [24] P. Kant et al., “HatHor for single top-quark production: Updated predictions and
uncertainty estimates for single top-quark production in hadronic collisions”, Comput. Phys. Commun. 191 (2015) 74, doi:10.1016/j.cpc.2015.02.001,
arXiv:1406.4403. [25] N. Kidonakis, “NNLL threshold resummation for top-pair and single-top production”,
Phys. Part. Nucl. 45 (2014) 714, doi:10.1134/S1063779614040091,
arXiv:1210.7813. [26] J. Allison et al., “Geant4 developments and applications”, IEEE Trans. Nucl. Sci. 53 (2006)
270, doi:10.1109/TNS.2006.869826. [27] M. Cacciari, G. P. Salam, and G. Soyez, “The anti-kt jet clustering algorithm”, JHEP 04
(2008) 063, doi:10.1088/1126-6708/2008/04/063, arXiv:0802.1189. [28] M. Cacciari, G. P. Salam, and G. Soyez, “FastJet user manual”, Eur. Phys. J. C 72 (2012)
1896, doi:10.1140/epjc/s10052-012-1896-2, arXiv:1111.6097. 32 References [29] CMS Collaboration, “The CMS experiment at the CERN LHC”, JINST 3 (2008) S08004,
doi:10.1088/1748-0221/3/08/S08004. [44] A. D. Martin, W. J. Stirling, R. S. Thorne, and G. Watt, “Parton distributions for the LHC”,
Eur. Phys. J. C 63 (2009) 189, doi:10.1140/epjc/s10052-009-1072-5,
arXiv:0901.0002. [45] B. Pecjak, D. Scott, X. Wang, and L. L. Yang, “Resummed Differential Cross Sections for
Top-Quark Pairs at the LHC”, Phys. Rev. Lett. 116 (2016) 202001,
doi:10.1103/PhysRevLett.116.202001, arXiv:1601.07020. [46] M. Czakon, D. Heymes, and A. Mitov, “High-Precision Differential Predictions for
Top-Quark Pairs at the LHC”, Phys. Rev. Lett. 116 (2016) 082003,
doi:10.1103/PhysRevLett.116.082003, arXiv:1511.00549. References [30] CMS Collaboration, “Particle-flow event reconstruction in CMS and performance for jets,
taus, and Emiss
T
”, CMS Physics Analysis Summary CMS-PAS-PFT-09-001, 2009. [31] CMS Collaboration, “Commissioning of the particle-flow event reconstruction with the
first LHC collisions recorded in the CMS detector”, CMS Physics Analysis Summary
CMS-PAS-PFT-10-001, 2010. [32] CMS Collaboration, “Performance of CMS muon reconstruction in pp collision events at
√s = 7 TeV”, JINST 7 (2012) P10002, doi:10.1088/1748-0221/7/10/P10002,
arXiv:1206.4071. [33] CMS Collaboration, “Measurement of inclusive W and Z boson production cross sections
in pp collisions at √s = 8 TeV”, Phys. Rev. Lett. 112 (2014) 191802,
doi:10.1103/PhysRevLett.112.191802, arXiv:1402.0923. [34] M. Cacciari and G. P. Salam, “Pileup subtraction using jet areas”, Phys. Lett. B 659 (2008)
119, doi:10.1016/j.physletb.2007.09.077, arXiv:0707.1378. [35] CMS Collaboration, “Performance of electron reconstruction and selection with the CMS
detector in proton-proton collisions at √s = 8 TeV”, JINST 10 (2015) P06005,
doi:10.1088/1748-0221/10/06/P06005, arXiv:1502.02701. [36] CMS Collaboration, “Jet energy scale and resolution in the CMS experiment in pp
collisions at 8 TeV”, (2016). arXiv:1607.03663. Submitted to JINST. [37] CMS Collaboration, “Identification of b quark jets at the CMS Experiment in the LHC
Run 2”, CMS Physics Analysis Summary CMS-PAS-BTV-15-001, 2016. [38] B. A. Betchart, R. Demina, and A. Harel, “Analytic solutions for neutrino momenta in
decay of top quarks”, Nucl. Instrum. Meth. A 736 (2014) 169,
doi:10.1016/j.nima.2013.10.039, arXiv:1305.1878. [39] G. D’Agostini, “A multidimensional unfolding method based on Bayes’ theorem”, Nucl. Instrum. Meth. A 362 (1995) 487, doi:10.1016/0168-9002(95)00274-X. [40] M. Guzzi, K. Lipka, and S.-O. Moch, “Top-quark pair production at hadron colliders:
differential cross section and phenomenological applications with DiffTop”, JHEP 01
(2015) 082, doi:10.1007/JHEP01(2015)082, arXiv:1406.0386. [41] S. Dulat et al., “New parton distribution functions from a global analysis of quantum
chromodynamics”, Phys. Rev. D 93 (2016) 033006,
doi:10.1103/PhysRevD.93.033006, arXiv:1506.07443. [42] N. Kidonakis, “NNNLO soft-gluon corrections for the top-antitop pair production cross
section”, Phys. Rev. D 90 (2014) 014006, doi:10.1103/PhysRevD.90.014006,
arXiv:1405.7046. [43] N. Kidonakis, “NNNLO soft-gluon corrections for the top-quark pT and rapidity
distributions”, Phys. Rev. D 91 (2015) 031501, doi:10.1103/PhysRevD.91.031501,
arXiv:1411.2633. 33 References [44] A. D. Martin, W. J. Stirling, R. S. Thorne, and G. Watt, “Parton distributions for the LHC”,
Eur. Phys. J. C 63 (2009) 189, doi:10.1140/epjc/s10052-009-1072-5,
arXiv:0901.0002. [45] B. Pecjak, D. Scott, X. Wang, and L. L. Yang, “Resummed Differential Cross Sections for
Top-Quark Pairs at the LHC”, Phys. Rev. Lett. 116 (2016) 202001,
doi:10.1103/PhysRevLett.116.202001, arXiv:1601.07020. [46] M. Czakon, D. Heymes, and A. References Mitov, “High-Precision Differential Predictions for
Top-Quark Pairs at the LHC”, Phys. Rev. Lett. 116 (2016) 082003,
doi:10.1103/PhysRevLett.116.082003, arXiv:1511.00549. A
Tables of parton level cross sections. 34 A
Tables of parton level cross sections. Table 4: Differential cross section at parton level as a function of pT(th). The values are shown
together with their statistical and systematic uncertainties. pT(th)
dσ
dpT(th)
pT(th)
dσ
dpT(th)
[GeV]
[fb GeV−1]
[GeV]
[fb GeV−1]
0–45
680±20±180
225–270
228± 6 ±32
45–90
1500±20±190
270–315
119± 5 ±18
90–135
1290±20±160
315–400
46± 2 ±7
135–180
790±10±100
400–800
5.1±0.3±0.8
180–225
420± 9 ±59
— Table 5: Differential cross section at parton level as a function of |y(th)|. The values are shown
together with their statistical and systematic uncertainties. |y(th)|
dσ
d|y(th)| [pb]
|y(th)|
dσ
d|y(th)| [pb]
0–0.2
142±2±14
1–1.3
100± 2 ±11
0.2–0.4
135±2±13
1.3–1.6
82± 2 ±11
0.4–0.7
129±2±13
1.6–2.5
44.0±0.9±6.4
0.7–1
114±2±12
— Table 6: Differential cross section at parton level as a function of pT(tℓ). The values are shown
together with their statistical and systematic uncertainties. pT(tℓ)
dσ
dpT(tℓ)
pT(tℓ)
dσ
dpT(tℓ)
[GeV]
[fb GeV−1]
[GeV]
[fb GeV−1]
0–45
690±10±100
225–270
218± 4 ±20
45–90
1470±20±190
270–315
115± 3 ±14
90–135
1300±10±150
315–400
47± 1 ±6
135–180
810±10±91
400–800
4.8±0.2±0.5
180–225
432± 7 ±44
— Table 6: Differential cross section at parton level as a function of pT(tℓ). The values are shown
together with their statistical and systematic uncertainties. 35 Table 7: Differential cross section at parton level as a function of |y(tℓ)|. The values are shown
together with their statistical and systematic uncertainties. |y(tℓ)|
dσ
d|y(tℓ)| [pb]
|y(tℓ)|
dσ
d|y(tℓ)| [pb]
0–0.2
135±2±14
1–1.3
101± 1 ±11
0.2–0.4
133±1±14
1.3–1.6
82± 1 ±9
0.4–0.7
128±1±14
1.6–2.5
45.5±0.9±5.1
0.7–1
118±1±13
— Table 7: Differential cross section at parton level as a function of |y(tℓ)|. The values are shown
together with their statistical and systematic uncertainties. Table 7: Differential cross section at parton level as a function of |y(tℓ)|. The values are shown
together with their statistical and systematic uncertainties. |y(tℓ)|
dσ
d|y(tℓ)| [pb]
|y(tℓ)|
dσ
d|y(tℓ)| [pb]
0–0.2
135±2±14
1–1.3
101± 1 ±11
0.2–0.4
133±1±14
1.3–1.6
82± 1 ±9
0.4–0.7
128±1±14
1.6–2.5
45.5±0.9±5.1
0.7–1
118±1±13
— Table 8: Differential cross section at parton level as a function of pT(tt). The values are shown
together with their statistical and systematic uncertainties. pT(tt)
dσ
dpT(tt)
pT(tt)
dσ
dpT(tt)
[GeV]
[fb GeV−1]
[GeV]
[fb GeV−1]
0–35
3050±70±870
140–200
220±10±30
35–80
1470±50±370
200–500
39± 1 ±5
80–140
570±20±90
— Table 8: Differential cross section at parton level as a function of pT(tt). A
Tables of parton level cross sections. The values are shown
together with their statistical and systematic uncertainties. (tt)
dσ
(tt)
dσ Differential cross section at parton level as a function of pT(tt). The values are shown
r with their statistical and systematic uncertainties. Table 9: Differential cross section at parton level as a function of M(tt). The values are shown
together with their statistical and systematic uncertainties. M(tt)
dσ
dM(tt)
M(tt)
dσ
dM(tt)
[GeV]
[fb GeV−1]
[GeV]
[fb GeV−1]
300–375
360±10±160
625–740
192± 6 ±31
375–450
990±20±130
740–850
84± 4 ±10
450–530
620±10±110
850–1100
35± 2 ±6
530–625
373± 9 ±48
1100–2000
3.6±0.3±0.4 Table 9: Differential cross section at parton level as a function of M(tt). The values are shown
together with their statistical and systematic uncertainties. Table 10: Differential cross section at parton level as a function of |y(tt)|. The values are shown
together with their statistical and systematic uncertainties. Table 10: Differential cross section at parton level as a function of |y(tt)|. The values are shown
together with their statistical and systematic uncertainties. |y(tt)|
dσ
d|y(tt)| [pb]
|y(tt)|
dσ
d|y(tt)| [pb]
0–0.2
166±3±17
0.6–0.9
137±2±15
0.2–0.4
157±3±17
0.9–1.3
103±2±12
0.4–0.6
149±3±16
1.3–2.3
48±1±6 Table 10: Differential cross section at parton level as a function of |y(tt)|. The values are shown
together with their statistical and systematic uncertainties. |y(tt)|
dσ
d|y(tt)| [pb]
|y(tt)|
dσ
d|y(tt)| [pb]
0–0.2
166±3±17
0.6–0.9
137±2±15
0 2 0 4
157±3±17
0 9 1 3
103±2±12 Table 11: Cross sections at parton level in bins of the number of additional jets. The values are
shown together with their statistical and systematic uncertainties. Additional jets
σ [pb]
Additional jets
σ [pb]
0
97±2±7
3
12.7±0.6±3.1
1
77±2±11
≥4
5.9±0.2±2.1
2
36±1±6
— Table 11: Cross sections at parton level in bins of the number of additional jets. The values are
shown together with their statistical and systematic uncertainties. Additional jets
σ [pb]
Additional jets
σ [pb]
0
97±2±7
3
12.7±0.6±3.1 Table 11: Cross sections at parton level in bins of the number of additional jets. The values are
shown together with their statistical and systematic uncertainties. Additional jets
σ [pb]
Additional jets
σ [pb]
0
97±2±7
3
12.7±0.6±3.1
1
77±2±11
≥4
5.9±0.2±2.1
2
36±1±6
— Table 11: Cross sections at parton level in bins of the number of additional jets. The va
shown together with their statistical and systematic uncertainties. A
Tables of parton level cross sections. A
Tables of parton level cross sections. 36 Table 12: Differential cross sections at parton level as a function of pT(th) in bins of the number
of additional jets. The values are shown together with their statistical and systematic uncer-
tainties. pT(th)
dσ
dpT(th)
pT(th)
dσ
dpT(th)
[GeV]
[fb GeV−1]
[GeV]
[fb GeV−1]
Additional jets: 0
0–45
340±20±100
225–270
71± 4 ±9
45–90
750±20±110
270–315
29± 3 ±5
90–135
610±20±70
315–400
11± 1 ±2
135–180
310±10±20
400–800
1.0±0.2±0.1
180–225
157± 7 ±16
—
Additional jets: 1
0–45
206± 6 ±30
225–270
79± 4 ±11
45–90
458± 9 ±60
270–315
42± 3 ±7
90–135
408± 8 ±69
315–400
17± 1 ±2
135–180
267± 6 ±52
400–800
1.8±0.2±0.4
180–225
138± 5 ±24
—
Additional jets: 2
0–45
92± 3 ±17
225–270
50± 2 ±9
45–90
210± 5 ±37
270–315
29± 2 ±6
90–135
196± 4 ±35
315–400
10.6±1.0±1.7
135–180
136± 4 ±25
400–800
1.1±0.2±0.2
180–225
82± 3 ±17
—
Additional jets: ≥3
0–45
40± 2 ±8
225–270
28± 2 ±8
45–90
90± 3 ±20
270–315
18± 1 ±5
90–135
94± 3 ±25
315–400
8.4±0.8±3.1
135–180
69± 2 ±21
400–800
1.2±0.2±0.3
180–225
45± 2 ±14
— 37 Table 13: Differential cross sections at parton level as a function of pT(tt) in bins of the number
of additional jets. The values are shown together with their statistical and systematic uncer-
tainties. pT(tt)
dσ
dpT(tt)
pT(tt)
dσ
dpT(tt)
[GeV]
[fb GeV−1]
[GeV]
[fb GeV−1]
Additional jets: 0
0–35
2220±60±530
140–200
14± 5 ±6
35–80
420±40±210
200–500
0.1±0.2±0.1
80–140
50±10±40
—
Additional jets: 1
0–35
610±40±160
140–200
100±10±20
35–80
670±30±90
200–500
9± 1 ±2
80–140
260±20±40
—
Additional jets: 2
0–35
150±10±40
140–200
68± 8 ±12
35–80
240±10±60
200–500
18± 1 ±3
80–140
180±10±40
—
Additional jets: ≥3
0–35
42± 6 ±22
140–200
54± 6 ±13
35–80
95± 8 ±29
200–500
14.3±0.8±3.4
80–140
77± 6 ±23
— A
Tables of parton level cross sections. 38 ble-differential cross section at parton level as a function of |y(th)|
own together with their statistical and systematic uncertainties. A
Tables of parton level cross sections. pT(th)
d2σ
dpT(th)d|y(th)|
pT(th)
d2σ
dpT(th)d|y(th)|
[GeV]
[fb GeV−1]
[GeV]
[fb GeV−1]
0 < |y(th)| < 0.5
0–45
370± 8 ±74
225–270
149± 4 ±19
45–90
830±10±120
270–315
81± 3 ±11
90–135
770±10±80
315–400
36± 2 ±6
135–180
493± 8 ±59
400–800
4.4± 0.3 ±0.6
180–225
268± 6 ±36
—
0.5 < |y(th)| < 1
0–45
340± 7 ±56
225–270
127± 4 ±22
45–90
730±10±110
270–315
65± 3 ±11
90–135
669±10±73
315–400
26± 1 ±3
135–180
425± 8 ±49
400–800
3.3± 0.3 ±0.6
180–225
238± 6 ±34
—
1 < |y(th)| < 1.5
0–45
278± 7 ±44
225–270
88± 3 ±11
45–90
600±10±70
270–315
48± 2 ±8
90–135
528± 9 ±65
315–400
19± 1 ±3
135–180
334± 7 ±46
400–800
1.5± 0.2 ±0.2
180–225
173± 5 ±25
—
1.5 < |y(th)| < 2.5
0–45
188± 7 ±24
225–270
46± 2 ±8
45–90
385± 9 ±50
270–315
20± 1 ±4
90–135
318± 7 ±44
315–400
6.3± 0.6 ±1.0
135–180
175± 5 ±22
400–800
0.50±0.09±0.09
180–225
91± 3 ±12
— Table 14: Double-differential cross section at parton level as a function of |y(th)| vs. pT(th). The
values are shown together with their statistical and systematic uncertainties. 39 Table 15: Double-differential cross section at parton level as a function of M(tt) vs. |y(tt)|. The
values are shown together with their statistical and systematic uncertainties. |y(tt)|
d2σ
dM(tt)d|y(tt)|
|y(tt)|
d2σ
dM(tt)d|y(tt)|
[fb GeV−1]
[fb GeV−1]
300 < M(tt) < 450 GeV
0–0.2
418±10±67
0.6–0.9
374± 7 ±53
0.2–0.4
418± 8 ±63
0.9–1.3
307± 7 ±46
0.4–0.6
409± 8 ±56
1.3–2.3
162± 5 ±25
450 < M(tt) < 625 GeV
0–0.2
359± 7 ±45
0.6–0.9
303± 6 ±43
0.2–0.4
343± 6 ±45
0.9–1.3
224± 5 ±36
0.4–0.6
331± 7 ±46
1.3–2.3
99± 3 ±15
625 < M(tt) < 850 GeV
0–0.2
123± 4 ±18
0.6–0.9
87± 3 ±13
0.2–0.4
108± 3 ±17
0.9–1.3
62± 3 ±13
0.4–0.6
92± 3 ±13
1.3–2.3
24± 2 ±5
850 < M(tt) < 2000 GeV
0–0.2
10.0±0.6±1.5
0.6–0.9
6.9±0.5±0.8
0.2–0.4
10.1±0.6±1.4
0.9–1.3
3.7±0.4±0.5
0.4–0.6
9.1±0.6±1.5
1.3–2.3
1.0±0.2±0.2 Table 15: Double-differential cross section at parton level as a function of M(tt) vs. |y(tt)|. The
values are shown together with their statistical and systematic uncertainties. A
Tables of parton level cross sections. 40 Table 16: Double-differential cross section at parton level as a function of pT(tt) vs. M(tt). The
values are shown together with their statistical and systematic uncertainties. A
Tables of parton level cross sections. A
Tables of parton level cross sections. are shown together with their statistical and systematic uncertainties. M(tt)
d2σ
dpT(tt)dM(tt)
M(tt)
d2σ
dpT(tt)dM(tt)
[GeV]
[fb GeV−2]
[GeV]
[fb GeV−2]
0 < pT(tt) < 35 GeV
300–375
4.8± 0.2 ±2.0
625–740
2.18± 0.09 ±0.63
375–450
13.7± 0.3 ±3.0
740–850
0.92± 0.06 ±0.18
450–530
8.5± 0.2 ±3.8
850–1100
0.36± 0.03 ±0.12
530–625
4.4± 0.1 ±1.3
1100–2000
0.039± 0.005 ±0.012
35 < pT(tt) < 80 GeV
300–375
2.25± 0.07 ±1.20
625–740
1.32± 0.04 ±0.22
375–450
6.6± 0.1 ±1.6
740–850
0.60± 0.03 ±0.07
450–530
4.30± 0.08 ±0.60
850–1100
0.23± 0.01 ±0.03
530–625
2.53± 0.06 ±0.29
1100–2000
0.022± 0.002 ±0.005
80 < pT(tt) < 140 GeV
300–375
0.76± 0.03 ±0.30
625–740
0.51± 0.02 ±0.07
375–450
2.24± 0.05 ±0.50
740–850
0.25± 0.01 ±0.04
450–530
1.52± 0.04 ±0.19
850–1100
0.100± 0.008 ±0.026
530–625
0.96± 0.03 ±0.10
1100–2000
0.011± 0.002 ±0.002
140 < pT(tt) < 500 GeV
300–375
0.095±0.005±0.025
625–740
0.068± 0.003 ±0.017
375–450
0.258±0.008±0.032
740–850
0.036± 0.002 ±0.004
450–530
0.185±0.006±0.024
850–1100
0.016± 0.001 ±0.003
530–625
0.122±0.005±0.034
1100–2000
0.0018±0.0003±0.0003 41 B
Tables of particle level cross sections. Table 17: Differential cross section at particle level as a function of pT(th). The values are shown
together with their statistical and systematic uncertainties. Table 17: Differential cross section at particle level as a function of pT(th). The values are shown
together with their statistical and systematic uncertainties. pT(th)
dσ
dpT(th)
pT(th)
dσ
dpT(th)
[GeV]
[fb GeV−1]
[GeV]
[fb GeV−1]
0–45
204±4±18
225–270
106± 2 ±9
45–90
461±5±40
270–315
61± 2 ±6
90–135
430±5±41
315–400
27.4±0.9±2.5
135–180
292±4±27
400–800
3.2±0.2±0.3
180–225
179±3±17
— Table 17: Differential cross section at particle level as a function of pT(th). The values are shown
together with their statistical and systematic uncertainties. pT(th)
dσ
dpT(th)
pT(th)
dσ
dpT(th)
[GeV]
[fb GeV−1]
[GeV]
[fb GeV−1]
0–45
204±4±18
225–270
106± 2 ±9
45–90
461±5±40
270–315
61± 2 ±6
90–135
430±5±41
315–400
27.4±0.9±2.5
135–180
292±4±27
400–800
3.2±0.2±0.3
180–225
179±3±17
— Table 18: Differential cross section at particle level as a function of |y(th)|. The values are shown
together with their statistical and systematic uncertainties. |y(th)|
dσ
d|y(th)| [pb]
|y(th)|
dσ
d|y(th)| [pb]
0–0.2
61.3±0.7±5.2
1–1.3
38.6±0.4±3.7
0.2–0.4
59.4±0.6±4.9
1.3–1.6
27.8±0.4±3.1
0.4–0.7
55.1±0.5±4.7
1.6–2.5
7.3±0.1±0.8
0.7–1
47.6±0.5±4.2
— Table 19: Differential cross section at particle level as a function of pT(tℓ). The values are shown
together with their statistical and systematic uncertainties. pT(tℓ)
dσ
dpT(tℓ)
pT(tℓ)
dσ
dpT(tℓ)
[GeV]
[fb GeV−1]
[GeV]
[fb GeV−1]
0–45
185±3±17
225–270
113± 2 ±9
45–90
425±4±41
270–315
67± 2 ±5
90–135
429±4±41
315–400
30.6±0.9±2.4
135–180
310±4±28
400–800
3.7±0.2±0.4
180–225
194±3±16
— Table 19: Differential cross section at particle level as a function of pT(tℓ). The values are shown
together with their statistical and systematic uncertainties. B
Tables of particle level cross sections. 42 Table 20: Differential cross section at particle level as a function of |y(tℓ)|. The values are shown
together with their statistical and systematic uncertainties. |y(tℓ)|
dσ
d|y(tℓ)| [pb]
|y(tℓ)|
dσ
d|y(tℓ)| [pb]
0–0.2
55.7±0.7±5.0
1–1.3
38.9±0.5±3.6
0.2–0.4
54.6±0.6±5.1
1.3–1.6
29.3±0.4±2.7
0.4–0.7
52.0±0.5±4.9
1.6–2.5
10.2±0.2±0.9
0.7–1
47.2±0.5±4.4
— Table 20: Differential cross section at particle level as a function of |y(tℓ)|. The values are shown
together with their statistical and systematic uncertainties. |y(tℓ)|
dσ
d|y(tℓ)| [pb]
|y(tℓ)|
dσ
d|y(tℓ)| [pb]
0–0.2
55.7±0.7±5.0
1–1.3
38.9±0.5±3.6
0.2–0.4
54.6±0.6±5.1
1.3–1.6
29.3±0.4±2.7
0.4–0.7
52.0±0.5±4.9
1.6–2.5
10.2±0.2±0.9
0.7–1
47.2±0.5±4.4
— Table 21: Differential cross section at particle level as a function of pT(tt). The values are shown
together with their statistical and systematic uncertainties. B
Tables of particle level cross sections. pT(tt)
dσ
dpT(tt)
pT(tt)
dσ
dpT(tt)
[GeV]
[fb GeV−1]
[GeV]
[fb GeV−1]
0–35
890±10±140
140–200
92± 3 ±10
35–80
577±10±62
200–500
18.4±0.5±1.7
80–140
219± 5 ±22
— Table 21: Differential cross section at particle level as a function of pT(tt). The values are shown
together with their statistical and systematic uncertainties. pT(tt)
dσ
d
(tt)
pT(tt)
dσ
d
(tt) Table 21: Differential cross section at particle level as a function of pT(tt). The values are shown
together with their statistical and systematic uncertainties. pT(tt)
dσ
dpT(tt)
pT(tt)
dσ
dpT(tt)
[GeV]
[fb GeV−1]
[GeV]
[fb GeV−1]
0–35
890±10±140
140–200
92± 3 ±10
35–80
577±10±62
200–500
18 4±0 5±1 7 e 21: Differential cross section at particle level as a function of pT(tt). The values are show
ther with their statistical and systematic uncertainties. Table 22: Differential cross section at particle level as a function of M(tt). The values are shown
together with their statistical and systematic uncertainties. Table 22: Differential cross section at particle level as a function of M(tt). The values are shown
together with their statistical and systematic uncertainties. M(tt)
dσ
dM(tt)
M(tt)
dσ
dM(tt)
[GeV]
[fb GeV−1]
[GeV]
[fb GeV−1]
300–375
124±4±14
625–740
91± 2 ±8
375–450
247±4±27
740–850
47± 2 ±4
450–530
200±4±22
850–1100
22.3±0.8±2.1
530–625
144±3±13
1100–2000
2.7±0.2±0.4 Table 23: Differential cross section at particle level as a function of |y(tt)|. The values are shown
together with their statistical and systematic uncertainties. Table 23: Differential cross section at particle level as a function of |y(tt)|. The values are shown
together with their statistical and systematic uncertainties. |y(tt)|
dσ
d|y(tt)| [pb]
|y(tt)|
dσ
d|y(tt)| [pb]
0–0.2
76.2±0.9±6.6
0.6–0.9
55.0±0.6±4.9
0.2–0.4
71.8±0.7±6.3
0.9–1.3
35.8±0.5±3.5
0.4–0.6
66.1±0.7±6.1
1.3–2.3
7.7±0.2±0.8 Table 24: Cross sections at particle level in bins of the number of additional jets. The values are
shown together with their statistical and systematic uncertainties. Additional jets
σ [pb]
Additional jets
σ [pb]
0
39.9±0.4±3.0
3
3.8± 0.1 ±0.6
1
25.6±0.3±2.7
≥4
1.75±0.07±0.36
2
10.6±0.2±1.3
— Table 24: Cross sections at particle level in bins of the number of additional jets. The va
shown together with their statistical and systematic uncertainties. 43 Table 25: Differential cross sections at particle level as a function of pT(th) in bins of the number
of additional jets. B
Tables of particle level cross sections. The values are shown together with their statistical and systematic uncertain-
ties pT(th)
dσ
dpT(th)
pT(th)
dσ
dpT(th)
[GeV]
[fb GeV−1]
[GeV]
[fb GeV−1]
Additional jets: 0
0–45
108± 3 ±7
225–270
44± 1 ±4
45–90
241± 4 ±16
270–315
22.7± 0.9 ±2.0
90–135
226± 3 ±16
315–400
9.7± 0.5 ±1.3
135–180
146± 3 ±10
400–800
1.09±0.09±0.15
180–225
84± 2 ±7
—
Additional jets: 1
0–45
60± 1 ±7
225–270
34.8± 0.9 ±3.6
45–90
136± 2 ±16
270–315
20.9± 0.7 ±2.6
90–135
129± 2 ±13
315–400
9.4± 0.4 ±0.8
135–180
92± 1 ±9
400–800
1.06±0.08±0.14
180–225
57± 1 ±6
—
Additional jets: 2
0–45
24.7±0.5±3.5
225–270
17.1± 0.5 ±2.1
45–90
55.8±0.9±7.7
270–315
10.7± 0.4 ±1.3
90–135
52.7±0.8±6.9
315–400
4.9± 0.3 ±0.6
135–180
38.4±0.7±4.6
400–800
0.60±0.05±0.08
180–225
26.0±0.6±3.1
—
Additional jets: ≥3
0–45
11.6±0.3±2.0
225–270
9.4± 0.4 ±1.4
45–90
25.9±0.6±4.4
270–315
6.5± 0.3 ±1.0
90–135
26.0±0.6±4.3
315–400
3.5± 0.2 ±0.6
135–180
19.2±0.5±2.8
400–800
0.47±0.05±0.07
180–225
13.5±0.4±1.8
— B
Tables of particle level cross sections. 44 B Table 26: Differential cross sections at particle level as a function of pT(tt) in bins of the number
of additional jets. The values are shown together with their statistical and systematic uncer-
tainties. pT(tt)
dσ
dpT(tt)
pT(tt)
dσ
dpT(tt)
[GeV]
[fb GeV−1]
[GeV]
[fb GeV−1]
Additional jets: 0
0–35
730±10±100
140–200
7± 1 ±2
35–80
268± 8 ±31
200–500
0.19±0.09±0.07
80–140
33± 3 ±8
—
Additional jets: 1
0–35
118± 5 ±19
140–200
45± 3 ±5
35–80
222± 5 ±26
200–500
6.6± 0.4 ±0.7
80–140
112± 4 ±12
—
Additional jets: 2
0–35
25± 2 ±5
140–200
26± 2 ±3
35–80
59± 3 ±10
200–500
6.8± 0.4 ±0.7
80–140
55± 2 ±8
—
Additional jets: ≥3
0–35
8.1±1.0±2.0
140–200
17± 1 ±4
35–80
23± 2 ±5
200–500
5.4± 0.3 ±0.8
80–140
22± 1 ±4
— 45 e 27: Double-differential cross section at particle level as a function of |y(th)| vs
values are shown together with their statistical and systematic uncertainties. B
Tables of particle level cross sections. pT(th)
d2σ
dpT(th)d|y(th)|
pT(th)
d2σ
dpT(th)d|y(th)|
[GeV]
[fb GeV−1]
[GeV]
[fb GeV−1]
0 < |y(th)| < 0.5
0–45
146± 2 ±12
225–270
78± 2 ±6
45–90
330± 4 ±28
270–315
46± 1 ±4
90–135
316± 4 ±26
315–400
21.8± 0.8 ±2.0
135–180
217± 3 ±18
400–800
2.7± 0.2 ±0.3
180–225
129± 2 ±11
—
0.5 < |y(th)| < 1
0–45
126± 2 ±13
225–270
63± 2 ±6
45–90
281± 3 ±25
270–315
36± 1 ±3
90–135
267± 3 ±23
315–400
16.4± 0.7 ±1.4
135–180
182± 3 ±15
400–800
2.2± 0.1 ±0.3
180–225
112± 2 ±10
—
1 < |y(th)| < 1.5
0–45
88± 2 ±9
225–270
44± 1 ±4
45–90
198± 3 ±21
270–315
25.3± 1.0 ±2.3
90–135
186± 3 ±18
315–400
11.1± 0.6 ±1.2
135–180
130± 2 ±12
400–800
0.99±0.09±0.11
180–225
77± 2 ±7
—
1.5 < |y(th)| < 2.5
0–45
21.9±0.8±3.3
225–270
12.9± 0.5 ±1.4
45–90
49± 1 ±6
270–315
7.0± 0.4 ±0.8
90–135
48± 1 ±5
315–400
2.9± 0.2 ±0.3
135–180
32.2±0.9±3.2
400–800
0.25±0.03±0.04
180–225
21.3±0.7±2.0
— Table 27: Double-differential cross section at particle level as a function of |y(th)| vs. pT(th). The values are shown together with their statistical and systematic uncertainties. B
Tables of particle level cross sections. 46 B Table 28: Double-differential cross section at particle level as a function of M(tt) vs. |y(tt)|. The
values are shown together with their statistical and systematic uncertainties. |y(tt)|
d2σ
dM(tt)d|y(tt)|
|y(tt)|
d2σ
dM(tt)d|y(tt)|
[fb GeV−1]
[fb GeV−1]
300 < M(tt) < 450 GeV
0–0.2
143± 3 ±12
0.6–0.9
124± 3 ±11
0.2–0.4
142± 3 ±12
0.9–1.3
96± 2 ±9
0.4–0.6
140± 3 ±12
1.3–2.3
25.7± 0.9 ±2.5
450 < M(tt) < 625 GeV
0–0.2
158± 3 ±15
0.6–0.9
118± 2 ±12
0.2–0.4
148± 3 ±15
0.9–1.3
75± 2 ±9
0.4–0.6
142± 3 ±14
1.3–2.3
15.5± 0.6 ±1.7
625 < M(tt) < 850 GeV
0–0.2
77± 2 ±6
0.6–0.9
47± 1 ±4
0.2–0.4
67± 2 ±6
0.9–1.3
27± 1 ±3
0.4–0.6
57± 2 ±5
1.3–2.3
4.3± 0.3 ±0.4
850 < M(tt) < 2000 GeV
0–0.2
8.4±0.4±0.9
0.6–0.9
4.7± 0.3 ±0.4
0.2–0.4
8.5±0.4±1.0
0.9–1.3
1.9± 0.1 ±0.2
0.4–0.6
6.7±0.3±0.7
1.3–2.3
0.20±0.03±0.03 Table 28: Double-differential cross section at particle level as a function of M(tt) vs. |y(tt)|. The
values are shown together with their statistical and systematic uncertainties. 47 Table 29: Double-differential cross section at particle level as a function of pT(tt) vs. M(tt). B
Tables of particle level cross sections. The
values are shown together with their statistical and systematic uncertainties. are shown together with their statistical and systematic uncertainties. M(tt)
d2σ
dpT(tt)dM(tt)
M(tt)
d2σ
dpT(tt)dM(tt)
[GeV]
[fb GeV−2]
[GeV]
[fb GeV−2]
0 < pT(tt) < 35 GeV
300–375
1.44± 0.05 ±0.09
625–740
0.88± 0.02 ±0.11
375–450
2.85± 0.06 ±0.41
740–850
0.48± 0.02 ±0.05
450–530
2.26± 0.05 ±0.40
850–1100
0.215± 0.010 ±0.040
530–625
1.47± 0.03 ±0.22
1100–2000
0.030± 0.003 ±0.012
35 < pT(tt) < 80 GeV
300–375
0.89± 0.02 ±0.09
625–740
0.66± 0.01 ±0.06
375–450
1.76± 0.03 ±0.20
740–850
0.36± 0.01 ±0.03
450–530
1.44± 0.02 ±0.16
850–1100
0.158± 0.006 ±0.020
530–625
1.03± 0.02 ±0.09
1100–2000
0.018± 0.001 ±0.004
80 < pT(tt) < 140 GeV
300–375
0.31± 0.01 ±0.03
625–740
0.249± 0.007 ±0.021
375–450
0.67± 0.02 ±0.08
740–850
0.137± 0.005 ±0.016
450–530
0.55± 0.01 ±0.06
850–1100
0.059± 0.003 ±0.007
530–625
0.395±0.010±0.036
1100–2000
0.0066±0.0007±0.0018
140 < pT(tt) < 500 GeV
300–375
0.035±0.002±0.007
625–740
0.039± 0.001 ±0.004
375–450
0.081±0.002±0.009
740–850
0.022± 0.001 ±0.003
450–530
0.077±0.002±0.008
850–1100
0.0107±0.0006±0.0009
530–625
0.061±0.002±0.008
1100–2000
0.0016±0.0002±0.0002 B
Tables of particle level cross sections. 48 B 49 Institut f¨ur Hochenergiephysik, Wien, Austria Institut f¨ur Hochenergiephysik, Wien, Austria
W. Adam, E. Asilar, T. Bergauer, J. Brandstetter, E. Brondolin, M. Dragicevic, J. Er¨o, M. Flechl,
M. Friedl, R. Fr¨uhwirth1, V.M. Ghete, C. Hartl, N. H¨ormann, J. Hrubec, M. Jeitler1, A. K¨onig,
I. Kr¨atschmer, D. Liko, T. Matsushita, I. Mikulec, D. Rabady, N. Rad, B. Rahbaran, H. Rohringer,
J. Schieck1, J. Strauss, W. Waltenberger, C.-E. Wulz1 Institute for Nuclear Problems, Minsk, Belarus
O. Dvornikov, V. Makarenko, V. Zykunov National Centre for Particle and High Energy Physics, Minsk, Belarus
V. Mossolov, N. Shumeiko, J. Suarez Gonzalez Universiteit Antwerpen, Antwerpen, Belgium Universiteit Antwerpen, Antwerpen, Belgium
S. Alderweireldt, E.A. De Wolf, X. Janssen, J. Lauwers, M. Van De Klundert, H. Van
Haevermaet, P. Van Mechelen, N. Van Remortel, A. Van Spilbeeck Vrije Universiteit Brussel, Brussel, Belgium
S. Abu Zeid, F. Blekman, J. D’Hondt, N. Daci, I. De Bruyn, K. Deroover, S. Lowette, S. Moortgat,
L. Moreels, A. Olbrechts, Q. Python, S. Tavernier, W. Van Doninck, P. Van Mulders, I. Van Parijs Universit´e Libre de Bruxelles, Bruxelles, Belgium Universit´e Libre de Bruxelles, Bruxelles, Belgium
H. Brun, B. Clerbaux, G. De Lentdecker, H. Delannoy, G. Fasanella, L. Favart, R. Goldouzian,
A. Grebenyuk, G. Karapostoli, T. Lenzi, A. L´eonard, J. Luetic, T. Maerschalk, A. Marinov,
A. Randle-conde, T. Seva, C. Vander Velde, P. Vanlaer, R. Yonamine, F. Zenoni, F. Zhang2 Ghent University, Ghent, Belgium
A. Cimmino, T. Cornelis, D. Dobur, A. Fagot, G. Garcia, M. Gul, I. Khvastunov, D. Poyraz,
S. Salva, R. Sch¨ofbeck, A. Sharma, M. Tytgat, W. Van Driessche, E. Yazgan, N. Zaganidis Universit´e Catholique de Louvain, Louvain-la-Neuve, Belgium
H. Bakhshiansohi, C. Beluffi3, O. Bondu, S. Brochet, G. Bruno, A. Caudron, S. De Visscher,
C. Delaere, M. Delcourt, B. Francois, A. Giammanco, A. Jafari, P. Jez, M. Komm, V. Lemaitre,
A
M
itt
i
A
M
t
M
M
i h
C
N tt
K
Pi t
k
ki
L
Q
t
t Universit´e Catholique de Louvain, Louvain-la-Neuve, Belgium
H. Bakhshiansohi, C. Beluffi3, O. Bondu, S. Brochet, G. Bruno, A. Caudron, S. De Visscher,
C. Delaere, M. Delcourt, B. Francois, A. Giammanco, A. Jafari, P. Jez, M. Komm, V. Lemaitre,
A. Magitteri, A. Mertens, M. Musich, C. Nuttens, K. Piotrzkowski, L. Quertenmont,
M. Selvaggi, M. Vidal Marono, S. Wertz A. Magitteri, A. Mertens, M. Musich, C. Nuttens, K. Piotrzkowski, L. Quertenmont,
M. Selvaggi, M. Vidal Marono, S. Wertz C
The CMS Collaboration Yerevan Physics Institute, Yerevan, Armenia
V. Khachatryan, A.M. Sirunyan, A. Tumasyan Universit´e de Mons, Mons, Belgium
N B li Centro Brasileiro de Pesquisas Fisicas, Rio de Janeiro, Brazil
W.L. Ald´a J´unior, F.L. Alves, G.A. Alves, L. Brito, C. Hensel, A. Moraes, M.E. Pol, P. Rebello
Teles Universidade do Estado do Rio de Janeiro, Rio de Janeiro, Brazil
E. Belchior Batista Das Chagas, W. Carvalho, J. Chinellato4, A. Cust´odio, E.M. Da Costa,
G.G. Da Silveira5, D. De Jesus Damiao, C. De Oliveira Martins, S. Fonseca De Souza,
L.M. Huertas Guativa, H. Malbouisson, D. Matos Figueiredo, C. Mora Herrera, L. Mundim,
H. Nogima, W.L. Prado Da Silva, A. Santoro, A. Sznajder, E.J. Tonelli Manganote4, A. Vilela
Pereira Universidade Estadual Paulista a, Universidade Federal do ABC b, S˜ao Paulo, Brazil
S. Ahujaa, C.A. Bernardesb, S. Dograa, T.R. Fernandez Perez Tomeia, E.M. Gregoresb, Universidade Estadual Paulista a, Universidade Federal do ABC b, S˜ao Paulo, Brazil
S. Ahujaa, C.A. Bernardesb, S. Dograa, T.R. Fernandez Perez Tomeia, E.M. Gregoresb, Universidade Estadual Paulista , Universidade Federal do ABC , Sao Paulo, Brazil
S. Ahujaa, C.A. Bernardesb, S. Dograa, T.R. Fernandez Perez Tomeia, E.M. Gregoresb, 50 C
The CMS Collaboration P.G. Mercadanteb, C.S. Moona, S.F. Novaesa, Sandra S. Padulaa, D. Romero Abadb, J.C. Ruiz
Vargas P.G. Mercadanteb, C.S. Moona, S.F. Novaesa, Sandra S. Padulaa, D. Romero Abadb, J.C. Ruiz
Vargas Institute for Nuclear Research and Nuclear Energy, Sofia, Bulgaria
A. Aleksandrov, R. Hadjiiska, P. Iaydjiev, M. Rodozov, S. Stoykova, G. Sultanov, M. Vutova University of Sofia, Sofia, Bulgaria
A. Dimitrov, I. Glushkov, L. Litov, B. Pavlov, P. Petkov University of Sofia, Sofia, Bulgaria
A. Dimitrov, I. Glushkov, L. Litov, B. Pavlov, P. Petkov Beihang University, Beijing, China
6 Institute of High Energy Physics, Beijing, China M. Ahmad, J.G. Bian, G.M. Chen, H.S. Chen, M. Chen, Y. Chen7, T. Cheng, C.H. Jiang,
D. Leggat, Z. Liu, F. Romeo, S.M. Shaheen, A. Spiezia, J. Tao, C. Wang, Z. Wang, H. Zhang,
J. Zhao State Key Laboratory of Nuclear Physics and Technology, Peking University, Beijing, China
Y. Ban, G. Chen, Q. Li, S. Liu, Y. Mao, S.J. Qian, D. Wang, Z. Xu Universidad de Los Andes, Bogota, Colombia
C. Avila, A. Cabrera, L.F. Chaparro Sierra, C. Florez, J.P. Gomez, C.F. Gonz´alez Hern´andez,
J.D. Ruiz Alvarez, J.C. Sanabria a, A. Cabrera, L.F. Chaparro Sierra, C. Florez, J.P. Gomez, C.F. Gonz´alez Hern´andez,
z Alvarez, J.C. Sanabria University of Split, Faculty of Electrical Engineering, Mechanical Engineering and Naval
Architecture, Split, Croatia
N. Godinovic, D. Lelas, I. Puljak, P.M. Ribeiro Cipriano, T. Sculac University of Split, Faculty of Science, Split, Croatia
Z Antunovic M Kovac University of Split, Faculty of Science, Split, Croatia
Z. Antunovic, M. Kovac University of Split, Faculty of Science, Split, Croatia
Z. Antunovic, M. Kovac Institute Rudjer Boskovic, Zagreb, Croatia
V. Brigljevic, D. Ferencek, K. Kadija, S. Micanovic, L. Sudic, T. Susa Institute Rudjer Boskovic, Zagreb, Croatia
V. Brigljevic, D. Ferencek, K. Kadija, S. Micanovic, L. Sudic, T. Susa University of Cyprus, Nicosia, Cyprus
A. Attikis, G. Mavromanolakis, J. Mousa, C. Nicolaou, F. Ptochos, P.A. Razis, H. Rykaczewski,
D. Tsiakkouri Charles University, Prague, Czech Republic
M. Finger8, M. Finger Jr.8 Charles University, Prague, Czech Republic
M. Finger8, M. Finger Jr.8 Universidad San Francisco de Quito, Quito, Ecuador
E. Carrera Jarrin Universidad San Francisco de Quito, Quito, Ecuador
E. Carrera Jarrin Academy of Scientific Research and Technology of the Arab Republic of Egypt, Egyptian
Network of High Energy Physics, Cairo, Egypt
Y. Assran9,10, T. Elkafrawy11, A. Mahrous12 National Institute of Chemical Physics and Biophysics, Tallinn, Estonia
B. Calpas, M. Kadastik, M. Murumaa, L. Perrini, M. Raidal, A. Tiko, C. Veelken Department of Physics, University of Helsinki, Helsinki, Finland
P. Eerola, J. Pekkanen, M. Voutilainen Helsinki Institute of Physics, Helsinki, Finland
J. H¨ark¨onen, T. J¨arvinen, V. Karim¨aki, R. Kinnunen, T. Lamp´en, K. Lassila-Perini, S. Lehti,
T. Lind´en, P. Luukka, J. Tuominiemi, E. Tuovinen, L. Wendland 51 Georgian Technical University, Tbilisi, Georgia
T. Toriashvili15 Georgian Technical University, Tbilisi, Georgia
T. Toriashvili15 Lappeenranta University of Technology, Lappeenranta, Finland
J. Talvitie, T. Tuuva IRFU, CEA, Universit´e Paris-Saclay, Gif-sur-Yvette, France M. Besancon, F. Couderc, M. Dejardin, D. Denegri, B. Fabbro, J.L. Faure, C. Favaro, F. Ferri,
S. Ganjour, S. Ghosh, A. Givernaud, P. Gras, G. Hamel de Monchenault, P. Jarry, I. Kucher,
E. Locci, M. Machet, J. Malcles, J. Rander, A. Rosowsky, M. Titov, A. Zghiche Laboratoire Leprince-Ringuet, Ecole Polytechnique, IN2P3-CNRS, Palaiseau, France
A. Abdulsalam, I. Antropov, S. Baffioni, F. Beaudette, P. Busson, L. Cadamuro, E. Chapon,
C. Charlot, O. Davignon, R. Granier de Cassagnac, M. Jo, S. Lisniak, P. Min´e, M. Nguyen,
C. Ochando, G. Ortona, P. Paganini, P. Pigard, S. Regnard, R. Salerno, Y. Sirois, T. Strebler,
Y. Yilmaz, A. Zabi Institut Pluridisciplinaire Hubert Curien, Universit´e de Strasbourg, Universit´e de Haute
Alsace Mulhouse, CNRS/IN2P3, Strasbourg, France ,
,
g,
J.-L. Agram13, J. Andrea, A. Aubin, D. Bloch, J.-M. Brom, M. Buttignol, E.C. Chabert,
N. Chanon, C. Collard, E. Conte13, X. Coubez, J.-C. Fontaine13, D. Gel´e, U. Goerlach, A.-C. Le
Bihan, K. Skovpen, P. Van Hove Centre de Calcul de l’Institut National de Physique Nucleaire et de Physique des Particules,
CNRS/IN2P3, Villeurbanne, France
S. Gadrat Centre de Calcul de l’Institut National de Physique Nucleaire et de Physique des Particules,
CNRS/IN2P3, Villeurbanne, France
S. Gadrat Universit´e de Lyon, Universit´e Claude Bernard Lyon 1, CNRS-IN2P3, Institut de Physique
Nucl´eaire de Lyon, Villeurbanne, France Universit´e de Lyon, Universit´e Claude Bernard Lyon 1, CNRS-IN2P3, Institut de Physique
Nucl´eaire de Lyon, Villeurbanne, France
S. Beauceron, C. Bernet, G. Boudoul, E. Bouvier, C.A. Carrillo Montoya, R. Chierici,
D C
d
B C
b
P D
H El M
i J F
J F
S G
M G
i h Universit´e de Lyon, Universit´e Claude Bernard Lyon 1, CNRS-IN2P3, Institut de Physique
Nucl´eaire de Lyon, Villeurbanne, France
S. Beauceron, C. Bernet, G. Boudoul, E. Bouvier, C.A. Carrillo Montoya, R. Chierici,
D. Contardo, B. Courbon, P. Depasse, H. El Mamouni, J. Fan, J. Fay, S. Gascon, M. Gouzevitch,
G. Grenier, B. Ille, F. Lagarde, I.B. Laktineh, M. Lethuillier, L. Mirabito, A.L. Pequegnot,
S. Perries, A. Popov14, D. Sabes, V. Sordini, M. Vander Donckt, P. Verdier, S. Viret S. Beauceron, C. Bernet, G. Boudoul, E. Bouvier, C.A. Carrillo Montoya, R. Chierici,
D. Contardo, B. Courbon, P. Depasse, H. El Mamouni, J. Fan, J. Fay, S. Gascon, M. Gouzevitch,
G. Grenier, B. Ille, F. Lagarde, I.B. Laktineh, M. Lethuillier, L. Mirabito, A.L. Pequegnot,
S. Perries, A. Popov14, D. Sabes, V. Sordini, M. Vander Donckt, P. Verdier, S. Viret Georgian Technical University, Tbilisi, Georgia
T. Toriashvili15 University of Hamburg, Hamburg, Germany y
g
g
y
V. Blobel, M. Centis Vignali, A.R. Draeger, T. Dreyer, E. Garutti, D. Gonzalez, J. Haller,
M. Hoffmann, A. Junkes, R. Klanner, R. Kogler, N. Kovalchuk, T. Lapsien, T. Lenz,
I. Marchesini, D. Marconi, M. Meyer, M. Niedziela, D. Nowatschin, F. Pantaleo16, T. Peiffer,
A. Perieanu, J. Poehlsen, C. Sander, C. Scharf, P. Schleper, A. Schmidt, S. Schumann,
J. Schwandt, H. Stadie, G. Steinbr¨uck, F.M. Stober, M. St¨over, H. Tholen, D. Troendle, E. Usai,
L. Vanelderen, A. Vanhoefer, B. Vormwald Tbilisi State University, Tbilisi, Georgia RWTH Aachen University, I. Physikalisches Institut, Aachen, Germany
C. Autermann, S. Beranek, L. Feld, A. Heister, M.K. Kiesel, K. Klein, M. Lipinski, A. Ostapchuk,
M. Preuten, F. Raupach, S. Schael, C. Schomakers, J. Schulz, T. Verlage, H. Weber RWTH Aachen University, I. Physikalisches Institut, Aachen, Germany
C. Autermann, S. Beranek, L. Feld, A. Heister, M.K. Kiesel, K. Klein, M. Lipinski, A. Ostapchuk,
M. Preuten, F. Raupach, S. Schael, C. Schomakers, J. Schulz, T. Verlage, H. Weber ,
,
,
,
,
,
p
,
M. Preuten, F. Raupach, S. Schael, C. Schomakers, J. Schulz, T. Verlage, H. Weber RWTH Aachen University, III. Physikalisches Institut A, Aachen, Germany
A. Albert, M. Brodski, E. Dietz-Laursonn, D. Duchardt, M. Endres, M. Erdmann, S. Erdweg,
T. Esch, R. Fischer, A. G¨uth, M. Hamer, T. Hebbeker, C. Heidemann, K. Hoepfner, S. Knutzen,
M. Merschmeyer, A. Meyer, P. Millet, S. Mukherjee, M. Olschewski, K. Padeken, T. Pook,
M. Radziej, H. Reithler, M. Rieger, F. Scheuch, L. Sonnenschein, D. Teyssier, S. Th¨uer RWTH Aachen University, III. Physikalisches Institut A, Aachen, Germany
A. Albert, M. Brodski, E. Dietz-Laursonn, D. Duchardt, M. Endres, M. Erdmann, S. Erdweg,
T. Esch, R. Fischer, A. G¨uth, M. Hamer, T. Hebbeker, C. Heidemann, K. Hoepfner, S. Knutzen,
M. Merschmeyer, A. Meyer, P. Millet, S. Mukherjee, M. Olschewski, K. Padeken, T. Pook,
M. Radziej, H. Reithler, M. Rieger, F. Scheuch, L. Sonnenschein, D. Teyssier, S. Th¨uer RWTH Aachen University, III. Physikalisches Institut B, Aachen, Germany
V. Cherepanov, G. Fl¨ugge, F. Hoehle, B. Kargoll, T. Kress, A. K¨unsken, J. Lingemann, T. M¨uller,
A. Nehrkorn, A. Nowack, I.M. Nugent, C. Pistone, O. Pooth, A. Stahl16 RWTH Aachen University, III. Physikalisches Institut B, Aachen, Germany
V. Cherepanov, G. Fl¨ugge, F. Hoehle, B. Kargoll, T. Kress, A. K¨unsken, J. Lingemann, T. M¨uller,
A. Nehrkorn, A. Nowack, I.M. Nugent, C. Pistone, O. Pooth, A. Stahl16 Deutsches Elektronen-Synchrotron, Hamburg, Germany
M. Aldaya Martin, T. Arndt, C. Asawatangtrakuldee, K. Beernaert, O. Behnke, U. Behrens,
A.A. Bin Anuar, K. Borras17, A. Campbell, P. Connor, C. Contreras-Campana, F. Costanza,
C. Diez Pardos, G. Dolinska, G. Eckerlin, D. Eckstein, T. Eichhorn, E. Eren, E. Gallo18, C
The CMS Collaboration 52 J. Garay Garcia, A. Geiser, A. Gizhko, J.M. Grados Luyando, P. Gunnellini, A. Harb,
J. Hauk, M. Hempel19, H. Jung, A. Kalogeropoulos, O. Karacheban19, M. Kasemann,
J. Keaveney, C. Kleinwort, I. Korol, D. Kr¨ucker, W. Lange, A. Lelek, J. Leonard, K. Lipka,
A. Tbilisi State University, Tbilisi, Georgia Lobanov, W. Lohmann19, R. Mankel, I.-A. Melzer-Pellmann, A.B. Meyer, G. Mittag, J. Mnich,
A. Mussgiller, E. Ntomari, D. Pitzl, R. Placakyte, A. Raspereza, B. Roland, M. ¨O. Sahin,
P. Saxena, T. Schoerner-Sadenius, C. Seitz, S. Spannagel, N. Stefaniuk, G.P. Van Onsem,
R. Walsh, C. Wissing g
y
p
P. Saxena, T. Schoerner-Sadenius, C. Seitz, S. Spannagel, N. Stefaniuk, G.P. Van Onsem,
R. Walsh, C. Wissing Indian Institute of Technology Madras, Madras, India
P.K. Behera Bhabha Atomic Research Centre, Mumbai, India
R. Chudasama, D. Dutta, V. Jha, V. Kumar, A.K. Mohanty16, P.K. Netrakanti, L.M. Pant,
P. Shukla, A. Topkar Tata Institute of Fundamental Research-A, Mumbai, India
T. Aziz, S. Dugad, G. Kole, B. Mahakud, S. Mitra, G.B. Mohan Tata Institute of Fundamental Research-A, Mumbai, India
T. Aziz, S. Dugad, G. Kole, B. Mahakud, S. Mitra, G.B. Mohanty, B. Parida, N. Sur, B. Sutar Tata Institute of Fundamental Research-B, Mumbai, India
S. Banerjee, S. Bhowmik25, R.K. Dewanjee, S. Ganguly, M. Guchait, Sa. Jain, S. Kumar,
M. Maity25, G. Majumder, K. Mazumdar, T. Sarkar25, N. Wickramage26 Indian Institute of Science Education and Research (IISER), Pune, India
S. Chauhan, S. Dube, V. Hegde, A. Kapoor, K. Kothekar, A. Rane, S. Sharma Institute for Research in Fundamental Sciences (IPM), Tehran, Iran
H. Behnamian, S. Chenarani27, E. Eskandari Tadavani, S.M. Etesami27, A. Fahim28, M. Khakzad,
M. Mohammadi Najafabadi, M. Naseri, S. Paktinat Mehdiabadi29, F. Rezaei Hosseinabadi,
B. Safarzadeh30, M. Zeinali Institut f¨ur Experimentelle Kernphysik, Karlsruhe, Germany Chawla, U.Bhawandeep, A.K. Kalsi, A. Kaur, M. Kaur,
R. Kumar, P. Kumari, A. Mehta, M. Mittal, J.B. Singh, G. Walia
University of Delhi, Delhi, India
Ashok Kumar, A. Bhardwaj, B.C. Choudhary, R.B. Garg, S. Keshri, S. Malhotra, M. Naimuddin,
N. Nishu, K. Ranjan, R. Sharma, V. Sharma Panjab University, Chandigarh, India
S. Bansal, S.B. Beri, V. Bhatnagar, R. Chawla, U.Bhawandeep, A.K. Kalsi, A. Kaur, M. Kaur,
R. Kumar, P. Kumari, A. Mehta, M. Mittal, J.B. Singh, G. Walia
U i
it
f D lhi D lhi I di Panjab University, Chandigarh, India
S. Bansal, S.B. Beri, V. Bhatnagar, R. Chawla, U.Bhawandeep, A.K. Kalsi, A. Kaur, M. Kaur,
R. Kumar, P. Kumari, A. Mehta, M. Mittal, J.B. Singh, G. Walia
University of Delhi, Delhi, India
Ashok Kumar, A. Bhardwaj, B.C. Choudhary, R.B. Garg, S. Keshri, S. Malhotra, M. Naimuddin,
N. Nishu, K. Ranjan, R. Sharma, V. Sharma University of Delhi, Delhi, India
Ashok Kumar, A. Bhardwaj, B.C. Choudhary, R.B. Garg, S. Keshri, S. Malhotra, M. Naimuddin,
N. Nishu, K. Ranjan, R. Sharma, V. Sharma University of Delhi, Delhi, India
Ashok Kumar, A. Bhardwaj, B.C. Choudhary, R.B. Garg, S. Keshri, S. Malhotra, M. Naimuddin,
N. Nishu, K. Ranjan, R. Sharma, V. Sharma Saha Institute of Nuclear Physics, Kolkata, India
R. Bhattacharya, S. Bhattacharya, K. Chatterjee, S. Dey, S. Dutt, S. Dutta, S. Ghosh,
N. Majumdar, A. Modak, K. Mondal, S. Mukhopadhyay, S. Nandan, A. Purohit, A. Roy, D. Roy,
S. Roy Chowdhury, S. Sarkar, M. Sharan, S. Thakur Institut f¨ur Experimentelle Kernphysik, Karlsruhe, Germany p
p y
y
M. Akbiyik, C. Barth, S. Baur, C. Baus, J. Berger, E. Butz, R. Caspart, T. Chwalek, F. Colombo,
W. De Boer, A. Dierlamm, S. Fink, B. Freund, R. Friese, M. Giffels, A. Gilbert, P. Goldenzweig,
D. Haitz, F. Hartmann16, S.M. Heindl, U. Husemann, I. Katkov14, S. Kudella, P. Lobelle Pardo,
H. Mildner, M.U. Mozer, Th. M¨uller, M. Plagge, G. Quast, K. Rabbertz, S. R¨ocker, F. Roscher,
M. Schr¨oder, I. Shvetsov, G. Sieber, H.J. Simonis, R. Ulrich, J. Wagner-Kuhr, S. Wayand,
M. Weber, T. Weiler, S. Williamson, C. W¨ohrmann, R. Wolf Institute of Nuclear and Particle Physics (INPP),
NCSR Demokritos, Aghia Paraskevi,
Greece G. Anagnostou, G. Daskalakis, T. Geralis, V.A. Giakoumopoulou, A. Kyriakis, D. Loukas,
I. Topsis-Giotis National and Kapodistrian University of Athens, Athens, Greece
S. Kesisoglou, A. Panagiotou, N. Saoulidou, E. Tziaferi National and Kapodistrian University of Athens, Athens, Greece
S Kesisoglou A Panagiotou N Saoulidou E Tziaferi National and Kapodistrian University of Athens, Athens, Greece National and Kapodistrian University of Athens, Athen
S. Kesisoglou, A. Panagiotou, N. Saoulidou, E. Tziaferi S. Kesisoglou, A. Panagiotou, N. Saoulidou, E. Tziaferi University of Io´annina, Io´annina, Greece
I. Evangelou, G. Flouris, C. Foudas, P. Kokkas, N. Loukas, N. Manthos, I. Papadopoulos,
E. Paradas MTA-ELTE Lend¨ulet CMS Particle and Nuclear Physics Group, E¨otv¨os Lor´and University,
Budapest, Hungary
N. Filipovic Wigner Research Centre for Physics, Budapest, Hungary
G. Bencze, C. Hajdu, P. Hidas, D. Horvath20, F. Sikler, V. Veszpremi, G. Vesztergombi21,
A.J. Zsigmond Institute of Nuclear Research ATOMKI, Debrecen, Hungary
N. Beni, S. Czellar, J. Karancsi22, A. Makovec, J. Molnar, Z. Szillasi University of Debrecen, Debrecen, Hungary
M. Bart´ok21, P. Raics, Z.L. Trocsanyi, B. Ujvari University of Debrecen, Debrecen, Hungary
M. Bart´ok21, P. Raics, Z.L. Trocsanyi, B. Ujvari National Institute of Science Education and Research, Bhubaneswar, India
S. Bahinipati, S. Choudhury23, P. Mal, K. Mandal, A. Nayak24, D.K. Sahoo, N. Sahoo, S.K. Swain National Institute of Science Education and Research, Bhubaneswar, India National Institute of Science Education and Research, Bhubaneswar, India
S. Bahinipati, S. Choudhury23, P. Mal, K. Mandal, A. Nayak24, D.K. Sahoo, N. Sahoo, S.K. Swain National Institute of Science Education and Research, Bhubaneswar, India
S. Bahinipati, S. Choudhury23, P. Mal, K. Mandal, A. Nayak24, D.K. Sahoo, N. Sahoo, S.K. Swain S. Bahinipati, S. Choudhury23, P. Mal, K. Mandal, A. Nayak24, D.K. Sahoo, N. Sahoo, S.K. Swain 53 Panjab University, Chandigarh, India
S. Bansal, S.B. Beri, V. Bhatnagar, R. University College Dublin, Dublin, Ireland
M. Felcini, M. Grunewald INFN Sezione di Bari a, Universit`a di Bari b, Politecnico di Bari c, Bari, Italy
M. Abbresciaa,b, C. Calabriaa,b, C. Caputoa,b, A. Colaleoa, D. Creanzaa,c, L. Cristellaa,b, N. De
Filippisa,c, M. De Palmaa,b, L. Fiorea, G. Iasellia,c, G. Maggia,c, M. Maggia, G. Minielloa,b,
S. Mya,b, S. Nuzzoa,b, A. Pompilia,b, G. Pugliesea,c, R. Radognaa,b, A. Ranieria, G. Selvaggia,b,
L. Silvestrisa,16, R. Vendittia,b, P. Verwilligena INFN Sezione di Bologna a, Universit`a di Bologna b, Bologna, Italy
G. Abbiendia, C. Battilana, D. Bonacorsia,b, S. Braibant-Giacomellia,b, L. Brigliadoria,b,
R. Campaninia,b, P. Capiluppia,b, A. Castroa,b, F.R. Cavalloa, S.S. Chhibraa,b, G. Codispotia,b,
M. Cuffiania,b, G.M. Dallavallea, F. Fabbria, A. Fanfania,b, D. Fasanellaa,b, P. Giacomellia,
C. Grandia, L. Guiduccia,b, S. Marcellinia, G. Masettia, A. Montanaria, F.L. Navarriaa,b,
A. Perrottaa, A.M. Rossia,b, T. Rovellia,b, G.P. Sirolia,b, N. Tosia,b,16 INFN Sezione di Catania a, Universit`a di Catania b, Catania, Italy
b
b
b
b INFN Sezione di Catania a, Universit`a di Catania b, Catania, Italy
S. Albergoa,b, M. Chiorbolia,b, S. Costaa,b, A. Di Mattiaa, F. Giordanoa,b, R. Potenzaa,b,
A. Tricomia,b, C. Tuvea,b g
A. Tricomia,b, C. Tuvea,b 54 C
The CMS Collaboration INFN Sezione di Firenze a, Universit`a di Firenze b, Firenze, Italy
G. Barbaglia, V. Ciullia,b, C. Civininia, R. D’Alessandroa,b, E. Focardia,b, V. Goria,b, P. Lenzia,b,
M. Meschinia, S. Paolettia, G. Sguazzonia, L. Viliania,b,16 INFN Sezione di Firenze a, Universit`a di Firenze b, Firenze, Italy INFN Sezione di Firenze a, Universita di Firenze b, Firenze, Italy
G. Barbaglia, V. Ciullia,b, C. Civininia, R. D’Alessandroa,b, E. Focardia,b, V. Goria,b, P. Lenzia,b,
M M
hi ia S P
l ttia G S
ia L Vili
ia b 16 y
G. Barbaglia, V. Ciullia,b, C. Civininia, R. D’Alessandroa,b, E. Focardia,b, V. Go M. Meschinia, S. Paolettia, G. Sguazzonia, L. Viliania,b,16 INFN Laboratori Nazionali di Frascati, Frascati, Italy
L. Benussi, S. Bianco, F. Fabbri, D. Piccolo, F. Primavera16 INFN Sezione di Genova a, Universit`a di Genova b, Genova, Italy
V. Calvellia,b, F. Ferroa, M. Lo Veterea,b, M.R. Mongea,b, E. Robuttia, S. Tosia,b INFN Sezione di Milano-Bicocca a, Universit`a di Milano-Bicocca b, Milano, Italy
L. Brianza16, M.E. Dinardoa,b, S. Fiorendia,b, S. Gennaia, A. Ghezzia,b, P. Govonia,b, M. Malberti,
S. Malvezzia, R.A. Manzonia,b,16, D. Menascea, L. Moronia, M. Paganonia,b, D. Pedrinia,
b
b INFN Sezione di Milano-Bicocca a, Universit`a di Milano-Bicocca b, Milano, Italy
L. Brianza16, M.E. Dinardoa,b, S. Fiorendia,b, S. Gennaia, A. Ghezzia,b, P. Govonia,b, M. Malberti, ,
,
,
y
L. University College Dublin, Dublin, Ireland
M. Felcini, M. Grunewald Brianza16, M.E. Dinardoa,b, S. Fiorendia,b, S. Gennaia, A. Ghezzia,b, P. Govonia,b, M. Malberti,
S. Malvezzia, R.A. Manzonia,b,16, D. Menascea, L. Moronia, M. Paganonia,b, D. Pedrinia,
S. Pigazzini, S. Ragazzia,b, T. Tabarelli de Fatisa,b S. Malvezzia, R.A. Manzonia,b,16, D. Menascea, L. Moronia, M. Paganonia,b, D. Pedrinia,
S. Pigazzini, S. Ragazzia,b, T. Tabarelli de Fatisa,b INFN Sezione di Napoli a, Universit`a di Napoli ’Federico II’ b, Napoli, Italy, Universit`a della
Basilicata c, Potenza, Italy, Universit`a G. Marconi d, Roma, Italy
S. Buontempoa, N. Cavalloa,c, G. De Nardo, S. Di Guidaa,d,16, M. Espositoa,b, F. Fabozzia,c,
F. Fiengaa,b, A.O.M. Iorioa,b, G. Lanzaa, L. Listaa, S. Meolaa,d,16, P. Paoluccia,16, C. Sciaccaa,b,
F Thyssen F. Fiengaa,b, A.O.M. Iorioa,b, G. Lanzaa, L. Listaa, S. Meolaa,d,16, P. Paoluccia,16, C. Sciaccaa,b,
F. Thyssen INFN Sezione di Padova a, Universit`a di Padova b, Padova, Italy, Universit`a di Trento c,
Trento, Italy
P. Azzia,16, N. Bacchettaa, L. Benatoa,b, D. Biselloa,b, A. Bolettia,b, R. Carlina,b, A. Car-
lh
A t
D
Oli
i
a b
P Ch
hi a
M
D ll’O
a b
T
D
i
a
U
D
llia INFN Sezione di Padova a, Universit`a di Padova b, Padova, Italy, Universit`a di Trento c,
Trento, Italy
P. Azzia,16, N. Bacchettaa, L. Benatoa,b, D. Biselloa,b, A. Bolettia,b, R. Carlina,b, A. Car-
valho Antunes De Oliveiraa,b, P. Checchiaa, M. Dall’Ossoa,b, T. Dorigoa, U. Dossellia,
F. Gasparinia,b, U. Gasparinia,b, A. Gozzelinoa, M. Gulminia,31, S. Lacapraraa, M. Margonia,b,
A.T. Meneguzzoa,b, J. Pazzinia,b, N. Pozzobona,b, P. Ronchesea,b, F. Simonettoa,b, E. Torassaa,
S Venturaa M Zanetti P Zottoa,b valho Antunes De Oliveira
, P. Checchia , M. Dall Osso
, T. Dorigo , U. Dosselli ,
F. Gasparinia,b, U. Gasparinia,b, A. Gozzelinoa, M. Gulminia,31, S. Lacapraraa, M. Margonia,b,
A.T. Meneguzzoa,b, J. Pazzinia,b, N. Pozzobona,b, P. Ronchesea,b, F. Simonettoa,b, E. Torassaa,
S. Venturaa, M. Zanetti, P. Zottoa,b p
p
p
g
A.T. Meneguzzoa,b, J. Pazzinia,b, N. Pozzobona,b, P. Ronchesea,b, F. Simonettoa,b, E. Torassaa,
S Venturaa M Zanetti P Zottoa,b S. Venturaa, M. Zanetti, P. Zottoa,b S. Venturaa, M. Zanetti, P. Zottoa,b INFN Sezione di Pavia a, Universit`a di Pavia b, Pavia, Italy
A. Braghieria, A. Magnania,b, P. Montagnaa,b, S.P. Rattia,b, V. Rea, C. Riccardia,b, P. Salvinia,
I. Vaia,b, P. Vituloa,b INFN Sezione di Perugia a, Universit`a di Perugia b, Perugia, Italy
L. Alunni Solestizia,b, G.M. Bileia, D. Ciangottinia,b, L. Fan`oa,b, P. Laricciaa,b, R. Leonardia,b,
G. Mantovania,b, M. Menichellia, A. Sahaa, A. Santocchiaa,b G. Mantovania,b, M. Menichellia, A. A. Romeroa,b, M. Ruspaa,c, R. Sacchia,b, K. Shchelinaa,b, V. Solaa, A. Solanoa,b, A. Staianoa,
P. Traczyka,b INFN Sezione di Trieste a, Universit`a di Trieste b, Trieste, Italy
S. Belfortea, M. Casarsaa, F. Cossuttia, G. Della Riccaa,b, A. Zanettia Kyungpook National University, Daegu, Korea
D.H. Kim, G.N. Kim, M.S. Kim, S. Lee, S.W. Lee, Y.D. Oh, S. Sekmen, D.C. Son, Y.C. Yang Chonbuk National University, Jeonju, Korea
A. Lee Chonnam National University, Institute for Universe and Elementary Particles, Kwangju,
Korea
H. Kim Hanyang University, Seoul, Korea
J.A. Brochero Cifuentes, T.J. Kim Hanyang University, Seoul, Korea
J.A. Brochero Cifuentes, T.J. Kim Korea University, Seoul, Korea
S. Cho, S. Choi, Y. Go, D. Gyun, S. Ha, B. Hong, Y. Jo, Y. Kim, B. Lee, K. Lee, K.S. Lee, S. Lee,
J. Lim, S.K. Park, Y. Roh Seoul National University, Seoul, Korea
J. Almond, J. Kim, H. Lee, S.B. Oh, B.C. Radburn-Smith, S.h. Seo, U.K. Yang, H.D. Yoo, G.B. Yu University of Seoul, Seoul, Korea
M. Choi, H. Kim, J.H. Kim, J.S.H. Lee, I.C. Park, G. Ryu, M.S. Ryu University of Seoul, Seoul, Korea
M. Choi, H. Kim, J.H. Kim, J.S.H. Lee, I.C. Park, G. Ryu, M.S. Ryu Sungkyunkwan University, Suwon, Korea
Y. Choi, J. Goh, C. Hwang, J. Lee, I. Yu C o , J Go , C
g, J
ee,
Vilnius University, Vilnius, Lithuania
V. Dudenas, A. Juodagalvis, J. Vaitkus
National Centre for Particle Physics, Universiti Malaya, Kuala Lumpur, Malaysia
I. Ahmed, Z.A. Ibrahim, J.R. Komaragiri, M.A.B. Md Ali34, F. Mohamad Idris35, W.A.T. Wan
Abdullah, M.N. Yusli, Z. Zolkapli
Centro de Investigacion y de Estudios Avanzados del IPN, Mexico City, Mexico
H. Castilla-Valdez, E. De La Cruz-Burelo, I. Heredia-De La Cruz36, A. Hernandez-Almada,
R. Lopez-Fernandez, R. Maga˜na Villalba, J. Mejia Guisao, A. Sanchez-Hernandez
Universidad Iberoamericana, Mexico City, Mexico
S. Carrillo Moreno, C. Oropeza Barrera, F. Vazquez Valencia
Benemerita Universidad Autonoma de Puebla, Puebla, Mexico
S. Carpinteyro, I. Pedraza, H.A. Salazar Ibarguen, C. Uribe Estrada
Universidad Aut´onoma de San Luis Potos´ı, San Luis Potos´ı, Mexico
A. Morelos Pineda
University of Auckland, Auckland, New Zealand
D. Krofcheck
University of Canterbury, Christchurch, New Zealand
P.H. Butler Vilnius University, Vilnius, Lithuania
V. Dudenas, A. Juodagalvis, J. Vaitkus National Centre for Particle Physics, Universiti Malaya, Kuala Lumpur, Malaysia
I. Ahmed, Z.A. Ibrahim, J.R. Komaragiri, M.A.B. Md Ali34, F. Mohamad Idris35, W.A.T. Wan
Abdullah, M.N. Yusli, Z. Zolkapli g
y
,
y,
H. Castilla-Valdez, E. De La Cruz-Burelo, I. Heredia-De La Cruz36, A. Hernandez-Almada,
R. Lopez-Fernandez, R. Maga˜na Villalba, J. Mejia Guisao, A. University College Dublin, Dublin, Ireland
M. Felcini, M. Grunewald Sahaa, A. Santocchiaa,b INFN Sezione di Pisa a, Universit`a di Pisa b, Scuola Normale Superiore di Pisa c, Pisa, Italy
K. Androsova,32, P. Azzurria,16, G. Bagliesia, J. Bernardinia, T. Boccalia, R. Castaldia,
M.A. Cioccia,32, R. Dell’Orsoa, S. Donatoa,c, G. Fedi, A. Giassia, M.T. Grippoa,32, F. Ligabuea,c,
T. Lomtadzea, L. Martinia,b, A. Messineoa,b, F. Pallaa, A. Rizzia,b, A. Savoy-Navarroa,33,
P. Spagnoloa, R. Tenchinia, G. Tonellia,b, A. Venturia, P.G. Verdinia INFN Sezione di Roma a, Universit`a di Roma b, Roma, Italy
L. Baronea,b, F. Cavallaria, M. Cipriania,b, D. Del Rea,b,16, M. Diemoza, S. Gellia,b, E. Longoa,b,
F Margarolia,b B Marzocchia,b P Meridiania G Organtinia,b R Paramattia F Preiatoa,b INFN Sezione di Roma a, Universit`a di Roma b, Roma, Italy
L. Baronea,b, F. Cavallaria, M. Cipriania,b, D. Del Rea,b,16, M. Diemoza, S. Gellia,b, E. Longoa,b,
F. Margarolia,b, B. Marzocchia,b, P. Meridiania, G. Organtinia,b, R. Paramattia, F. Preiatoa,b,
S. Rahatloua,b, C. Rovellia, F. Santanastasioa,b INFN Sezione di Torino a, Universit`a di Torino b, Torino, Italy, Universit`a del Piemonte
Orientale c, Novara, Italy
N. Amapanea,b, R. Arcidiaconoa,c,16, S. Argiroa,b, M. Arneodoa,c, N. Bartosika, R. Bellana,b,
C. Biinoa, N. Cartigliaa, F. Cennaa,b, M. Costaa,b, R. Covarellia,b, A. Deganoa,b, N. Demariaa,
L. Fincoa,b, B. Kiania,b, C. Mariottia, S. Masellia, E. Migliorea,b, V. Monacoa,b, E. Monteila,b,
M.M. Obertinoa,b, L. Pachera,b, N. Pastronea, M. Pelliccionia, G.L. Pinna Angionia,b, F. Raveraa,b, INFN Sezione di Torino a, Universit`a di Torino b, Torino, Italy, Universit`a del Piemonte
Orientale c, Novara, Italy
b
16
b
b C. Biino , N. Cartiglia , F. Cenna
, M. Costa
, R. Covarelli
, A. Degano
, N. Demaria ,
L. Fincoa,b, B. Kiania,b, C. Mariottia, S. Masellia, E. Migliorea,b, V. Monacoa,b, E. Monteila,b,
M M Ob
ti
a b L P
h
a b N P
t
a M P lli
i
ia G L Pi
A
i
ia b F R
a b g
M.M. Obertinoa,b, L. Pachera,b, N. Pastronea, M. Pelliccionia, G.L. Pinna Angionia,b, F. Rave 55 A. Romeroa,b, M. Ruspaa,c, R. Sacchia,b, K. Shchelinaa,b, V. Solaa, A. Solanoa,b, A. Staianoa,
P. Traczyka,b A. Romeroa,b, M. Ruspaa,c, R. Sacchia,b, K. Shchelinaa,b, V. Solaa, A. Solanoa,b, A. Staianoa,
P. Traczyka,b Sanchez-Hernandez H. Castilla-Valdez, E. De La Cruz-Burelo, I. Heredia-De La Cruz36, A. Hernandez-Almada,
R. Lopez-Fernandez, R. Maga˜na Villalba, J. Mejia Guisao, A. Sanchez-Hernandez Universidad Iberoamericana, Mexico City, Mexico
S. Carrillo Moreno, C. Oropeza Barrera, F. Vazquez Valencia Universidad Iberoamericana, Mexico City, Mexico
S. Carrillo Moreno, C. Oropeza Barrera, F. Vazquez Valencia Benemerita Universidad Autonoma de Puebla, Puebla, Mexico
S. Carpinteyro, I. Pedraza, H.A. Salazar Ibarguen, C. Uribe Estrada Benemerita Universidad Autonoma de Puebla, Puebla, Mexico
S. Carpinteyro, I. Pedraza, H.A. Salazar Ibarguen, C. Uribe Estrada Universidad Aut´onoma de San Luis Potos´ı, San Luis Potos´ı, Mexico
A. Morelos Pineda Universidad Aut´onoma de San Luis Potos´ı, San Luis Potos´ı, Mexico
A. Morelos Pineda University of Auckland, Auckland, New Zealand
D. Krofcheck University of Auckland, Auckland, New Zealand
D. Krofcheck University of Canterbury, Christchurch, New Zealand
P.H. Butler University of Canterbury, Christchurch, New Zealand
P.H. Butler 56 C
The CMS Collaboration National Centre for Physics, Quaid-I-Azam University, Islamabad, Pakistan
A. Ahmad, M. Ahmad, Q. Hassan, H.R. Hoorani, W.A. Khan, A. Saddique, M.A. Shah,
M. Shoaib, M. Waqas National Centre for Nuclear Research, Swierk, Poland National Centre for Nuclear Research, Swierk, Poland
H. Bialkowska, M. Bluj, B. Boimska, T. Frueboes, M. G´orski, M. Kazana, K. Nawrocki,
K. Romanowska-Rybinska, M. Szleper, P. Zalewski Institute of Experimental Physics, Faculty of Physics, University of Warsaw, Warsaw, Poland
K. Bunkowski, A. Byszuk37, K. Doroba, A. Kalinowski, M. Konecki, J. Krolikowski, M. Misiura,
M. Olszewski, M. Walczak Laborat´orio de Instrumenta¸c˜ao e F´ısica Experimental de Part´ıculas, Lisboa, Portugal Laborat´orio de Instrumenta¸c˜ao e F´ısica Experimental de Part´ıculas, Lisboa, Portugal
P. Bargassa, C. Beir˜ao Da Cruz E Silva, A. Di Francesco, P. Faccioli, P.G. Ferreira Parracho,
M. Gallinaro, J. Hollar, N. Leonardo, L. Lloret Iglesias, M.V. Nemallapudi, J. Rodrigues
Antunes, J. Seixas, O. Toldaiev, D. Vadruccio, J. Varela, P. Vischia Laboratorio de Instrumenta¸cao e Fısica Experimental de Partıculas, Lisboa, Portugal
P. Bargassa, C. Beir˜ao Da Cruz E Silva, A. Di Francesco, P. Faccioli, P.G. Ferreira Parracho,
M Gallinaro
J Hollar N Leonardo
L Lloret Iglesias
M V Nemallapudi
J Rodrigues P. Bargassa, C. Beirao Da Cruz E Silva, A. Di Francesco, P. Faccioli, P.G. Ferreira Parracho,
M. Gallinaro, J. Hollar, N. Leonardo, L. Lloret Iglesias, M.V. Nemallapudi, J. Rodrigues
Antunes, J. Seixas, O. Toldaiev, D. Vadruccio, J. Varela, P. Vischia Joint Institute for Nuclear Research, Dubna, Russia Joint Institute for Nuclear Research, Dubna, Russia V. Alexakhin, A. Golunov, I. Golutvin, N. Gorbounov, A. Kamenev, V. Karjavin, V. Korenkov,
A. Lanev, A. Malakhov, V. Matveev38,39, V.V. Mitsyn, V. Palichik, V. Perelygin, S. Shmatov,
N. Skatchkov, V. Smirnov, E. Tikhonenko, A. Zarubin Petersburg Nuclear Physics Institute, Gatchina (St. Petersburg), Russia
L. Chtchipounov, V. Golovtsov, Y. Ivanov, V. Kim40, E. Kuznetsova41, V. Murzin, V. Oreshkin,
V. Sulimov, A. Vorobyev Institute for Nuclear Research, Moscow, Russia
Yu. Andreev, A. Dermenev, S. Gninenko, N. Golubev, A. Karneyeu, M. Kirsanov, N. Krasnikov,
A. Pashenkov, D. Tlisov, A. Toropin Institute for Theoretical and Experimental Physics, Moscow, Russia
V. Epshteyn,
V. Gavrilov,
N. Lychkovskaya,
V. Popov,
I. Pozdnyakov,
G. Safronov,
A. Spiridonov, M. Toms, E. Vlasov, A. Zhokin
Moscow Institute of Physics and Technology V. Epshteyn,
V. Gavrilov,
N. Lychkovskaya,
V. Popov,
I. Pozdnyakov,
G. Safronov,
A. Spiridonov, M. Toms, E. Vlasov, A. Zhokin
Moscow Institute of Physics and Technology
A. Bylinkin39 Moscow Institute of Physics and Technology
A. Bylinkin39 Moscow Institute of Physics and Technology
A. Bylinkin39 A. Bylinkin
National Research Nuclear University ’Moscow Engineering Physics Institute’ (MEPhI),
Moscow, Russia
O. Markin, E. Popova, E. Tarkovskii
P.N. Lebedev Physical Institute, Moscow, Russia
V. Andreev, M. Azarkin39, I. Dremin39, M. Kirakosyan, A. Leonidov39, S.V. Rusakov,
A. Terkulov
Skobeltsyn Institute of Nuclear Physics, Lomonosov Moscow State University, Moscow,
Russia
A. Baskakov, A. Belyaev, E. Boos, V. Bunichev, M. Dubinin42, L. Dudko, A. Ershov, V. Klyukhin,
O. Kodolova, N. Korneeva, I. Lokhtin, I. Miagkov, S. Obraztsov, M. Perfilov, V. Savrin
Novosibirsk State University (NSU), Novosibirsk, Russia
V. Blinov43, Y.Skovpen43 National Research Nuclear University ’Moscow Engineering Physics Institute’ (MEPhI),
Moscow, Russia
O. Markin, E. Popova, E. Tarkovskii P.N. Lebedev Physical Institute, Moscow, Russia
V. Andreev, M. Azarkin39, I. Dremin39, M. Kirakosyan, A. Leonidov39, S.V. Rusakov,
A. Terkulov Skobeltsyn Institute of Nuclear Physics, Lomonosov Moscow State University, Moscow,
Russia Novosibirsk State University (NSU), Novosibirsk, Russia
V. Blinov43, Y.Skovpen43 57 Centro de Investigaciones Energ´eticas Medioambientales y Tecnol´ogicas (CIEMAT),
Madrid, Spain Centro de Investigaciones Energ´eticas Medioambientales y Tecnol´ogicas (CIEMAT),
Madrid, Spain J. Alcaraz Maestre, M. Barrio Luna, E. Calvo, M. Cerrada, M. Chamizo Llatas, N. Colino, B. De
La Cruz, A. Delgado Peris, A. Escalante Del Valle, C. Fernandez Bedoya, J.P. Fern´andez Ramos,
J. Flix, M.C. Fouz, P. Garcia-Abia, O. Gonzalez Lopez, S. Goy Lopez, J.M. Hernandez, M.I. Josa,
E. Navarro De Martino, A. P´erez-Calero Yzquierdo, J. Puerta Pelayo, A. Quintario Olmeda,
I. Redondo, L. Romero, M.S. Soares I. Redondo, L. Romero, M.S. Soares Universidad Aut´onoma de Madrid, Madrid, Spain
J.F. de Troc´oniz, M. Missiroli, D. Moran Universidad de Oviedo, Oviedo, Spain
J. Cuevas, J. Fernandez Menendez, I. Gonzalez Caballero, J.R. Gonz´alez Fern´andez, E. Palencia
Cortezon, S. Sanchez Cruz, I. Su´arez Andr´es, J.M. Vizan Garcia Instituto de F´ısica de Cantabria (IFCA), CSIC-Universidad de Cantabria, Santander, Spain
I.J. Cabrillo, A. Calderon, J.R. Casti˜neiras De Saa, E. Curras, M. Fernandez, J. Garcia-Ferrero,
G. Gomez, A. Lopez Virto, J. Marco, C. Martinez Rivero, F. Matorras, J. Piedra Gomez,
T. Rodrigo, A. Ruiz-Jimeno, L. Scodellaro, N. Trevisani, I. Vila, R. Vilar Cortabitarte CERN, European Organization for Nuclear Research, Geneva, Switzerland
D. Abbaneo, E. Auffray, G. Auzinger, M. Bachtis, P. Baillon, A.H. Ball, D. Barney, P. Bloch,
A. Bocci, A. Bonato, C. Botta, T. Camporesi, R. Castello, M. Cepeda, G. Cerminara,
M. D’Alfonso, D. d’Enterria, A. Dabrowski, V. Daponte, A. David, M. De Gruttola, A. De Roeck,
E. Di Marco45, M. Dobson, B. Dorney, T. du Pree, D. Duggan, M. D¨unser, N. Dupont, A. Elliott-
Peisert, S. Fartoukh, G. Franzoni, J. Fulcher, W. Funk, D. Gigi, K. Gill, M. Girone, F. Glege,
D. Gulhan, S. Gundacker, M. Guthoff, J. Hammer, P. Harris, J. Hegeman, V. Innocente, P. Janot,
J. Kieseler, H. Kirschenmann, V. Kn¨unz, A. Kornmayer16, M.J. Kortelainen, K. Kousouris,
M. Krammer1, C. Lange, P. Lecoq, C. Lourenc¸o, M.T. Lucchini, L. Malgeri, M. Mannelli,
A. Martelli, F. Meijers, J.A. Merlin, S. Mersi, E. Meschi, F. Moortgat, S. Morovic, M. Mulders,
H. Neugebauer, S. Orfanelli, L. Orsini, L. Pape, E. Perez, M. Peruzzi, A. Petrilli, G. Petrucciani,
A. Pfeiffer, M. Pierini, A. Racz, T. Reis, G. Rolandi46, M. Rovere, M. Ruan, H. Sakulin,
J.B. Sauvan, C. Sch¨afer, C. Schwick, M. Seidel, A. Sharma, P. Silva, P. Sphicas47, J. Steggemann,
M. Stoye, Y. Takahashi, M. Tosi, D. Treille, A. Triossi, A. Tsirou, V. Veckalns48, G.I. Veres21,
N. Wardle, A. Zagozdzinska37, W.D. Zeuner N. Wardle, A. Zagozdzinska37, W.D. State Research Center of Russian Federation, Institute for High Energy Physics, Protvino,
Russia I. Azhgirey, I. Bayshev, S. Bitioukov, D. Elumakhov, V. Kachanov, A. Kalinin, D. Konstantinov,
V. Krychkine, V. Petrov, R. Ryutin, A. Sobol, S. Troshin, N. Tyurin, A. Uzunian, A. Volkov University of Belgrade, Faculty of Physics and Vinca Institute of Nuclear Sciences, Belgrade,
Serbia P. Adzic44, P. Cirkovic, D. Devetak, M. Dordevic, J. Milosevic, V. Rekovic National Central University, Chung-Li, Taiwan National Central University, Chung-Li, Taiwan V. Candelise, T.H. Doan, Sh. Jain, R. Khurana, M. Konyushikhin, C.M. Kuo, W. Lin, Y.J. Lu,
A. Pozdnyakov, S.S. Yu National Taiwan University (NTU), Taipei, Taiwan
Arun Kumar, P. Chang, Y.H. Chang, Y.W. Chang, Y. Chao, K.F. Chen, P.H. Chen, C. Dietz,
F. Fiori, W.-S. Hou, Y. Hsiung, Y.F. Liu, R.-S. Lu, M. Mi˜nano Moya, E. Paganis, A. Psallidas,
J.f. Tsai, Y.M. Tzeng Chulalongkorn University, Faculty of Science, Department of Physics, Bangkok, Thailand
B. Asavapibhop, G. Singh, N. Srimanobhas, N. Suwonjandee B. Asavapibhop, G. Singh, N. Srimanobhas, N. Suwonjandee Cukurova University, Adana, Turkey Cukurova University, Adana, Turkey
A. Adiguzel, M.N. Bakirci51, S. Cerci52, S. Damarseckin, Z.S. Demiroglu, C. Dozen,
I. Dumanoglu, S. Girgis, G. Gokbulut, Y. Guler, I. Hos, E.E. Kangal53, O. Kara, A. Kayis Topaksu, Cukurova University, Adana, Turkey
A. Adiguzel, M.N. Bakirci51, S. Cerci52, S. Damarseckin, Z.S. Demiroglu, C. Dozen,
I. Dumanoglu, S. Girgis, G. Gokbulut, Y. Guler, I. Hos, E.E. Kangal53, O. Kara, A. Kayis Topaksu,
U. Kiminsu, M. Oglakci, G. Onengut54, K. Ozdemir55, B. Tali52, S. Turkcapar, I.S. Zorbakir,
C Zorbilmez Cukurova University, Adana, Turkey
A. Adiguzel, M.N. Bakirci51, S. Cerci52, S. Damarseckin, Z.S. Demiroglu, C. Dozen,
I. Dumanoglu, S. Girgis, G. Gokbulut, Y. Guler, I. Hos, E.E. Kangal53, O. Kara, A. Kayis Topaksu, g
g
I. Dumanoglu, S. Girgis, G. Gokbulut, Y. Guler, I. Hos, E.E. Kangal53, O. Kara, A. Kayis Topaksu,
U. Kiminsu, M. Oglakci, G. Onengut54, K. Ozdemir55, B. Tali52, S. Turkcapar, I.S. Zorbakir,
C. Zorbilmez U. Kiminsu, M. Oglakci, G. Onengut54, K. Ozdemir55, B. Tali52, S. Turkcapar, I.S. Zorbakir,
C. Zorbilmez Middle East Technical University, Physics Department, Ankara, Turkey
B. Bilin, S. Bilmis, B. Isildak56, G. Karapinar57, M. Yalvac, M. Zeyrek Bogazici University, Istanbul, Turkey
E. G¨ulmez, M. Kaya58, O. Kaya59, E.A. Yetkin60, T. Yetkin61 Istanbul Technical University, Istanbul, Turkey
A. Cakir, K. Cankocak, S. Sen62 Istanbul Technical University, Istanbul, Turkey
A. Cakir, K. Cankocak, S. Sen62 Institute for Scintillation Materials of National Academy of Science of Ukraine, Kharkov,
Ukraine
B. Grynyov National Scientific Center, Kharkov Institute of Physics and Technology, Kharkov, Ukraine
L. Levchuk, P. Sorokin University of Bristol, Bristol, United Kingdom
R. Aggleton, F. Ball, L. Beck, J.J. Brooke, D. Burns, E. Clement, D. Cussans, H. Flacher,
J. Goldstein, M. Grimes, G.P. Heath, H.F. Heath, J. Jacob, L. Kreczko, C. Lucas, D.M. Newbold63,
S. Paramesvaran, A. Poll, T. Sakuma, S. Seif El Nasr-storey, D. Centro de Investigaciones Energ´eticas Medioambientales y Tecnol´ogicas (CIEMAT),
Madrid, Spain Zeuner Paul Scherrer Institut, Villigen, Switzerland
W. Bertl, K. Deiters, W. Erdmann, R. Horisberger, Q. Ingram, H.C. Kaestli, D. Kotlinski,
U. Langenegger, T. Rohe g
W. Bertl, K. Deiters, W. Erdmann, R. Horisberger, Q. Ingram, H.C. Kaestli, D. Kotlinski,
U. Langenegger, T. Rohe W. Bertl, K. Deiters, W. Erdmann, R. Horisberger, Q. Ingram, H.C. Kaestli, D. Kotlinski,
U. Langenegger, T. Rohe Institute for Particle Physics, ETH Zurich, Zurich, Switzerland Institute for Particle Physics, ETH Zurich, Zurich, Switzerland Institute for Particle Physics, ETH Zurich, Zurich, Switzerland
F. Bachmair, L. B¨ani, L. Bianchini, B. Casal, G. Dissertori, M. Dittmar, M. Doneg`a, C. Grab,
C. Heidegger, D. Hits, J. Hoss, G. Kasieczka, P. Lecomte†, W. Lustermann, B. Mangano,
M. Marionneau, P. Martinez Ruiz del Arbol, M. Masciovecchio, M.T. Meinhard, D. Meister, Institute for Particle Physics, ETH Zurich, Zurich, Switzerland
F. Bachmair, L. B¨ani, L. Bianchini, B. Casal, G. Dissertori, M. Dittmar, M. Doneg`a, C. Grab,
C. Heidegger, D. Hits, J. Hoss, G. Kasieczka, P. Lecomte†, W. Lustermann, B. Mangano,
M. Marionneau, P. Martinez Ruiz del Arbol, M. Masciovecchio, M.T. Meinhard, D. Meister, F. Bachmair, L. B¨ani, L. Bianchini, B. Casal, G. Dissertori, M. Dittmar, M. Doneg`a, C. Grab,
C. Heidegger, D. Hits, J. Hoss, G. Kasieczka, P. Lecomte†, W. Lustermann, B. Mangano,
M Marionneau P Martinez Ruiz del Arbol M Masciovecchio M T Meinhard D Meister g
C. Heidegger, D. Hits, J. Hoss, G. Kasieczka, P. Lecomte†, W. Lustermann, B. Mangano, 58 C
The CMS Collaboration F. Micheli, P. Musella, F. Nessi-Tedaldi, F. Pandolfi, J. Pata, F. Pauss, G. Perrin, L. Perrozzi,
M. Quittnat, M. Rossini, M. Sch¨onenberger, A. Starodumov49, V.R. Tavolaro, K. Theofilatos,
R. Wallny Universit¨at Z¨urich, Zurich, Switzerland Universit¨at Z¨urich, Zurich, Switzerland
T.K. Aarrestad, C. Amsler50, L. Caminada, M.F. Canelli, A. De Cosa, C. Galloni, A. Hinzmann,
T. Hreus, B. Kilminster, J. Ngadiuba, D. Pinna, G. Rauco, P. Robmann, D. Salerno, Y. Yang,
A. Zucchetta T.K. Aarrestad, C. Amsler50, L. Caminada, M.F. Canelli, A. De Cosa, C. Galloni, A. Hin ,
,
,
,
,
,
,
T. Hreus, B. Kilminster, J. Ngadiuba, D. Pinna, G. Rauco, P. Robmann, D. Salerno, Y. Yang,
A. Zucchetta University of California, Davis, Davis, USA R. Breedon, G. Breto, D. Burns, M. Calderon De La Barca Sanchez, S. Chauhan, M. Chertok,
J. Conway, R. Conway, P.T. Cox, R. Erbacher, C. Flores, G. Funk, M. Gardner, W. Ko, R. Lander, R. Breedon, G. Breto, D. Burns, M. Calderon De La Barca Sanchez, S. Chauhan, M. Chertok,
J. Conway, R. Conway, P.T. Cox, R. Erbacher, C. Flores, G. Funk, M. Gardner, W. Ko, R. Lander,
C. Mclean, M. Mulhearn, D. Pellett, J. Pilot, S. Shalhout, J. Smith, M. Squires, D. Stolp,
M Tripathi S Wilbur R Yohay R. Breedon, G. Breto, D. Burns, M. Calderon De La Barca Sanchez, S. Chauhan, M. Chertok,
J. Conway, R. Conway, P.T. Cox, R. Erbacher, C. Flores, G. Funk, M. Gardner, W. Ko, R. Lander,
C. Mclean, M. Mulhearn, D. Pellett, J. Pilot, S. Shalhout, J. Smith, M. Squires, D. Stolp,
M. Tripathi, S. Wilbur, R. Yohay y
y
C. Mclean, M. Mulhearn, D. Pellett, J. Pilot, S. Shalhout, J. Smith, M. Squires, D. Stolp,
M. Tripathi, S. Wilbur, R. Yohay University of California, Los Angeles, USA
C. Bravo, R. Cousins, P. Everaerts, A. Florent, J. Hauser, M. Ignatenko, N. Mccoll, D. Saltzberg,
C. Schnaible, E. Takasugi, V. Valuev, M. Weber University of California, Riverside, Riverside, USA
K. Burt, R. Clare, J. Ellison, J.W. Gary, S.M.A. Ghiasi Shirazi, G. Hanson, J. Heilman, P. Jandir,
E. Kennedy, F. Lacroix, O.R. Long, M. Olmedo Negrete, M.I. Paneva, A. Shrinivas, W. Si, H. Wei,
S. Wimpenny, B. R. Yates S. Wimpenny, B. R. Yates University of California, San Diego, La Jolla, USA
J.G. Branson, G.B. Cerati, S. Cittolin, M. Derdzinski, R. Gerosa, A. Holzner, D. Klein,
V. Krutelyov, J. Letts, I. Macneill, D. Olivito, S. Padhi, M. Pieri, M. Sani, V. Sharma, S. Simon,
M. Tadel, A. Vartak, S. Wasserbaech66, C. Welke, J. Wood, F. W¨urthwein, A. Yagil, G. Zevi Della
Porta University of California, Santa Barbara - Department of Physics, Santa Barbara, USA
h
d
d
ll
ld
h University of California, Santa Barbara - Department of Physics, Santa Barbara, USA
N. Amin, R. Bhandari, J. Bradmiller-Feld, C. Campagnari, A. Dishaw, V. Dutta, K. Flowers,
M. Franco Sevilla, P. Geffert, C. George, F. Golf, L. Gouskos, J. Gran, R. Heller, J. Incandela,
S.D. Mullin, A. Ovcharova, J. Richman, D. Stuart, I. Suarez, J. Yoo California Institute of Technology, Pasadena, USA
D. Anderson, A. Apresyan, J. Bendavid, A. Bornheim, J. Bunn, Y. Chen, J. Brunel University, Uxbridge, United Kingdom Brunel University, Uxbridge, United Kingdom
J.E. Cole, P.R. Hobson, A. Khan, P. Kyberd, D. Leslie, I.D. Reid, P. Symonds, L. Teodorescu,
M. Turner Baylor University, Waco, USA A. Borzou, K. Call, J. Dittmann, K. Hatakeyama, H. Liu, N. Pastika The University of Alabama, Tuscaloosa, USA
O. Charaf, S.I. Cooper, C. Henderson, P. Rumerio, C. West Boston University, Boston, USA
D. Arcaro, A. Avetisyan, T. Bose, D. Gastler, D. Rankin, C. Richardson, J. Rohlf, L. Sulak, D. Zou Boston University, Boston, USA Brown University, Providence, USA
G. Benelli, E. Berry, D. Cutts, A. Garabedian, J. Hakala, U. Heintz, J.M. Hogan, O. Jesus,
K.H.M. Kwok, E. Laird, G. Landsberg, Z. Mao, M. Narain, S. Piperov, S. Sagir, E. Spencer,
R. Syarif University of California, Davis, Davis, USA National Central University, Chung-Li, Taiwan Smith, V.J. Smith Rutherford Appleton Laboratory, Didcot, United Kingdom
K.W. Bell, A. Belyaev64, C. Brew, R.M. Brown, L. Calligaris, D. Cieri, D.J.A. Cockerill,
J.A. Coughlan, K. Harder, S. Harper, E. Olaiya, D. Petyt, C.H. Shepherd-Themistocleous,
A. Thea, I.R. Tomalin, T. Williams Imperial College, London, United Kingdom
M. Baber, R. Bainbridge, O. Buchmuller, A. Bundock, D. Burton, S. Casasso, M. Citron,
D. Colling, L. Corpe, P. Dauncey, G. Davies, A. De Wit, M. Della Negra, R. Di Maria, P. Dunne, 59 A. Elwood, D. Futyan, Y. Haddad, G. Hall, G. Iles, T. James, R. Lane, C. Laner, R. Lucas63,
L. Lyons, A.-M. Magnan, S. Malik, L. Mastrolorenzo, J. Nash, A. Nikitenko49, J. Pela, B. Penning,
M Pesaresi D M Raymond A Richards A Rose C Seez S Summers A Tapper K Uchida A. Elwood, D. Futyan, Y. Haddad, G. Hall, G. Iles, T. James, R. Lane, C. Laner, R. Lucas63,
L. Lyons, A.-M. Magnan, S. Malik, L. Mastrolorenzo, J. Nash, A. Nikitenko49, J. Pela, B. Penning,
M. Pesaresi, D.M. Raymond, A. Richards, A. Rose, C. Seez, S. Summers, A. Tapper, K. Uchida,
M Vazquez Acosta65 T Virdee16 J Wright S C Zenz A. Elwood, D. Futyan, Y. Haddad, G. Hall, G. Iles, T. James, R. Lane, C. Laner, R. Lucas ,
L. Lyons, A.-M. Magnan, S. Malik, L. Mastrolorenzo, J. Nash, A. Nikitenko49, J. Pela, B. Penning,
M. Pesaresi, D.M. Raymond, A. Richards, A. Rose, C. Seez, S. Summers, A. Tapper, K. Uchida,
M. Vazquez Acosta65, T. Virdee16, J. Wright, S.C. Zenz y
,
g
,
,
, J
,
, J
,
g,
M. Pesaresi, D.M. Raymond, A. Richards, A. Rose, C. Seez, S. Summers, A. Tapper, K. Uchida,
M. Vazquez Acosta65, T. Virdee16, J. Wright, S.C. Zenz D. Winn Fermi National Accelerator Laboratory, Batavia, USA Fermi National Accelerator Laboratory, Batavia, USA
S. Abdullin, M. Albrow, G. Apollinari, S. Banerjee, L.A.T. Bauerdick, A. Beretvas, J. Berryhill,
P.C. Bhat, G. Bolla, K. Burkett, J.N. Butler, H.W.K. Cheung, F. Chlebana, S. Cihangir†,
M. Cremonesi, V.D. Elvira, I. Fisk, J. Freeman, E. Gottschalk, L. Gray, D. Green, S. Gr¨unendahl,
O. Gutsche, D. Hare, R.M. Harris, S. Hasegawa, J. Hirschauer, Z. Hu, B. Jayatilaka, S. Jindariani,
M. Johnson, U. Joshi, B. Klima, B. Kreis, S. Lammel, J. Linacre, D. Lincoln, R. Lipton,
T. Liu, R. Lopes De S´a, J. Lykken, K. Maeshima, N. Magini, J.M. Marraffino, S. Maruyama,
D. Mason, P. McBride, P. Merkel, S. Mrenna, S. Nahn, C. Newman-Holmes†, V. O’Dell, K. Pedro,
O. Prokofyev, G. Rakness, L. Ristori, E. Sexton-Kennedy, A. Soha, W.J. Spalding, L. Spiegel,
S. Stoynev, N. Strobbe, L. Taylor, S. Tkaczyk, N.V. Tran, L. Uplegger, E.W. Vaandering,
C. Vernieri, M. Verzocchi, R. Vidal, M. Wang, H.A. Weber, A. Whitbeck Cornell University, Ithaca, USA J. Alexander, J. Chaves, J. Chu, S. Dittmer, K. Mcdermott, N. Mirman, G. Nicolas Kaufman,
J.R. Patterson, A. Rinkevicius, A. Ryd, L. Skinnari, L. Soffi, S.M. Tan, Z. Tao, J. Thom, J. Tucker,
P. Wittich, M. Zientek University of California, Davis, Davis, USA Duarte, J.M. Lawhorn,
A. Mott, H.B. Newman, C. Pena, M. Spiropulu, J.R. Vlimant, S. Xie, R.Y. Zhu California Institute of Technology, Pasadena, USA C
The CMS Collaboration 60 Carnegie Mellon University, Pittsburgh, USA
M.B. Andrews, V. Azzolini, T. Ferguson, M. Paulini, J. Russ, M. Sun, H. Vogel, I. Vorobiev,
M. Weinberg Fairfield University, Fairfield, USA Fairfield University, Fairfield, USA University of Colorado Boulder, Boulder, USA University of Colorado Boulder, Boulder, USA
J.P. Cumalat, W.T. Ford, F. Jensen, A. Johnson, M. Krohn, T. Mulholland, K. Stenson,
S.R. Wagner University of Florida, Gainesville, USA Maksimovic, C. Martin, M. Osherson, J. Roskes, U. Sarica, M. Swartz, M. Xiao, Y. Xin, C. You I. Anderson, B. Blumenfeld, A. Cocoros, N. Eminizer, D. Fehling, L. Feng, A.V. Gritsan,
P. Maksimovic, C. Martin, M. Osherson, J. Roskes, U. Sarica, M. Swartz, M. Xiao, Y. Xin, C. You The University of Kansas, Lawrence, USA
A. Al-bataineh, P. Baringer, A. Bean, S. Boren, J. Bowen, C. Bruner, J. Castle, L. Forthomme,
R.P. Kenny III, A. Kropivnitskaya, D. Majumder, W. Mcbrayer, M. Murray, S. Sanders,
R. Stringer, J.D. Tapia Takaki, Q. Wang A. Al-bataineh, P. Baringer, A. Bean, S. Boren, J. Bowen, C. Bruner, J. Castle, L. Forthomme,
R.P. Kenny III, A. Kropivnitskaya, D. Majumder, W. Mcbrayer, M. Murray, S. Sanders,
R. Stringer, J.D. Tapia Takaki, Q. Wang Kansas State University, Manhattan, USA Kansas State University, Manhattan, USA
A. Ivanov, K. Kaadze, S. Khalil, Y. Maravin, A. Mohammadi, L.K. Saini, N. Skhirtladze, S. Toda y,
,
A. Ivanov, K. Kaadze, S. Khalil, Y. Maravin, A. Mohammadi, L.K. Saini, N. Skhirtladze, S. Toda Lawrence Livermore National Laboratory, Livermore, USA
F. Rebassoo, D. Wright Lawrence Livermore National Laboratory, Livermore, USA
F. Rebassoo, D. Wright University of Maryland, College Park, USA
C. Anelli, A. Baden, O. Baron, A. Belloni, B. Calvert, S.C. Eno, C. Ferraioli, J.A. Gomez,
N.J. Hadley, S. Jabeen, R.G. Kellogg, T. Kolberg, J. Kunkle, Y. Lu, A.C. Mignerey, F. Ricci-Tam,
Y.H. Shin, A. Skuja, M.B. Tonjes, S.C. Tonwar Massachusetts Institute of Technology, Cambridge, USA D. Abercrombie, B. Allen, A. Apyan, R. Barbieri, A. Baty, R. Bi, K. Bierwagen, S. Brandt,
W. Busza, I.A. Cali, Z. Demiragli, L. Di Matteo, G. Gomez Ceballos, M. Goncharov, D. Hsu,
Y. Iiyama, G.M. Innocenti, M. Klute, D. Kovalskyi, K. Krajczar, Y.S. Lai, Y.-J. Lee, A. Levin,
P.D. Luckey, B. Maier, A.C. Marini, C. Mcginn, C. Mironov, S. Narayanan, X. Niu, C. Paus,
C. Roland, G. Roland, J. Salfeld-Nebgen, G.S.F. Stephans, K. Sumorok, K. Tatar, M. Varma,
D. Velicanu, J. Veverka, J. Wang, T.W. Wang, B. Wyslouch, M. Yang, V. Zhukova University of Minnesota, Minneapolis, USA
A.C. Benvenuti, R.M. Chatterjee, A. Evans, A. Finkel, A. Gude, P. Hansen, S. Kalafut, S.C. Kao,
Y. Kubota, Z. Lesko, J. Mans, S. Nourbakhsh, N. Ruckstuhl, R. Rusack, N. Tambe, J. Turkewitz University of Mississippi, Oxford, USA
J.G. Acosta, S. Oliveros University of Florida, Gainesville, USA y
D. Acosta, P. Avery, P. Bortignon, D. Bourilkov, A. Brinkerhoff, A. Carnes, M. Carver,
D. Curry, S. Das, R.D. Field, I.K. Furic, J. Konigsberg, A. Korytov, P. Ma, K. Matchev, H. Mei,
P. Milenovic67, G. Mitselmakher, D. Rank, L. Shchutska, D. Sperka, L. Thomas, J. Wang,
S. Wang, J. Yelton . Acosta, P. Avery, P. Bortignon, D. Bourilkov, A. Brinkerhoff, A. Carnes, M. Ca
. Curry, S. Das, R.D. Field, I.K. Furic, J. Konigsberg, A. Korytov, P. Ma, K. Matchev, H. P. Milenovic67, G. Mitselmakher, D. Rank, L. Shchutska, D. Sperka, L. Thomas, J. Wang,
S. Wang, J. Yelton S. Wang, J. Yelton Florida International University, Miami, USA
S. Linn, P. Markowitz, G. Martinez, J.L. Rodriguez Florida State University, Tallahassee, USA
A. Ackert, J.R. Adams, T. Adams, A. Askew, S. Bein, B. Diamond, S. Hagopian, V. Hagopian,
K.F. Johnson, A. Khatiwada, H. Prosper, A. Santra Florida Institute of Technology, Melbourne, USA
M.M. Baarmand, V. Bhopatkar, S. Colafranceschi68, M. Hohlmann, D. Noonan, T. Roy,
F. Yumiceva University of Illinois at Chicago (UIC), Chicago, USA
M.R. Adams, L. Apanasevich, D. Berry, R.R. Betts, I. Bucinskaite, R. Cavanaugh, O. Evdokimov,
L. Gauthier, C.E. Gerber, D.J. Hofman, K. Jung, P. Kurt, C. O’Brien, I.D. Sandoval Gonzalez,
P. Turner, N. Varelas, H. Wang, Z. Wu, M. Zakaria, J. Zhang M.R. Adams, L. Apanasevich, D. Berry, R.R. Betts, I. Bucinskaite, R. Cavanaugh, O. Evdokimov,
L. Gauthier, C.E. Gerber, D.J. Hofman, K. Jung, P. Kurt, C. O’Brien, I.D. Sandoval Gonzalez,
P. Turner, N. Varelas, H. Wang, Z. Wu, M. Zakaria, J. Zhang The University of Iowa, Iowa City, USA
B. Bilki69, W. Clarida, K. Dilsiz, S. Durgut, R.P. Gandrajula, M. Haytmyradov, V. Khristenko,
J.-P. Merlo, H. Mermerkaya70, A. Mestvirishvili, A. Moeller, J. Nachtman, H. Ogul, Y. Onel,
F. Ozok71, A. Penzo, C. Snyder, E. Tiras, J. Wetzel, K. Yi 61 Johns Hopkins University, Baltimore, USA J
p
y,
,
I. Anderson, B. Blumenfeld, A. Cocoros, N. Eminizer, D. Fehling, L. Feng, A.V. Gritsan,
P. Maksimovic, C. Martin, M. Osherson, J. Roskes, U. Sarica, M. Swartz, M. Xiao, Y. Xin, C. You I. Anderson, B. Blumenfeld, A. Cocoros, N. Eminizer, D. Fehling, L. Feng, A.V. Gritsan,
P M k i
i
C M
i
M O h
J R
k
U S
i
M S
M Xi
Y Xi
C Y I. Anderson, B. Blumenfeld, A. Cocoros, N. Eminizer, D. Fehling, L. Feng, A.V. Gritsan,
P. University of Mississippi, Oxford, USA
J.G. Acosta, S. Oliveros University of Nebraska-Lincoln, Lincoln, USA E. Avdeeva, R. Bartek, K. Bloom, D.R. Claes, A. Dominguez, C. Fangmeier, R. Gonzalez Suarez,
R. Kamalieddin, I. Kravchenko, A. Malta Rodrigues, F. Meier, J. Monroy, J.E. Siado, G.R. Snow,
B. Stieger E. Avdeeva, R. Bartek, K. Bloom, D.R. Claes, A. Dominguez, C. Fangmeier, R. Gonzalez Suarez, E. Avdeeva, R. Bartek, K. Bloom, D.R. Claes, A. Dominguez, C. Fangmeier, R. Gonzalez Suarez,
R Kamalieddin I Kravchenko A Malta Rodrigues F Meier J Monroy J E Siado G R Snow ,
,
,
,
g
,
g
,
,
R. Kamalieddin, I. Kravchenko, A. Malta Rodrigues, F. Meier, J. Monroy, J.E. Siado, G.R. Snow,
B. Stieger State University of New York at Buffalo, Buffalo, USA M. Alyari, J. Dolen, J. George, A. Godshalk, C. Harrington, I. Iashvili, J. Kaisen, A. Kharchilava,
A. Kumar, A. Parker, S. Rappoccio, B. Roozbahani Northeastern University, Boston, USA
G. Alverson, E. Barberis, A. Hortiangtham, A. Massironi, D.M. Morse, D. Nash, T. Orimoto,
R. Teixeira De Lima, D. Trocino, R.-J. Wang, D. Wood Northwestern University, Evanston, USA
S. Bhattacharya, K.A. Hahn, A. Kubik, A. Kumar, J.F. Low, N. Mucia, N. Odell, B. Pollack,
M.H. Schmitt, K. Sung, M. Trovato, M. Velasco University of Notre Dame, Notre Dame, USA
N. Dev, M. Hildreth, K. Hurtado Anampa, C. Jessop, D.J. Karmgard, N. Kellams, K. Lannon, 62 C
The CMS Collaboration N. Marinelli, F. Meng, C. Mueller, Y. Musienko38, M. Planer, A. Reinsvold, R. Ruchti, G. Smith,
S. Taroni, M. Wayne, M. Wolf, A. Woodard The Ohio State University, Columbus, USA
J. Alimena, L. Antonelli, J. Brinson, B. Bylsma, L.S. Durkin, S. Flowers, B. Francis, A. Hart,
C. Hill, R. Hughes, W. Ji, B. Liu, W. Luo, D. Puigh, B.L. Winer, H.W. Wulsin Princeton University, Princeton, USA Princeton University, Princeton, USA
S. Cooperstein, O. Driga, P. Elmer, J. Hardenbrook, P. Hebda, D. Lange, J. Luo, D. Marlow, J. Mc
Donald, T. Medvedeva, K. Mei, M. Mooney, J. Olsen, C. Palmer, P. Pirou´e, D. Stickland, C. Tully,
A. Zuranski Purdue University, West Lafayette, USA Purdue University, West Lafayette, USA
A. Barker, V.E. Barnes, S. Folgueras, L. Gutay, M.K. Jha, M. Jones, A.W. Jung, D.H. Miller,
N. Neumeister, J.F. Schulte, X. Shi, J. Sun, A. Svyatkovskiy, F. Wang, W. Xie, L. Xu Purdue University Calumet, Hammond, USA
N. Parashar, J. Stupak Rice University, Houston, USA y,
,
A. Adair, B. Akgun, Z. Chen, K.M. Ecklund, F.J.M. Geurts, M. Guilbaud, W. Li, B. Michlin,
M Northup B P Padley R Redjimi J Roberts J Rorie Z Tu J Zabel A. Adair, B. Akgun, Z. Chen, K.M. Ecklund, F.J.M. Geurts, M. Guilbaud, W. Li, B. M Adair, B. Akgun, Z. Chen, K.M. Ecklund, F.J.M. Geurts, M. Guilbaud, W. Li, B. Michli
Northup, B.P. Padley, R. Redjimi, J. Roberts, J. Rorie, Z. Tu, J. Zabel g
J
M. Northup, B.P. Padley, R. Redjimi, J. Roberts, J. Rorie, Z. Tu, J. Zabel M. Northup, B.P. Padley, R. Redjimi, J. Roberts, J. Rorie, Z. Tu, J. Zab Texas Tech University, Lubbock, USA Vanderbilt University, Nashville, USA University of Rochester, Rochester, USA B. Betchart, A. Bodek, P. de Barbaro, R. Demina, Y.t. Duh, T. Ferbel, M. Galanti, A. Garcia-
Bellido, J. Han, O. Hindrichs, A. Khukhunaishvili, K.H. Lo, P. Tan, M. Verzetti Rutgers, The State University of New Jersey, Piscataway, USA
A. Agapitos,
J.P. Chou,
E. Contreras-Campana,
Y. Gershtein,
T.A. G´omez Espinosa,
E. Halkiadakis, M. Heindl, D. Hidas, E. Hughes, S. Kaplan, R. Kunnawalkam Elayavalli,
S. Kyriacou, A. Lath, K. Nash, H. Saka, S. Salur, S. Schnetzer, D. Sheffield, S. Somalwar, R. Stone,
S. Thomas, P. Thomassen, M. Walker S. Thomas, P. Thomassen, M. Walker University of Tennessee, Knoxville, USA
A.G. Delannoy, M. Foerster, J. Heideman, G. Riley, K. Rose, S. Spanier, K. Thapa Texas A&M University, College Station, USA
O. Bouhali72, A. Celik, M. Dalchenko, M. De Mattia, A. Delgado, S. Dildick, R. Eusebi,
J. Gilmore, T. Huang, E. Juska, T. Kamon73, R. Mueller, Y. Pakhotin, R. Patel, A. Perloff,
L. Perni`e, D. Rathjens, A. Rose, A. Safonov, A. Tatarinov, K.A. Ulmer L. Perni`e, D. Rathjens, A. Rose, A. Safonov, A. Tatarinov, K.A. Ulmer Texas Tech University, Lubbock, USA
N. Akchurin, C. Cowden, J. Damgov, F. De Guio, C. Dragoiu, P.R. Dudero, J. Faulkner,
E. Gurpinar, S. Kunori, K. Lamichhane, S.W. Lee, T. Libeiro, T. Peltola, S. Undleeb, I. Volobouev,
Z. Wang †: Deceased 1: Also at Vienna University of Technology, Vienna, Austria 2: Also at State Key Laboratory of Nuclear Physics and Technology, Peking University, Beijing,
China 2: Also at State Key Laboratory of Nuclear Physics and Technology, Peking University, Beijing,
China 3: Also at Institut Pluridisciplinaire Hubert Curien, Universit´e de Strasbourg, Universit´e de
H
t Al
M lh
CNRS/IN2P3 St
b
F 3: Also at Institut Pluridisciplinaire Hubert Curien, Universit´e de Strasbourg, Universit´e de Haute Alsace Mulhouse, CNRS/IN2P3, Strasbourg, France Haute Alsace Mulhouse, CNRS/IN2P3, Strasbourg, France g
4: Also at Universidade Estadual de Campinas, Campinas, Brazil 4: Also at Universidade Estadual de Campinas, Campinas, Brazil 5: Also at Universidade Federal de Pelotas, Pelotas, Brazil 5: Also at Universidade Federal de Pelotas, Pelotas, Brazil 6: Also at Universit´e Libre de Bruxelles, Bruxelles, Belgium 6: Also at Universit´e Libre de Bruxelles, Bruxelles, Belgium 7: Also at Deutsches Elektronen-Synchrotron, Hamburg, Germany 7: Also at Deutsches Elektronen-Synchrotron, Hamburg, Germany 8: Also at Joint Institute for Nuclear Research, Dubna, Russia 8: Also at Joint Institute for Nuclear Research, Dubna, Russia 8: Also at Joint Institute for Nuclear Rese 9: Also at Suez University, Suez, Egypt 10: Now at British University in Egypt, Cairo, Egypt 10: Now at British University in Egypt, Cairo, Egypt 11: Also at Ain Shams University, Cairo, Egypt 11: Also at Ain Shams University, Cairo, Egypt 12: Now at Helwan University, Cairo, Egypt 12: Now at Helwan University, Cairo, Egypt 13: Also at Universit´e de Haute Alsace, Mulhouse, France 13: Also at Universit´e de Haute Alsace, Mulhouse, France 14: Also at Skobeltsyn Institute of Nuclear Physics, Lomonosov Moscow State University,
Moscow, Russia 14: Also at Skobeltsyn Institute of Nuclear Physics, Lomonosov Moscow State University,
Moscow, Russia 29: Also at Yazd University, Yazd, Iran 29: Also at Yazd University, Yazd, Iran
30: Also at Plasma Physics Research Center, Science and Research Branch, Islamic Azad 30: Also at Plasma Physics Research Center, Science and Research Branch, Islami University, Budapest, Hungary 22: Also at University of Debrecen, Debrecen, Hungary 22: Also at University of Debrecen, Debrecen, Hungary 23: Also at Indian Institute of Science Education and Re 23: Also at Indian Institute of Science Education and Research, Bhopal, India 24: Also at Institute of Physics, Bhubaneswar, India 24: Also at Institute of Physics, Bhubaneswar, India 25: Also at University of Visva-Bharati, Santiniketan, India 25: Also at University of Visva-Bharati, Santiniketan, India 26: Also at University of Ruhuna, Matara, Sri Lanka 26: Also at University of Ruhuna, Matara, Sri Lanka 27: Also at Isfahan University of Technology, Isfahan, Iran 27: Also at Isfahan University of Technology, Isfahan, Iran 28: Also at University of Tehran, Department of Engineering Science, Tehran, Iran 28: Also at University of Tehran, Department of Engineerin 28: Also at University of Tehran, Depart Moscow, Russia 15: Also at Tbilisi State University, Tbilisi, Georgia 15: Also at Tbilisi State University, Tbilisi, Georgia 16: Also at CERN, European Organization for Nuclear Research, Geneva, Switzerland p
g
17: Also at RWTH Aachen University, III. Physikalisches Institut A, Aachen, Germany 18: Also at University of Hamburg, Hamburg, Germany 18: Also at University of Hamburg, Hamburg, Germany 19: Also at Brandenburg University of Technology, Cottbus, Germany 19: Also at Brandenburg University of Technology, Cottbus, Germany 20: Also at Institute of Nuclear Research ATOMKI, Debrecen, Hungary 20: Also at Institute of Nuclear Research ATOMKI, Debrecen, Hungary f Nuclear Research ATOMKI, Debrecen, Hun 21: Also at MTA-ELTE Lend¨ulet CMS Particle and Nuclear Physics Group, E¨otv¨os Lor´and
University, Budapest, Hungary University, Tehran, Iran y
31: Also at Laboratori Nazionali di Legnaro dell’INFN, Legnaro, Italy 31: Also at Laboratori Nazionali di Legnaro dell’INFN, Legnaro, Italy 32: Also at Universit`a degli Studi di Siena, Siena, Italy 32: Also at Universit`a degli Studi di Siena, Siena, Italy Vanderbilt University, Nashville, USA Vanderbilt University, Nashville, USA
S. Greene, A. Gurrola, R. Janjam, W. Johns, C. Maguire, A. Melo, H. Ni, P. Sheldon, S. Tuo,
J. Velkovska, Q. Xu University of Virginia, Charlottesville, USA
M.W. Arenton, P. Barria, B. Cox, J. Goodell, R. Hirosky, A. Ledovskoy, H. Li, C. Neu,
T. Sinthuprasith, X. Sun, Y. Wang, E. Wolfe, F. Xia 63 Wayne State University, Detroit, USA
C. Clarke, R. Harr, P.E. Karchin, J. Sturdy University of Wisconsin - Madison, Madison, WI, USA y
,
,
,
D.A. Belknap, C. Caillol, S. Dasu, L. Dodd, S. Duric, B. Gomber, M. Grothe, M. Herndon,
A. Herv´e, P. Klabbers, A. Lanaro, A. Levine, K. Long, R. Loveless, I. Ojalvo, T. Perry, G.A. Pierro,
G. Polese, T. Ruggles, A. Savin, N. Smith, W.H. Smith, D. Taylor, N. Woods D.A. Belknap, C. Caillol, S. Dasu, L. Dodd, S. Duric, B. Gomber, M. Grothe, M. Herndon,
A. Herv´e, P. Klabbers, A. Lanaro, A. Levine, K. Long, R. Loveless, I. Ojalvo, T. Perry, G.A. Pierro,
G. Polese, T. Ruggles, A. Savin, N. Smith, W.H. Smith, D. Taylor, N. Woods 55: Also at Piri Reis University, Istanbul, Turkey 55: Also at Piri Reis University, Istanbul, Turkey 56: Also at Ozyegin University, Istanbul, Turkey 57: Also at Izmir Institute of Technology, Izmir, Turkey 57: Also at Izmir Institute of Technology, Izmir, Turkey 58: Also at Marmara University, Istanbul, Turkey 58: Also at Marmara University, Istanbul, Turkey 59: Also at Kafkas University, Kars, Turkey 60: Also at Istanbul Bilgi University, Istanbul, Turkey 60: Also at Istanbul Bilgi University, Istanbul, Turkey 61: Also at Yildiz Technical University, Istanbul, Turkey 61: Also at Yildiz Technical University, Istanbul, Turkey 62: Also at Hacettepe University, Ankara, Turkey 62: Also at Hacettepe University, Ankara, Turkey 63: Also at Rutherford Appleton Laboratory, Didcot, United Kingdom 63: Also at Rutherford Appleton Laboratory, Didcot, United Kingdom 64: Also at School of Physics and Astronomy, University of Southampton, Southampton,
United Kingdom 64: Also at School of Physics and Astronomy, University of Southampton, Southampton,
U i
d Ki
d United Kingdom 65: Also at Instituto de Astrof´ısica de Canarias, La Laguna, Spain lso at Instituto de Astrof´ısica de Canarias, La : Also at Utah Valley University, Orem, USA 67: Also at University of Belgrade, Faculty of Physics and Vinca Institute of Nuclear Sciences,
Belgrade, Serbia 33: Also at Purdue University, West Lafayette, USA Petersburg, Russia 41: Also at University of Florida, Gainesville, USA gy
43: Also at Budker Institute of Nuclear Physics, Novosibirsk, Russi 43: Also at Budker Institute of Nuclear Physics, Novosibirsk, Russia y
44: Also at Faculty of Physics, University of Belgrade, Belgrade, Serbia y
44: Also at Faculty of Physics, University of Belgrade, Belgrade, Serbia y
y
y
g
g
45: Also at INFN Sezione di Roma; Universit`a di Roma, Roma, Italy 45: Also at INFN Sezione di Roma; Universit`a di Roma, Roma, Italy 46: Also at Scuola Normale e Sezione dell’INFN, Pisa, Italy 46: Also at Scuola Normale e Sezione dell’INFN, Pisa, Italy y
47: Also at National and Kapodistrian University of Athens, Athens, Greece 47: Also at National and Kapodistrian University of Athens, Athens, Greece 48: Also at Riga Technical University, Riga, Latvia 49: Also at Institute for Theoretical and Experimental Physics, Moscow, Russia 9: Also at Institute for Theoretical and Experimental Physics, Moscow, Russia 50: Also at Albert Einstein Center for Fundamental Physics, Bern, Switzerland 50: Also at Albert Einstein Center for Fundamental Physics, Bern, Switzerland 51: Also at Gaziosmanpasa University, Tokat, Turkey 51: Also at Gaziosmanpasa University, Tokat, Turkey 52: Also at Adiyaman University, Adiyaman, Turkey 53: Also at Mersin University, Mersin, Turkey 54: Also at Cag University, Mersin, Turkey 33: Also at Purdue University, West Lafayette, USA 33: Also at Purdue University, West Lafayette, USA 34: Also at International Islamic University of Malaysia, Kuala Lumpur, Malaysia 34: Also at International Islamic University of Malaysia, Kuala Lumpur, Malaysia 35: Also at Malaysian Nuclear Agency, MOSTI, Kajang, Malaysia 35: Also at Malaysian Nuclear Agency, MOSTI, Kajang, Malaysia 36: Also at Consejo Nacional de Ciencia y Tecnolog´ıa, Mexico city, Mexico 36: Also at Consejo Nacional de Ciencia y Tecnolog´ıa, Mexico city, Mexico C
The CMS Collaboration 64 Also at Warsaw University of Technology, Institute of Electronic Systems, Warsaw, Poland 37: Also at Warsaw University of Technology, Institute of Electronic Systems, Warsaw, Poland
38: Also at Institute for Nuclear Research, Moscow, Russia 37: Also at Warsaw University of Technology, Institute of Electronic Systems, Warsaw, Poland
38: Also at Institute for Nuclear Research, Moscow, Russia y
gy
y
38: Also at Institute for Nuclear Research, Moscow, Russia y
gy
38: Also at Institute for Nuclear Research, Moscow, Russia 38: Also at Institute for Nuclear Research, Moscow, Russia 39:
Now
at
National
Research
Nuclear
University
’Moscow
Engineering
Physics
Institute’ (MEPhI), Moscow, Russia 39:
Now
at
National
Research
Nuclear
University
’Moscow
Engineering
Physics
Institute’ (MEPhI), Moscow, Russia (
)
40: Also at St. Petersburg State Polytechnical University, St. Petersburg, Russia 40: Also at St. Petersburg State Polytechnical University, St. Belgrade, Serbia g
68: Also at Facolt`a Ingegneria, Universit`a di Roma, Roma, Italy 68: Also at Facolt`a Ingegneria, Universit`a di Roma, Roma, Italy 69: Also at Argonne National Laboratory, Argonne, USA 69: Also at Argonne National Laboratory, Argonne, USA 70: Also at Erzincan University, Erzincan, Turkey 70: Also at Erzincan University, Erzincan, Turkey y
y
71: Also at Mimar Sinan University, Istanbul, Istanbul, Turkey 71: Also at Mimar Sinan University, Istanbul, Istanbul, Turkey 72: Also at Texas A&M University at Qatar, Doha, Qatar 72: Also at Texas A&M University at Qatar, Doha, Qatar 73: Also at Kyungpook National University, Daegu, Korea 73: Also at Kyungpook National University, Daegu, Korea
|
https://openalex.org/W2944090517
|
https://curis.ku.dk/ws/files/218606554/s41467_019_10130_2.pdf
|
English
| null |
Theoretical analysis of Polycomb-Trithorax systems predicts that poised chromatin is bistable and not bivalent
|
Nature communications
| 2,019
|
cc-by
| 22,340
|
Citation for published version (APA):
Sneppen, K., & Ringrose, L. (2019). Theoretical analysis of Polycomb-Trithorax systems predicts that poised
chromatin is bistable and not bivalent. Nature Communications, 10, [2133]. https://doi.org/10.1038/s41467-019-
10130-2 Download date: 24. Oct. 2024 university of copenhagen university of copenhagen
Theoretical analysis of Polycomb-Trithorax systems predicts that poised chromatin is
bistable and not bivalent
Sneppen, Kim; Ringrose, Leonie university of copenhagen Theoretical analysis of Polycomb-Trithorax systems predicts that poised chromatin is
bistable and not bivalent Sneppen, Kim; Ringrose, Leonie Published in:
Nature Communications Document version
Publisher's PDF, also known as Version of record Document license:
CC BY Citation for published version (APA):
Sneppen, K., & Ringrose, L. (2019). Theoretical analysis of Polycomb-Trithorax systems predicts that poised
chromatin is bistable and not bivalent. Nature Communications, 10, [2133]. https://doi.org/10.1038/s41467-019-
10130-2 Download date: 24. Oct. 2024 ARTICLE 1 Center for Models of Life, Niels Bohr Institute, University of Copenhagen, Blegdamsvej 17, 2100 Copenhagen, Denmark. 2 Integrated Research Institute for
Life Sciences, Humboldt-Universität zu Berlin, Philippstrasse 13, Haus 22, 10115 Berlin, Germany. 3These authors contributed equally: Kim Sneppen, Leonie
Ringrose. Correspondence and requests for materials should be addressed to K.S. (email: ksneppen@gmail.com)
or to L.R. (email: leonie.ringrose@hu-berlin.de) Theoretical analysis of Polycomb-Trithorax
systems predicts that poised chromatin is bistable
and not bivalent Kim Sneppen
1,3 & Leonie Ringrose
2,3 Polycomb (PcG) and Trithorax (TrxG) group proteins give stable epigenetic memory of silent
and active gene expression states, but also allow poised states in pluripotent cells. Here we
systematically address the relationship between poised, active and silent chromatin, by
integrating 73 publications on PcG/TrxG biochemistry into a mathematical model comprising
144 nucleosome modification states and 8 enzymatic reactions. Our model predicts that
poised chromatin is bistable and not bivalent. Bivalent chromatin, containing opposing active
and silent modifications, is present as an unstable background population in all system states,
and different subtypes co-occur with active and silent chromatin. In contrast, bistability, in
which the system switches frequently between stable active and silent states, occurs under a
wide range of conditions at the transition between monostable active and silent system
states. By proposing that bistability and not bivalency is associated with poised chromatin,
this work has implications for understanding the molecular nature of pluripotency. 1 Center for Models of Life, Niels Bohr Institute, University of Copenhagen, Blegdamsvej 17, 2100 Copenhagen, Denmark. 2 Integrated Research Institute for
Life Sciences, Humboldt-Universität zu Berlin, Philippstrasse 13, Haus 22, 10115 Berlin, Germany. 3These authors contributed equally: Kim Sneppen, Leonie
Ringrose. Correspondence and requests for materials should be addressed to K.S. (email: ksneppen@gmail.com)
or to L.R. (email: leonie.ringrose@hu-berlin.de) 1 NATURE COMMUNICATIONS | (2019) 10:2133 | https://doi.org/10.1038/s41467-019-10130-2 | www.nature.com/naturecommunications ARTICLE NATURE COMMUNICATIONS | https://doi.org/10.1038/s41467-019-10130-2 indeed exist in several cell types29,30,32–34. These studies have
also given more accurate estimations of the true number of
bivalent loci, and how their distribution changes in different
cell types29,30,34. However, the central question of whether
bivalent chromatin is important for defining the poised state
and whether it is required for pluripotency is extremely difficult
to address experimentally19,20. By definition, bivalent chroma-
tin comprises chromatin modifications that are each involved in
activation and silencing of many hundreds of genes. Thus any
experiment that perturbs bivalent chromatin has a large impact
on the PcG/TrxG system as a whole, rendering results difficult
to interpret. B B
istable systems can adopt two mutually exclusive stable
states1–3. Bistability is central to many epigenetic gene
regulatory systems4. Epigenetic bistability can be visualised
as variegation in reporter assays5–7. Theoretical analysis of Polycomb-Trithorax
systems predicts that poised chromatin is bistable
and not bivalent Reporters carrying cis- reg-
ulatory elements that recruit epigenetic regulators5,7, or in which
the reporter is placed near to heterochromatin6, show early
random switching and later stable maintenance, resulting in
lineages of cells in which the reporter is either off or on, and in
which each state is mitotically inherited. Several theoretical
models address the potential of bistable systems to maintain
epigenetic memory via histone modifications that are dynamic
and reversible8–11. In these models, a given nucleosome can
change its state repeatedly over time, but the system as a whole is
stably in an active or silent state, and can survive the disruptions
of simulated replication, provided sufficient feedbacks between
nucleosomes are present12,13. p
To understand the relevance of bivalent chromatin and how
it relates to other properties of the PcG/TrxG system, theoretical
approaches
have
immense
potential. However,
previous
theoretical models for PcG/TrxG regulation have not considered
bivalent states and the full biochemical complexity of the sys-
tem8–11,35. The greatest obstacle to the integration of individual
experimental observations into a coherent whole has been the
lack of a comprehensive theoretical framework. In this study, we
curated current literature on PcG/TrxG biochemistry, comprising
73 publications, and formalised the resulting information as a
dynamic stochastic mathematical model with 144 nucleosome
states. Surprisingly, despite the possibility to adopt 144 states, the
model has strongly bistable properties, preferring to occupy only
the most extreme active or silent states. The model predicts that
bivalent states also exist, but that they are present as a dynamic,
unstable background population in all system states. Distinct
forms of bivalent chromatin preferentially co- occur with active
and silent system states. Importantly, midway in the transition
between active and silent states, poised chromatin is not bivalent
in the model, but is robustly bistable, and differs from monostable
modes only in its higher frequency of switching. Furthermore, we
show that several published observations strongly support the
model predictions. We propose a central role for bistability in
PcG/TrxG function, not only for ensuring epigenetic memory but
also as a central feature of poised chromatin. Thus, this work has
profound implications for understanding the molecular nature of
pluripotency and the stability and reversibility of epigenetic states. The Polycomb (PcG) and Trithorax (TrxG) group proteins are
essential epigenetic regulators that can maintain stable epigenetic
memory of silent states (via PcG) and active states (via TrxG) of
their target genes over many cell generations14. Theoretical analysis of Polycomb-Trithorax
systems predicts that poised chromatin is bistable
and not bivalent Reporter genes
carrying Polycomb/Trithorax Response Elements (PRE/TREs)
can show variegation, depending on genomic location and DNA
sequence of the PRE/TRE, indicating that the system has bistable
properties5,15. A potential biochemical basis for this bistability is
beginning to emerge (reviewed in16–18). The PcG and TrxG
proteins modify chromatin, and their biochemistry is exquisitely
complex. In the last two decades, over 70 publications have
documented biochemical properties of PcG and TrxG proteins
and complexes (see Tables 1–3 and Methods). At least ten specific
chromatin modifications catalysed by PcG and TrxG proteins
have been identified, and the enzymes that add and remove them
are well characterised. Furthermore, several self- reinforcing and
antagonistic interactions exist, suggesting a potential molecular
basis for bistability in the PcG/TrxG system18 (for full list, see
Tables 1-3 and Methods). However the immense complexity of
the system means that it also has the potential to adopt more than
two states. Indeed, it has been proposed that a third, bivalent, state is
essential in pluripotent stem cells. Bivalent chromatin contains
histone modifications catalysed by both PcG and TrxG proteins,
and is present in mouse and human ESCs and several other
vertebrate cell types19,20. Bivalent chromatin is thought to
represent a poised or undecided form that is resolved to a stably
active or silent form by removal of one or other type of mod-
ification upon appropriate signals21–25. The best - studied biva-
lent
chromatin
form
is
that
containing
H3K27me3
and
H3K4me3 (histone H3 tritmethylated at lysine 27 or lysine 4). Other bivalent forms include H2A K119 ubiquitination (H2Aub),
catalysed by the PRC1 complex, in addition to H3K4me3, RNA
Polymerase, and/ or H3K27me326–28. Genes residing in bivalent
chromatin are typically silent or expressed at low levels21–24,28–30. These genes remain silent upon loss of H3K4me3, and become
activated upon removal of H3K27me3 or H2AUb21,22,25,26,28,31. Results Later studies have used co- ChIP, re-ChIP,
mass spectrometry and imaging approaches to confirm that
bivalent chromatin containing H3K27me3 and H3K4me3 on
the same or adjacent nucleosomes on the same allele does NATURE COMMUNICATIONS | (2019) 10:2133 | https://doi.org/10.1038/s41467-019-10130-2 | www.nature.com/naturecommunications 2 ARTICLE ARTICLE NATURE COMMUNICATIONS | https://doi.org/10.1038/s41467-019-10130-2 Table 1 Enzymes and histone modifications represented in the model
Activity
Enzyme or complex (Drosophila)
Enzyme or complex (vertebrate)
Model name
H3K27me3 addition
PRC2 (subunit E(Z)) (H3K27me3)79,80
PRC2 (subunit EZH2) (H3K27me3)77,78
PRC2
H3K27me3 removal
dUTX85
UTX (KDM6A) and JMJD3 (KDM6B)46,82–84
UTX
H2Aub addition
PRC1 subunit dRING (monoubiquitinates H2AK118)88
PRC1 subunits RING1A and RING1B (monoubiquitinate
H2AK119)88–90
PRC1
H2Aub removal
PR-DUB subunit BAP-193
PR-DUB subunit BAP-194
PR-DUB
H3K27ac addition
CBP (acetylates several residues including H3K27)96
CBP and p300 (acetylate several residues including
H3K27)40
CBP
H3K27ac removal
RPD3 (deacetylates several residues including
H3K27)96
NuRD (deacetylates several residues including H3K27)97
NURD
H3K4me addition
TRX (H3K4me1)42 TRX (H3K4me2)105 SET1
(H3K4me3)103,104
MLL1 (H3K4me1)42 MLL1 (H3K4me2)105 SET1
(H3K4me3)101,102
TRXG
H3K4me removal
Lid (KDM5 homolog)112,113. dKDM2115
JARID1A (synonyms: RBP2, KDM5A)111. KDM2B86
KDM
H3K36me addition
ASH1 (H3K36me2)43,107,108
ASH1L (H3K36me2)109,110
TRXG
H3K36me removal
Fbxl10 (synonyms: dKDM2, JHDM1A)91
Fbxl10 (synonyms: KDM2B, JHDM1A)114
KDM Table 1 Enzymes and histone modifications represented in the model Table 2 Self- reinforcing interactions
Molecule
Activity
Drosophila
Vertebrate
H3K27me3
Binds PRC1
Pc chromodomain binds H3K27me3116–118
CBX2 and CBX7 chromodomains (PC homologs)
bind H3K27me3119
H3K27me3
Binds and
stimulates PRC2
Genetic evidence for similar mechanism in
Drosophila121. H3K27me3 binds PRC2120,121. H3K27me3 stimulates
PRC2 activity121. H2Aub
Binds and
stimulates PRC2
Binds PRC2122. Binds PRC2122–125. Binding stimulates activity122
H3K4me
Stimulates CBP
H3K4me1 stimulates CBP in H3K27 acteylation42,96
H3K27ac
Enhances TRXG binding
to chromatin
FSH1 (homolog of BRD4) binds and colocalises with
acetylated histones and interacts with ASH1127,128
TrxG protein BRD4 binds acetylated histones in vitro
and in vivo142
ASH1 and TRX
Interact physically and
functionally
In vitro binding and in vivo colocalisation, genetic
interaction37. TRX chromatin association depends
on ASH136
In vivo colocalisation on single genes38. Colocalisation of MLL and ASH1L-catalysed
H3K36me2 genome wide. MLL chromatin binding
depends on H3K36me2 binding39
TRXG and CBP
Interact physically and
functionally
ASH1 and CBP interact physically and
functionally138. The TrxG protein BRM is associated
with CBP and stimulates the activity of CBP in
acetylating H3K27139
KDM
and NURD
Interact physically and
functionally
H3K4 demethylase Lid interacts physically and
functionally with RPD3136. H3K4/K6 demethylase LSD1 is a component of the
NuRD complex at active enhancers137. Results A comprehensive model of PcG/TrxG regulation. In the current
work, we aimed to investigate emergent properties of the PcG/
TrxG system by taking account of its full complexity. To achieve
this, we curated all available current literature comprising 73
publications (see Tables 1–3 and “Methods” section), and for-
malised the observations in a dynamic stochastic model (Fig. 1). The model contains all biochemical properties of the system thus
far reported, including the possibility for bivalent states. We
nevertheless introduced various simplifications as follows (the
model is explained in detail in Methods): p
The model contains a reduced number of histone modifica-
tions. We use a dynamic stochastic model, formulated in terms
of known PcG and TrxG enzymes and nucleosome modifica-
tions. We consider four nucleosome modifications: PcG-
catalysed H3K27 methylation and H2AK119 ubiquitination,
and TrxG-catalysed H3K27 acetylation and H3K4/K36 methy-
lation14. The rationale for fusing H3K4 and HK36 methylation,
which are catalysed by different TrxG proteins, is based on
observations that these proteins interact physically and func-
tionally, frequently colocalise, and that they can recruit each
other (36–39, see Tables 1–3 and “Methods” section for more
detail). The model consists of an array of nucleosomes, each of
which can carry combinations of these four modifications
(Fig. 1a). The proposition that bivalent chromatin may function as
poised chromatin for the PcG/TrxG system is based on indirect
evidence. The first studies to identify bivalent chromatin gen-
ome- wide did so using separate single – antibody ChIP
experiments for each of the modifications in question21–25. This
approach does not distinguish between true bivalent chromatin
and a mixture of different states in different cells. The idea that
bivalent marks are essential to pluripotency and resolve upon
differentiation
was
based
upon
these
single-antibody
experiments21,22. Results Table 2 Self- reinforcing interactions H3K4/K6 demethylase LSD1 is a component of the
NuRD complex at active enhancers137. Table 3 Antagonistic interactions
Molecule
Activity
Drosophila
Vertebrate
H3K27me3
Inhibits TRXG
binding
Human SET1 and MLL1 complexes bind poorly to
H3K27me3 histones. Catalytic activity is not
prevented134
H2Aub
Inhibits TRXG
activity
Histone H2A ubiquitination inhibits the enzymatic
activity of H3 lysine 36 methyltransferases45. H3K4/K36me
Inhibits PRC2 activity
H3K4 and H3K36 methylation inhibit PRC2
H3K27 methylation activity43,44
H3K4 and H3K36 methylation inhibit PRC2 H3K27
methylation activity32,44. PRC1 and CBP
PC inhibits CBP
activity
Polycomb (PC, subunit of PRC1) inhibits histone
acetylation mediated by CBP by binding directly to
the CBP catalytic domain133
PRC2 and KDM
Interact physically
and functionally
PRC2 recruits RBP2 (H3K4 demethylase)111 and LSD1
(H3K4 demethylase)135
PRC1 and KDM
Interact physically
and functionally
Non-canonical PRC1 complex dRAF contains
dRING (ubiquitin ligase) and dKDM2 (H3K4 and
H3K36 demethylase)115,91
Non-canonical PRC1 complex PRC1.1 contains dRING
(ubiquitin ligase) and KDM2B (H3K4 and H3K36
demethylase)86,114,92
TRXG and UTX
Interact physically
and functionally
UTX (H3K27 demethylase) is associated with CBP
and the TrxG protein BRM139
UTX (H3K27 demethylase) is associated with MLL 2/3
(vertebrate homologs of TRX)46 Table 3 Antagonistic interactions Vertebrate PRC2 recruits RBP2 (H3K4 demethylase)111 and LSD1
(H3K4 demethylase)135
Non-canonical PRC1 complex PRC1.1 contains dRING
(ubiquitin ligase) and KDM2B (H3K4 and H3K36
demethylase)86,114,92 UTX (H3K27 demethylase) is associated with MLL 2/3
(vertebrate homologs of TRX)46 NATURE COMMUNICATIONS | (2019) 10:2133 | https://doi.org/10.1038/s41467-019-10130-2 | www.nature.com/naturecommunications 3 NATURE COMMUNICATIONS | https://doi.org/10.1038/s41467-019-10130-2 ARTICLE del is based on half-nucleosomes, and combines them
ole nucleosomes. The model is formulated in terms of
t unit possible, namely the half-nucleosome containing
py of H3 and H2A. Whole nucleosomes each comprise
ucleosomes which are paired for the duration of the
scored as all whole nucleosomes containing a mix of op
modifications (Fig. 2). The model allows combinatorial modifications. Eac
nucleosome in the model can be unmodified or can car
combination of three modifications but not all four s
KDM
TRXG
H3K4/36me
(1,2,5,6,9,10)
H2Aub
(2,3,6,7,10,11)
H3K27ac
(1–4)
H3K27me3
(9–12)
b
c
Read half-nucleosome 1
d
PRC1
PRC2
PRC2:KDM
PRC2:KDM:NURD
PRC1
PRC2
PRC2:KDM
PRC2:KDM:NURD
NURD
PRC2:KDM:NURD
PRC2:KDM
CBP
CBP
PRC2
PRC2:KDM
PRC2:KDM:NURD
TRXG:CBP
TRXG:UTX
PRC2:KDM
PRC2:KDM:NURD
TRXG
TRXG:CBP
TRXG:UTX
TRXG
TRXG:CBP
TRXG:UTX
TRXG:CBP
TRXG:UTX
PRC2
PRC2:KDM
PRC2:KDM:NURD
H3K27ac
H3K4/36me
H3K27ac
H3K4/36me
H2Aub
2
1
4
3
H3K27ac
H3K27ac
H2Aub
H3K4/36me
H3K4/36me
H2Aub
H2Aub
H3K4/36me
H3K27me3
8
7
5
6
H3K4/36me
H3K27me3
H3K27me3
H3K27me3
H2Aub
H2Aub
12
11
9
10
Active
Silent
Bivalent
Unmodified
H3K27me3
H2Aub
Write to half-nucleosome 2
PR-DUB
PRC1
PR-DUB
PRC1
PR-DUB
PRC1
PR-DUB
PRC1
PR-DUB
PRC1
PR-DUB
PRC1
KDM
KDM
KDM
KDM
TRXG
KDM
TRXG
CBP
UTX
NURD
CBP
NURD
CBP
NURD
CBP
UTX
PRC2
UTX
PRC2
UTX
PRC2:KDM:NURD
PRC2:KDM
PRC1
PRC2:KDM:NURD
PRC2:KDM
PRC1
5
6
7
8
9
10
11
12
1
2
3
4
PRC1
ub
H2AK118/119 ubiquitylation
H3K27 methylation
H3K4/K36 methylation
H3K27 acetylation
a
Active
Silent
Bivalent
Unmodified
H3K27ac
H3K4/36me
H3K27me3
H2Aub
CBP
TRXG
PRC2
PRC1
NURD
KDM
UTX
PR-DUB
Modification
Writer
Eraser
NURD
PR-DUB
High
Low
KDM
Low
High
High
Low
UTX
Low
High
LE
NATURE COMMUNICATIONS | https://doi.org/10.1038/s41467-019- del is based on half-nucleosomes, and combines them
ole nucleosomes. The model is formulated in terms of
t unit possible namely the half nucleosome containing
scored as all whole nucleosomes containing a mix of
modifications (Fig. 2). NATURE COMMUNICATIONS | https://doi.org/10.1038/s41467-019-10130-2 The model allows combinatorial modifications
E
KDM
TRXG
H3K4/36me
(1,2,5,6,9,10)
H2Aub
(2,3,6,7,10,11)
H3K27ac
(1–4)
H3K27me3
(9–12)
b
c
Read half-nucleosome 1
d
PRC1
PRC2
PRC2:KDM
PRC2:KDM:NURD
PRC1
PRC2
PRC2:KDM
PRC2:KDM:NURD
NURD
PRC2:KDM:NURD
PRC2:KDM
CBP
CBP
PRC2
PRC2:KDM
PRC2:KDM:NURD
TRXG:CBP
TRXG:UTX
PRC2:KDM
PRC2:KDM:NURD
TRXG
TRXG:CBP
TRXG:UTX
TRXG
TRXG:CBP
TRXG:UTX
TRXG:CBP
TRXG:UTX
PRC2
PRC2:KDM
PRC2:KDM:NURD
H3K27ac
H3K4/36me
H3K27ac
H3K4/36me
H2Aub
2
1
4
3
H3K27ac
H3K27ac
H2Aub
H3K4/36me
H3K4/36me
H2Aub
H2Aub
H3K4/36me
H3K27me3
8
7
5
6
H3K4/36me
H3K27me3
H3K27me3
H3K27me3
H2Aub
H2Aub
12
11
9
10
Active
Silent
Bivalent
Unmodified
H3K27me3
H2Aub
Write to half-nucleosome 2
PR-DUB
PRC1
PR-DUB
PRC1
PR-DUB
PRC1
PR-DUB
PRC1
PR-DUB
PRC1
PR-DUB
PRC1
KDM
KDM
KDM
KDM
TRXG
KDM
TRXG
CBP
UTX
NURD
CBP
NURD
CBP
NURD
CBP
UTX
PRC2
UTX
PRC2
UTX
PRC2:KDM:NURD
PRC2:KDM
PRC1
PRC2:KDM:NURD
PRC2:KDM
PRC1
5
6
7
8
9
10
11
12
1
2
3
4
PRC1
ub
H2AK118/119 ubiquitylation
H3K27 methylation
H3K4/K36 methylation
H3K27 acetylation
a
Active
Silent
Bivalent
Unmodified
H3K27ac
H3K4/36me
H3K27me3
H2Aub
CBP
TRXG
PRC2
PRC1
NURD
KDM
UTX
PR-DUB
Modification
Writer
Eraser
NURD
PR-DUB
High
Low
KDM
Low
High
High
Low
UTX
Low
High a H3K4/36me
(1,2,5,6,9,10)
H2Aub
(2,3,6,7,10,11)
H3K27ac
(1–4)
H3K27me3
(9–12)
b
Active
Silent
Bivalent
Unmodified
5
6
7
8
9
10
11
12
1
2
3
4
H2AK118/119 ubiquitylation
H3K27 methylation
H3K4/K36 methylation
H3K27 acetylation
NURD
PR-DUB
High
Low
KDM
Low
High
High
Low
UTX
Low
High H3K4/36me
(1,2,5,6,9,10)
H2Aub
(2,3,6,7,10,11)
H3K27ac
(1–4)
H3K27me3
(9–12)
b
5
6
7
8
9
10
11
12
1
2
3
4
H2AK118/119 ubiquitylation
NURD
PR-DUB
High
Low
KDM
Low
High
High
Low
UTX
Low
High b c el is based on half-nucleosomes, and combines them
le nucleosomes. The model is formulated in terms of
unit possible, namely the half-nucleosome containing
y of H3 and H2A. Whole nucleosomes each comprise
cleosomes, which are paired for the duration of the
except during replication. Bivalent nucleosomes are
scored as all whole nucleosomes containing a
modifications (Fig. 2). NATURE COMMUNICATIONS | (2019) 10:2133 | https://doi.org/10.1038/s41467-019-10130-2 | www.nature.com/naturecommunicatio NATURE COMMUNICATIONS | https://doi.org/10.1038/s41467-019-10130-2 Each of these 12 half-
nucleosomes can be paired with any other, giving 144 possible
whole-nucleosome modification states (Fig. 2). that inhibits it (Fig. 1d, grey arrows; see Methods for details). Third, several physical interactions between proteins with
different enzymatic activities have been described, for example
TRXG proteins recruit the H3K27me3 demethylase UTX46. The
model explicitly describes all compound complexes for which we
found evidence (e.g., TRXG:UTX Fig. 1d; see Tables 1–3 and
“Methods” section). In reality these interactions may be highly
regulated19,20. In the model their presence or absence can be
simulated by adaptation of specific parameters. Finally, we do not include transcription. The model deals only
with the enzymatic modifications of nucleosomes themselves, and
does not include the interactions of these histone modifications
with polymerase, or with transcription itself10,11,26,28. However,
for
simplicity,
we
designate
PcG-catalysed
modifications
(H3K27me and H2Aub) as silent, and those catalysed by TrxG
(H3K4/K36me and H3K27ac) as active. We further designate
system states (the average modification state of the whole array)
as active or silent when dominated by one or other type of
modification. This nomenclature does not imply any claim to the
transcriptional state of the locus, and whether transcription is a
cause or consequence of histone modification. y
p
p
p
Finally, there is ample evidence that recruitment of PcG and
TrxG proteins occurs not only via existing histone modifications
but can also be achieved by independent means, including direct
specific and non- specific DNA binding (reviewed in ref. 17, for
non- specific DNA binding see refs. 47–50). In addition, histone
modifications can be erased not only by enzymes but also by any
process that removes or remodels nucleosomes themselves8. For
simplicity we designate all such events in the model as direct
conversions, meaning that they are independent of recruitment of
an enzyme by existing histone modifications. The rate of direct
conversion can be adjusted for each transition separately. However, in the absence of quantitative data, we use a single
parameter ‘beta’ for most direct conversions (see “Methods”
section). Remarkably, we found that evidence exists for every
enzymatic reaction (Table 1a), and for the majority of self-
reinforcing and antagonistic interactions (Table 1b) in the model
in both flies and vertebrates. Thus the model is potentially equally
relevant to both (reviewed in17, see “Methods” section for detailed
references). NATURE COMMUNICATIONS | https://doi.org/10.1038/s41467-019-10130-2 In summary, the model unifies current literature on
the biochemistry of Polycomb/Trithorax regulation into a single
coherent framework. Each of the modifications described in the model is reversible,
and the complexes or enzymes that catalyse their addition
(writers) and removal (erasers) have been identified in flies and
vertebrates (Fig. 1a, right;14, see Methods and Tables 1–3 for
detailed references). Figure 1b represents the 12 possible half-
nucleosome states in a three-dimensional geometry, in which
adjacent states differ from one another by a single modification. Figure 1d shows this same geometry viewed from above, showing
the writers and erasers for transitions between states. The
exquisite complexity of PcG and TrxG chromatin comprises
several layers. First, there are numerous examples of nucleosome
modifications that stimulate or recruit specific writers and erasers. For example, H3K27me3 can enhance the binding of both PRC1
and PRC2 (14, see “Methods” section and Tables 1–3 for full list). We formalised these observations in the model by imposing
appropriate positive feedbacks, whereby existing nucleosome
states in the array affect the probability of new modifications to
the array. These feedbacks are implemented at the level of cross-
talk between two half- nucleosomes: Half-nucleosome 1 is
selected at random from the array (Fig. 1c). Specific modifications
on half-nucleosome 1 can recruit readers (text above each state in
Fig. 1d). Half-nucleosome 2 is selected at random. A reader
recruited by half-nucleosome 1 can add (write) or remove (erase)
a single modification on half-nucleosome 2, changing its state by
one step (see Fig. 1c for example). Any half-nucleosome in the
array can stimulate the conversion of any other, except itself. Bivalent nucleosomes fall into several categories. Bivalent
nucleosomes have been defined experimentally as those that
contain opposite modifications on the same nucleosome, but not
necessarily on the same histone29,32,34. In order to allow eva-
luation of full nucleosome modification states in the model, we
assigned each half- nucleosome in the array to a partner, so that
each pair of nucleosomes represents a full nucleosome with two
H2A and two H3 tails (Fig. 2a). Figure 2b shows the 144 full-
nucleosome modification states that emerge from all possible
pairings of half-nucleosomes. We classify nucleosomes into
categories according to the proportion of active and silent mod-
ifications they contain, giving 15 forms that each contain only
active (dark green, Fig. 2b) or only silent (dark red) modifications,
and one that is unmodified (grey, Fig. NATURE COMMUNICATIONS | https://doi.org/10.1038/s41467-019-10130-2 Fig. 1 A comprehensive model of Polycomb/Trithorax regulation. a Left: The model considers four histone modifications catalysed by TrxG and PcG
proteins. Middle: The model consists of an array of half-nucleosomes, each of which can carry one or more of these modifications, and are combined to Fig. 1 A comprehensive model of Polycomb/Trithorax regulation. a Left: The model considers four histone modifications catalysed by TrxG and PcG
proteins. Middle: The model consists of an array of half-nucleosomes, each of which can carry one or more of these modifications, and are combined to
give whole nucleosomes (see Fig. 2b). Right: the writers and erasers for each of these modifications are included in the model (see also Table 1). A
simplified nomenclature is used throughout as indicated. TRXG indicates both H3K4 and H3K36 methyltransferases. KDM indicates both H3K4 and H3K36
demethylases. b There are 12 different possible modification states for each half-nucleosome (containing one copy of H3 and one of H2A), shown in a
three-dimensional geometry. Adjacent states in horizontal and vertical directions are different from one another by one modification. Modifications in
common for given half-nucleosome states are indicated. Grey arrows indicate the direction in which the system state is pushed by raising or lowering the
activities of the complexes indicated. c Model logic. Half-nucleosome 1 is selected at random from the array, and recruits a complex (reader) depending on
modifications. The recruited complex attempts to modify (write to) a second randomly selected half-nucleosome. d Full model. The diagram shown in (a) is
viewed from above. Half-nucleosome states are numbered 1-12 with modifications as shown. Writers and erasers: Coloured arrows indicate transitions
between half-nucleosome states for half-nucleosome 2. Red: towards silencing, green: towards activation. Complexes responsible for each transition are
indicated on the arrows. Grey arrows: if the complex responsible for this transition is inhibited by an existing modification on half-nucleosome 2, the
transition occurs only via the direct conversion parameter beta (see Methods). Readers: Complexes that can be recruited by half-nucleosome 1 in a given
state are shown above each state, with red or green text indicating complexes that favour silencing or activation respectively. Compound complexes (e.g.,
TRXG:CBP) are included where evidence exists for physical interaction (see Methods and Tables 1–3 for details) mutually exclusive on the same H3 tail40–42. The model thus
allows 12 possible modification states for each half-nucleosome,
including the unmodified state (Fig. 1b, d). ARTICLE ARTICLE NATURE COMMUNICATIONS | https://doi.org/10.1038/s41467-019-10130-2 The model allows combinatorial modificat
nucleosome in the model can be unmodified o
combination of three modifications, but not
neously,
because
H3K27
methylation
and
KDM
TRXG
H3K27me3
(9–12)
d
PRC1
PRC2
PRC2:KDM
PRC2:KDM:NURD
PRC1
PRC2
PRC2:KDM
PRC2:KDM:NURD
NURD
PRC2:KDM:NURD
PRC2:KDM
CBP
CBP
PRC2
PRC2:KDM
PRC2:KDM:NURD
TRXG:CBP
TRXG:UTX
PRC2:KDM
PRC2:KDM:NURD
TRXG
TRXG:CBP
TRXG:UTX
TRXG
TRXG:CBP
TRXG:UTX
TRXG:CBP
TRXG:UTX
PRC2
PRC2:KDM
PRC2:KDM:NURD
H3K27ac
H3K4/36me
H3K27ac
H3K4/36me
H2Aub
2
1
4
3
H3K27ac
H3K27ac
H2Aub
H3K4/36me
H3K4/36me
H2Aub
H2Aub
H3K4/36me
H3K27me3
8
7
5
6
H3K4/36me
H3K27me3
H3K27me3
H3K27me3
H2Aub
H2Aub
12
11
9
10
Unmodified
H2Aub
PR-DUB
PRC1
PR-DUB
PRC1
PR-DUB
PRC1
PR-DUB
PRC1
PR-DUB
PRC1
PR-DUB
PRC1
KDM
KDM
KDM
KDM
TRXG
KDM
TRXG
CBP
UTX
NURD
CBP
NURD
CBP
NURD
CBP
UTX
PRC2
UTX
PRC2
UTX
PRC2:KDM:NURD
PRC2:KDM
PRC1
PRC2:KDM:NURD
PRC2:KDM
PRC1
9
10
11
12
UTX
Low d scored as all whole nucleosomes containing a mix of opposing
modifications (Fig. 2). scored as all whole nucleosomes containing a mix of opposing
modifications (Fig. 2). The model is based on half-nucleosomes, and combines them
to give whole nucleosomes. The model is formulated in terms of
the smallest unit possible, namely the half-nucleosome containing
a single copy of H3 and H2A. Whole nucleosomes each comprise
two half-nucleosomes, which are paired for the duration of the
simulation except during replication. Bivalent nucleosomes are The model allows combinatorial modifications. Each half-
nucleosome in the model can be unmodified or can carry any
combination of three modifications, but not all four simulta-
neously,
because
H3K27
methylation
and
acetylation
are NATURE COMMUNICATIONS | (2019) 10:2133 | https://doi.org/10.1038/s41467-019-10130-2 | www.nature.com/naturecommunications 4 NATURE COMMUNICATIONS | (2019) 10:2133 | https://doi.org/10.1038/s41467-019-10130-2 | www.nature.com/naturecommunications NATURE COMMUNICATIONS | https://doi.org/10.1038/s41467-019-10130-2 3d–f). Situations with active or silent state
dominance each show a single peak. When active, the system
gravitates towards fully modified whole nucleosomes, in which
both H3 tails contain both H3K27ac and H3K4/36me (extreme
green corner in Fig. 3d). When silent, the system tends towards
fully silent nucleosomes, containing H3K37me3 and H2Aub on
both half- nucleosomes (extreme red corner in Fig. 3f). This
tendency towards fully modified nucleosome states is caused by
the strong feedbacks in the system. The intermediate situation
with bistability is reflected in an average landscape with two peaks
at the extreme corners (Fig. 3e). This analysis demonstrates that a
two-fold change in a single enzyme rate is sufficient to flip the
system state and affects the distribution of all nucleosome
modifications (even those that do not depend directly on that
enzyme). We observed similar switching behaviour and category
distributions upon two-fold variation of parameters representing
each of the eight enzymes in the model (Supplementary Fig. 2a). Thus the system has bistable properties and can be flipped to
either extreme state (dominated by fully active or silent
nucleosomes) by multiple different small perturbations. We did
not observe any condition in the above analysis in which any of
the bivalent nucleosome categories 2–4 dominated (Fig. 3g–i,
Supplementary Fig. 2b). % red modifications
0
20–40
50
60–80
100
0 red,
0 green
Nucleosome
category
1
2
3
4
5
6
Active
Silent
Bivalent
Unmodified Fig. 2 12 half-nucleosomes combine to give 144 whole nucleosomes. a Each
half-nucleosome contains a single H3 and a single H2A tail, each of which
can be modified as shown in Fig. 1d, giving 12 possible different half-
nucleosome modification states. Each of these half-nucleosomes can be
combined with one of 12 opposite halves, giving 144 possible whole-
nucleosome modification states. b Half-nucleosomes are shown above and
beside the figure with numbers corresponding to Fig. 1b, d. Modifications
are indicated as closed circles. Red: modifications associated with silencing;
green: modifications associated with activation. Whole nucleosomes are
colour coded and assigned to categories according to the proportion of
active and silent modifications they contain. For simplicity, this is
represented as % red modifications in the colour chart below the plot. NATURE COMMUNICATIONS | https://doi.org/10.1038/s41467-019-10130-2 A change of only two-fold in the rate of
H3K27 deacetylation in either direction is sufficient to switch the
system to a stably active (Fig. 3a) or stably silent (Fig. 3c) state. To examine the occupation of the 144 possible individual
nucleosome modification states (Fig. 2b), we plotted the time-
averaged occupation of each nucleosome state on a landscape,
showing frequently occupied states on the peaks and rare states in
the valleys (Fig. 3d–f). Situations with active or silent state
dominance each show a single peak. When active, the system
gravitates towards fully modified whole nucleosomes, in which
both H3 tails contain both H3K27ac and H3K4/36me (extreme
green corner in Fig. 3d). When silent, the system tends towards
fully silent nucleosomes, containing H3K37me3 and H2Aub on
both half- nucleosomes (extreme red corner in Fig. 3f). This
tendency towards fully modified nucleosome states is caused by
the strong feedbacks in the system. The intermediate situation
with bistability is reflected in an average landscape with two peaks
at the extreme corners (Fig. 3e). This analysis demonstrates that a
two-fold change in a single enzyme rate is sufficient to flip the
system state and affects the distribution of all nucleosome
modifications (even those that do not depend directly on that
enzyme). We observed similar switching behaviour and category
distributions upon two-fold variation of parameters representing
each of the eight enzymes in the model (Supplementary Fig. 2a). Thus the system has bistable properties and can be flipped to
either extreme state (dominated by fully active or silent
nucleosomes) by multiple different small perturbations. We did
not observe any condition in the above analysis in which any of
the bivalent nucleosome categories 2–4 dominated (Fig. 3g–i,
Supplementary Fig. 2b). T
i
b h
i
f
id
f Poised chromatin is robustly bistable and minimally bivalent. To address bivalency and bistability we simulated changes in the
composition of the nucleosome array over time (see “Methods”
section). Since quantitative information on enzymatic rates is not
available for the majority of reactions in our system, we tested the
model under a range of different parameter combinations. We
first varied the rate of H3K27 deacetylation by NURD in 2-fold
steps and plotted a time course for each condition (Fig. 3a–c). NATURE COMMUNICATIONS | https://doi.org/10.1038/s41467-019-10130-2 We
observed that intermediate situations between active (category 1:
dark green) and silent (category 5: dark red) cases are bistable in
the sense that the system periodically transits between the two
extremes (Fig. 3b). This bistable intermediate is not dominated by
any of the bivalent categories 2–4 (light green, yellow or orange),
but consists of either one or the other extreme modification state
for long periods of time. A change of only two-fold in the rate of
H3K27 deacetylation in either direction is sufficient to switch the
system to a stably active (Fig. 3a) or stably silent (Fig. 3c) state. h
i
l
i
k (
li h
Fi
b)
H3K27ac
H3K4/36me
H3K27me
H2Aub
H3K27ac
H3K4/36me
H3K27me
H2Aub
b
% red modifications
0
20–40
50
60–80
100
0 red,
0 green
H3
1
4
5
2
3
6
9
10
7
12
11
1
4
5
2
3
6
9
10
7
12
11
8
8
12 half-nuc
144 whole-nucleosome
modification states
Nucleosome
category
1
2
3
4
5
6
Active
Silent
Bivalent
Unmodified
Fig. 2 12 half-nucleosomes combine to give 144 whole nucleosomes. a Each
half-nucleosome contains a single H3 and a single H2A tail, each of which
can be modified as shown in Fig. 1d, giving 12 possible different half-
nucleosome modification states. Each of these half-nucleosomes can be
combined with one of 12 opposite halves, giving 144 possible whole-
nucleosome modification states. b Half-nucleosomes are shown above and
beside the figure with numbers corresponding to Fig. 1b, d. Modifications
are indicated as closed circles. Red: modifications associated with silencing;
green: modifications associated with activation. Whole nucleosomes are
colour coded and assigned to categories according to the proportion of
active and silent modifications they contain. For simplicity, this is
represented as % red modifications in the colour chart below the plot. Note
that when considering whole nucleosomes, we define bivalent nucleosomes
as all those in categories 2, 3, and 4, which carry any combination of
opposing marks, on the same or different half- nucleosomes H3K27ac
H3K4/36me
H3K27me
H2Aub
H3K
H3K4/
H3K
H2
b b system to a stably active (Fig. 3a) or stably silent (Fig. 3c) state. To examine the occupation of the 144 possible individual
nucleosome modification states (Fig. 2b), we plotted the time-
averaged occupation of each nucleosome state on a landscape,
showing frequently occupied states on the peaks and rare states in
the valleys (Fig. NATURE COMMUNICATIONS | https://doi.org/10.1038/s41467-019-10130-2 2b). The remaining 113
forms contain a mixture of active and silent modifications and are
thus bivalent (light green, yellow, and orange, Fig. 2b). Interest-
ingly, these bivalent nucleosomes fall into three categories: those Second, some histone modifications have been shown to inhibit
the activity of specific complexes (PRC2 is inhibited by H3K4/
K36me32,43,44, TRXG enzymes are inhibited by H2Aub45, see
Tables 1–3). This was implemented in the model so that a given
writer cannot modify a half nucleosome carrying a modification TURE COMMUNICATIONS | (2019) 10:2133 | https://doi.org/10.1038/s41467-019-10130-2 | www.nature.com/naturecommunications 5 ARTICLE NATURE COMMUNICATIONS | https://doi.org/10.1038/s41467-019-10130-2 NATURE COMMUNICATIONS | https://doi.org/10.1038/s41467-019-10130-2 H3K27ac
H3K4/36me
H3K27me
H2Aub
H3K27ac
H3K4/36me
H3K27me
H2Aub
b
% red modifications
0
20–40
50
60–80
100
0 red,
0 green
a
H3
H2A
H3
H2A
H3
H2A
1
4
5
2
3
6
9
10
7
12
11
1
4
5
2
3
6
9
10
7
12
11
8
8
12 half-nucleosome modification states
12 half-nucleosome modification states
H3
H2A
144 whole-nucleosome
modification states
Nucleosome
category
1
2
3
4
5
6
Active
Silent
Bivalent
Unmodified
Fig. 2 12 half-nucleosomes combine to give 144 whole nucleosomes. a Eac
half-nucleosome contains a single H3 and a single H2A tail, each of whic
can be modified as shown in Fig. 1d, giving 12 possible different half-
nucleosome modification states. Each of these half-nucleosomes can be
combined with one of 12 opposite halves, giving 144 possible whole-
nucleosome modification states. b Half-nucleosomes are shown above an a
H3
H2A
H3
H2A
H3
H2A
12 half-nucleosome modification states
12 half-nucleosome modification states
H3
H2A
144 whole-nucleosome
modification states modifications and pairs of modifications occur preferentially but
not exclusively in each of these bivalent categories (Supplemen-
tary Fig. 1). For clarity, we henceforth refer to these bivalent
categories as active bivalent, silent bivalent, and balanced bivalent,
without implying any assumptions as to their effects on
transcription. a a H3
H2A H3
H2A
H3
H2A
12 half-nucleosome modification states
12 half-nucleosome modification states
H3
H2A
144 whole-nucleosome
modification states H3
H2A
H3
H2A
12 half-nucleosome modification states
12 half-nucleosome modification states
H3
H2A
144 whole-nucleosome
modification states Poised chromatin is robustly bistable and minimally bivalent. To address bivalency and bistability we simulated changes in the
composition of the nucleosome array over time (see “Methods”
section). Since quantitative information on enzymatic rates is not
available for the majority of reactions in our system, we tested the
model under a range of different parameter combinations. We
first varied the rate of H3K27 deacetylation by NURD in 2-fold
steps and plotted a time course for each condition (Fig. 3a–c). We
observed that intermediate situations between active (category 1:
dark green) and silent (category 5: dark red) cases are bistable in
the sense that the system periodically transits between the two
extremes (Fig. 3b). This bistable intermediate is not dominated by
any of the bivalent categories 2–4 (light green, yellow or orange),
but consists of either one or the other extreme modification state
for long periods of time. NATURE COMMUNICATIONS | https://doi.org/10.1038/s41467-019-10130-2 3a and b). NATURE COMMUNICATIONS | https://doi.org/10.1038/s41467-019-10130-2 ARTICLE a
Category occupation
Active (NURD = 0.05)
0
0.2
0.4
0.6
0.8
1.0
Time
20,000
30,00
10,000 a
b
Bistable (NURD = 0.1)
Category occupation
Active (NURD = 0.05)
c
Silent (NURD = 0.2)
0
0.2
0.4
0.6
0.8
1.0
0
0.2
0.4
0.6
0.8
1.0
0
0.2
0.4
0.6
0.8
1.0
Time
20,000
30,000
10,000
Time
20,000
30,000
10,000
Time
20,000
30,000
10,000 a
d
b
e
f
Bistable (NURD = 0.1)
Category occupation
Active (NURD = 0.05)
c
Silent (NURD = 0.2)
0
0.2
11
12
4
7
5
10
2
9
3
Half-nucleosome states
6
6
3
9
2
10
5
7
4
12
1
11
0
0.2
0
0.1
0
0.2
0.4
0.6
0.8
1.0
0
0.2
0.4
0.6
0.8
1.0
0
0.2
0.4
0.6
0.8
1.0
1
Half-nucleosome states
11
12
4
7
5
10
2
9
3
Half-nucleosome states
6
6
3
9
2
10
5
7
4
12
1
11
Half-nucleosome states
1
11
12
4
7
5
10
2
9
3
Half-nucleosome states
6
6
3
9
2
10
5
7
4
12
1
11
Half-nucleosome states
1
Average state occupation:
whole nucleosomes
0
0.2
0.4
0.6
0.8
Average category
occupation
0
0.2
0.4
0.6
0.8
0
0.2
0.4
0.6
0.8
g
h
i
Whole nucleosome categories
1
2
3
4
5
6
1
2
3
4
5
6
1
2
3
4
5
6
With replication
j
k
Without replication
Time
20,000
30,000
10,000
Time
20,000
30,000
10,000
Time
20,000
30,000
10,000
a–i: Whole nucleosome categories
Active
Silent
Bivalent
Unmod c
Silent (NURD = 0.2)
0
0.2
0.4
0.6
0.8
1.0
00
Time
20,000
30,000
10,000 b a c Active (NURD = 0.05) on the most stably silent states (bottom right of red zones in
Fig. 3j, k, and Supplementary Fig. 3a and b). Thus in the model,
sufficient feedbacks exist for epigenetic memory of both active
and silent chromatin states under a wide range of conditions. Surprisingly however, silent states near the transition zone were
weakened by replication, with a correspondingly wider parameter
regime for bistable transitions between the extreme states (Fig. 2k,
Supplementary Fig. 3b). In general the bistable mode wa
favoured by replication under these conditions, with sever
parameter combinations that gave no bistability in the absence o
replication, becoming bistable when replication was adde
(compare white areas at bottom left of plots in Supplementar
Fig. NATURE COMMUNICATIONS | https://doi.org/10.1038/s41467-019-10130-2 At higher replication rates we observed a simila
survival of extreme active and silent states, and loss of silent state
d
e
f
Category occupation
0
0.2
11
12
4
7
5
10
2
9
3
Half-nucleosome states
6
6
3
9
2
10
5
7
4
12
1
11
0
0.2
0
0.1
0
0.2
0.4
0.6
0.8
1.0
0
0.2
0.4
0.6
0.8
1.0
0
0.2
0.4
0.6
0.8
1.0
1
Half-nucleosome states
11
12
4
7
5
10
2
9
3
Half-nucleosome states
6
6
3
9
2
10
5
7
4
12
1
11
Half-nucleosome states
1
11
12
4
7
5
10
2
9
3
Half-nucleosome states
6
6
3
9
2
10
5
7
4
12
1
11
Half-nucleosome states
1
Average state occupation:
whole nucleosomes
0
0.2
0.4
0.6
0.8
Average category
occupation
0
0.2
0.4
0.6
0.8
0
0.2
0.4
0.6
0.8
g
h
i
Whole nucleosome categories
1
2
3
4
5
6
1
2
3
4
5
6
1
2
3
4
5
6
With replication
j
k
Without replication
0.1
1
0.1
1
PR-DUB
NURD
0.1
1
PR-DUB
0.1
1
NURD
Time
20,000
30,000
10,000
Time
20,000
30,000
10,000
Time
20,000
30,000
10,000
a–i: Whole nucleosome categories
j–k: Average dominant categories
Bistable (categories 1 and 5)
% red modifications
0
20–40
50
60–80
100
0 red,
0 green
1
2
3
4
5
6
Active
Silent
Bivalent
Unmod
>50% category 1
>50% category 2
>50% category 3
>50% category 4
>50% category 5
No category dominates
NATURE COMMUNICATIONS | (2019) 10:2133 | https://doi.org/10.1038/s41467-019-10130-2 | www.nature.com/naturecommunications e
f
11
12
7
0
0.2
0
0.1
tates
11
12
4
7
5
10
2
9
3
Half-nucleosome states
6
6
3
9
2
10
5
7
4
12
1
11
Half-nucleosome states
1
11
12
4
7
5
10
2
9
3
Half-nucleosome states
6
6
3
9
2
10
5
7
4
12
1
11
Half-nucleosome states
1 d
e
f
0
0.2
11
12
4
7
5
10
2
9
3
Half-nucleosome states
6
6
3
9
2
10
5
7
4
12
1
11
0
0.2
0
0.1
1
Half-nucleosome states
11
12
4
7
5
10
2
9
3
Half-nucleosome states
6
6
3
9
2
10
5
7
4
12
1
11
Half-nucleosome states
1
11
12
4
7
5
10
2
9
3
Half-nucleosome states
6
6
3
9
2
10
5
7
4
12
1
11
Half-nucleosome states
1
Average state occupation:
whole nucleosomes f
0
0.2
11
12
4
7
5
10
2
9
3
Half-nucleosome states
6
6
3
9
2
10
5
7
4
12
1
11
Half-nucleosome states
1 f d e 0
11
12
4
7
5
10
2
9
3
Half-nucleosome states
6
6
3
9
2
10
5
7
4
12
1
11
0
0
1
Half-nucleosome states
11
12
4
7
5
10
2
9
3
Half-nucleosome states
6
6
3
9
2
10
5
7
4
12
1
11
Half-nucleosome states
1
11
12
4
7
5
10
2
9
3
Half-nucleosome states
6
6
3
9
2
10
5
7
4
12
1
11
Half-nucleosome states
1
Average sta
whole nu
0
0.2
0.4
0.6
0.8
Average category
occupation
0
0.2
0.4
0.6
0.8
0
0.2
0.4
0.6
0.8
g
h
i
Whole nucleosome categories
1
2
3
4
5
6
1
2
3
4
5
6
1
2
3
4
5
6 s
0
0.2
0.4
0.6
0.8
Average category
occupation
0
0.2
0.4
0.6
0.8
0
0.2
0.4
0.6
0.8
g
h
i
Whole nucleosome categories
1
2
3
4
5
6
1
2
3
4
5
6
1
2
3
4
5
6 0
0.2
0.4
0.6
0.8
Average category
occupation
0
0.2
0.4
0.6
0.8
g
h
Whole nucleosome categories
1
2
3
4
5
6
1
2
3
4
5
6 0 8
g j j With replication
k
0.1
1
PR-DUB
0.1
1
NURD j
Without replication
0.1
1
0.1
1
PR-DUB
NURD a–i: Whole nucleosome categories
j–k: Average dominant categories
Bistable (categories 1 and 5)
% red modifications
0
20–40
50
60–80
100
0 red,
0 green
1
2
3
4
5
6
Active
Silent
Bivalent
Unmod
>50% category 1
>50% category 2
>50% category 3
>50% category 4
>50% category 5
No category dominates a–i: Whole nucleosome categories k Unmod on the most stably silent states (bottom right of red zones in
Fig. NATURE COMMUNICATIONS | https://doi.org/10.1038/s41467-019-10130-2 Note
that when considering whole nucleosomes, we define bivalent nucleosomes
as all those in categories 2, 3, and 4, which carry any combination of
opposing marks, on the same or different half- nucleosomes pp
y
g
To examine system behaviour for a wider range of parameters
we systematically varied pairs of parameters and examined the
time-averaged system state for each condition (Fig. 3j, k,
Supplementary Fig. 3). This analysis showed that for a wide
range of parameter combinations, the system recapitulates the
features observed above (Fig. 3a–c) with an intermediate between
an active and a silenced system that is bistable, and an absence of
dominant bivalent states (Fig. 3j, Supplementary Figs. 3a and 4). To challenge the system further, we simulated the effects of
replication as described previously8, by resetting each half-
nucleosome to the unmodified state 8 with 50% probability at
regular intervals (Fig. 3k, Supplementary Fig. 3b). Interestingly,
replication had no effect on the stably active states (compare
green zones in Fig. 3j, k, and Supplementary Fig. 3a and b), nor that contain mostly active marks (category 2, light green, Fig. 2b),
those that contain mostly silent marks (category 4, orange,
Fig. 2b), and those that contain a balanced mix of both active and
silent marks (category 3, yellow, Fig. 2b). Specific individual NATURE COMMUNICATIONS | (2019) 10:2133 | https://doi.org/10.1038/s41467-019-10130-2 | www.nature.com/naturecommunications 6 TURE COMMUNICATIONS | (2019) 10:2133 | https://doi.org/10.1038/s41467-019-10130-2 | www.nature.com/naturecommunications NATURE COMMUNICATIONS | https://doi.org/10.1038/s41467-019-10130-2 5h–n). We did not find a parameter
combination in which category 4 or category 2 stably dominated
the system (Supplementary Figs. 4 and 5). We conclude that
except under extreme conditions, the model avoids dominant
bivalent states, preferring to pass via bistability in the transition
between active and silent states. Thus in the model, poised
chromatin is robustly bistable and minimally bivalent. In contrast, across the active parameter regime, the dominant
monovalent active nucleosomes (category 1, Fig. 4e), are
consistently accompanied by a background of nucleosomes
carrying H2Aub and H3K27ac. These are depleted in the silent
parameter regime (Fig. 4g). The silent system states (category 5,
Fig. 4h) contain a background of methylated H3K27 and H3K4/
K36, which are depleted from the active parameter regime
(Fig. 4j). Similar distributions of these bivalent nucleosome types
with respect to active and silent system states were consistently
seen across all parameter combinations tested (Supplementary
Fig. 6). Active and silent chromatin contain distinct bivalent sub-
populations. Bivalent chromatin containing opposite modifica-
tions on the same nucleosome has nevertheless been observed at
many
PcG
target
loci
in
sequential
ChIP
and
co-ChIP
experiments29,30,33,51 and by single molecule imaging34. Thus
we asked whether our model is consistent with these experimental
observations. Although the model predicts that dominant bivalent
nucleosome states are difficult to maintain, we did observe a
substantial background of bivalent nucleosomes in all simula-
tions, representing up to 40 % of total nucleosomes an any given
time point (see light green and orange curves in Fig. 3. a-c). Interestingly we found that the prevalent bivalent nucleosome
category changed systematically with system state (Fig. 4). The
active system state was accompanied by a background of the
active bivalent nucleosome category 2 (Figs. 3g, 4e, f). In contrast,
the bivalent background in the silent system mainly comprised g
Each bivalent form can also be depleted by extreme parameter
conditions that render the system fully active or silent and thus
disfavour bivalent forms. For example, bivalent nucleosomes
containing H2Aub and H3K27ac are progressively lost from
active chromatin as PR-DUB levels increase, because H2Aub is
removed (Supplementary Fig. 6b, panels 2 and 5). Likewise,
nucleosomes containing H3K4/36 and H3K27me are progres-
sively lost from silent chromatin as KDM levels increase, because
H3K4/K36 is removed (Supplementary Fig. 6c, panels 3 to 5). Bivalent chromatin containing H3K27me3 and H3K4me3, with
or without additional H2Aub, has been extensively studied
experimentally21–26,28–30. NATURE COMMUNICATIONS | https://doi.org/10.1038/s41467-019-10130-2 Importantly, bivalent states (nucleosome categories
2–4) remained rare under all of the conditions tested, and never
became stably dominant in the transition zone (Supplementary
Figs. 3 and 4). Thus in the model under these conditions, the
transition between stable active and silent states typically passes
through bistability and not bivalency. the silent bivalent nucleosome category 4 (Figs. 3i, 4h, i). This
preference of active and silent system states for specific active and
silent bivalent subtypes held true regardless of which parameter
was used to switch the system (Supplementary Fig. 2b). The same
subtypes co-occur with active and silent states within the bistable
regime and are thus seen as an average in Fig. 3h. g
g
g
To evaluate whether active and silent system states show a
preference for specific bivalent modifications, and to enable
comparison with experimental data on pairwise combinations of
modifications, we examined four possible pairwise combinations
of opposing marks on the same nucleosome (Fig. 4, see also
Supplementary Fig. 1). This analysis showed that bivalent
H3K27me/H3K27ac nucleosomes are rare, and are depleted
from both active and silent system states (Fig. 4c). This is
consistent with experimental observations29,32,40,52. Further-
more, bivalent nucleosomes containing H2Aub and H3K4/
K36me are abundant, are enriched at the border of the active
zone, and present in both active and silent system states across a
wide range of parameters (Fig. 4d). This is consistent with the
observation that this combination of marks can accompany a
wide range of gene expression levels26,28. In summary in the
model, neither of these types occurs exclusively with active or
silent system states. g
y
y
Indeed it was very difficult to find parameters for which the
bivalent states became dominant. We found dominant bivalent
states when beta, NURD and PR-DUB were all very small (of
order 0.005; Supplementary Fig. 5). Examination of categories
and specific modifications showed that under these conditions,
the system essentially becomes blocked in a state containing
H3K27ac and H2Aub, as the rates of removal of these
modifications are very low (Supplementary Fig. 4b–g). Similar
simulations in which beta was reduced to 0.005 and the rate of
removal of H3K4 and H3K27 methylation by low KDM and UTX
rates (of order 0.005) resulted in a dynamic bistable system that
switches rapidly between silent (category 5) and silent bivalent
(category 4) system states with a predominance of bivalent
nucleosomes containing H3K4/K36 and H3K27 methylation
(Supplementary Fig. NATURE COMMUNICATIONS | https://doi.org/10.1038/s41467-019-10130-2 Fig. 3 The model adopts active, silent, or bistable modes. a–c Simulated time course of dynamics of a system of N = 20 half nucleosomes with 2-fold
variation in the rate of removal of H3K27ac by NURD. For all three plots, the rate of all transitions in the model is 1.0, with the exception of NURD as
indicated, beta = 0.1, and PR-DUB = 0.15. Time is measured as number of recruitment attempts per half-nucleosome (see “Methods” section for more
detail). Y-axis shows proportion of whole nucleosomes in each of categories 1–5, according to the colour scale given on the bottom right of the figure. d–f
Time-averaged landscapes for the three cases, shown in terms of the probability to find a whole-nucleosome in any of the 144 modification states,
averaged over a simulation of 50,000 time units. Half- and whole-nucleosome modification states are arranged and colour coded as in Fig. 2b. The vertical
scale gives average proportion of whole array in a given modification state. e The bistability shown in (b) is reflected in the two peaks at the extreme
corners of the landscape. g–i The data from (d) to (f) are summarised in terms of total average occupation of each of categories 1–6. j System behaviour
upon change in the rate constants NURD and PR-DUB with all other parameters fixed as in (a–f). Each simulation was performed for 50,000 time units and
average system state over the entire time course was calculated. Red dots: category 5 (silent) dominates (average occupancy of category 5 nucleosomes
larger than 50%). Green dots: category 1 (active) dominates. Note that all five modification categories were scored (see legend) but categories 2–4 did not
dominate under any condition. Black circles: bistability, defined as multiple transitions between situations with more than 60% category 5 nucleosomes
and those with more than 60% category 1 nucleosomes, each of which has an average lifetime of 40 time units. Grey arrows: positions for PR-DUB and
NURD values as in (a–i). k As for (j) but with replication at every 20 time units, simulated by resetting each half-nucleosome to the unmodified state with
50% probability near the transition zone, and a softer transition in which
bistability was lost, and the system switched more frequently
between short-lived active and silent states (Supplementary
Fig. 3c). NATURE COMMUNICATIONS | https://doi.org/10.1038/s41467-019-10130-2 3j, k, and Supplementary Fig. 3a and b). Thus in the model,
sufficient feedbacks exist for epigenetic memory of both active
and silent chromatin states under a wide range of conditions. Surprisingly however, silent states near the transition zone were
weakened by replication, with a correspondingly wider parameter
regime for bistable transitions between the extreme states (Fig. 2k, Supplementary Fig. 3b). In general the bistable mode was
favoured by replication under these conditions, with several
parameter combinations that gave no bistability in the absence of
replication, becoming bistable when replication was added
(compare white areas at bottom left of plots in Supplementary
Fig. 3a and b). At higher replication rates we observed a similar
survival of extreme active and silent states, and loss of silent states 7 NATURE COMMUNICATIONS | https://doi.org/10.1038/s41467-019-10130-2 In
contrast,
the
co-occurence
of
H3K27ac and H2Aub on the same nucleosome has not been
investigated experimentally to our knowledge. TURE COMMUNICATIONS | (2019) 10:2133 | https://doi.org/10.1038/s41467-019-10130-2 | www.nature.com/naturecommunications 8 ARTICLE NATURE COMMUNICATIONS | https://doi.org/10.1038/s41467-019-10130-2 0
0.2
0.4
0.6
0.8
1
0.1
1
0.1
1
PR-DUB
0.1
1
Nucleosome categories
% red modifications
0
20–40
50
60–80
100
1
2
3
4
5
Active
Silent
Bivalent
Average occupancy
0.1
1
0.1
1
Category 3
0.1
1
0.1
1
0.1
1
0.1
1
0.1
1
0.1
1
0.1
1
Category 1
Category 5
Category 4
Category 2
H3K4/K36me & H3K27me
a
e
H2Aub & H3K27ac
h
NURD
>50% category 1
>50% category 5
No category dominates
Bistable (categories 1 and 5)
0.1
1
H3K27me & H3K27ac
0.1
1
0.1
1
0.1
1
0.1
1
0.1
1
0.1
1
0.1
1
H2Aub & H3K4/K36me
b
f
i
c
g
j
d
nd silent chromatin contain distinct bivalent chromatin types. a Data from Fig. 3j. b–j For the same parameter values, average occ
ated time course for different whole-nucleosome categories as indicated are shown on the same colour scale (right). (c, d, g, j) Prob
hole nucleosomes carrying both modifications as indicated 0.1
1
0.1
1
a a Nucleosome categories
% red modifications
0
20–40
50
60–80
100
1
2
3
4
5
Active
Silent
Bivalent
>50% category 1
>50% category 5
No category dominates
Bistable (categories 1 and 5) 0.1
1
0.1
1
H2Aub & H3K4/K36me
d b 0.1
1
0.1
1
H3K27me & H3K27ac
c b 0.1
1
0 1
1
Category 3
b c d 1 0.1
1
0.1
1
Category 1
e
0.1
1 0.1
1
H2Aub & H3K27ac
0.1
1
g f 0.1
1
Category 2
0.1
1
f g 1 0.1
1
Category 5
h
0.1
1 0.1
1
H3K4/K36me & H3K27me
0.1
1
j 0.1
1
Category 4
NURD
0.1
1
i i j h Fig. 4 Active and silent chromatin contain distinct bivalent chromatin types. a Data from Fig. 3j. b–j For the same parameter values, average occupancies
over the simulated time course for different whole-nucleosome categories as indicated are shown on the same colour scale (right). NATURE COMMUNICATIONS | https://doi.org/10.1038/s41467-019-10130-2 (c, d, g, j) Probability to
find bivalent whole nucleosomes carrying both modifications as indicated In summary this analysis shows that in the model, distinct
forms of bivalent chromatin exist as subpopulations within active
and silent chromatin as a result of dynamic transitions within the
system. represent the poised situation than bivalency (Fig. 5a, b). Stable
active and silent system states are nevertheless dynamic, and each
contain a background of active or silent bivalent nucleosomes
(Fig. 5c). Our model unites a large body of literature into a coherent
unifying framework (Fig. 1), comprising 144 different potential
nucleosome modification states. The model is thus more complex
than several previous models for PcG/TrxG regulation or epige-
netic regulation in general, which typically use the simplest model
structure
that
recapitulates
experimental
observations
and NATURE COMMUNICATIONS | (2019) 10:2133 | https://doi.org/10.1038/s41467-019-10130-2 | www.nature.com/naturecommunications Discussion We have modelled the PcG/TrxG system based on currently
available knowledge of its biochemical properties. On the basis of
this analysis, we propose that bistability is more likely to 9 9 ARTICLE NATURE COMMUNICATIONS | https://doi.org/10.1038/s41467-019-10130-2 NATURE COMMUNICATIONS | https://doi.org/10.1038/s41467-019-10130-2 f
Stable
active
Stable
silent
Bistable
b
able model prediction
Experim Active
Active bivalent
Silent bivalent
Silent
c a c b ves bistability8–11,35. In these examples, the system was
ed to 3 states (active, silent and intermediate). This sim-
ng approach has been designated bottom- down modelling:
n a model that captures main features of the data, one
d persistently strive to simplify it while still capturing the
omenon.”12. These simplified models are immensely pow-
for understanding unifying principles, but they necessarily
assumptions about the number of states that exist, and the
t f
bi t bilit
H
t
k
b tt
h
to ask: given all the known individual features of the sys
without any assumptions about bistability, what p
emerge, and do they help to understand data that is not
by
the
simplified
models? PcG/TrxG
regulation
much attention in the literature, with over 600 publica
year since 2013. In the absence of a coherent theoretic
work, it has been difficult to place this large amount
mation into the context of the system as a whole. Ou
bl
thi
Bistable cell 2
Bistable cell 1
Stable silent
Stable active
a
d
e
Crosslinking
time window
Single
ChIP
Population
mRNA
f
High
Low
i
Citations
20,000
10,000
5000
1000
Stable
active
Stable
silent
Bivalent
Stable
active
Stable
silent
Bistable
Active
Bivalent Ub/Ac
Silent
Time
b
Active - silent
mix
Mainly active
Mainly silent
Re-ChIP
for bivalent
Single-cell
RNA
High
Low
High
Low
1
2
3
4
5
6
7
8
9
10
11
12
Inter-
mediate
Bivalent K4/27me
Ub/Ac
Ub/Ac
K4/K27me
K4/K27me
H2AK118/119 ubiquitylation
H3K27 methylation
H3K4/K36 methylation
H3K27 acetylation
Bistable model prediction
Experimental outcomes under bistable model
Silent
Bivalent K4/27me
Active
Bivalent Ub/Ac
Active
Active bivalent
Silent bivalent
Silent
c
j
8
1
2
3
4
5
6
7
9
10
11
12
1
0
Average state
occupation
Active
Bivalent
Silent
Unmodified Active bivalent Silent bivalent d f e Experimental outcomes under bistable model Experimental outcomes under bistable model i
Citations
20,000
10,000
5000
1000
Time
1
2
3
4
5
6
7
8
9
10
11
12
H2AK118/119 ubiquitylation
H3K27 methylation
H3K4/K36 methylation
H3K27 acetylation
j
8
1
2
3
4
5
6
7
9
10
11
12
1
0
Average state
occupation
Active
Bivalent
Silent
Unmodified i achieves bistability8–11,35. In these examples, the system was
reduced to 3 states (active, silent and intermediate). ARTICLE ARTICLE Fig. 5 Reconciling bivalent and bistable chromatin. a Previous models propose that bivalent chromatin, containing H3K27me3 and H3K4me3, is poised for
activation or silencing in pluripotent cells, and is resolved to an active or silent state upon differentiation. b We propose that the poised, pluripotent state is
bistable, not bivalent. c Dynamic bistability. In the model, in active chromatin, nucleosomes switch dynamically between fully active and active bivalent
configurations. In silent chromatin, nucleosomes alternate between fully silent and silent bivalent configurations. d Schematic time course plots of bistable
model in different modes. Crosslinking window: a ChIP experiment can capture bistable chromatin in different states in different cells and will thus
represent a mixture of the active and silent states. e Predictions of the bistable model for a single locus. The model predicts that distinct types of bivalent
nucleosomes co-occur with active or silent chromatin. f Single ChIP and population mRNA analysis would give identical results for both the bivalent and
bistable models. g, h Re-ChIP (i.e. sequential or concurrent ChIP with two antibodies) and single-cell mRNA analysis would give different results under the
bivalent (g) and bistable (h) models. i Epigeneticist landscape. Citation frequency for each of the 12 half-nucleosome states shown in Fig. 1d is plotted as –
log of total citations of original publications that address a state or transition (see “Methods” section for details and literature). Rarely cited and less
abundant publications are shaded light and occur on the hills. Highly cited publications are shaded dark and occur in the valleys. Publications relating to
silent or bivalent states that contain H3K27me3 (half-nucleosome states 9–12) are the most abundant and highly cited. j For comparison, a similar
landscape plotting – log of average occupancy of each of the 12 half- nucleosome states under bistable conditions as in Fig. 3b. Rare states are shaded light
and occur on the hills. Frequently occurring states are shaded dark and occur in the valleys. The density of research relating to each state is
disproportionate to the predicted occurrence of each in the model system literature, and can be applied to any organism for which there is
sufficient information. There are several potential limitations that could complicate
the interpretation of the model predictions. ARTICLE First, despite its
complexity, the model is still vastly simplified: for example, we do
not distinguish between different degrees of lysine methylation
(mono- di or tri)53, we do not consider the interaction of
nucleosome states with transcriptional processes10,11 or the dis-
tance between nucleosomes in the array8,54,55. These processes
could all potentially affect the outcome of simulations. In parti-
cular, we treat each half-nucleosome in the array as equally likely
to interact with its nearest neighbour as with any other in the
array. We reason that since the model array is 2kb in size, this is a
realistic approximation. PRC1-bound chromatin arrays have been
shown to be highly compacted over similar distances56 and ChIP-
seq peaks of PcG and TrxG protein binding and the modifications
they catalyse are typically at least 1-2kb in size25,57,58. However,
some proteins and modifications do spread over much longer
distances of several tens of kilobases25,59. Several theoretical and
experimental studies have examined the effects of distance
between nucleosomes in large modelled arrays, and the phe-
nomena of spreading and looping8,54,55,60,61. It will be interesting
in future to extend the model to address larger domains, for
example by modulating the probability that one nucleosome
affects others in the array. We show that the PcG/TrxG model system is highly bistable. Remarkably, despite the possibility to adopt 144 different states,
and despite that fact that we made no assumptions about bist-
ability in the model, one of the most important properties to
emerge from the model system is robust bistability. Under a wide
range of parameter combinations, the model gravitates toward
one or other of the extreme active or silent states (the fully
modified active and silent states). There is also a broad parameter
regime between these two extremes in which the system is bis-
table, transitioning rapidly between active and silent states, and
can be readily pushed towards one or other extreme stable state
by a small change in a single parameter (Fig. 3, Supplementary
Fig. 2). Bistability emerges from the model because of multiple
cooperative interactions that stabilise each state, and is reinforced
by the antagonistic relationships between opposite states (Fig. 1,
Tables 1–3). If the PcG /TrxG system is simply bistable, why is it so com-
plex? The system contains multiple feedbacks that contribute to
its robustness. We propose that this complexity contributes both
to stability and flexibility. ARTICLE Under conditions that place the system
in a stably active or silent state, there is a wide range of values
(over several orders of magnitude) for any given parameter pair,
that do not destabilise that state, even in the face of rapid repli-
cation (Supplementary Fig. 3). Thus, the system has potential for
extremely robust memory of both active and silent states that can
withstand
substantial
fluctuations
in
the
activities
of
its
components. y
A second potential limitation is the lack of available quanti-
tative information. Although the affinities of several components
for their binding partners has been measured (reviewed in
refs. 18,62 see also ref. 50) and for a few components the absolute
concentrations have been measured in specific cell types63,64, the
catalytic conversion rates of the enzymes involved are largely
unknown. If quantitative data become available, this will allow
selection of parameter values. However even in the absence of
such data, the model has immense value in defining system
properties across a large parameter space, and in predicting the
effect of experimentally tractable interventions, for example
varying the quantity of any specific enzyme in the system (Sup-
plementary Figs. 2 and 3). p
However, the complexity of the system also offers opportu-
nities for flexible regulation. We predict that several different
single perturbations can flip the system towards activation or
silencing if it is in or near the transition zone (Supplementary
Figs. 2 and 3). The activities of different system components may
vary globally in different cell types, or locally due to recruitment
to specific loci. Each component may be highly regulated20 and
recruitment will also depend on local DNA sequence14,66. These
observations have important implications for understanding
epigenetic memory14, the consequences of misregulation of PcG/
TrxG proteins in disease67, and the effects and side effects of
therapeutic interventions based on inhibition of enzymatic
activities68. Finally the model is unlikely to be complete: the high rate of
research in the PcG/TrxG field means that new modifications,
interactions or enzymes will undoubtedly be discovered in the
future. Indeed, although no readers of bivalent marks have yet
been reported in Drosophila or vertebrates, the first bivalent
reader (H3K27me3/H3K4me3) has recently been reported in
Arabidopsis65. Unfortunately many other components of the
system are not as well characterised in plants as they are in
Drosophila and vertebrates. NATURE COMMUNICATIONS | https://doi.org/10.1038/s41467-019-10130-2 This sim-
plifying approach has been designated bottom- down modelling:
“Given a model that captures main features of the data, one
should persistently strive to simplify it while still capturing the
phenomenon.”12. These simplified models are immensely pow-
erful for understanding unifying principles, but they necessarily
make assumptions about the number of states that exist, and the
requirement for bistability. Here, we took a bottom-up approach, to ask: given all the known individual features of the system, and
without any assumptions about bistability, what properties
emerge, and do they help to understand data that is not captured
by
the
simplified
models? PcG/TrxG
regulation
receives
much attention in the literature, with over 600 publications per
year since 2013. In the absence of a coherent theoretical frame-
work, it has been difficult to place this large amount of infor-
mation into the context of the system as a whole. Our model
enables this. NATURE COMMUNICATIONS | (2019) 10:2133 | https://doi.org/10.1038/s41467-019-10130-2 | www.nature.com/naturecommunications 10 NATURE COMMUNICATIONS | (2019) 10:2133 | https://doi.org/10.1038/s41467-019-10130-2 | www.nature.com/naturecommunications ARTICLE However, we reason that within the
framework we have established, new components can readily be
included as they appear in the literature. Thus, the model is a
valuable and evolvable tool for formalising the PcG/TrxG The model predicts that poised chromatin is not bivalent but is
robustly bistable (Fig. 5a, b). We used the model to examine the
nature of poised chromatin in the transition between active and
silent states. Except under extreme conditions (Supplementary
Fig. 5) bistability and not bivalency is systematically predicted in
the transition regime in which the system switches between active 11 ARTICLE NATURE COMMUNICATIONS | https://doi.org/10.1038/s41467-019-10130-2 and silent states (Supplementary Fig. 3). Thus we propose that
bistability may be an essential feature of poised chromatin, dif-
fering only from the extreme monostable states in its higher
frequency of switching. Previous theoretical studies of bistable
epigenetic systems have focused on the importance of bistability
for long- term epigenetic memory of stable chromatin states in
determined cells8,9. Our work raises the intriguing possibility that
bistable chromatin may in fact be a key molecular feature of
poised PcG/TrxG targets in pluripotent cells. system state. In the silent system state, (conditions as in Fig. 3c),
9% of all H3K27me tails are also methylated on H3K4/36. This
prediction is consistent with the observation of frequent co-
occurrence of H3K36me and H3K27me on the same tail, but
much higher than the observed co-occurrence of H3K4me3 and
H3K27me3. We do not separate H3K4 and H3K36 methylation
in the model. The fusion of these two modifications is justified by
the biochemical data available, and has benefits in reducing the
complexity of the model: the cube shown in Fig. 1b would be a
four- dimensional structure if H3K4 and H3K36 were separated. However the cost of this reductionism is seen when comparing
the model predictions to experimental data that show different
behaviours of the two modifications. Further refinement of the
model to treat H3K4 and H3K36 methylation separately will be
interesting in future to address this discrepancy. p
g
p
p
How can this be tested? ChIP experiments cannot detect bis-
table chromatin, as they examine average enrichments for a
population of cells at a given time point (Fig. 5d-f). Instead,
analysis of transcriptional noise in single cells may give some
insights. Although we are cautious not to equate chromatin state
directly with transcriptional state, we do expect some correlation. ARTICLE The
inverse relationship between cell cycle length and stability of
epigenetic
memory
has
been
noted
in
several
theoretical
studies8,10. Here we propose that the fast-switching bistability
induced by replication may in fact be an essential feature of
chromatin plasticity in rapidly cycling cells such as pluripotent
and cancer cells, by conferring a naïve state on PcG /TrxG
target genes. To what extent are these predictions consistent with existing
experimental
observations? Surprisingly,
the
predominant
active bivalent form that we predict (co-occurrence of H3K27ac
and H2Aub on the same nucleosome) has not been addressed
experimentally to our knowledge. It would be highly informa-
tive to determine whether this form exists, and whether it does
indeed preferentially occur with active states. In contrast,
bivalent chromatin containing H3K4me3 and H3K27me3,
which we predict to be a substantial background in silent
chromatin has been extensively characterised in vertebrates, but
rarely observed in Drosophila74,75. Several vertebrate studies
have noted a correlation between K4/K27me bivalent chroma-
tin and silent gene expression states in ESC cells and also in
differentiated cell types, which is consistent with the model
prediction21–25,29,32–34,51. The model predicts that bivalent chromatin is difficult to
maintain as a stable dominant state, except under extreme per-
turbations (Supplementary Fig. 5). This is consistent with recent
theoretical studies, showing that in a system with two opposing
modifications, bivalent chromatin emerges only upon perturba-
tion of the system71, or when parameters that favour bistability
are set to 072. However, our model also predicts that without
perturbation, and when all states are allowed, bivalent chromatin
is present as a relatively abundant sub-population for specific
parameter regimes in both active and silent system states, due to
the dynamic nature of the system (Fig. 5c). p
Although the active (H3K27ac/H2Aub) form of bivalent
chromatin that we predict here has not been studied, two recent
studies have addressed heterogeneity within bivalent chromatin
carrying H3K27me3 and H3K4me330,76. Both studies identified
bivalent chromatin carrying higher amounts of H3K4me3 than
H3K27me3 associated with active loci, and the converse (higher
H3K27me3 than H3K4me3) associated with silent loci. Con-
sistent with these observations, our model does indeed predict
that the ratio of H3K27me to H3K4me in the population of
bivalent nucleosomes carrying both marks is reduced as the
system switches from the silent to the active regime (Supple-
mentary Fig. 7e). ARTICLE In this context, the model makes an important testable prediction,
namely that a bistable locus would show fluctuations or bursts of
transcription over time, to a greater extent than one that is stably
active or silent. This would manifest in high cell-to-cell variation
in mRNA levels (Fig. 5f). This prediction is consistent with sev-
eral recent studies, showing that specific classes of mammalian
PcG targets do indeed show high cell-to-cell variation in single-
cell RNA-seq experiments69,70, that this variation is greater than
that of non- PcG target genes69 and greater than that of stably
active or stably silent PcG target genes70. Furthermore, several
recent studies based on single-cell imaging of PcG- regulated
reporter gene activity have shown that regulation occurs in an all-
or none fashion, so that the fraction of cells rather than the
amount of gene expression is quantitatively regulated7,9,35. Fur-
ther exploration of the interplay between chromatin states and
transcriptional output will be of great interest in future. It will
also be important to assess bistability using techniques that
capture dynamics in real time35. Interestingly, two different bivalent subtypes strongly correlate
with active or silent system states. Active bivalent nucleosomes,
containing containing H2A ubiquitylation and H3K27 acetyla-
tion, preferentially co-occur as a subpopulation accompanying
the active system state (Fig. 4g, Fig. 5, Supplementary Fig. 6). Similarly, silent bivalent nucleosomes containing both H3K4/
K36me and H3K27me co-occur with the silent system state
(Fig. 4, Fig. 5, Supplementary Fig. 6). We interpret this bias
towards active or silent bivalent nucleosomes as a result of the
dynamic bistable nature of the system. These active and silent
bivalent nucleosomes are only a few modifications away from
their monovalent counterparts (Fig. 2b), and thus the system
readily occupies the nearby states even when in a stable dominant
active or silent mode. Thus in each stable mode, although
monovalent states dominate on average, each nucleosome in the
system periodically alternates between monovalent modes and
the closest bivalent modes (Fig. 5c). The recent development of
in vivo sensors for chromatin modifications and particularly for
bivalent chromatin73 may offer a direct means of testing these
predictions in living cells. y
The model further predicts that poised bistable chromatin is
influenced by replication. In the model, the transition regime is
broadened upon replication, and bistable switching becomes
more frequent with replication, with bistability being lost only
upon very rapid replication (Supplementary Fig. 3b, c). NATURE COMMUNICATIONS | (2019) 10:2133 | https://doi.org/10.1038/s41467-019-10130-2 | www.nature.com/naturecommunicatio Methods
M d l For simplicity in the model the
methylated H3K4 and H3K36 were fused to a single species (H3K4/36), (see Fig. 1d,
half-nucleosome states 1, 2, 5, 6, 9 and 10). Likewise the activities of ASH1 and TRX
were fused in the model to a single activity (named TRXG in Fig. 1d). We note that
this simplification has costs, as situations in which the two proteins act independently
of each other do exist29,59. However the cost to the model of considering them
separately would be large: The 3 dimensional model of Fig. 1b would become four-
dimensional, with a corresponding increase in unknown parameters. Furthermore any
version of such a model that includes the interdependency of ASH1 and TRX would
behave identically to one in which they are fused. Thus, at present, we consider that
fusing the two is justified for the purposes of modelling but we do not wish to imply by
this that they act in tandem in all possible situations. separately would be large: The 3 dimensional model of Fig. 1b would become four-
dimensional, with a corresponding increase in unknown parameters. Furthermore any
version of such a model that includes the interdependency of ASH1 and TRX would
behave identically to one in which they are fused. Thus, at present, we consider that
fusing the two is justified for the purposes of modelling but we do not wish to imply by
this that they act in tandem in all possible situations. Histone modifications, writers and erasers. The model contains four key histone
modifications or groups of modifications that are important for PcG/TrxG reg-
ulation. The PcG mediated modifications in the model are histone H3 lysine 27
trimethylation (H3K27me3) and histone H2A lysine 118 or 119 monoubiquityla-
tion (H2Aub). The TrxG mediated modifications in the model are histone H3
lysine 4 and/ or 36 methylation (H3K4/36me) and histone H3 lysine 27 acetylation
(H3K27ac). Acetylation and methylation on H3K27 are mutually exclusive. For
each of these four modifications, the enzymes or complexes responsible for its
addition and for its removal have been identified and are included in the model,
giving a total of eight catalytic conversions (Table 1). All of the enzymes and
modifications, and most of the feedbacks and inhibitions included in the model
have been documented genetically and/ or biochemically for both Drosophila and
vertebrates, thus the model is potentially applicable to both. NATURE COMMUNICATIONS | https://doi.org/10.1038/s41467-019-10130-2 In conclusion, our work prompts a re-evaluation of the
potential biological role of bivalent chromatin. The proposition
that bivalent chromatin is poised, and is resolved to active or silent
states upon differentiation was largely based on ChIP experiments
performed with a single-antibody against each of these marks
(H3K4me3 or H3K27me3)21–25. We propose instead that bivalent
H3K4/K27me nucleosomes are a subpopulation of silent chro-
matin in a dynamic system (Fig. 5c), and that true poised chro-
matin is bistable. As shown in Fig. 5f, the loss of H3K4me or
H3K27me marks upon differentiation observed in single ChIP
experiments would be consistent with a resolution of poised bis-
table chromatin, as we propose here. Sequential or co-ChIP
experiments would give different results if poised chromatin is
indeed bistable as we propose (Fig. 5f). If the bistable model is
correct, then both K4/K27me nucleosomes and Ub/Ac nucleo-
somes should exist in bistable chromatin due to the mix of active
and silent states each containing the preferred background of one
of these bivalent forms. Each of these forms should then become
preferentially detectable at stable active or silent loci (Fig. 4g, j, 5f). Finally, we propose that future research will profit from a shift
in focus toward studying modification states that are currently
poorly characterised or have not been studied at all. Research to
date has been strongly biased towards silent chromatin states and
bivalent states containing H3K27me3 (Fig. 5i, j see “Methods”
section for full reference list). We propose that a consideration of
the less popular states will prove very fruitful. H2Aub: monoubiquitylated Histone H2A is associated with gene silencing and
is catalyzed by the Polycomb repressive complex 1 (PRC1). Histone H2A is
monoubiquitylated on lysine 118 in Drosophila by the PRC1 subunit dRING88 and
on lysine 119 in vertebrates by the PRC1 subunits RING1A and RING1B88–90. In
both flies and vertebrates, non- canonical PRC1 complexes exist that contain the
RING ubiquitin ligase but lack the Polycomb (Drosophila) or CBX proteins
(vertebrates)91,92 reviewed in17. In the model the transition from histone H2A in
which lysine 119 or 118 is unmodified to one carrying a ubiquitin modification at
this site is represented by a single reaction catalysed by PRC1 (Fig. 1d, half-
nucleosome states 2,3,6,7,10 and 11). De- ubiquitylation of H2Aub K118/119 is catalysed in both flies and vertebrates
by the BAP-1 subunit of the PR-DUB complex (fly93, vertebrate94). Methods
M d l The model includes
several simplifications for each modification, which are illustrated in Fig. 1c and
outlined below. Several demethylases act specifically on methylated H3K4 and H3K36 (reviewed
in ref. 86). For example JARID1A (synonyms: RBP2, KDM5A) demethylates
H3K4me3 and me2 in vertebrates86,111. The Drosophila KDM5 homolog Lid (Little
imaginal discs) demethylates H3K4me3 and me2 in flies112,113. Fbxl10 (synonyms:
KDM2B, JHDM1A) demethylates H3K36me1 and me2 in vertebrates114 and
Drosophila91. KDM2B is also reported to be a H3K4me3 demethylase in flies115
and in vertebrates86. In the model, the demethylation of H3K4 and K36 is
represented by a single reaction, catalysed by KDM, giving a histone H3 tail that is
unmodified at K36 and K4 (Fig. 1d). The unmodified state: the half-nucleosome state designated as unmodified in
the model (state 8, Fig. 1d) is unmodified on histone H3K4, K36 and K27, and on
histone H2AK118 (flies) or K119 (vertebrates). Other residues that do not affect
PcG/TrxG regulation are not relevant in the model and are not considered. The
model makes the assumption that half-nucleosomes that are incorporated into
newly replicated chromatin are in state 8 (see also model implementation below). H3K27me3: trimethylation of H3K27 is associated with gene silencing and is
catalyzed by the Polycomb repressive complex 2 (PRC2) by the catalytic subunit
EZH2 in vertebrates77,78, and in Drosophila by the catalytic subunit E(Z)79,80. In
addition, there exist mono- and dimethylated states of H3K27, which potentially
have different roles in gene regulation and are also catalyzed by PRC281, but are not
included in the model, which considers only the trimethylated state of H3K27
(Fig. 1b, d, half-nucleosome states 9–12). In the model the transition from a histone
H3 that is unmodified on K27 to one that carries K27me3 is represented by PRC2
(Fig. 1d). Model structure. Nucleosome states and transitions: the model contains 12 half-
nucleosome modification states, comprising the unmodified state plus all
possible combinations of the four modifications except the simultaneous presence
of H3K27me3 and H3K27ac, which are mutually exclusive on the same histone
tail40–42. The states are arranged in the model in a three-dimensional three- layered
structure in which each layer contains four states (Fig. 1a). NATURE COMMUNICATIONS | https://doi.org/10.1038/s41467-019-10130-2 PR-DUB
removes monoubiquitylation from H2AK118 or 119 but not from H2AK13/1595
nor from H2B93, thus it is specific for PRC1 mediated monoubiquitylation. No
other enzymes that catalyse this reaction have so far been described. The de-
ubiquitylation of H2Aub is represented in the model by the PR-DUB mediated
transitions (Fig. 1d). H3K27ac: acetylation of histone H3 on lysine 27 is associated with gene
activation. The histone acetyltransferase CBP catalyses the acetylation of H3K27
and other residues in Drosophila96. In vertebrates, both CBP and p300 catalyse this
reaction40. In the model the transition from a histone H3 tail in which K27 is
unmodified to one in which K27 is acetylated is represented by a single reaction
catalysed by CBP (Fig. 1d, half-nucleosome states 1-4). The deacetylation of H3K27 (among many other residues) is catalysed by the
NuRD (nucleosome remodelling and histone deacetylase) complex in vertebrates97
and by the histone deacetylase RPD3 in Drosophila96. In the model the
deacetylation of H3K27 is represented by a single reaction, catalysed by NuRD,
giving a histone H3 tail that is unmodified at K27 (Fig. 1d). Finally, we propose that future research will profit from a shift
in focus toward studying modification states that are currently
poorly characterised or have not been studied at all. Research to
date has been strongly biased towards silent chromatin states and
bivalent states containing H3K27me3 (Fig. 5i, j see “Methods”
section for full reference list). We propose that a consideration of
the less popular states will prove very fruitful. g
g
g
H3K4/K36me: methylation of histone H3 on lysines 4 and 36 is associated with
gene activation. The specificity of TRX and its vertebrate homolog MLL1 is
disputed. Both have previously been reported to methylate H3K4me398–100. However it was later shown that Most H3K4me3 is not catalysed by MLL1 in
vertebrates101,102 and TRX in Drosophila103,104 but by the SET1/COMPASS
complex. Two recent publications disagree on the specificity of TRX and MLL1. The first42 reports that both catalyse H3K4me1 only. The second105 reports that
both TRX and MLL1 are specific for H3K4me2. Some of these discrepancies may
be attributable to anti H3K4me antibody specificity106. However, for the purposes
of our model, we are interested in any H3K4 methylation that genetically or
biochemically antagonises PcG proteins, thus in the absence of a consensus and in
the interests of simplification we refer to H3K4me. Methods
M d l Model summary. The model consists of an array of half-nucleosomes, each con-
taining a single copy of histones H2A and H3 (H2B and H4 are not considered in
the model). Each pair of half-nucleosomes constitutes a whole nucleosome. The
size of the system is given by the parameter N, indicating the number of half-
nucleosomes. Thus if N = 20, there are 10 whole nucleosomes, representing
approximately 2kb of DNA. Half-nucleosome 1 is selected at random from the
array. Specific modifications on half-nucleosome 1 can recruit readers. Half-
nucleosome 2 is selected at random. A reader recruited by half-nucleosome 1 can
add a single modification to half-nucleosome 2, changing its state by one step. Rates of recruitment and modification are the same for all half- nucleosomes:
halves that belong to the same nucleosome and those that belong to different
nucleosomes obey the same rules, and the linear distance between nucleosomes is
not taken into account. The system state is evaluated in terms of whole nucleo-
somes, which are categorized as shown in Fig. 2b. p
Drosophila ASH1 and vertebrate ASH1L dimethylate H3K36. (Fly43,107,108,
vertebrate109,110. Both H3K4me and H3K36me are associated with active genes, and
fly and vertebrate ASH1 (ASH1L in vertebrates) and TRX (MLL1) show highly similar
localisation on chromatin at many loci and interact directly with each other (in fly:37,
in vertebrates:38,39). TRX binding to chromatin is dependent on ASH1 at in flies36 and
on H3K36me2 at several loci in vertebrates39. For simplicity in the model the
methylated H3K4 and H3K36 were fused to a single species (H3K4/36), (see Fig. 1d,
half-nucleosome states 1, 2, 5, 6, 9 and 10). Likewise the activities of ASH1 and TRX
were fused in the model to a single activity (named TRXG in Fig. 1d). We note that
this simplification has costs, as situations in which the two proteins act independently
of each other do exist29,59. However the cost to the model of considering them Drosophila ASH1 and vertebrate ASH1L dimethylate H3K36. (Fly43,107,108,
vertebrate109,110. Both H3K4me and H3K36me are associated with active genes, and
fly and vertebrate ASH1 (ASH1L in vertebrates) and TRX (MLL1) show highly similar
localisation on chromatin at many loci and interact directly with each other (in fly:37,
in vertebrates:38,39). TRX binding to chromatin is dependent on ASH1 at in flies36 and
on H3K36me2 at several loci in vertebrates39. ARTICLE It would be highly interesting in future to apply
the approaches of30,76 to the H3K27ac/H2Aub form, which we
predict to display the opposite behaviour (i.e. the proportion of
H3K27ac to H2Aub is expected to decrease as the system
becomes silent (Supplementary Fig. 7d).) The
evaluation
of
modifications
on
whole-
and
half-
nucleosomes allows comparison of the model predictions with
experimental observations of bivalent modifications on the same
or opposite H3 tails32,34,52. In52 and32, H3K4me3 and H3K27me3
were not detected on the same histone tail by mass spectrometry. Using single molecule imaging,34 detected H3K4me3 on 0.5% of
H3K27me3 tails. The reported co-occurrence of H3K36me2 with
H3K27me3 is higher. In52 3.9% of H3K27me3 tails also carry
H3K36me2. Our model predicts that the percentage of total
H3K37me tails that are also methylated on H3K4/36 depends on TURE COMMUNICATIONS | (2019) 10:2133 | https://doi.org/10.1038/s41467-019-10130-2 | www.nature.com/naturecommunications 12 ARTICLE The following states can enhance binding and/ or stimulate specific complexes
and are implemented as recruitments in the model: ● Human SET1 and MLL1 complexes bind poorly to H3K27me3 histones. Catalytic activity is not prevented134. In the model, half- nucleosome states
9–12 do not recruit TRXG (Fig. 1d). However, if the first half-nucleosome can
recruit TRXG, but the second half-nucleosome contains H3K27me3 and not
H2Aub (state 12), then addition of H3K4/K36me to this second half-
nucleosome by TRXG is allowed. ●
H3K27me3 can bind PRC1 (Fly:116–118, vertebrate:119). In the model half-
nucleosome states 9–12 recruit PRC1 (Fig. 1d). ●
H3K27me3 can bind vertebrate PRC2120,121. Binding occurs via the EED
subunit of PRC2 and stimulates PRC2 methyltransferase activity121. Genetic
evidence for a similar mechanism in Drosophila was provided in121. In the
model, half-nucleosome states 11 and 12 recruit PRC2 (Fig. 1d). y
Compound complexes. Several complexes or proteins have been found in
direct association with each other, and are formally represented as separate
compound species in the model, indicated above the states, with double
names, e.g., PRC2:KDM. This feature of the model enables recruitment and
activity to be separated in specific cases. For example, the enzymatic activity of
PRC2 in methylating H3K27, but not its binding to H3K27me3, is inhibited by
the
presence
of
H3K4
and
H3K36
methylation
on
nucleosomal
substrates32,43,44. This is implemented in the model as follows: a PRC2
complex that attempts to bind to a half-nucleosome in state 9 or 10 (which
have both H3K27me3 and H3K4/36me) will be able to bind via the
H3K27me3 modification, but will not be able to modify another half-
nucleosome in the array, because its activity is inhibited by H3K4/K36me. The
binding of PRC2 would however, recruit other enzymes that are associated
with PRC2, and whose activity is not inhibited by the H3K4/K36 modification. An example is the PRC2:KDM dual complex (see below for more detail). In
the model, the PRC2 part of PRC2:KDM can recruit to a first half-nucleosome,
and the KDM part can demethylate H3K4/K36 on the second half-nucleosome
chosen from the array, if it is present. The following compound complexes are
represented in the model, following similar logic: (N.B. for each compound
complex in the model, only the last enzyme listed has catalytic activity)
111 g
●
H2Aub can bind PRC2 (Vertebrate:122–125, fly122). Binding of vertebrate PRC2
to H2Aub stimulates PRC2 methyltransferase activity in vitro122). ARTICLE NATURE COMMUNICATIONS | https://doi.org/10.1038/s41467-019-10130-2 half-nucleosome by one modification (Fig. 1b, d). Forward and reverse transitions
are catalyzed by the complexes described above, with the additional constraints for
recruitment and inhibition of specific complexes described below (see also
Tables 1–3). Half- nucleosomes are paired to make whole nucleosomes for the
purposes of evaluation (Fig. 2). half-nucleosome contains H3K4/K36me. In the model, this affects the
conversion of state 5 to 9 and state 6 to 10, which are governed only by
beta (indicated by grey arrows on Fig. 1d). (
y g y
g
)
PRC2
activity
is
inhibited
by
nucleosomes
containing
H3K4me3
or
H3K36me3 on both tails (symmetric), but not if only one tail is modified
(asymmetric)32. This is represented in the model by the relationship between
half- and whole nucleosomes (Fig. 2b). A half-nucleosome that already carries
H3K4/K36me is not a substrate for PRC2 as explained above. This means that
a nucleosome in which both halves carry H3K4/K36me (symmetric) will not
be modified by PRC2. In contrast, the presence of H3K4/K36me on only one
H3 tail of a nucleosome (asymmetric) does not prevent modification of the
other H3 tail of that nucleosome. Histone-modification dependent recruited conversions: numerous studies have
reported histone modifications that can enhance binding to chromatin or stimulate
the activity of specific PcG or TrxG complexes. These molecular interactions are
included in the model as follows: a complex that can bind to a given histone
modification acts as a reader of that modification on half-nucleosome 1, and can in
turn modify (write) to half-nucleosome 2 (Fig. 1c). These interactions are referred
to as recruited conversions in the model. Names of complexes that are recruited to
each of the 12 half-nucleosome states in the model are shown above the half-
nucleosome state containing that modification in Fig. 1d. Polycomb (subunit of PRC1) inhibits histone acetylation mediated by CBP by
binding directly to the CBP catalytic domain133. In the model, half-
nucleosome states 9 and 10 do not recruit CBP despite the presence of
H3K4me, because they recruit PRC1 (Fig. 1d). Polycomb (subunit of PRC1) inhibits histone acetylation mediated by CBP by
binding directly to the CBP catalytic domain133. In the model, half-
nucleosome states 9 and 10 do not recruit CBP despite the presence of
H3K4me, because they recruit PRC1 (Fig. 1d). ARTICLE For example, the Drosophila H3K4 demethylase Lid interacts physically and
functionally with RPD3136. The vertebrate H3K4/K6 demethylase LSD1 is a
component of the NuRD complex at active enhancers137. These interactions
are summarized in the model by the compound complex PRC2:KDM:NURD,
which is recruited via PRC2 binding to H3K27me3 or H2Aub, to all states
containing one or both of these modifications (half-nucleosome states 2, 3, 6,
7, 9–12; Fig. 1d). In addition, histone loss due to turnover, transcription and replication leads to
loss of modifications. Replication is explicitly modelled (see Model implementation
below). For all other histone modification-independent recruitments or removals,
the term ‘beta’, applied to all transitions, describes the rate of conversion between
states that is independent of recruitment. For simplicity we use a single value of
beta for all histone modification-independent transitions. If more quantitative
information becomes available in future, beta can be adjusted separately for specific
transitions, for example all those catalyzed by PRC1, to reflect the recruitment of
canonical and non- canonical PRC1 to chromatin independently of
H3K27me3118,130–132 (reviewed in17). Histone modification-mediated inhibition: recently, several studies have
identified histone modifications that inhibit the activity of specific PcG or TrxG
complexes. These inhibitions are formalised in the model and simplified in some
cases, as described below. TRXG:CBP. ASH1 and CBP interact physically and functionally138. The TrxG
protein BRM is associated with CBP and stimulates the activity of CBP in
acetylating H3K27139. These interactions are represented in the model by the
compound complex TRXG:CBP (Fig. 1d). ●
Histone H2A ubiquitination inhibits the enzymatic activity of H3 lysine 36
methyltransferases45. This is represented in the model in two ways. Firstly, by
the fact that half-nucleosome states 2 and 3, containing H2Aub, do not recruit
TRXG, despite the presence of H3K27ac, which would normally cause
recruitment (Fig. 1c). In the model, this means that a half-nucleosome in state
2 or 3 will not be able to recruit TRXG and will not be able to add H3K4/
K36me to a second half-nucleosome. Secondly, if the first half-nucleosome can
recruit TRXG (ie it has H3K27ac but not H2Aub; state 1 or 4), but the second
half-nucleosome chosen does contain H2AUb, then addition of H3K4/K36me
to this second half-nucleosome by TRXG is prevented. In the model, this
affects the conversion of state 11 to 10, state 7 to 6 and state 3 to 2. ARTICLE In the
model, half-nucleosome states 3, 7 and 11 recruit PRC2 (Fig. 1d). ( g
)
●
H3K4 monomethylation stimulates the activity of Drosophila CBP in
acetylating H3K2742,96. In the model, half-nucleosome states 1,2,5 and 6
recruit CBP or TRX:CBP via this interaction (Fig. 1d). Although half-
nucleosome states 9 and 10 also contain H3K4me, they do not recruit CBP due
to inhibition. ●
H3K27ac enhances binding of TrxG proteins to chromatin. The vertebrate
TrxG protein BRD4 (fly homolog: FSH(1)) binds acetylated histones126. Drosophila FSH binds acetylated histones, interacts physically with ASH1 and
may recruit or stabilise ASH1 at some loci127,128. Since ASH1 and TRX
interact directly with each other (in fly: ref. 37, in vertebrates: ref. 38,39), we
simplify these interactions in the model so that both are represented by TRXG. Half-nucleosome states 1-4 contain H3K27ac and recruit TRXG and/or
TRXG:UTX and TRXG:CBP. Histone modification-independent conversions: there is ample evidence that
not all recruitment of PcG and TrxG proteins depends on pre-existing histone
modifications. Documented mechanisms of histone – independent recruitment
include specific DNA sequences, non- specific and specific DNA binding proteins,
demethylated DNA, non-coding RNAs, and RNA polymerase (reviewed in
refs. 17,18. For example, both PRC249,50 and PRC147,129 bind DNA non- specifically
with higher affinity than for specific histone tail modifications. p
y
y
y
y
PRC2:KDM. Vertebrate PRC2 recruits RBP2 (H3K4 demethylase)111 and
LSD1 (H3K4 demethylase)135. Recruited via H3K27me3 and H2Aub to all
states that contain these modifications (half-nucleosome states 2, 3, 6, 7, 9–12,
(Fig. 1d)). ● PRC2:KDM:NURD. The inclusion of this complex in the model is based on
the
above
observations
that
PRC2
associates
with
H3K4/K36
lysine
demethylases, and that several H3K4/K36 lysine demethylases in turn
associate with the deacetylases NuRD (vertebrates) or RPD3 (Drosophila). For example, the Drosophila H3K4 demethylase Lid interacts physically and
functionally with RPD3136. The vertebrate H3K4/K6 demethylase LSD1 is a
component of the NuRD complex at active enhancers137. These interactions
are summarized in the model by the compound complex PRC2:KDM:NURD,
which is recruited via PRC2 binding to H3K27me3 or H2Aub, to all states
containing one or both of these modifications (half-nucleosome states 2, 3, 6,
7, 9–12; Fig. 1d). PRC2:KDM:NURD. The inclusion of this complex in the model is based on
the
above
observations
that
PRC2
associates
with
H3K4/K36
lysine
demethylases, and that several H3K4/K36 lysine demethylases in turn
associate with the deacetylases NuRD (vertebrates) or RPD3 (Drosophila). Methods
M d l Each horizontal tran-
sition (within a layer) or vertical transition (between layers) changes the state of a Demethylation of di and trimethylated H3K27me3 is catalyzed by the histone
demethylases UTX (KDM6A) and JMJD3 (KDM6B) in vertebrates46,82–84 and by
dUTX in Drosophila85, for review see ref. 86. In vertebrates the dual demethylase
KDM7 demethylates monomethylated H3K27 and H3K987. In the model a single
transition from trimethylated H3K27me3 to unmodified H3K27 is represented by
UTX (Fig. 1d). 13 TURE COMMUNICATIONS | (2019) 10:2133 | https://doi.org/10.1038/s41467-019-10130-2 | www.nature.com/naturecommunications ARTICLE t2 smallest): Select a random half-nucleosome,
and select with equal chance which of the modification positions (H3K27,
K4/36, H2A) should be changed. Recruitment: Citations are added to the state that recruits, and to the state
containing the modification that is added or erased by the recruited enzyme. Inhibition: Citations are added to the state that inhibits, and to those
containing the modification that is inhibited. Recruitment: Citations are added to the state that recruits, and to the state
containing the modification that is added or erased by the recruited enzyme. Inhibition: Citations are added to the state that inhibits, and to those
containing the modification that is inhibited. If position H3K27 is selected, evaluate the state of the modification on this
position of the half-nucleosome. If this state is H3K27ac or H3K27me then it is
changed to the non-modified state of H3K27. If H3K27 is unmodified then it is
changed with equal probability to either an acetylated or to a methylated state. If position K4/36 is selected, reverse the present state of the half-nucleosome
between the methylated and the non-methylated state. Bivalent chromatin: Citations are added to the bivalent half-nucleosome states
that are addressed by the paper (yellow dots on Fig. 5i). This analysis generated the following total citations for each state:
State
Citations This analysis generated the following total citations for each state:
S
Ci
i State
Citations
1
6344
2
17261
3
10671
4
1415
5
5232
6
15808
7
9176
8
1391
9
33089
10
28568
11
22232
12
15988 If position H2A is selected, reverse the present state of the half-nucleosome
between the ubiquitinated and the non-ubiquitinated state. ●
Direct move due to non-recruited NURD activity (3, i.e., t3 smallest): Select a
random half-nucleosome, and if this has modification H3K27ac, then its state
is changed to non-modified H3K27. ●
Direct move due to non-recruited NURD activity (3, i.e., t3 smallest): Select a
random half-nucleosome, and if this has modification H3K27ac, then its state
is changed to non-modified H3K27. ●
Direct move due to non-recruited PR-DUB activity (4, i.e. t4 smallest): Select a
random half-nucleosome, and if this has modification H2Aub, then its state is
changed to non-modified H2A. ●
Direct move due to non-recruited PR-DUB activity (4, i.e. t4 smallest): Select a
random half-nucleosome, and if this has modification H2Aub, then its state is
changed to non-modified H2A. ARTICLE Citation density landscape. The citation density landscape shown in Fig. 5i
gives an indication of the popularity of each of the 12 states considered in our
model, as indicated by citation frequency of papers relating to each state. The
landscape plots -log(total citations) for each state. The total citations for each
state were calculated for the 73 publications used to formulate the
model32,37,38,40–47,49,50,77–100,105,107–109,111–125,127–139,142, and 17 publications
relating to bivalent chromatin21–27 29–34 51 75 76 143 The bivalent chromatin Citation density landscape. The citation density landscape shown in Fig. 5i
gives an indication of the popularity of each of the 12 states considered in our
model, as indicated by citation frequency of papers relating to each state. The
landscape plots -log(total citations) for each state. The total citations for each
state were calculated for the 73 publications used to formulate the
model32,37,38,40–47,49,50,77–100,105,107–109,111–125,127–139,142, and 17 publications
relating to bivalent chromatin21–27,29–34,51,75,76,143. The bivalent chromatin
publications were selected to represent the first reports of bivalent chromatin
in mouse and human ESCs21–24,26,27, and in differentiating systems25, but do
not cover publications on iPSCs since these are highly cited for reasons other 1. A recruitment type process with rate equal to one, set to occur at time
increment t1=-ln(ran1) ecruitment type process with rate equal to one, set to occur increment t1=-ln(ran1). 2. Direct conversion with rate beta, t2=-ln(ran2)/beta. 3. A direct acetylation move with rate NURD (dn), t3=-ln(ran3)/dn. 4. A direct de-ubiquination move with rate PR-DUB (dp), t4=-ln(ran4)/dp. relating to bivalent chromatin21 27,29 34,51,75,76,143. The bivalent chromatin
publications were selected to represent the first reports of bivalent chromatin
in mouse and human ESCs21–24,26,27, and in differentiating systems25, but do
not cover publications on iPSCs, since these are highly cited for reasons other
than bivalent chromatin itself. Some papers also report bivalent chromatin
containing H2AUb26,27,31. We found a single publication addressing H3K4/
K27me bivalent chromatin in Drosophila75. Finally, we include recent re-
evaluations based on mass spec32, imaging34, and co-ChIP or sequential
ChIP29,30,33,51, demonstrating the co- occurrence of opposite modifications on
the same or adjacent nucleosomes. The number of citations for each reference
was calculated according to Google Scholar on 26.01.2019. In these, time steps ran1, ran2, ran3 and ran4 are four different random
numbers that each are selected from uniform distribution between 0 and 1. ARTICLE The
next attempted move is pinpointed by the move type with the smallest time
increment among t1, t2, t3 and t4. ●
Recruitment move (1, i.e., t1 smallest): In case that a recruitment move is
selected (t1 smallest) then a recruiting half-nucleosome is chosen randomly
among the N half- nucleosomes. The state of this half-nucleosome specifies
which recruitment processes are allowed to act. In case the recruiting half-
nucleosome is in state 8 there is no recruitment possible and no further
action is taken for this update step. Otherwise one selects randomly, with
equal weight, one of the allowed read-write or read- erase enzymes that can
be recruited by the state of the recruiting half-nucleosome. Subsequently a
target half-nucleosome is selected at random. If the selected read-write
enzyme can modify the state of the target, then this target half-nucleosome
changes state as specified by the write or erase processes on the directed
arrows in Fig. 1d. If the enzyme cannot change the state of the target half-
nucleosome then no further action is taken for this update step. Otherwise
the target half-nucleosome is assigned the new state specified by the
enzyme. Notice that each read-write or read- erase enzyme uniquely
specifies which change it can make to any of the 12 half-nucleosome states
in the model (Figure 1d). To generate the citation density landscape, total citations were added to each of
the 12 half-nucleosome states (Fig. 1d) according to the following rules: ●
●
●
●
●
●
●
Re
the ●
●
●
●
●
●
● ●
Addition of a modification: Citations of papers demonstrating addition of a
modification are
added
to all half-nucleosome states that
carry the
modification. ●
Removal of a modification: Citations are added to the half-nucleosome state
from which the modification is removed. ●
Exclusivity of H3K27ac and H3K27me3: Citations of papers documenting this
mutually exclusive relationship are added to all states that contain either
H3K27me3 or H3K27ac, and to the unmodified state 8. PRC1 can bind independently of H3K27me3: Citations are added to all
H2Aub states (ie 2, 3, 6, 7, 10, 11). Random direct move (2 i.e. t2 smallest): Select a random half-nucleosome,
and select with equal chance which of the modification positions (H3K27,
K4/36, H2A) should be changed. Random direct move (2 i.e. Data availability In simulations, we also consider replication, when accumulated time since
last replication reaches a generation time of 20 multiplied by the system size N. At replication we select each half-nucleosome and with 50% probability replace
it with one in the unmodified state 8. Thus on average 50% of the half-
nucleosomes are replaced, but stochastic partitioning variations are allowed. In
figures, we plot the dynamics in units of time counter divided by number of half-
nucleosomes (i.e. number of recruitment attempts per half-nucleosome). In
some simulations, we vary the strength of a subset of the recruitment reactions. When a recruitment reaction is lowered, this is done by accepting an otherwise
acceptable move with a reduced probability. When a recruitment reaction is set
to be larger than one, we select all recruitment reactions with larger rate, and
reduce the acceptance of the non-selected ones. For evaluation of the output of
simulations, half-nucleosomes are assigned to pairs according to their position
in the array. Each whole nucleosome thus created is categorized according to
Fig. 2b. All relevant data supporting the key findings of this study are available within the article
and its Supplementary Information files or from the corresponding author upon
reasonable request. A reporting summary for this Article is available as a Supplementary
Information file. ARTICLE Notice that change is only executed when the above conditions are met, and
that there will be situations where the attempted move is not successful. In any case
the overall time-counter is updated by adding a time increment = min (t1,t2,t3,t4). Thus in each update step, a maximum of one of the above changes is executed. Subsequently the next update step is started by again generating ran1- ran4. Reporting Summary. Further information on research design is available in
the Nature Research Reporting Summary linked to this article. Code availability All computer code used in this study is available on request. All computer code used in this study is available on request. Received: 28 August 2018 Accepted: 12 April 2019 Received: 28 August 2018 Accepted: 12 April 2019 We emphasize that the above implementation is hugely simplified. In
particular it employs a nucleosome-centered view of recruitment, where each
half-nucleosome selects each of its read-write or read- erase enzymes with equal
rate (set equal to one). This will cause selection of different read-write enzymes
to occur with different frequency. There is no reason to refine this approach ARTICLE These
transitions are governed only by beta, describing the rate of non-recruited
conversions (indicated by grey arrows on Fig. 1d). p
p
g
●
TRXG:UTX. Vertebrate UTX (H3K27 demethylase) is associated with MLL 2/
3 (vertebrate homologs of TRX)46. Drosophila UTX (H3K27 demethylase) is
associated with CBP and the TrxG protein BRM139. These interactions are
represented in the model by the compound complex TRXG:UTX (Fig. 1d). p
y
p
p
g
PRC1:KDM. Not explicitly represented in the model are the non-canonical
PRC1-KDM complexes dRAF and PRC1.1, which contain the ubiquitin ligase
dRING (flies) or RING1A/ RING1B (vertebrates) and H3K36 demethylase
dKDM2 (flies91) or KDM2B (vertebrates92). These complexes lack the
chromodomain containing subunits PC (flies) or CBX (vertebrates) and thus
are not recruited to chromatin via the interaction between the PC
chromodomain and H3K27me3 (for reviews, see17,140,141). No recruitment
via existing histone modifications has yet been reported for non-canonical
PRC1, thus at present these complexes are covered by the recruitment-
independent parameter beta. y g y
g
●
H3K4 and H3K36 methylation inhibit PRC2 H3K27 methylation activity
(Drosophila PRC243,44, vertebrate PRC232,44. This is represented in the model
in two ways as described above. Firstly, half-nucleosome states 9 and 10
contain H3K4/36me and thus do not recruit PRC2 despite the presence of
modifications that would normally recruit PRC2 (H3K27me3 in both states,
and H2Aub in state 10; Fig. 1d). Secondly, any PRC2- mediated transition of
the second half-nucleosome towards H3K27me3 is prevented if that second Model implementation. The model is implemented as an agent based model on
N = 20 units, that each can be in any of the 12 half-nucleosome states shown in 14 NATURE COMMUNICATIONS | (2019) 10:2133 | https://doi.org/10.1038/s41467-019-10130-2 | www.nature.com/naturecommunications ARTICLE ARTICLE NATURE COMMUNICATIONS | https://doi.org/10.1038/s41467-019-10130-2 Fig. 1b, d. The model lets these units interact and modify each other as specified by
seven read - write or read - erase enzymes (1: TRX; 2: UTX = TRX:UTX; 3: CBP =
TRX:CBP; 4: PRC1; 5: PRC2; 6: PRC2:KDM; 7: PRC2:KDM:NURD), each of which
has one or zero ways to modify each of the 12 allowed states. before we have information about actual recruitment rates. At present, the
proof-of-concept model that we present here represents a baseline of current
knowledge for exploring systemic features of nucleosome modification
dynamics. before we have information about actual recruitment rates. At present, the
proof-of-concept model that we present here represents a baseline of current
knowledge for exploring systemic features of nucleosome modification
dynamics. before we have information about actual recruitment rates. At present, the
proof-of-concept model that we present here represents a baseline of current
knowledge for exploring systemic features of nucleosome modification
dynamics. has one or zero ways to modify each of the 12 allowed states. The update uses an event-based Gillespie type algorithm. At each update step
one selects the next time for potential occurrence of one of four types of moves: The update uses an event-based Gillespie type algorithm. At each update step
one selects the next time for potential occurrence of one of four types of moves: Citation density landscape. The citation density landscape shown in Fig. 5i
gives an indication of the popularity of each of the 12 states considered in our
model, as indicated by citation frequency of papers relating to each state. The
landscape plots -log(total citations) for each state. The total citations for each
state were calculated for the 73 publications used to formulate the
model32,37,38,40–47,49,50,77–100,105,107–109,111–125,127–139,142, and 17 publications
relating to bivalent chromatin21–27,29–34,51,75,76,143. The bivalent chromatin
publications were selected to represent the first reports of bivalent chromatin
in mouse and human ESCs21–24,26,27, and in differentiating systems25, but do
not cover publications on iPSCs, since these are highly cited for reasons other
than bivalent chromatin itself. Some papers also report bivalent chromatin
containing H2AUb26,27,31. We found a single publication addressing H3K4/
K27me bivalent chromatin in Drosophila75. Finally, we include recent re-
evaluations based on mass spec32, imaging34, and co-ChIP or sequential
ChIP29,30,33,51, demonstrating the co- occurrence of opposite modifications on
the same or adjacent nucleosomes. The number of citations for each reference
was calculated according to Google Scholar on 26.01.2019. ARTICLE ARTICLE NATURE COMMUNICATIONS | https://doi.org/10.1038/s41467-019-10130-2 2. Ferrell, J. E. & Xiong, W. Bistability in cell signaling: How to make continuous
processes discontinuous, and reversible processes irreversible. Chaos. 11,
227–236 (2001). 34. Shema, E. et al. Single-molecule decoding of combinatorially modified
nucleosomes. Science 352, 717–721 (2016). 35. Bintu, L. et al. Dynamics of epigenetic regulation at the single-cell level. Science 351, 720–724 (2016). 3. Maamar, H., Raj, A. & Dubnau, D. Noise in gene expression determines cell
fate in Bacillus subtilis. Science 317, 526–529 (2007). 36. Kuzin, B., Tillib, S., Sedkov, Y., Mizrokhi, L. & Mazo, A. The Drosophila
trithorax gene encodes a chromosomal protein and directly regulates the
region-specific homeotic gene fork head. Gene Dev. 8, 2478–2490 (1994). 4. Sneppen, K. & Dodd, I. B. A simple histone code opens many paths to
epigenetics. PLoS Comput. Biol. 8, e1002643 (2012). 37. Rozovskaia, T. et al. Trithorax and ASH1 interact directly and associate with
the trithorax group-responsive bxd region of the Ultrabithorax promoter. Mol. Cell Biol. 19, 6441–6447 (1999). 5. Chan, C. S., Rastelli, L., Pirrotta, V. & Polycomb, A. response element in the
Ubx gene that determines an epigenetically inherited state of repression. EMBO J. 13, 2553–2564 (1994). 6. Allshire, R. C., Javerzat, J. P., Redhead, N. J. & Cranston, G. Position effect
variegation at fission yeast centromeres. Cell 76, 157–169 (1994). 38. Gregory, G. D. et al. Mammalian ASH1L is a histone methyltransferase that
occupies the transcribed region of active genes. Mol. Cell Biol. 27, 8466–8479
(2007). 7. Berry, S., Hartley, M., Olsson, T. S., Dean, C. & Howard, M., Local chromatin
environment of a Polycomb target gene instructs its own epigenetic
inheritance. eLife 4, 07205 (2015). 39. Zhu, L. et al. ASH1L links histone H3 lysine 36 dimethylation to MLL
leukemia. Cancer Discov. 6, 770–783 (2016). f
8. Dodd, I. B., Micheelsen, M. A., Sneppen, K. & Thon, G. Theoretical analysis of
epigenetic cell memory by nucleosome modification. Cell 129, 813–822
(2007). 40. Pasini, D. et al. Characterization of an antagonistic switch between histone H3
lysine 27 methylation and acetylation in the transcriptional regulation of
Polycomb group target genes. Nucleic Acids Res. 38, 4958–4969 (2010). (
)
9. Angel, A., Song, J., Dean, C., Howard, M. & Polycomb-based, A. switch
underlying quantitative epigenetic memory. Nature 476, 105–108 (2011). y
g
p
g
g
41. Jung, H. R., Pasini, D., Helin, K. & Jensen, O. N. ARTICLE Quantitative mass
spectrometry of histones H3.2 and H3.3 in Suz12-deficient mouse embryonic
stem cells reveals distinct, dynamic post-translational modifications at Lys-27
and Lys-36. Mol. Cell Proteom. 9, 838–850 (2010). 9. Angel, A., Song, J., Dean, C., Howard, M. & Polycomb-based, A. switch
underlying quantitative epigenetic memory. Nature 476, 105–108 (2011). 10. Zerihun, M. B., Vaillant, C. & Jost, D. Effect of replication on epigenetic
memory and consequences on gene transcription. Phys. Biol. 12, 026007
(2015). y
g q
p g
y
10. Zerihun, M. B., Vaillant, C. & Jost, D. Effect of replication on epigenetic
memory and consequences on gene transcription. Phys. Biol. 12, 026007
(2015). 42. Tie, F. et al. Trithorax monomethylates histone H3K4 and interacts directly
with CBP to promote H3K27 acetylation and antagonize Polycomb silencing. Development 141, 1129–1139 (2014). 11. Berry, S., Dean, C. & Howard, M. Slow chromatin dynamics allow polycomb
target genes to filter fluctuations in transcription factor activity. Cell Syst. 4,
445–457 (2017). p
43. Yuan, W. et al. H3K36 methylation antagonizes PRC2-mediated H3K27
methylation. J. Biol. Chem. 286, 7983–7989 (2011). 12. Sneppen, K. Models of life: epigenetics, diversity and cycles. Reports on
progress in physics. Phys. Soc. 80, 042601 (2017). y
44. Schmitges, F. W. et al. Histone methylation by PRC2 is inhibited by active
chromatin marks. Mol. Cell 42, 330–341 (2011). y
44. Schmitges, F. W. et al. Histone methylation by PRC2 is inhibited by active
h
k
l
ll
(
) chromatin marks. Mol. Cell 42, 330–341 (2011). 13. Ringrose, L. & Howard, M. Dissecting chromatin-mediated gene regulation
and epigenetic memory through mathematical modelling. Curr. Opin. Syst. Biol. 3, 7–14 (2017). 45. Yuan, G. et al. Histone H2A ubiquitination inhibits the enzymatic activity of
H3 lysine 36 methyltransferases. J. Biol. Chem. 288, 30832–30842 (2013). 14. Steffen, P. A. & Ringrose, L. What are memories made of? How Polycomb and
Trithorax proteins mediate epigenetic memory. Nat. Rev. Mol. cell Biol. 15,
340–356 (2014). 46. Lee, M. G. et al. Demethylation of H3K27 regulates polycomb recruitment and
H2A ubiquitination. Science 318, 447–450 (2007). q
47. Francis, N. J., Saurin, A. J., Shao, Z. & Kingston, R. E. Reconstitution of a
functional core polycomb repressive complex. Mol. Cell 8, 545–556 (2001). 15. Fauvarque, M. O. & Dura, J. M. polyhomeotic regulatory sequences induce
developmental regulator-dependent variegation and targeted P-element
insertions in Drosophila. Genes Dev. 7, 1508–1520 (1993). 48. Krajewski, W. ARTICLE A., Nakamura, T., Mazo, A. & Canaani, E. A motif within SET-
domain proteins binds single-stranded nucleic acids and transcribed and
supercoiled DNAs and can interfere with assembly of nucleosomes. Mol. Cell
Biol. 25, 1891–1899 (2005). 16. Klose, R. J., Cooper, S., Farcas, A. M., Blackledge, N. P. & Brockdorff, N. Chromatin sampling--an emerging perspective on targeting polycomb
repressor proteins. PLoS Genet 9, e1003717 (2013). 49. Poepsel, S., Kasinath, V. & Nogales, E. Cryo-EM structures of
PRC2 simultaneously engaged with two functionally distinct nucleosomes. Nat. Struct. Mol. Biol. 25, 154–162 (2018). 17. Bauer, M., Trupke, J. & Ringrose, L. The quest for mammalian Polycomb
response elements: are we there yet? Chromosoma 125, 471–496 (2016). 50. Choi, J. et al. DNA binding by PHF1 prolongs PRC2 residence time on
chromatin and therebyy promotes H3K27 methylation. Nat. Struct. Mol. Biol. 24, 1039–1047 (2017). 18. Ringrose, L., Noncoding rnas in polycomb and trithorax regulation: a
quantitative perspective. Annu. Rev. Genet. 51, 385–411 (2017). 19. Voigt, P., Tee, W. W. & Reinberg, D. A double take on bivalent promoters. Genes Dev. 27, 1318–1338 (2013). 51. Lorzadeh, A. et al. Nucleosome density chip-seq identifies distinct chromatin
modification signatures associated with mnase accessibility. Cell Rep. 17,
2112–2124 (2016). 20. Harikumar, A. & Meshorer, E. Chromatin remodeling and bivalent histone
modifications in embryonic stem cells. EMBO Rep. 16, 1609–1619 (2015). 52. Young, N. L. et al. High throughput characterization of combinatorial histone
codes. Mol. Cell Proteom. 8, 2266–2284 (2009). 21. Azuara, V. et al. Chromatin signatures of pluripotent cell lines. Nat. Cell Biol. 8, 532–538 (2006). 22. Bernstein, B. E. et al. A bivalent chromatin structure marks key developmental
genes in embryonic stem cells. Cell 125, 315–326 (2006). 53. Luo, M. Chemical and biochemical perspectives of protein lysine methylation. Chem. Rev. 118, 6656–6705 (2018). 54. Erdel, F. & Greene, E. C. Generalized nucleation and looping model for
epigenetic memory of histone modifications. Proc. Natl Acad. Sci. USA 113,
E4180–E4189 (2016). 23. Pan, G. et al. Whole-genome analysis of histone H3 lysine 4 and lysine 27 23. Pan, G. et al. Whole-genome analysis of histone H3 lysine 4 and lysine 27
methylation in human embryonic stem cells. Cell Stem Cell 1, 299–312 (2007). 24. Zhao, X. D. et al. Whole-genome mapping of histone H3 Lys4 and 27
trimethylations reveals distinct genomic compartments in human embryonic
stem cells. Cell Stem Cell 1, 286–298 (2007). 55. Obersriebnig, M. ARTICLE J., Pallesen, E. M., Sneppen, K., Trusina, A. & Thon, G. Nucleation and spreading of a heterochromatic domain in fission yeast. Nat. Commun. 7, 11518 (2016). 25. Mikkelsen, T. S. et al. Genome-wide maps of chromatin state in pluripotent
and lineage-committed cells. Nature 448, 553–560 (2007). 56. Francis, N. J., Kingston, R. E. & Woodcock, C. L. Chromatin compaction by a
polycomb group protein complex. Science 306, 1574–1577 (2004). 26. Stock, J. K. et al. Ring1-mediated ubiquitination of H2A restrains poised RNA
polymerase II at bivalent genes in mouse ES cells. Nat. Cell Biol. 9, 1428–1435
(2007). 57. Peng, J. C. et al. Jarid2/Jumonji coordinates control of PRC2 enzymatic
activity and target gene occupancy in pluripotent cells. Cell 139, 1290–1302
(2009). 27. Ku, M. et al. Genomewide analysis of PRC1 and PRC2 occupancy identifies
two classes of bivalent domains. PLoS Genet 4, e1000242 (2008). 58. Schuettengruber, B. et al. Functional anatomy of polycomb and trithorax
chromatin landscapes in Drosophila embryos. PLoS Biol. 7, e13 (2009). 28. Brookes, E. et al. Polycomb associates genome-wide with a specific RNA 28. Brookes, E. et al. Polycomb associates genome-wide with a specific RNA
polymerase II variant, and regulates metabolic genes in ESCs. Cell Stem Cell
10, 157–170 (2012). 59. Schwartz, Y. B. et al. Alternative epigenetic chromatin states of polycomb
target genes. PLoS Genet 6, e1000805 (2010). 29. Weiner, A. et al. Co-ChIP enables genome-wide mapping of histone mark co-
occurrence at single-molecule resolution. Nat. Biotechnol. 34, 953–961
(2016). 60. Hathaway, N. A. et al. Dynamics and memory of heterochromatin in living
cells. Cell 149, 1447–1460 (2012). 61. Erdel, F., How communication between nucleosomes enables spreading and
epigenetic memory of histone modifications. BioEssays 39, 1700053 https://
doi.org/10.1002/bies.201700053 (2017). 30. Kinkley, S. et al. reChIP-seq reveals widespread bivalency of H3K4me3 and
H3K27me3 in CD4(+) memory T cells. Nat. Commun. 7, 12514 (2016). g
62. Steffen, P. A., Fonseca, J. P. & Ringrose, L. Epigenetics meets mathematics:
towards a quantitative understanding of chromatin biology. BioEssays 34,
901–913 (2012). y
31. Surface, L. E. et al. H2A.Z.1 monoubiquitylation antagonizes BRD2 to
maintain poised chromatin in ESCs. Cell Rep. 14, 1142–1155 (2016). 32. Voigt, P. et al. Asymmetrically modified nucleosomes. Cell 151, 181–193
(2012). 63. Fonseca, J. P. et al. In vivo Polycomb kinetics and mitotic chromatin
binding distinguish stem cells from differentiated cells. Genes Dev. 26,
857–871 (2012). 33. Sen, S., Block, K. References References
. Gardner, T. S., Cantor, C. R. & Collins, J. J. Construction of a genetic toggle
switch in Escherichia coli. Nature 403, 339–342 (2000). . Gardner, T. S., Cantor, C. R. & Collins, J. J. Construction of a genetic toggle
switch in Escherichia coli. Nature 403, 339–342 (2000). 15 (2019) 10:2133 | https://doi.org/10.1038/s41467-019-10130-2 | www.nature.com/naturecommunications NATURE COMMUNICATIONS | (2019) 10:2133 | https://doi.org/10.1038/s41467-019-10130-2 | www.nature.com/naturecommunications ARTICLE Drosophila enhancer of Zeste/ESC complexes have a histone
H3 methyltransferase activity that marks chromosomal Polycomb sites. Cell
111, 185–196 (2002). regulatory mechanism. J. Biol. Chem. 286, 8369–8374 (2011) 110. Miyazaki, H. et al. Ash1l methylates Lys36 of histone H3 independently of
transcriptional elongation to counteract polycomb silencing. PLoS Genet 9,
e1003897 (2013). 80. Muller, J. et al. Histone methyltransferase activity of a Drosophila Polycomb
group repressor complex. Cell 111, 197–208 (2002). 111. Pasini, D. et al. Coordinated regulation of transcriptional repression by the
RBP2 H3K4 demethylase and Polycomb-Repressive Complex 2. Genes Dev. 22, 1345–1355 (2008). 81. Ferrari, K. J. et al. Polycomb-dependent H3K27me1 and H3K27me2 regulate
active transcription and enhancer fidelity. Mol. cell 53, 49–62 (2014). 82. Agger, K. et al. UTX and JMJD3 are histone H3K27 demethylases involved in
HOX gene regulation and development. Nature 449, 731–734 (2007). 112. Lee, N. et al. The trithorax-group protein Lid is a histone H3 trimethyl-Lys4
demethylase. Nat. Struct. Mol. Biol. 14, 341–343 (2007). 83. Hong, S. et al. Identification of JmjC domain-containing UTX and JMJD3 as
histone H3 lysine 27 demethylases. Proc. Natl Acad Sci. USA 104,
18439–18444 (2007). 113. Eissenberg, J. C. et al. The trithorax-group gene in Drosophila little imaginal
discs encodes a trimethylated histone H3 Lys4 demethylase. Nat. Struct. Mol. Biol. 14, 344–346 (2007). 84. Lan, F. et al. A histone H3 lysine 27 demethylase regulates animal posterior
development. Nature 449, 689–694 (2007). 114. He, J., Kallin, E. M., Tsukada, Y. & Zhang, Y. The H3K36 demethylase
Jhdm1b/Kdm2b regulates cell proliferation and senescence through p15
(Ink4b). Nat. Struct. Mol. Biol. 15, 1169–1175 (2008). 85. Smith, E. R. et al. Drosophila UTX is a histone H3 Lys27 demethylase that
colocalizes with the elongating form of RNA polymerase II. Mol. Cell Biol. 28,
1041–1046 (2008). 115. Kavi, H. H. & Birchler, J. A. Drosophila KDM2 is a H3K4me3 demethylase
regulating nucleolar organization. BMC Res. notes 2, 217 (2009). 86. Kooistra, S. M. & Helin, K. Molecular mechanisms and potential functions of
histone demethylases. Nat. Rev. Mol. Cell Biol. 13, 297–311 (2012). 116. Min, J., Zhang, Y. & Xu, R. M. Structural basis for specific binding of
Polycomb chromodomain to histone H3 methylated at Lys 27. Genes Dev. 17,
1823–1828 (2003). 87. Tsukada, Y., Ishitani, T. & Nakayama, K. I. KDM7 is a dual demethylase for
histone H3 Lys 9 and Lys 27 and functions in brain development. Genes Dev. ARTICLE ARTICLE NATURE COMMUNICATIONS | https://doi.org/10.1038/s41467-019-10130-2 95. Sahtoe, D. D., van Dijk, W. J., Ekkebus, R., Ovaa, H. & Sixma, T. K. BAP1/
ASXL1 recruitment and activation for H2A deubiquitination. Nat. Commun. 7, 10292 (2016). 64. Steffen, P. A. et al. Quantitative in vivo analysis of chromatin binding of
Polycomb and Trithorax group proteins reveals retention of ASH1 on mitotic
chromatin. Nucleic Acids Res. 41, 5235–5250 (2013). 65. Yang, Z. et al. EBS is a bivalent histone reader that regulates floral phase
transition in Arabidopsis. Nat. Genet. 50, 1247–1253 (2018). 96. Tie, F. et al. CBP-mediated acetylation of histone H3 lysine 27 antagonizes
Drosophila Polycomb silencing. Development 136, 3131–3141 (2009). 66. Okulski, H., Druck, B., Bhalerao, S. & Ringrose, L. Quantitative analysis of
polycomb response elements (PREs) at identical genomic locations
distinguishes contributions of PRE sequence and genomic environment. Epigenet. Chromatin 4, 4 (2011). 97. Reynolds, N. et al. NuRD-mediated deacetylation of H3K27 facilitates
recruitment of Polycomb Repressive Complex 2 to direct gene repression. EMBO J. 31, 593–605 (2012). 98. Smith, S. T. et al. Modulation of heat shock gene expression by the TAC1
chromatin-modifying complex. Nat. Cell Biol. 6, 162–167 (2004). p g
(
)
67. Laugesen, A. & Helin, K. Chromatin repressive complexes in stem cells,
development, and cancer. Cell Stem Cell 14, 735–751 (2014). 99. Milne, T. A. et al. MLL targets SET domain methyltransferase activity to Hox
gene promoters. Mol. Cell 10, 1107–1117 (2002). p
68. McCabe, M. T. et al. EZH2 inhibition as a therapeutic strategy for lymphoma
with EZH2-activating mutations. Nature 492, 108–112 (2012). 100. Nakamura, T. et al. ALL-1 is a histone methyltransferase that assembles a
supercomplex of proteins involved in transcriptional regulation. Mol. Cell 10,
1119–1128 (2002). 69. Kumar, R. M. et al. Deconstructing transcriptional heterogeneity in
pluripotent stem cells. Nature 516, 56–61 (2014). p
p
70. Kar, G. et al. Flipping between Polycomb repressed and active transcriptional
states introduces noise in gene expression. Nat. Commun. 8, 36 (2017). 101. Lee, J. H., Tate, C. M., You, J. S. & Skalnik, D. G. Identification and
characterization of the human Set1B histone H3-Lys4 methyltransferase
complex. J. Biol. Chem. 282, 13419–13428 (2007). 71. Mukhopadhyay, S. & Sengupta, A. M. The role of multiple marks in epigenetic
silencing and the emergence of a stable bivalent chromatin state. PLoS
Comput. Biol. 9, e1003121 (2013). 102. Wu, M. et al. ARTICLE Molecular regulation of H3K4 trimethylation by Wdr82, a
component of human Set1/COMPASS. Mol. Cell Biol. 28, 7337–7344
(2008). p
72. Ku, W. L., Girvan, M., Yuan, G. C., Sorrentino, F. & Ott, E. Modeling the
dynamics of bivalent histone modifications. PloS ONE 8, e77944 (2013). 103. Ardehali, M. B. et al. Drosophila Set1 is the major histone H3 lysine 4
trimethyltransferase with role in transcription. EMBO J. 30, 2817–2828
(2011). 73. Delachat, A. M. et al. Engineered multivalent sensors to detect coexisting
histone modifications in living stem cells. Cell Chem. Biol. 25, 51–56 e56
(2018). 104. Hallson, G. et al. dSet1 is the main H3K4 di- and tri-methyltransferase
throughout Drosophila development. Genetics 190, 91–100 (2012). 74. Kharchenko, P. V. et al. Comprehensive analysis of the chromatin landscape
in Drosophila melanogaster. Nature 471, 480–485 (2011). 105. Rickels, R. et al. An evolutionary conserved epigenetic mark of polycomb
response elements implemented by Trx/MLL/COMPASS. Mol. Cell 63,
318–328 (2016). 75. Schertel, C. et al. A large-scale, in vivo transcription factor screen defines
bivalent chromatin as a key property of regulatory factors mediating
Drosophila wing development. Genome Res. 25, 514–523 (2015). 106. Shah, R. N. et al. Examining the roles of H3K4 methylation states with
systematically characterized antibodies. Mol. Cell 72, 162–177 e167
(2018). p
g
p
76. Grzybowski, A. T., Chen, Z. & Ruthenburg, A. J. Calibrating ChIP-Seq with
nucleosomal internal standards to measure histone modification density
genome wide. Mol. cell 58, 886–899 (2015). 107. Tanaka, Y., Katagiri, Z., Kawahashi, K., Kioussis, D. & Kitajima, S. Trithorax-
group protein ASH1 methylates histone H3 lysine 36. Gene 397, 161–168
(2007). g
77. Kuzmichev, A., Nishioka, K., Erdjument-Bromage, H., Tempst, P. & Reinberg,
D. Histone methyltransferase activity associated with a human multiprotein
complex containing the Enhancer of Zeste protein. Genes Dev. 16, 2893–2905
(2002). 108. Dorighi, K. M. & Tamkun, J. W. The trithorax group proteins Kismet and
ASH1 promote H3K36 dimethylation to counteract Polycomb group
repression in Drosophila. Development 140, 4182–4192 (2013). 78. Cao, R. et al. Role of histone H3 lysine 27 methylation in Polycomb-group
silencing. Science 298, 1039–1043 (2002). repression in Drosophila. Development 140, 4182–4192 (2013 109. An, S., Yeo, K. J., Jeon, Y. H. & Song, J. J. Crystal structure of the human
histone methyltransferase ASH1L catalytic domain and its implications for the
regulatory mechanism. J. Biol. Chem. 286, 8369–8374 (2011). g
79. Czermin, B. et al. ARTICLE F., Pasini, A., Baylin, S. B. & Easwaran, H. Genome-wide
positioning of bivalent mononucleosomes. BMC Med. Genom. 9, 60 (2016). 16 NATURE COMMUNICATIONS | (2019) 10:2133 | https://doi.org/10.1038/s41467-019-10130-2 | www.nature.com/naturecommunicatio Competing interests: The authors declare no competing interests. 133. Tie, F. et al. Polycomb inhibits histone acetylation by CBP by binding directly
to its catalytic domain. Proc. Natl Acad. Sci. USA 113, E744–E753 (2016). Reprints and permission information is available online at http://npg.nature.com/
reprintsandpermissions/ y
134. Kim, D. H. et al. Histone H3K27 trimethylation inhibits H3 binding and
function of SET1-like H3K4 methyltransferase complexes. Mol. Cell Biol. 33,
4936–4946 (2013). Journal peer review information: Nature Communications thanks Lacramioara Bintu
and other anonymous reviewer(s) for their contribution to the peer review of this work. Peer reviewer reports are available. Journal peer review information: Nature Communications thanks Lacramioara Bintu
and other anonymous reviewer(s) for their contribution to the peer review of this work. Peer reviewer reports are available. 135. Jin, Y. et al. LSD1 collaborates with EZH2 to regulate expression of interferon-
stimulated genes. Biomed. Pharmacother. 88, 728–737 (2017). 136. Gajan, A., Barnes, V. L., Liu, M., Saha, N. & Pile, L. A. The histone
demethylase dKDM5/LID interacts with the SIN3 histone deacetylase complex
and shares functional similarities with SIN3. Epigenetics chromatin 9, 4 (2016). Publisher’s note: Springer Nature remains neutral with regard to jurisdictional claims in
published maps and institutional affiliations. Publisher’s note: Springer Nature remains neutral with regard to jurisdictional claims in
published maps and institutional affiliations. 137. Whyte, W. A. et al. Enhancer decommissioning by LSD1 during embryonic
stem cell differentiation. Nature 482, 221–225 (2012). 138. Bantignies, F., Goodman, R. H. & Smolik, S. M. Functional interaction
between the coactivator Drosophila CREB-binding protein and ASH1, a
member of the trithorax group of chromatin modifiers. Mol. Cell Biol. 20,
9317–9330 (2000). Open Access This article is licensed under a Creative Commons
Attribution 4.0 International License, which permits use, sharing,
adaptation, distribution and reproduction in any medium or format, as long as you give
appropriate credit to the original author(s) and the source, provide a link to the Creative
Commons license, and indicate if changes were made. The images or other third party
material in this article are included in the article’s Creative Commons license, unless
indicated otherwise in a credit line to the material. If material is not included in the
article’s Creative Commons license and your intended use is not permitted by statutory
regulation or exceeds the permitted use, you will need to obtain permission directly from
the copyright holder. To view a copy of this license, visit http://creativecommons.org/
licenses/by/4.0/. 139. Tie, F., Banerjee, R., Conrad, P. NATURE COMMUNICATIONS | https://doi.org/10.1038/s41467-019-10130-2 NATURE COMMUNICATIONS | https://doi.org/10.1038/s41467-019-10130-2 124. Cooper, S. et al. Targeting polycomb to pericentric heterochromatin in
embryonic stem cells reveals a role for H2AK119u1 in PRC2 recruitment. Cell
Rep. 7, 1456–1470 (2014). 143. Ku, M. et al. H2A.Z landscapes and dual modifications in pluripotent and
multipotent stem cells underlie complex genome regulatory functions. Genome Biol. 13, R85 (2012). p
125. Cooper, S. et al. Jarid2 binds mono-ubiquitylated H2A lysine 119 to mediate
crosstalk between Polycomb complexes PRC1 and PRC2. Nat. Commun. 7,
13661 (2016). Acknowledgements 126. Dey, A., Chitsaz, F., Abbasi, A., Misteli, T. & Ozato, K. The double
bromodomain protein Brd4 binds to acetylated chromatin during interphase
and mitosis. Proc. Natl Acad. Sci. USA 100, 8758–8763 (2003). L.R. thanks Marc Rehmsmeier for discussions. KS acknowledges support by the Danish
National Research Foundation through the Center for Models of Life, and by the Eur-
opean Research Council under the European Union’s Seventh Framework Programme
(FP/2007 2013)/ERC Grant Agreement n. 740704. L.R. acknowledges the support of the
German National Research Foundation (DFG) Exzellenzinitiativ IRI-Lifesciences https://
www.iri-ls.hu-berlin.de/en and of the EU MSCA ITN ‘PEP-NET’ Grant Agreement n. 813282. 127. Kellner, W. A. et al. Distinct isoforms of the Drosophila Brd4 homologue are
present at enhancers, promoters and insulator sites. Nucleic acids Res. 41,
9274–9283 (2013). 128. Kockmann, T. et al. The BET protein FSH functionally interacts with ASH1 to
orchestrate global gene activity in Drosophila. Genome Biol. 14, R18 (2013). 129. Francis, N. J., Follmer, N. E., Simon, M. D., Aghia, G. & Butler, J. D. Polycomb
proteins remain bound to chromatin and DNA during DNA replication
in vitro. Cell 137, 110–122 (2009). Author contributions Both authors contributed to model design, conceptual project design and writing the
manuscript. L.R. performed literature curation. K.S. performed modelling. 130. Schoeftner, S. et al. Recruitment of PRC1 function at the initiation of X
inactivation independent of PRC2 and silencing. EMBO J. 25, 3110–3122 (2006). Additional information 131. Dietrich, N. et al. REST-mediated recruitment of polycomb repressor
complexes in mammalian cells. PLoS Genet 8, e1002494 (2012). Supplementary Information accompanies this paper at https://doi.org/10.1038/s41467-
019-10130-2. 132. Yu, M. et al. Direct recruitment of polycomb repressive complex 1 to
chromatin by core binding transcription factors. Mol. cell 45, 330–343 (2012). Competing interests: The authors declare no competing interests. Competing interests: The authors declare no competing interests. ARTICLE 24, 432–437 (2010). 117. Fischle, W. et al. Molecular basis for the discrimination of repressive methyl-
lysine marks in histone H3 by Polycomb and HP1 chromodomains. Genes
Dev. 17, 1870–1881 (2003). 88. Wang, H. et al. Role of histone H2A ubiquitination in Polycomb silencing. Nature 431, 873–878 (2004). 89. de Napoles, M. et al. Polycomb group proteins Ring1A/B link ubiquitylation
of histone H2A to heritable gene silencing and X inactivation. Dev. Cell 7,
663–676 (2004). 118. Wang, L. et al. Hierarchical recruitment of polycomb group silencing
complexes. Mol. Cell 14, 637–646 (2004). 119. Bernstein, E. et al. Mouse polycomb proteins bind differentially to methylated
histone H3 and RNA and are enriched in facultative heterochromatin. Mol. Cell Biol. 26, 2560–2569 (2006). 90. Cao, R., Tsukada, Y. & Zhang, Y. Role of Bmi-1 and Ring1A in H2A
ubiquitylation and Hox gene silencing. Mol. Cell 20, 845–854 (2005). y
g
g
91. Lagarou, A. et al. dKDM2 couples histone H2A ubiquitylation to histone H3
demethylation during Polycomb group silencing. Genes Dev. 22, 2799–2810
(2008). 120. Hansen, K. H. et al. A model for transmission of the H3K27me3 epigenetic
mark. Nat. Cell Biol. 10, 1291–1300 (2008). 121. Margueron, R. et al. Role of the polycomb protein EED in the propagation of
repressive histone marks. Nature 461, 762–767 (2009). 92. Gao, Z. et al. PCGF homologs, CBX proteins, and RYBP define functionally
distinct PRC1 family complexes. Mol. cell 45, 344–356 (2012). 122. Kalb, R. et al. Histone H2A monoubiquitination promotes histone H3
methylation in Polycomb repression. Nat. Struct. Mol. Biol. 21, 569–571
(2014). 93. Scheuermann, J. C. et al. Histone H2A deubiquitinase activity of the Polycomb
repressive complex PR-DUB. Nature 465, 243–247 (2010). 94. Baymaz, H. I. et al. MBD5 and MBD6 interact with the human PR-DUB
complex through their methyl-CpG-binding domain. Proteomics 14,
2179–2189 (2014). 123. Blackledge, N. P. et al. Variant PRC1 complex-dependent H2A ubiquitylation
drives PRC2 recruitment and polycomb domain formation. Cell 157,
1445–1459 (2014). 17 NATURE COMMUNICATIONS | (2019) 10:2133 | https://doi.org/10.1038/s41467-019-10130-2 | www.nature.com/naturecommunications ARTICLE ARTICLE © The Author(s) 2019 Competing interests: The authors declare no competing interests. A., Scacheri, P. C. & Harte, P. J. Histone
demethylase UTX and chromatin remodeler BRM bind directly to CBP and
modulate acetylation of histone H3 lysine 27. Mol. Cell Biol. 32, 2323–2334
(2012). 140. Gil, J. & O'Loghlen, A. PRC1 complex diversity: where is it taking us? Trends
cell Biol. 24, 632–641 (2014). 141. Simon, J. A. & Kingston, R. E. Occupying chromatin: Polycomb mechanisms
for getting to genomic targets, stopping transcriptional traffic, and staying put. Mol. Cell 49, 808–824 (2013). 142. Devaiah, B. N. et al. BRD4 is an atypical kinase that phosphorylates serine2 of
the RNA polymerase II carboxy-terminal domain. Proc. Natl Acad. Sci. USA
109, 6927–6932 (2012). © The Author(s) 2019 NATURE COMMUNICATIONS | (2019) 10:2133 | https://doi.org/10.1038/s41467-019-10130-2 | www.nature.com/naturecommunications 18 NATURE COMMUNICATIONS | (2019) 10:2133 | https://doi.org/10.1038/s41467-019-10130-2 | www.nature.com/naturecommunicatio
|
https://openalex.org/W4229511412
|
https://www.qeios.com/read/WD9IQ9/pdf
|
English
| null |
Hair Cell
|
Definitions
| 2,020
|
cc-by
| 79
|
Hair Cell National Cancer Institute National Cancer Institute Qeios · Definition, February 2, 2020 Open Peer Review on Qeios Open Peer Review on Qeios Qeios ID: WD9IQ9 · https://doi.org/10.32388/WD9IQ9 Source National Cancer Institute. Hair Cell. NCI Thesaurus. Code C12629. A type of sensory epithelial cell, located in the inner ear, that use hair-like projections
called stereocilia to detect sound waves and transform them into neural signals that can
be interpreted by the brain. Qeios ID: WD9IQ9 · https://doi.org/10.32388/WD9IQ9 1/1
|
https://openalex.org/W3103023628
|
https://europepmc.org/articles/pmc5271141?pdf=render
|
English
| null |
Physical and electrochemical area determination of electrodeposited Ni, Co, and NiCo thin films
|
Nano convergence
| 2,016
|
cc-by
| 6,965
|
© 2016 Gira et al. This article is distributed under the terms of the Creative Commons Attribution 4.0 International License (http://
creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium, provided
you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate
if changes were made. Physical and electrochemical area
determination of electrodeposited Ni, Co,
and NiCo thin films Matthew J. Gira1,2, Kevin P. Tkacz1,3 and Jennifer R. Hampton1* Abstract The surface area of electrodeposited thin films of Ni, Co, and NiCo was evaluated using electrochemical double-layer
capacitance, electrochemical area measurements using the [Ru(NH3)6]3+/[Ru(NH3)6]2+ redox couple, and topographic
atomic force microscopy (AFM) imaging. These three methods were compared to each other for each composition
separately and for the entire set of samples regardless of composition. Double-layer capacitance measurements were
found to be positively correlated to the roughness factors determined by AFM topography. Electrochemical area
measurements were found to be less correlated with measured roughness factors as well as applicable only to two of
the three compositions studied. The results indicate that in situ double-layer capacitance measurements are a practi-
cal, versatile technique for estimating the accessible surface area of a metal sample. Keywords: Electrodeposition, Ni, Co, NiCo, Alloy , Capacitance, Area, Atomic force microscopy Gira et al. Nano Convergence (2016) 3:6
DOI 10.1186/s40580-016-0063-0 Open Access 1 Background One technique for area measurement is based on the
physical absorption of gas molecules to a surface follow-
ing the theory presented by Brunauer, Emmet, and Teller
(BET) [2–4]. Although this is a well-understood and reg-
ularly-used method, BET measurements have limitations,
specifically the effects that heat treatments may have on
the sample being characterized as well as the larger sam-
ple sizes needed to achieve the desired sensitivity [5]. Nanoporous materials are of increasing scientific and
technological interest due to a variety of useful prop-
erties such as low mass density, high surface area, high
strength, and enhanced optical, electrical, thermal, and
catalytic behavior. Potential applications of metals with
nanoporous morphology include batteries, capacitors,
magnetic storage media, lightweight structures, sensors,
and water filtration devices [1]. The enhanced surface
area and size-dependent reactivity of nanoporous metals
also make them a promising area of study for a number of
catalytic applications. Electrochemical techniques for determining surface
area have the advantage of being in situ and can be per-
formed just previous to or after any electrochemical reac-
tivity measurements of interest. These techniques fall
into two general categories. The first type uses a surface-
limited chemical reaction to quantify the surface area of
the electrode. In contrast, the second type measures a
physical characteristic that is proportional to the surface
area. An important factor in evaluating the reactivity of a
porous metal is the surface area available for reaction. Both increased surface area and changes in intrinsic reac-
tivity can have significant effects on the overall behavior
of a target material. Thus, straightforward and practical
area measurement procedures are an essential aspect of
catalysis research. Using a surface-limited chemical reaction such as
adsorption of hydrogen or carbon monoxide [6–15],
underpotential deposition of a new metallic species [5,
13, 16, 17], or surface oxide formation [9, 17–22] to
quantify the surface area of the electrode can be quite
sensitive. However, a disadvantage is that a particular *Correspondence: hampton@hope.edu
1 Department of Physics, Hope College, Holland, MI 49423, USA
Full list of author information is available at the end of the article *Correspondence: hampton@hope.edu
1 Department of Physics, Hope College, Holland, MI 49423, USA
Full list of author information is available at the end of the article 2.1 Electrochemistryh The electrodeposition and electrochemical characteriza-
tion were performed using an Epsilon electrochemical
workstation (Bioanalytical Systems, Inc., West Lafayette,
IN, USA) and a custom-built Teflon cell with a work-
ing electrode area of 0.032 cm2 defined with a Kalrez
o-ring [24]. The counter electrode was a coil of platinum
wire (Alfa Aesar, Ward Hill, MA, USA) and the reference
electrode was an Ag/AgCl (3 M NaCl) electrode (Bio-
analytical Systems, Inc., West Lafayette, IN, USA). All
of the potentials recorded are with respect to this refer-
ence electrode. The electrolyte solutions were created
using water that was purified through successive reverse
osmosis, deionization, and UV purification stages. All of
the chemicals used for these electrolytes were purchased
from Sigma-Aldrich (St. Louis, MO, USA) and used
as received. Every experiment was carried out at room
temperature. p
q
Topographic measurements of samples with a scanning
tunneling microscope (STM) or atomic force microscope
(AFM) can also be used to quantify the surface area of a
sample [8, 34, 35]. These methods have the advantage of
providing direct quantitative measurements of surface
morphology. For AFM in particular, topographic meas-
urements are not sensitive to the nature of the surface
being probed. However, scanning probe techniques are
local rather than ensemble measurements. Thus, a num-
ber of images must be taken for any surface in question
to ensure the images are representative of the sample as
a whole. For materials with porous morphology, scanning
probe microscope measurements are limited, because the
local probe can only measure structures which are acces-
sible from the top of the sample. Similarly, if a surface has
features smaller than that of the scanning probe tip itself,
those features will not be imaged accurately by the tech-
nique. However, for materials with simpler morphology,
scanning probe measurements provide a nice comple-
ment to the other methods described here. Page 2 of 8 Gira et al. Nano Convergence (2016) 3:6 two electrochemical methods, double-layer capacitance
measurements and area determination using a ruthe-
nium-based redox probe. These measurements were
compared to the roughness factors extracted from ex situ
AFM images of the samples. Correlations between these
three measurements were explored, both for the samples
with the same composition and for the entire set of sam-
ples regardless of composition. reaction may be specific to the material being assessed. For example, gold oxide formation has been used exten-
sively as a probe of gold electrode surface area, but this
method can not be applied directly to an electrode of a
different composition without considering the extent and
potential range of oxide formation on that new material. Rather than a chemical reaction, a electrochemical
characterization using a physical characteristic can be
used to quantify the surface area of a working electrode. The current due to a well-characterized redox reaction,
such as the reduction of [Fe(CN)6]3− to Fe(CN)6]4−, is
one such measurement [19, 23–25]. Similarly, the elec-
trochemical double-layer capacitance of an electrode,
which can be measured either by cyclic voltammetry or
by electrochemical impedance spectroscopy, is propor-
tional to its surface area [9, 19–22, 25–33]. These tech-
niques depend on the conducting nature of the electrode
rather than its chemical identity, so to first approxima-
tion they do not depend on the nature of the material
being studied. However, the potential range necessary
for these measurements must be considered, because the
characterization technique itself may affect the structure
or composition of the material in question. 2.2 Depositioni All thin films were deposited from solutions containing
0.5 M H3BO3 and 1 M Na2SO4 along with 0.1 M NiSO4
for the nickel thin films, 0.1 M CoSO4 for the cobalt
thin films, or 0.75 mM NiSO4 and 0.25 mM CoSO4 for
the nickel–cobalt thin films. The working electrode sub-
strates were cleaved from a silicon wafer plated with
1000 Å of gold over a 50 Å titanium adhesion layer (Plat-
ypus Technologies, LLC, Madison, WI, USA). Controlled
potential electrolysis was used to step the potential of
the working electrode from open circuit to −1000 mV. The deposition was stopped once the desired amount of
charge, ranging from 200 to 1000 mC, was achieved in
order to vary the thickness of the deposited films. In this work we compare electrochemical methods
for determining the surface area of electrodeposited
metal thin films with AFM topographic measurements
of the same samples. Electrodeposited nickel, cobalt,
and nickel–cobalt were chosen for the study because of
the interest in these materials as catalysts. The thickness
of these films was varied by controlling the total charge
during the deposition process. In this way, the resulting
roughness, and therefore surface area, of the material
was varied. The resulting films were characterized using 2.4 Electrochemical characterization Electrochemical characterization consisted of double-
layer capacitance and active area measurements. Elec-
trochemical capacitance was measured using cyclic
voltammetry (CV) in 0.5 M KOH by sweeping from −50
to −350 mV and back to −50 mV. The scan rates were
varied between 25 and 400 mV/s. The electrochemically
active area was also measured using CV. The electrolyte
solution was 5 mM Ru(NH3)6Cl3 and 1 M KCl. The poten-
tial was swept from 100 to −600 mV and back to 100 mV
with varying scan rates in the range of 100–901 mV/s. A
minimum of three trials of both experiments were per-
formed for each sample. 2.3 Physical characterization Physical characterization of the samples consisted of
roughness and composition measurements. Atomic
force microscope topography was used to measure the
roughness of each thin film. This was completed using a
Dimension Icon AFM (Bruker, Santa Barbara, CA, USA)
using the ScanAsyst mode and SCANASYST-AIR canti-
levers. A minimum of three 10 µm AFM images (512 pix-
els × 512 pixels) were taken of each sample. Nanoscope Gira et al. Nano Convergence (2016) 3:6 Page 3 of 8 Fig. 1 Example 10 µm × 10 µm AFM topographic measurements for
a Ni, b Co and c NiCo thin films. Each sample had deposited charge
of 1000 mC. The scale bar is 2 µm for all the images. The vertical scale
is indicated to the right and is different for each image. The roughness
factors for these images are (a) 1.12, (b) 1.41, and (c) 1.05 (a) Ni (b) Co
2.5
m
(a) Ni
1.0
m
(c) NiCo
250
nm (a) Ni Analysis software (Bruker, Santa Barbara, CA, USA) was
used to find the three-dimensional area of each image. For the NiCo thin films the elemental composition was
measured. Scanning electron microscopy (SEM) and
energy dispersive X-ray spectroscopy (EDS) measure-
ments were completed using a TM3000 Tabletop SEM
(Hitachi, Tokyo, Japan) and a Quantax 70 EDS attach-
ment (Bruker, Madison, WI, USA). Images and EDS data
were taken at ×60 magnification, and Quantax 70 soft-
ware was used to obtain the Ni and Co compositions
from the EDS spectra. (b) Co (b) Co 3 Results and discussion (c) NiCo The goals of this work were to explore the correlations
between the AFM-based and electrochemical measure-
ments for samples with different roughnesses and there-
fore different areas. The roughness of each of the samples
was determined using AFM topographic measurements. Example AFM images are shown in Fig. 1 for samples
with a deposited charge of 1000 mC. The Ni and Co films
exhibit similar crystallite formation, with the resulting
Co features larger and taller than the corresponding Ni
ones for the same deposited charge. In contrast, the NiCo
film has a distinct texture with smaller, less compact
crystallites. For each image, the data were flattened using a first
order filter to remove sample tilt. Afterwards, the rough-
ness factor, RF, was calculated as RF = AAFM/Aproj,
where AAFM is the surface area calculated from the image
using the Nanoscope Analysis software and Aproj is the Gira et al. Nano Convergence (2016) 3:6 Page 4 of 8 deposited charge. For the entire set of samples, regardless
of composition, the roughness factors ranged from about
1.05 to 1.4. That is the samples had measured surface
areas ranging from 5 to 40 % higher than the correspond-
ing projected area. projected (flat) area of the measured region, 100 µm2 in
this case. From this calculation, the roughness factor is
proportional to the surface area of the sample measured
using AFM, but is not specific to the image sized used. i
The average RF for the three types of films are graphed
in Fig. 2 as a function of the deposited charge, Q. The
approximate average thickness, t, of the films corre-
sponding to each deposited charge is shown on the upper
horizontal axis of the figure. The conversion from depos-
ited charge to thickness was calculated assuming 100 %
current efficiency from t = Q/(neAρ∗), where n = 2 is
the number of electrons in the Ni or Co deposition reac-
tion, e is the charge on the electron, A is the defined area
of the working electrode, and ρ∗ is the number density
of the deposit. The bulk densities (in g/cm3) and molar
masses (in g/mol) of Ni and Co were used to calculate
a value of ρ∗ for each metal. 3 Results and discussion Because the values for Ni
and Co are so similar, 9.14 × 1022 and 9.09 × 1022 cm−3
respectively, an average value of ρ∗ was used to calculate
the axis in the figure, corresponding to the assumption of
an equal-component alloy. The systematic error for this
assumption compared to using the value of ρ∗ for pure Ni
or pure Co is approximately 0.2 %. The compositions of the NiCo thin films were meas-
ured from EDS spectra taken at ×60 magnification and
are shown in Fig. 3 as a function of the average rough-
ness of the samples. The Ni composition of the deposited
alloys was generally between 60 and 70 at.%. The fact
that the samples have a smaller Ni composition than that
of the deposition solution (75 at.%) is attributed to the
anomalous codeposition phenomenon which is common
for iron group metals [36–39]. The decrease in Ni com-
position with increasing roughness is consistent with the
data in Fig. 2 where NiCo samples are generally rougher
than the Ni samples but smoother than the Co samples. Electrochemical double-layer capacitance measure-
ments were made on all the electrodeposited samples
using CV in KOH electrolyte. Example measurements
for a variety of scan rates are shown in the inset of Fig. 4,
showing the featureless current response expected of
a capacitor. For these metals in alkaline electrolytes, a
more complex pseudocapactive response corresponding
to metal oxide and/or hydroxide redox reactions is often
seen [40–42]. For the measurements here, however, the
potential window used is significantly negative of that
needed for these redox reactions to occur. As a result, the
featureless CVs shown in the inset are measured instead. For a given scan rate, v, the average currents during the
forward and reverse sweeps were calculated, and half
of the difference between these two values was taken as As seen qualitatively in Fig. 2, for the same film thick-
ness, the Ni samples generally are the smoothest, the Co
samples have the roughest topography, and the NiCo
alloy samples have intermediate roughness factors. 3 Results and discussion For
the Ni and Co samples, the roughness factor generally
increases as the thickness of the samples increases, while
for the NiCo samples, the roughness fluctuates with 0
200
400
600
800
1000
1200
1.0
1.1
1.2
1.3
1.4
Ni
Co
NiCo
RF
0
2
4
6
8
10
12
t / µm f
0
200
400
600
800
1000
1200
1.0
1.1
1.2
1.3
1.4
Ni
Co
NiCo
RF
Q / mC
0
2
4
6
8
10
12
t / µm
Fig. 2 Average roughness factor, RF, of each sample as a function
of the deposited charge, Q. The second horizontal axis indicates the
approximate average thickness, t, of the samples. Error bars represent
the standard error of the mean for the measurements
1.0
1.1
1.2
1.3
50
55
60
65
70
75
NiCo Samples
Solution
Ni / at. %
RF
Fig. 3 Ni composition for the NiCo samples as a function of the
average roughness factor, RF, of the samples. Composition error bars
represent the typical EDS uncertainty. The dashed line indicates the Ni
composition in the deposition solution 1.0
1.1
1.2
1.3
50
55
60
65
70
75
NiCo Samples
Solution
Ni / at. %
RF
Fig. 3 Ni composition for the NiCo samples as a function of the
average roughness factor, RF, of the samples. Composition error bars
represent the typical EDS uncertainty. The dashed line indicates the Ni
composition in the deposition solution 1.0
1.1
1.2
1.3
50
55
60
65
70
75
NiCo Samples
Solution
Ni / at. %
RF
Fig. 3 Ni composition for the NiCo samples as a function of the
average roughness factor, RF, of the samples. Composition error bars
represent the typical EDS uncertainty. The dashed line indicates the Ni
composition in the deposition solution 1.0
1.1
1.2
1.3
50
55
60
65
70
75
NiCo Samples
Solution
Ni / at. %
RF RF Fig. 2 Average roughness factor, RF, of each sample as a function
of the deposited charge, Q. The second horizontal axis indicates the
approximate average thickness, t, of the samples. Error bars represent
the standard error of the mean for the measurements Fig. 2 Average roughness factor, RF, of each sample as a function
of the deposited charge, Q. The second horizontal axis indicates the
approximate average thickness, t, of the samples. Error bars represent
the standard error of the mean for the measurements Fig. 3 Results and discussion 3 Ni composition for the NiCo samples as a function of the
average roughness factor, RF, of the samples. Composition error bars
represent the typical EDS uncertainty. The dashed line indicates the Ni
composition in the deposition solution Page 5 of 8 Gira et al. Nano Convergence (2016) 3:6 results, average RF, in Fig. 6a, b respectively. Because Aec
could not be measured for the Co samples, no data for Co
are included in Fig. 6b. the capacitive current, Idl, for that scan rate. This current
was linearly dependent on the scan rate, as seen in Fig. 4,
indicating that the films acted as simple capacitors in this
potential scan range. The measured capacitance, Cdl, was
calculated using the time derivative of the definition of
capacitance, Idl = Cdlv, as the slope of the linear fit [43]. The results in Fig. 6a for all three types of samples show
that there is a clear trend towards larger capacitance for
rougher samples. There is some fluctuation in this cor-
relation between capacitance and roughness, which
increases for the rougher samples. Within this level of
fluctuation, however, the observed trend between capaci-
tance and roughness factor is the same for the group of
samples as a whole, regardless of the sample composi-
tion or the morphological differences seen in the AFM
topography (Fig. 1). This was of particular interest for this
study because of the practical importance of determining
surface area of materials with a variety of compositions
and structures. For these reasons, the results indicate
that electrochemical double-layer capacitance is useful as
a semi-quantitative measure of the surface area of elec-
trodeposited samples. i
Electrochemical area measurements were made on
Ni and NiCo samples with CV using the [Ru(NH3)6]3+/
[Ru(NH3)6]2+ redox couple. Example measurements for
a variety of scan rates are shown in the inset of Fig. 5,
which show the expected current response for a revers-
ible redox reaction [43]. The ruthenium-based probe was
chosen because the potential window for the CV experi-
ment generally does not interfere with the deposited film. These area measurements could not be made on the Co
samples, however, because the CV measurements did
not result in reversible redox behavior and the scans in
that potential range affected the structure of the film. The
magnitude of the peak cathodic current, Ip, as a func-
tion of the scan rate, v, is shown in Fig. 3 Results and discussion 5 for an example
measurement. The electrochemical area, Aec, of the sam-
ple was calculated using the Randles–Sevcik equation,
Ip = 0.4463nFAecC(nF/RT)1/2v1/2D1/2, where n = 1 is
the number of electrons involved in the redox reaction,
F is Faraday’s constant, C is the bulk concentration of the
analyte, R is the molar gas constant, T is the tempera-
ture, and D is the diffusion constant of the analyte [43]. For [Ru(NH3)6]3+, the measured diffusion constant is
7.1 × 10−6 cm2/s [44–46]. In contrast to the capacitance results, the correla-
tion between area measurements and roughness factor,
shown in Fig. 6b for the Ni and NiCo samples, is less
clear. In particular, although the smoother Ni samples
generally have lower capacitance values than the rougher
NiCo samples, they have higher measured electrochemi-
cal areas. To explore these observations further, the ratio of aver-
age capacitance to average area, Cdl/Aec, was calculated
for each of the Ni and NiCo samples. Fig. 6c shows this
ratio as a function of the average RF of the samples. For
the Ni samples, the capacitance-to-area ratio fluctuates The results of these two electrochemical measure-
ments, the average Cdl and Aec for each sample, are
graphed as a function of the AFM-based measurement 0
200
400
600
800
1000
0
20
40
60
80
100
Ip / µA
v / mV s
-1
-750 -500 -250
0
250
-100
-50
0
50
100
Current / µA
Potential / mV
Fig. 5 Example electrochemical area measurements for a NiCo thin
film. The sample had a deposited charge of 1000 mC. The inset shows
CV measurements in 5 mM Ru(NH3)6Cl3 and 1 M KCl at 200, 400, and
800 mV/s. The magnitude of the peak cathodic current, Ip, is fit to a
square root function vs. scan rate, v, to determine the area, Aec, of the
sample using the Randles–Sevcik equation 0
200
400
600
800
1000
0
20
40
60
80
100
Ip / µA
v / mV s
-1
-750 -500 -250
0
250
-100
-50
0
50
100
Current / µA
Potential / mV
Fig. 5 Example electrochemical area measurements for a NiCo thin
film. The sample had a deposited charge of 1000 mC. The inset shows
CV measurements in 5 mM Ru(NH3)6Cl3 and 1 M KCl at 200, 400, and
800 mV/s. 3 Results and discussion The magnitude of the peak cathodic current, Ip, is fit to a
square root function vs. scan rate, v, to determine the area, Aec, of the
sample using the Randles–Sevcik equation 0
100
200
300
400
0.00
0.25
0.50
0.75
1.00
1.25
Idl / µA
v / mV s
-1
-400 -300 -200 -100
0
-2
0
2
Current / µA
Potential / mV
Fig. 4 Example double-layer capacitance measurements for a NiCo
thin film. The sample had a deposited charge of 1000 mC. The inset
shows CV measurements in 1 M KOH at 75, 225, and 350 mV/s. The
slope of the linear fit to the capacitive current, Idl, vs. scan rate, v, is the
measured double-layer capacitance, Cdl, for the sample
0
200
400
600
800
1000
0
20
40
60
80
100
Ip / µA
v / mV s
-1
-750 -500 -250
0
250
-100
-50
0
50
100
Current / µA
Potential / mV
Fig. 5 Example electrochemical area measurements for a NiCo thin
film. The sample had a deposited charge of 1000 mC. The inset shows
CV measurements in 5 mM Ru(NH3)6Cl3 and 1 M KCl at 200, 400, and
800 mV/s. The magnitude of the peak cathodic current, Ip, is fit to a
square root function vs. scan rate, v, to determine the area, Aec, of the
sample using the Randles–Sevcik equation 0
200
400
600
800
1000
0
20
40
60
80
100
Ip / µA
v / mV s
-1
-750 -500 -250
0
250
-100
-50
0
50
100
Current / µA
Potential / mV 0
100
200
300
400
0.00
0.25
0.50
0.75
1.00
1.25
Idl / µA
v / mV s
-1
-400 -300 -200 -100
0
-2
0
2
Current / µA
Potential / mV
Fig. 4 Example double-layer capacitance measurements for a NiCo
thin film. The sample had a deposited charge of 1000 mC. The inset
shows CV measurements in 1 M KOH at 75, 225, and 350 mV/s. The
slope of the linear fit to the capacitive current, Idl, vs. scan rate, v, is the
measured double-layer capacitance, Cdl, for the sample 0
100
200
300
400
0.00
0.25
0.50
0.75
1.00
1.25
Idl / µA
v / mV s
-1
-400 -300 -200 -100
0
-2
0
2
Current / µA
Potential / mV Fig. 3 Results and discussion This
value is larger than, but on the order of 20 µF/cm2, the
specific capacitance value typically used in the literature
for a variety of metals and alloys [9, 13, 21, 22, 26, 28, 30–
32, 47–52]. In contrast, the NiCo films have even larger
capacitance-to-area ratios, between 100 and 500 µF/cm2,
and the ratio tends to increase with increasing roughness
factor. The larger ratios for the NiCo films may be the
result of the area measurements being smaller than they between 40 and 75 µF/cm2 for all roughness factors. This
value is larger than, but on the order of 20 µF/cm2, the
specific capacitance value typically used in the literature
for a variety of metals and alloys [9, 13, 21, 22, 26, 28, 30–
32, 47–52]. In contrast, the NiCo films have even larger
capacitance-to-area ratios, between 100 and 500 µF/cm2,
and the ratio tends to increase with increasing roughness
factor. The larger ratios for the NiCo films may be the
result of the area measurements being smaller than they g
One explanation for the electrochemical areas of the
NiCo samples being underestimated is that in addition
to the NiCo films generally being rougher than the Ni
films, they display a distinct morphology (Fig. 1c). For
rougher, more complex morphologies, the assumption of
planar diffusion which leads to the Randles–Sevcik equa-
tion may not be accurate. Specifically, the thickness of the
diffusion layer can be as large as 10s of µm for the scan
ranges and rates used in the area measurements [43]. Thus, for the samples here, with topographic features
on the scale of 100s of nm to a few µm, some portions
of the sample area would not contribute as strongly to
the measured current compared to that expected from
the simple planar diffusion model. On the other hand,
double-layer capacitance measurements do not depend
on the geometry and extent of the diffusion layer. Instead,
during capacitive charging and discharging, non-specifi-
cally adsorbing ions such as K+ and OH− can approach an
electrode surface as close as the outer Helmholtz plane,
generally a distance of 5–10 Å [43]. Thus, area meas-
urements may be underestimated in the case of rough,
complex topography compared to capacitance measure-
ments of the same sample. This, in turn, would lead to the
observed higher capacitance-to-area ratios as well as to
the lack of correlation between area and roughness meas-
urements. 3 Results and discussion A similar, but smaller, effect may also explain
capacitance-to-area ratios for the smoother Ni samples
being slightly higher than is typical in the literature. 1
10
Ni
Co
NiCo
Cdl / µF
(a)
0.01
0.02
0.03
0.04
Ni
NiCo
(b)
Aec / cm
2
1
1.1
1.2
1.3
1.4
RF
0
100
200
300
400
500
Ni
NiCo
(c)
Cdl/Aec / µF cm
-2
Fig. 6 a Average capacitance, Cdl, b average area, Aec, and c ratio of
capacitance to area, Cdl/Aec, of each sample as a function of the aver-
age roughness factor, RF. Error bars represent the standard error of the
mean for the measurements 1
10
Ni
Co
NiCo
Cdl / µF
(a) (a) (a) (b) 3 Results and discussion 5 Example electrochemical area measurements for a NiCo thin
film. The sample had a deposited charge of 1000 mC. The inset shows
CV measurements in 5 mM Ru(NH3)6Cl3 and 1 M KCl at 200, 400, and
800 mV/s. The magnitude of the peak cathodic current, Ip, is fit to a
square root function vs. scan rate, v, to determine the area, Aec, of the
sample using the Randles–Sevcik equation Fig. 5 Example electrochemical area measurements for a NiCo thin
film. The sample had a deposited charge of 1000 mC. The inset shows
CV measurements in 5 mM Ru(NH3)6Cl3 and 1 M KCl at 200, 400, and
800 mV/s. The magnitude of the peak cathodic current, Ip, is fit to a
square root function vs. scan rate, v, to determine the area, Aec, of the
sample using the Randles–Sevcik equation Fig. 5 Example electrochemical area measurements for a NiCo thin
film. The sample had a deposited charge of 1000 mC. The inset shows
CV measurements in 5 mM Ru(NH3)6Cl3 and 1 M KCl at 200, 400, and
800 mV/s. The magnitude of the peak cathodic current, Ip, is fit to a
square root function vs. scan rate, v, to determine the area, Aec, of the
sample using the Randles–Sevcik equation Fig. 4 Example double-layer capacitance measurements for a NiCo
thin film. The sample had a deposited charge of 1000 mC. The inset
shows CV measurements in 1 M KOH at 75, 225, and 350 mV/s. The
slope of the linear fit to the capacitive current, Idl, vs. scan rate, v, is the
measured double-layer capacitance, Cdl, for the sample Fig. 4 Example double-layer capacitance measurements for a NiCo
thin film. The sample had a deposited charge of 1000 mC. The inset
shows CV measurements in 1 M KOH at 75, 225, and 350 mV/s. The
slope of the linear fit to the capacitive current, Idl, vs. scan rate, v, is the
measured double-layer capacitance, Cdl, for the sample Page 6 of 8 Gira et al. Nano Convergence (2016) 3:6 should be. Additional evidence for this interpretation is
seen in Fig. 6b, where the NiCo area measurements are
generally smaller than the Ni area measurements of sam-
ples with similar roughness factors. between 40 and 75 µF/cm2 for all roughness factors. References 1. B.C. Tappan, S.A. Steiner, E.P. Luther, Nanoporous metal foams. Angew. Chem. Int. Ed. 49(27), 4544–4565 (2010) 1. B.C. Tappan, S.A. Steiner, E.P. Luther, Nanoporous metal foams. Angew. Chem. Int. Ed. 49(27), 4544–4565 (2010) 24. N.R. Wozniak, A.A. Frey, L.W. Osterbur, T.S. Boman, J.R. Hampton, An elec-
trochemical cell for the efficient turn around of wafer working electrodes. Rev. Sci. Instrum. 81(3), 034102 (2010) 2. S. Brunauer, P.H. Emmett, E. Teller, Adsorption of gases in multimolecular
layers. J. Am. Chem. Soc. 60(2), 309–319 (1938) 2. S. Brunauer, P.H. Emmett, E. Teller, Adsorption of gases in multimolecular
layers. J. Am. Chem. Soc. 60(2), 309–319 (1938) layers. J. Am. Chem. Soc. 60(2), 25. A.N. Patel, M.G. Collignon, M.A. O-Connell, W.O.Y. Hung, K. McKelvey, J.V. Macpherson, P.R. Unwin, A new view of electrochemistry at highly ori-
ented pyrolytic graphite. J. Am. Chem. Soc. 134(49), 20117–20130 (2012) 3. C. Ji, P.C. Searson, Synthesis and characterization of nanoporous gold
nanowires. J. Phys. Chem. B 107(19), 4494–4499 (2003) 3. C. Ji, P.C. Searson, Synthesis and characterization of nanoporous gold
nanowires. J. Phys. Chem. B 107(19), 4494–4499 (2003) nanowires. J. Phys. Chem. B 107(19), 4494–4499 (2003) 4. J. Snyder, P. Asanithi, A.B. Dalton, J. Erlebacher, Stabilized nanoporous
metals by dealloying ternary alloy precursors. Adv. Mater. 20(24),
4883–4886 (2008) 4. J. Snyder, P. Asanithi, A.B. Dalton, J. Erlebacher, Stabilized na 4. J. Snyder, P. Asanithi, A.B. Dalton, J. Erlebacher, Stabilized nanoporous
metals by dealloying ternary alloy precursors. Adv. Mater. 20(24),
4883–4886 (2008) 26. P. Zoltowski, The capacity of monocrystalline nickel electrode in potas-
sium hydroxide solution at low hydrogen overpotentials. Electrochim. Acta 38(14), 2129–2133 (1993) 5. Y. Liu, S. Bliznakov, N. Dimitrov, Comprehensive study of the application
of a Pb underpotential deposition-assisted method for surface area
measurement of metallic nanoporous materials. J. Phys. Chem. C 113(28),
12362–12372 (2009) 27. R. Šimpraga, L. Bai, B.E. Conway, Real area and electrocatalysis factors in
hydrogen evolution kinetics at electrodeposited Ni-Mo and Ni-Mo-Cd
composites: effect of Cd content and nature of substrate. J. Appl. Electro-
chem. 25(7), 628–641 (1995) 6. A.N. Correia, L.H. Mascaro, S.A.S. Machado, L.A. Avaca, Active surface area
determination of Pd-Si alloys by H-adsorption. Electrochim. Acta 42(3),
493–495 (1997) 28. R. Šimpraga, G. Tremiliosi-Filho, S.Y. Qian, B.E. Conway, In situ determina-
tion of the ‘real area factor’ in H2 evolution electrocatalysis at porous Ni-Fe
composite electrodes. J. Electroanal. Chem. 424(1–2), 141–151 (1997) 7. R. Liu, K. Triantafillou, L. 4 Conclusions For the metal thin films studied here, the results indicate
that in situ electrochemical measurements of double-
layer capacitance are correlated with the roughness fac-
tors extracted from ex situ topographic AFM images. In
addition, these measurements can be adapted to a wide
variety of metal systems by choosing an appropriate
potential range where only capacitive behavior is evident,
thus minimizing any permanent effects on the sample. In contrast, the area measurements using a ruthenium-
based redox probe are both less correlated with rough-
ness measurements and less broadly applicable.hl 1
1.1
1.2
1.3
1.4
RF
0
100
200
300
400
500
Ni
NiCo
(c)
Cdl/Aec / µF cm
-2 (c) The fluctuations present in the capacitance vs. rough-
ness data do place some limitations on the quantitative
nature of the results. Nevertheless, the versatility and sim-
plicity of capacitance measurements make the technique
useful as a semi-quantitative measure of the electrochem-
ically accessible surface area of a sample. Ongoing work
in our lab aims to explore this method further by look-
ing at additional metals and alloys as well as at the more Fig. 6 a Average capacitance, Cdl, b average area, Aec, and c ratio of
capacitance to area, Cdl/Aec, of each sample as a function of the aver-
age roughness factor, RF. Error bars represent the standard error of the
mean for the measurements Page 7 of 8 Gira et al. Nano Convergence (2016) 3:6 12. T. Vidaković, M. Christov, K. Sundmacher, The use of CO stripping for
in situ fuel cell catalyst characterization. Electrochim. Acta 52(18),
5606–5613 (2007) complex morphologies with higher roughness factors,
such as those produced by electrodeposition through self-
assembled colloidal sphere masks. Double-layer capaci-
tance provides a simple, practical, and reliable measure
of the accessible surface area of metal and alloy thin films
which can be used to quantify the intrinsic reactivity of
these systems towards a variety of catalytic reactions. 13. L.-L. Fang, Q. Tao, M.-F. Li, L.-W. Liao, D. Chen, Y.-x. Chen, Determination of
the real surface area of palladium electrode. Chin. J. Chem. Phys. 23(5),
543–548 (2010) 14. R.W. Lindström, K. Kortsdottir, M. Wesselmark, A. Oyarce, C. Lagergren,
G. Lindbergh, Active area determination of porous Pt electrodes used
in polymer electrolyte fuel cells: Temperature and humidity effects. J. Electrochem. Soc. 157(12), 1795–1801 (2010) Electrochem. Soc. 157(12), 1795–1801 (2010) 15. H. Schulenburg, J. Durst, E. Müller, A. Wokaun, G.G. 4 Conclusions Scherer, Real surface
area measurements of Pt3Co/C catalysts. J. Electroanal. Chem. 642(1),
52–60 (2010) Author details
1 17. E. Rouya, S. Cattarin, M.L. Reed, R.G. Kelly, G. Zangari, Electrochemical
characterization of the surface area of nanoporous gold films. J. Electro-
chem. Soc. 159(4), 97–102 (2012) 1 Department of Physics, Hope College, Holland, MI 49423, USA. 2 Depart-
ment of Chemistry, Hope College, Holland, MI 49423, USA. 3 Present Address:
Department of Chemical Engineering and Materials Science, University
of California, Irvine, CA 92697, USA. 18. S.A.S. Machado, L.A. Avaca, The hydrogen evolution reaction on nickel
surfaces stabilized by H-absorption. Electrochim. Acta 39(10), 1385–1391
(1994) Competing interests The authors declare that they have no competing interests. 22. J. van Drunen, B. Kinkead, M.C.P. Wang, E. Sourty, B.D. Gates, G. Jerkiewicz,
Comprehensive structural, surface-chemical and electrochemical char-
acterization of nickel-based metallic foams. ACS Appl. Mater. Interfaces
5(14), 6712–6722 (2013) Received: 19 December 2014 Accepted: 14 July 2015 Received: 19 December 2014 Accepted: 14 July 2015 23. J.M. Gregoire, M. Kostylev, M.E. Tague, P.F. Mutolo, R.B. van Dover, F.J. DiSalvo, H.D. Abruña, High-throughput evaluation of dealloyed Pt-Zn
composition-spread thin film for methanol-oxidation catalysis. J. Electro-
chem. Soc. 156(1), 160–166 (2009) Authors’ contributions MJG and KPT carried out the experiments and contributed to the data analy-
sis. JRH coordinated the study and helped analyze the data. All authors helped
to draft the manuscript. All authors read and approved the final manuscript. 16. A.H. Creus, P. Carro, S. González, R.C. Salvarezza, A.J. Arvia, A new electro-
chemical method for determining the fractal dimension of the surface of
rough metal electrodeposits. J. Electrochem. Soc. 139(4), 1064–1070 (1992) Acknowledgements 19. G. Jarzabek, Z. Borkowska, On the real surface area of smooth solid elec-
trodes. Electrochim. Acta 42(19), 2915–2918 (1997) We thank Professors Beth Anderson and TJ Mullen for helpful discussions. This
material is based upon work supported by the United States National Science
Foundation under Grants No. RUI-DMR-1104725, REU-PHY/DMR-1004811,
MRI-CHE-1126462, MRI-CHE-0959282, and ARI-PHY-0963317 as well as by the
Jacob E. Nyenhuis Faculty Development Fund of Hope College. We thank Professors Beth Anderson and TJ Mullen for helpful discussions. This
material is based upon work supported by the United States National Science
Foundation under Grants No. RUI-DMR-1104725, REU-PHY/DMR-1004811, 20. M. Łukaszewski, A. Czerwiński, Electrochemical preparation and char-
acterization of thin deposits of Pd-noble metal alloys. Thin Solid Films
518(14), 3680–3689 (2010) MRI-CHE-1126462, MRI-CHE-0959282, and ARI-PHY-0963317 as well as by the
Jacob E. Nyenhuis Faculty Development Fund of Hope College. 21. M. Grdeń, M. Alsabet, G. Jerkiewicz, Surface science and electrochemi-
cal analysis of nickel foams. ACS Appl. Mater. Interfaces 4(6), 3012–3021
(2012) References Liu, C. Pu, C. Smith, E.S. Smotkin, A coulometric
normalization procedure for comparing high surface area methanol
anode catalysts by rotating disk electrode voltammetry. J. Electrochem. Soc. 144(6), 148–150 (1997) 29. R.P. Šimpraga, B.E. Conway, The real-area scaling factor in electrocatalysis
and in charge storage by supercapacitors. Electrochim. Acta 43(19–20),
3045–3058 (1998) 30. A. Kellenberger, N. Vaszilcsin, W. Brandl, Roughness factor evaluation of
thermal arc sprayed skeleton nickel electrodes. J. Solid State Electrochem. 11(1), 84–89 (2005) 8. V. Lakshminarayanan, R. Srinivasan, D. Chu, S. Gilman, Area determination
in fractal surfaces of Pt and PtRu electrodes. Surf. Sci. 392(1–3), 44–51
(1997) 31. M.-L. Tremblay, M.H. Martin, C. Lebouin, A. Lasia, D. Guay, Determination
of the real surface area of powdered materials in cavity microelectrodes
by electrochemical impedance spectroscopy. Electrochim. Acta 55(21),
6283–6291 (2010) 9. R.K. Shervedani, A. Lasia, Evaluation of the surface roughness of micropo-
rous Ni-Zn-P electrodes by in situ methods. J. Appl. Electrochem. 29(8),
979–986 (1999) 10. J.P.I. de Souza, T. Iwasita, F.C. Nart, W. Vielstich, Performance evaluation of
porous electrocatalysts via normalization of the active surface. J. Appl. Electrochem. 30(1), 43–48 (2000) 32. I. Herraiz-Cardona, E. Ortega, J.G. Antón, V. Pérez-Herranz, Assessment of
the roughness factor effect and the intrinsic catalytic activity for hydro-
gen evolution reaction on Ni-based electrodeposits. Int. J. Hydrogen
Energy 36(16), 9428–9438 (2011) 11. M. Søgaard, M. Odgaard, E.M. Skou, An improved method for the determi-
nation of the electrochemical active area of porous composite platinum
electrodes. Solid State Ionics 145(1–4), 31–35 (2001) 33. G. Hasegawa, K. Kanamori, K. Nakanishi, T. Abe, New insights into the
relationship between micropore properties, ionic sizes, and electric Page 8 of 8 Gira et al. Nano Convergence (2016) 3:6 double-layer capacitance in monolithic carbon electrodes. J. Phys. Chem. C 116(50), 26197–26203 (2012) 43. A.J. Bard, L.R. Faulkner, Electrochemical Methods: Fundamentals and Appli-
cations, 2nd edn. (Wiley-Interscience, New York; London, 2001) double-layer capacitance in monolithic carbon electrodes. J. Phys. Chem. C 116(50), 26197–26203 (2012) 34. E. Lust, A. Jänes, V. Sammelselg, P. Miidla, K. Lust, Surface roughness of
bismuth, antimony and cadmium electrodes. Electrochim. Acta 44(2–3),
373–383 (1998) 44. A.J. Bard, J.A. Crayston, G.P. Kittlesen, T. Varco Shea, M.S. Wrighton, Digital
simulation of the measured electrochemical response of reversible redox
couples at microelectrode arrays: consequences arising from closely
spaced ultramicroelectrodes. Anal. Chem. 58(11), 2321–2331 (1986) 35. H.M. Saffarian, R. Srinivasan, D. Chu, S. References Gilman, Area determination in frac-
tal surfaces of Pt and Pt-Ru catalysts for methanol oxidation. Electrochim. Acta 44(8–9), 1447–1454 (1998) 45. S. Licht, V. Cammarata, M.S. Wrighton, Direct measurements of the physi-
cal diffusion of redox active species: microelectrochemical experiments
and their simulation. J. Phys. Chem. 94(15), 6133–6140 (1990) 36. A. Brenner, Electrodeposition of Alloys: Principles and Practice, vol. 1 (Aca-
demic Press, New York, 1963) y
46. S.O. Engblom, J.C. Myland, K.B. Oldham, A.L. Taylor, Large amplitude AC
voltammetry—a comparison between theory and experiment. Electroa-
nalysis 13(8–9), 626–630 (2001) 37. T. Akiyama, H. Fukushima, Recent study on the mechanism of the electro-
deposition of iron-group metal-alloys. ISIJ Int. 32(7), 787–798 (1992) 47. S. Trasatti, O.A. Petrii, Real surface area measurements in electrochemistry. Pure Appl. Chem. 63(5), 711–734 (1991) 38. K.Y. Sasaki, J.B. Talbot, Electrodeposition of iron-group metals and binary
alloys from sulfate baths I. Experimental study. J. Electrochem. Soc. 145(3), 981–990 (1998) 48. L. Chen, A. Lasia, Study of the kinetics of hydrogen evolution reaction
on nickel–zinc alloy electrodes. J. Electrochem. Soc. 138(11), 3321–3328
(1991) 39. N. Zech, E.J. Podlaha, D. Landolt, Anomalous codeposition of iron group
metals I. Experimental results. J. Electrochem. Soc. 146(8), 2886–2891
(1999) 49. L. Chen, A. Lasia, Study of the kinetics of hydrogen evolution reaction on
nickel–zinc powder electrodes. J. Electrochem. Soc. 139(11), 3214–3219
(1992) 40. C.-C. Hu, C.-Y. Cheng, Ideally pseudocapacitive behavior of amorphous
hydrous cobalt–nickel oxide prepared by anodic deposition. Electro-
chem. Solid State Lett. 5(3), 43–46 (2002) 50. R.K. Shervedani, A. Lasia, Studies of the hydrogen evolution reaction on
Ni-P electrodes. J. Electrochem. Soc. 144(2), 511–519 (1997) 41. J.-K. Chang, C.-M. Wu, I.-W. Sun, Nano-architectured Co(OH)2 electrodes
constructed using an easily-manipulated electrochemical protocol for
high-performance energy storage applications. J. Mater. Chem. 20(18),
3729–3735 (2010) 51. R.K. Shervedani, A. Lasia, Kinetics of hydrogen evolution reaction on
nickel–zinc–phosphorous electrodes. J. Electrochem. Soc. 144(8),
2652–2657 (1997) 52. R.K. Shervedani, A. Lasia, Study of the hydrogen evolution reaction on
Ni-Mo-P electrodes in alkaline solutions. J. Electrochem. Soc. 145(7),
2219–2225 (1998) 42. D.-S. Kong, J.-M. Wang, H.-B. Shao, J.-Q. Zhang, C.-n. Cao, Electrochemical
fabrication of a porous nanostructured nickel hydroxide film electrode
with superior pseudocapacitive performance. J. Alloys Compd. 509(18),
5611–5616 (2011)
|
https://openalex.org/W4256188974
|
https://acp.copernicus.org/articles/18/6483/2018/acp-18-6483-2018.pdf
|
English
| null |
High-resolution inversion of methane emissions in the Southeast US using SEAC&lt;sup&gt;4&lt;/sup&gt;RS aircraft observations of atmospheric methane: anthropogenic and wetland sources
| null | 2,017
|
cc-by
| 8,268
|
Correspondence: Jian-Xiong Sheng (jsheng@seas.harvard.edu) Correspondence: Jian-Xiong Sheng (jsheng@seas.harvard.edu) Received: 8 December 2017 – Discussion started: 18 December 2017
Revised: 29 March 2018 – Accepted: 6 April 2018 – Published: 7 May 2018 Received: 8 December 2017 – Discussion started: 18 December 2017 Revised: 29 March 2018 – Accepted: 6 April 2018 – Published: 7 May 2018 Abstract. We use observations of boundary layer methane
from the SEAC4RS aircraft campaign over the Southeast
US in August–September 2013 to estimate methane emis-
sions in that region through an inverse analysis with up to
0.25◦× 0.3125◦(25 × 25 km2) resolution and with full er-
ror characterization. The Southeast US is a major source
region for methane including large contributions from oil
and gas production and wetlands. Our inversion uses state-
of-the-art emission inventories as prior estimates, includ-
ing a gridded version of the anthropogenic EPA Greenhouse
Gas Inventory and the mean of the WetCHARTs ensemble
for wetlands. Inversion results are independently verified by
comparison with surface (NOAA/ESRL) and column (TC-
CON) methane observations. Our posterior estimates for the
Southeast US are 12.8±0.9 Tga−1 for anthropogenic sources
(no significant change from the gridded EPA inventory) and
9.4 ± 0.8 Tga−1 for wetlands (27 % decrease from the mean
in the WetCHARTs ensemble). The largest source of error in
the WetCHARTs wetlands ensemble is the land cover map
specification of wetland areal extent. Our results support the
accuracy of the EPA anthropogenic inventory on a regional
scale but there are significant local discrepancies for oil and
gas production fields, suggesting that emission factors are
more variable than assumed in the EPA inventory. Atmos. Chem. Phys., 18, 6483–6491, 2018
https://doi.org/10.5194/acp-18-6483-2018
© Author(s) 2018. This work is distributed under
the Creative Commons Attribution 4.0 License. Atmos. Chem. Phys., 18, 6483–6491, 2018
https://doi.org/10.5194/acp-18-6483-2018
© Author(s) 2018. This work is distributed under
the Creative Commons Attribution 4.0 License. 1
Introduction Methane is an important greenhouse gas (Myhre et al., 2013). Individual countries must report their national emissions
to the United Nations Framework Convention on Climate
Change (UNFCCC; United Nation, 1992). Observations of
atmospheric methane reviewed by Brandt et al. (2014) have
implied that the US national inventory reported by the Envi-
ronmental Protection Agency (EPA) may be greatly underes-
timated. Here we use aircraft observations from the NASA
SEAC4RS aircraft campaign over the Southeast US (Toon
et al., 2016), together with a newly gridded version of the
EPA inventory (Maasakkers et al., 2016), in a fine-resolution
inversion with detailed error characterization to better quan-
tify methane emissions over this major source region. The EPA (2016) reports a national anthropogenic emission
total of 29.2 TgCH4 a−1 for 2014, with no significant trend
over the past decade and less than ±3 % interannual vari-
ability. Major contributors are livestock (32 %), the oil and
gas industry (32 %), waste (22 %), and coal mining (8 %). The EPA (2016) inventory is consistent with Lyon et al. (2015) for oil and gas systems and Wolf et al. (2017) for
livestock, and 8 % higher than the previous versions (EPA,
2013, 2014), largely due to updated oil and gas emissions. There is also a highly uncertain natural source from wet-
lands, estimated at 4.5–14 Tga−1 for the contiguous US in
the WETCHIMP compilation of inventories (Melton et al., Jian-Xiong Sheng1, Daniel J. Jacob1, Alexander J. Turner1,a, Joannes D. Maasakkers1, Melissa P. Sulprizio1,
A. Anthony Bloom2, Arlyn E. Andrews3, and Debra Wunch4 1School of Engineering and Applied Sciences, Harvard University, Cambridge, MA, USA
2Jet Propulsion Laboratory, California Institute of Technology, Pasadena, CA, USA
3NOAA Earth System Research Laboratory, Boulder, Colorado, USA
4Department of Physics, University of Toronto, Toronto, Canada
anow at: Department of Earth and Planetary Sciences, University of California at Berkeley, CA J.-X. Sheng et al.: Methane emissions in the Southeast US using SEAC4RS aircraft observations Unlike
adjoint-based inversions, analytical solution provides direct
error characterization of the optimal estimate ˆx through its
error covariance matrix ˆS. The related averaging kernel ma-
trix A describes the sensitivity of the optimal estimate ˆx to
the true emissions x. The trace of A quantifies the degrees
of freedom for signal (DOFS), i.e., the number of pieces
of information in the observing system for constraining the
methane emissions (DOFS ≤n). where In is the identity matrix with n being the dimension of
the state vector x. Inversions of atmospheric methane obser-
vations may solve ∇xJ(x) = 0 either analytically, or numer-
ically using an adjoint method (Jacob et al., 2016). Unlike
adjoint-based inversions, analytical solution provides direct
error characterization of the optimal estimate ˆx through its
error covariance matrix ˆS. The related averaging kernel ma-
trix A describes the sensitivity of the optimal estimate ˆx to
the true emissions x. The trace of A quantifies the degrees
of freedom for signal (DOFS), i.e., the number of pieces
of information in the observing system for constraining the
methane emissions (DOFS ≤n). The SEAC4RS aircraft campaign conducted in August–
September 2013 offers an opportunity for better estimating
methane emissions in the Southeast US, a region that ac-
counts for about half of total anthropogenic methane emis-
sions in the US, according to the gridded EPA inventory
(Maasakkers et al., 2016), and also has extensive wetlands. The aircraft flights provided extensive boundary layer mea-
surements of methane across the region. We conduct an
inverse analysis of the SEAC4RS data with the GEOS-
Chem chemical transport model (CTM) at 0.25◦× 0.3125◦
resolution, using state-of-the-art prior estimates from the
gridded EPA inventory of Maasakkers et al. (2016) and
the WetCHARTs extended ensemble wetlands inventory of
Bloom et al. (2017). This allows us to evaluate the EPA in-
ventory with better accuracy and resolution than has been
done before and also to gain better understanding of US wet-
land emissions. The Jacobian matrix K for the inversion is constructed
with the GEOS-Chem CTM (http://www.geos-chem.org, last
access: 1 November 2017), which relates methane emissions
to atmospheric concentrations through simulation of atmo-
spheric transport. We use a nested version of GEOS-Chem
as described by Kim et al. J.-X. Sheng et al.: Methane emissions in the Southeast US using SEAC4RS aircraft observations (2015) with 0.25◦× 0.3125◦hor-
izontal resolution over the North America window and ad-
jacent oceans (9.75–60◦N, 130–60◦W), driven by GEOS-
FP assimilated meteorological data from the NASA Global
Modeling and Assimilation Office (GMAO). The same ver-
sion of the GEOS-Chem has been applied to simulation
of other chemical observations from the SEAC4RS cam-
paign (Kim et al., 2015; Fisher et al., 2016; Marais et al.,
2016; Travis et al., 2016; Zhu et al., 2016; Yu et al., 2016;
Chan Miller et al., 2017). The boundary conditions for the
nested-grid simulation are from a 4◦× 5◦global simula-
tion by Turner et al. (2015) using methane emissions opti-
mized with three years of GOSAT satellite data. The GOSAT-
optimized emissions have been independently evaluated with
atmospheric methane observations from the NOAA surface
network (Turner et al., 2015). The GEOS-Chem uses a 3-
D archive of monthly average OH concentrations from Park
et al. (2004) to compute the methane sink, with a lifetime
of 8.9 years in the troposphere consistent with observational
constraints (Prather et al., 2012; Turner et al., 2017). The sink
is irrelevant for our North American simulation since ventila-
tion of the domain is much faster (Wecht et al., 2014). Since
we treat OH concentrations as decoupled from methane in
the inversion, the relationship between emissions and con-
centrations is strictly linear, so that K fully describes the
GEOS-Chem model for the purpose of the inversion. Published by Copernicus Publications on behalf of the European Geosciences Union. Published by Copernicus Publications on behalf of the European Geosciences Union. 6484 J.-X. Sheng et al.: Methane emissions in the Southeast US using SEAC4RS aircraft observations 2013). Inverse analyses of atmospheric methane observa-
tions have suggested that the EPA bottom-up inventory (EPA,
2013, 2014) is too low by about 30 % (Miller et al., 2013;
Turner et al., 2015; Alexe et al., 2015). However, Turner et al. (2015) and Alexe et al. (2015) relied on prior estimates from
the global EDGAR v4.2 inventory (European Commission,
2011) that have large errors in source patterns particularly
for oil and gas systems (Maasakkers et al., 2016; Sheng et al.,
2017). Errors in source patterns used as prior estimates can
greatly bias inversion results (Jacob et al., 2016), though this
depends on the constraint from observations and on the un-
certainty assigned to the prior estimates. Miller et al. (2013)
used a geostatistical inversion that did not rely on any prior
estimates, but had little constraints in the Southeast US. 2013). Inverse analyses of atmospheric methane observa-
tions have suggested that the EPA bottom-up inventory (EPA,
2013, 2014) is too low by about 30 % (Miller et al., 2013;
Turner et al., 2015; Alexe et al., 2015). However, Turner et al. (2015) and Alexe et al. (2015) relied on prior estimates from
the global EDGAR v4.2 inventory (European Commission,
2011) that have large errors in source patterns particularly
for oil and gas systems (Maasakkers et al., 2016; Sheng et al.,
2017). Errors in source patterns used as prior estimates can
greatly bias inversion results (Jacob et al., 2016), though this
depends on the constraint from observations and on the un-
certainty assigned to the prior estimates. Miller et al. (2013)
used a geostatistical inversion that did not rely on any prior
estimates, but had little constraints in the Southeast US. 4 Analytical solution of ∇xJ(x) = 0 yields the optimal esti-
mate ˆx, the posterior error covariance matrix ˆS, and the asso-
ciated averaging kernel matrix A as follows (Rodgers, 2000;
Brasseur and Jacob, 2017): ˆx = xA + ST
A K(KST
A K + SO)−1(y −KxA),
(2)
ˆS−1 = KT SO−1K + S−1
A ,
(3)
A = In −ˆSS−1
A ,
(4) (2) (2)
(3)
(4) (4) where In is the identity matrix with n being the dimension of
the state vector x. Inversions of atmospheric methane obser-
vations may solve ∇xJ(x) = 0 either analytically, or numer-
ically using an adjoint method (Jacob et al., 2016). 2
Methods We derive an optimized estimate of spatially resolved
methane emissions in the Southeast US (domain of Fig. 1)
through Bayesian inverse analysis of atmospheric methane
observations from the SEAC4RS aircraft campaign. Let the
vector x represent a gridded ensemble of methane emissions
in the region (state vector for the inversion).The inversion
minimizes the cost function J(x) by solving ∇xJ(x) = 0 as
follows: J(x) = 1
2(x −xA)T S−1
A (x −xA)
(1)
+ 1
2(y −Kx)T S−1
O (y −Kx). (1) Here the methane observations are assembled as a vector y,
xA is the prior emission estimate, K is the Jacobian matrix
describing the sensitivity of concentrations to emissions, and
SA and SO are the prior and observational error covariance
matrices, respectively. The observational error includes con-
tributions from both the instrument error and the model trans-
port error. Here the methane observations are assembled as a vector y,
xA is the prior emission estimate, K is the Jacobian matrix
describing the sensitivity of concentrations to emissions, and
SA and SO are the prior and observational error covariance
matrices, respectively. The observational error includes con-
tributions from both the instrument error and the model trans-
port error. The prior emission estimates for the inversion are taken
from the 0.1◦× 0.1◦gridded version of the EPA an-
thropogenic greenhouse gas emission inventory for 2012 www.atmos-chem-phys.net/18/6483/2018/ Atmos. Chem. Phys., 18, 6483–6491, 2018 J.-X. Sheng et al.: Methane emissions in the Southeast US using SEAC4RS aircraft observations J.-X. Sheng et al.: Methane emissions in the Southeast US using SEAC4RS aircraft observations 6485 Figure 1. Methane emissions in the Southeast US in August–September 2013. Panel (a) shows the prior anthropogenic and wetland methane
emissions, and panel (b) shows the inversion results including posterior emissions, scaling factors (posterior / prior emission ratios), and the
diagonal elements of the averaging kernel matrix representing the sensitivity of the inversion results to the observations. The sum of these
diagonal elements over the domain (trace of the averaging kernel matrix) quantifies the degrees of freedom for signal (DOFS) of the inversion. Numbers inset in the emission panels are the regional totals expressed as annual means for clarity (i.e., assuming that August–September
emission rates hold for the rest of the year). Values in parentheses are the totals for the region with averaging kernel sensitivities larger than
0.05 (stippled areas in lower left panel). Figure 1. 2
Methods Methane emissions in the Southeast US in August–September 2013. Panel (a) shows the prior anthropogenic and wetland methane
emissions, and panel (b) shows the inversion results including posterior emissions, scaling factors (posterior / prior emission ratios), and the
diagonal elements of the averaging kernel matrix representing the sensitivity of the inversion results to the observations. The sum of these
diagonal elements over the domain (trace of the averaging kernel matrix) quantifies the degrees of freedom for signal (DOFS) of the inversion. Numbers inset in the emission panels are the regional totals expressed as annual means for clarity (i.e., assuming that August–September
emission rates hold for the rest of the year). Values in parentheses are the totals for the region with averaging kernel sensitivities larger than
0.05 (stippled areas in lower left panel). altitude averaged over the 0.25◦×0.3125◦GEOS-Chem grid
for individual flights. This represents 652 individual methane
observations. (Maasakkers et al., 2016) and the mean wetland emissions
from the 0.5◦×0.5◦monthly WetCHARTs extended ensem-
ble for 2013 (Bloom et al., 2017). Figure 1 (top panels) shows
the distribution of these prior methane emissions over the in-
version domain for August–September 2013. Emissions to-
tal 13.3 Tga−1 for anthropogenic sources and 13.0 Tga−1
for wetlands over these two months (expressed on an an-
nual basis). Anthropogenic emissions in the EPA inventory
have little seasonal or interannual variability (Turner et al.,
2015; Maasakkers et al., 2016), while wetland emissions
have a large seasonal variation. (Maasakkers et al., 2016) and the mean wetland emissions
from the 0.5◦×0.5◦monthly WetCHARTs extended ensem-
ble for 2013 (Bloom et al., 2017). Figure 1 (top panels) shows
the distribution of these prior methane emissions over the in-
version domain for August–September 2013. Emissions to-
tal 13.3 Tga−1 for anthropogenic sources and 13.0 Tga−1
for wetlands over these two months (expressed on an an-
nual basis). Anthropogenic emissions in the EPA inventory
have little seasonal or interannual variability (Turner et al.,
2015; Maasakkers et al., 2016), while wetland emissions
have a large seasonal variation. We use the residual error method (Heald et al., 2004) to
estimate the diagonal elements of the observational error
covariance matrix SO. The method assumes that the mean
model bias between the observations and the model is due
to error in prior emissions to be corrected by the inversion. The residuals are the differences between observed and mod-
eled values after removing the mean model bias. 2
Methods The resid-
ual error standard deviation (RSD) is taken to represent the
observational error including contributions from the instru-
ment and the transport model. Figure 3 shows the vertical
profile of the RSD for the ensemble of the SEAC4RS data
over the Southeast US. The RSD is about 60 ppb below 2 km
and 20 ppb in the free troposphere above. Subsetting the data
by latitudinal bands gives similar results. Thus we use 60 ppb
for the standard deviation of the observational error (diagonal
elements in SO). The instrument precision is better than 2 ppb
(Simpson et al., 2002), therefore most of that observational
error is from the transport model. We take SO to be diagonal
since error correlations between boundary layer observations
on the GEOS-Chem grid are not significant (Wecht et al.,
2014). g
The SEAC4RS DC-8 aircraft conducted 21 flights over
the Southeast between 6 August and 21 September 2013. Methane was measured by gas chromatography from whole
air flask samples and calibrated to the NOAA standard. Fig-
ure 2 (left panel) shows the SEAC4RS flight tracks and the
spatial distribution of the methane flask measurements below
2 km altitude. The mean observed vertical profile is shown in
the right panel of Fig. 2, and compared to the GEOS-Chem
profile using the prior emissions. The model is unbiased in
the free troposphere above 2 km, implying a successful rep-
resentation of background methane by the boundary con-
ditions. Model overestimation in the boundary layer below
2 km suggests that the prior US emissions are too high. For
the inversion, we use the SEAC4RS observations below 2 km Atmos. Chem. Phys., 18, 6483–6491, 2018 Atmos. Chem. Phys., 18, 6483–6491, 2018 J.-X. Sheng et al.: Methane emissions in the Southeast US using SEAC4RS aircraft observations Figure 2. Boundary layer methane concentrations over the Southeast US measured during the SEAC4RS aircraft campaign (6 August–
21 September 2013). Panel (a) shows the flight tracks in grey and the methane measurements at 0–2 km altitude. The three NOAA/ESRL
sites at SGP (Southern Great Plains, Oklahoma; 36.6◦N, 97.5◦W), WKT (Moody, Texas; 31.3◦N, 97.3◦W), and SCT (Beech Island, South
Carolina; 33.4◦, 81.8◦W) are indicated. SGP is co-located with the TCCON site at Lamont, Oklahoma. Panel (b) shows the mean methane
vertical profiles over the Southeast US domain measured from the aircraft and simulated by GEOS-Chem using the prior and posterior
emissions. Figure 2. Boundary layer methane concentrations over the Southeast US measured during the SEAC4RS aircraft campaign (6 August–
21 September 2013). Panel (a) shows the flight tracks in grey and the methane measurements at 0–2 km altitude. The three NOAA/ESRL
sites at SGP (Southern Great Plains, Oklahoma; 36.6◦N, 97.5◦W), WKT (Moody, Texas; 31.3◦N, 97.3◦W), and SCT (Beech Island, South
Carolina; 33.4◦, 81.8◦W) are indicated. SGP is co-located with the TCCON site at Lamont, Oklahoma. Panel (b) shows the mean methane
vertical profiles over the Southeast US domain measured from the aircraft and simulated by GEOS-Chem using the prior and posterior
emissions. Figure 3. Residual standard deviation (RSD) of the difference be-
tween SEAC4RS methane observations and the GEOS-Chem model
with prior emissions, for 1 km altitude bins. The RSD defines the
observational error standard deviations for the inversion as de-
scribed in the text. Values are shown for two latitudinal bands. dominated by wetlands have resolution of 100–200 km in the
GMM because they are broadly distributed. The anthropogenic inventory of Maasakkers et al. (2016)
and the wetlands inventory of Bloom et al. (2017) both in-
clude gridded error estimates that serve as the diagonal ele-
ments of the prior error covariance matrix SA. Maasakkers
et al. (2016) found no significant spatial error correlation at
0.1◦× 0.1◦resolution in their inventory while a variogram
analysis across the elements of the WetCHARTS ensemble
indicates a spatial error correlation length scale of 130 km. Here we ignore this correlation and take SA to be diagonal. www.atmos-chem-phys.net/18/6483/2018/ J.-X. Sheng et al.: Methane emissions in the Southeast US using SEAC4RS aircraft observations J.-X. Sheng et al.: Methane emissions in the Southeast US using SEAC4RS aircraft observations 6486 J.-X. Sheng et al.: Methane emissions in the Southeast US using SEAC4RS aircraft observations Error bars (one standard deviation) on sectoral emis-
sions are from the prior and posterior error variances of our inver-
sion. Methane emissions in the subdomain with averaging kernel
sensitivities larger than 0.05 (Fig. 1) are also indicated. 35 % lower than the prior estimate. The posterior errors are
18–30 % over these regions. The scaling factors show large
downward corrections of prior emissions in Louisiana and
Mississippi, and along the Gulf Coast, where wetlands are
the dominant sources. There are also downward corrections
in southern West Virginia, where coal mines are dominant,
and in the Haynesville Shale gas production region of north-
ern Louisiana and southern Arkansas. On the other hand,
there are significant upward corrections for the coal mines of
southern Illinois and for the Floyd Shale in central Alabama. We can attribute the 0.25◦× 0.3125◦scaling factors from
the inversion to specific methane source sectors by using the
sector-resolved spatial patterns in the prior emission invento-
ries, as described by Turner et al. (2015) but here with the im-
proved anthropogenic source patterns from Maasakkers et al. (2016) and wetland emissions from (Bloom et al., 2017). An-
thropogenic and wetland sources are well separated spatially
in these inventories. Figure 5 compares our results with the
prior emission totals for the different sectors in the Southeast
US. We find a significant 27 % reduction in regional wet-
land emissions relative to the prior estimate (mean of the
WetCHARTs extended ensemble). For the subdomain with
averaging kernel sensitivity larger than 0.05 that reduction
is 42 %. By contrast, we find no significant regional bias in
the EPA anthropogenic inventory for any of the major source 35 % lower than the prior estimate. The posterior errors are
18–30 % over these regions. The scaling factors show large
downward corrections of prior emissions in Louisiana and
Mississippi, and along the Gulf Coast, where wetlands are
the dominant sources. There are also downward corrections
in southern West Virginia, where coal mines are dominant,
and in the Haynesville Shale gas production region of north-
ern Louisiana and southern Arkansas. On the other hand,
there are significant upward corrections for the coal mines of
southern Illinois and for the Floyd Shale in central Alabama. Figure 4. Evaluation of the SEAC4RS inversion of methane emis-
sions in the Southeast US for the 6 August–21 September 2013
period. J.-X. Sheng et al.: Methane emissions in the Southeast US using SEAC4RS aircraft observations Panels (a, b) compare GEOS-Chem methane concentra-
tions with the SEAC4RS observations, using prior emissions (left)
and posterior emissions (right). Panels (c, d) compare GEOS-
Chem methane concentrations with independent observations from
the three NOAA/ESRL surface sites in the inversion domain (see
Fig. 2 and caption). Panels (e, f) compare GEOS-Chem methane
columns with TCCON hourly column observations at Lamont, Ok-
lahoma (Wennberg et al., 2017), after correction for stratospheric
bias in the model (Turner et al., 2015). The 1 : 1 lines (dashed) and
the reduced-major-axis(black solid line) linear regressions are also
shown, along with the coefficients of determination (R2) and the
slopes (±1σ) derived from the bootstrap method. y
We can attribute the 0.25◦× 0.3125◦scaling factors from
the inversion to specific methane source sectors by using the
sector-resolved spatial patterns in the prior emission invento-
ries, as described by Turner et al. (2015) but here with the im-
proved anthropogenic source patterns from Maasakkers et al. (2016) and wetland emissions from (Bloom et al., 2017). An-
thropogenic and wetland sources are well separated spatially
in these inventories. Figure 5 compares our results with the
prior emission totals for the different sectors in the Southeast
US. We find a significant 27 % reduction in regional wet-
land emissions relative to the prior estimate (mean of the
WetCHARTs extended ensemble). For the subdomain with
averaging kernel sensitivity larger than 0.05 that reduction
is 42 %. By contrast, we find no significant regional bias in
the EPA anthropogenic inventory for any of the major source
sectors for the period of August–September 2013. However,
there are large local biases that tend to cancel each other on
a regional scale (e.g., Haynesville Shale vs. Floyd Shale for
natural gas). This suggests that methane emission factors for
the oil and gas sector are more variable than assumed in the
EPA (2016) inventory. of independent data but the 2 month period is too sparse for
useful evaluation (Wecht et al., 2014). Total posterior emissions over the SEAC4RS domain are
15 % (4 Tga−1) lower than the prior estimate (Fig. 1). The
inversion is able to constrain about 10 pieces of informa-
tion in the spatial distribution of methane emissions as mea-
sured by the DOFS. It is strongly sensitive to the Gulf Coast
and to large anthropogenic source areas such as the Floyd
Shale in central Alabama. J.-X. Sheng et al.: Methane emissions in the Southeast US using SEAC4RS aircraft observations J.-X. Sheng et al.: Methane emissions in the Southeast US using SEAC4RS aircraft observations 6487 Figure 4. Evaluation of the SEAC4RS inversion of methane emis-
sions in the Southeast US for the 6 August–21 September 2013
period. Panels (a, b) compare GEOS-Chem methane concentra-
tions with the SEAC4RS observations, using prior emissions (left)
and posterior emissions (right). Panels (c, d) compare GEOS-
Chem methane concentrations with independent observations from
the three NOAA/ESRL surface sites in the inversion domain (see
Fig. 2 and caption). Panels (e, f) compare GEOS-Chem methane
columns with TCCON hourly column observations at Lamont, Ok-
lahoma (Wennberg et al., 2017), after correction for stratospheric
bias in the model (Turner et al., 2015). The 1 : 1 lines (dashed) and
the reduced-major-axis(black solid line) linear regressions are also
shown, along with the coefficients of determination (R2) and the
slopes (±1σ) derived from the bootstrap method. Figure 5. Prior and posterior methane emissions in the Southeast
US (domain of Fig. 1) for August–September 2013. The prior an-
thropogenic emissions are from the EPA national inventory for 2012
(EPA, 2016; Maasakkers et al., 2016) and the prior wetland emis-
sions are the means of the WetCHARTs extended ensemble (Bloom
et al., 2017). Error bars (one standard deviation) on sectoral emis-
sions are from the prior and posterior error variances of our inver-
sion. Methane emissions in the subdomain with averaging kernel
sensitivities larger than 0.05 (Fig. 1) are also indicated. Figure 5. Prior and posterior methane emissions in the Southeast
US (domain of Fig. 1) for August–September 2013. The prior an-
thropogenic emissions are from the EPA national inventory for 2012
(EPA, 2016; Maasakkers et al., 2016) and the prior wetland emis-
sions are the means of the WetCHARTs extended ensemble (Bloom
et al., 2017). Error bars (one standard deviation) on sectoral emis-
sions are from the prior and posterior error variances of our inver-
sion. Methane emissions in the subdomain with averaging kernel
sensitivities larger than 0.05 (Fig. 1) are also indicated. Figure 5. Prior and posterior methane emissions in the Southeast
US (domain of Fig. 1) for August–September 2013. The prior an-
thropogenic emissions are from the EPA national inventory for 2012
(EPA, 2016; Maasakkers et al., 2016) and the prior wetland emis-
sions are the means of the WetCHARTs extended ensemble (Bloom
et al., 2017). J.-X. Sheng et al.: Methane emissions in the Southeast US using SEAC4RS aircraft observations For the regions with averaging
kernel sensitivity larger than 0.05, posterior emissions are Atmos. Chem. Phys., 18, 6483–6491, 2018 Atmos. Chem. Phys., 18, 6483–6491, 2018 3
Results and discussion Figure 3. Residual standard deviation (RSD) of the difference be-
tween SEAC4RS methane observations and the GEOS-Chem model
with prior emissions, for 1 km altitude bins. The RSD defines the
observational error standard deviations for the inversion as de-
scribed in the text. Values are shown for two latitudinal bands. Figure 1b shows the results of the inversion including the
optimized posterior emissions, the corrections to the prior
emissions, and the DOFS as measured by the diagonal ele-
ments of the averaging kernel matrix. Figure 4 (top panels)
compare the observed boundary layer methane concentra-
tions to the values simulated by GEOS-Chem with prior and
posterior emissions (Fig. 1). The simulation with prior emis-
sions has a positive bias that is effectively corrected when
using posterior emissions. The coefficient of determination
(R2) between model and observations increases from 0.30 to
0.50 when using posterior emissions. Figure 4 also evaluates
the SEAC4RS inversion results with independent surface air
observations from the three NOAA/ESRL surface network
sites in the region (Andrews et al., 2014) and with methane
column observations from the TCCON site in Lamont, Okla-
homa (Wunch et al., 2011; Wennberg et al., 2017). The pos-
terior emissions improve the simulation of these independent
data sets. GOSAT satellite observations are another source The inversion can, in principle, optimize emissions at the
0.25◦× 0.3125◦grid resolution of the GEOS-Chem model,
representing 3004 grid cells over the inversion domain. But
the aircraft observations do not have sufficient information
to constrain emissions at that resolution. In order to reduce
the dimensionality of the state vector, we project the 3004
grid cells onto 216 elements of a Gaussian mixture model
(GMM) with radial basis functions based on spatial proxim-
ity and source type patterns (Turner and Jacob, 2015). The
use of the GMM allows us to retain high resolution of up
to 25 km for major localized sources while degrading reso-
lution in areas of weak or broadly distributed sources. Areas www.atmos-chem-phys.net/18/6483/2018/ Atmos. Chem. Phys., 18, 6483–6491, 2018 J.-X. Sheng et al.: Methane emissions in the Southeast US using SEAC4RS aircraft observations J.-X. Sheng et al.: Methane emissions in the Southeast US using SEAC4RS aircraft observations J.-X. Sheng et al.: Methane emissions in the Southeast US using SEAC4RS aircraft observations 6488 The WetCHARTs extended ensemble includes 18 wetland
methane emission models intended to encompass the un-
certainties in estimating wetland emissions (Bloom et al.,
2017). The different models (ensemble members) use dif-
ferent datasets for wetland extent fraction A [m2 wet-
lands m−2 surface area], heterotrophic respiration rate R
[mgC day−1 m−2 of wetland area], temperature-dependent
factor qT /10
10
of C respired as CH4 [mgCH4 mg−1 C] where T
is the surface skin temperature, and global scaling factors s. The wetland methane emission flux E [mgCH4 m−2 day−1]
at a time t and location x for each of these members is given
by the following equation: E(t,x) = sA(t,x)R(t,x)qT (t,x)/10
10
. (5) (5) Figure 6. Range of wetland emission estimates in the Southeast
US (domain of Fig. 1) for August–September 2013. The figure
shows the spread of the WetCHARTs extended ensemble and com-
pares with the posterior emission estimate from our inversion in
terms of emission total and root-mean-square error (RMSE) on the
0.25◦× 0.3125◦spatial grid. The RMSE of the posterior emission
estimate with itself is zero by definition. WetCHARTs ensemble
members use wetland areal extent data from either the GLOB-
COVER (Bontemps et al., 2011) or GLWD (Lehner and Dölla,
2004) databases, as well as different estimates of temperature sen-
sitivity q10 and global scaling factors s (see Eq. 5 and text). The
posterior wetland emission estimate from our inversion is shown
as black solid diamond with its associated error standard deviation. The mean of the WetCHARTs ensemble used as prior for our inver-
sion is shown as blue solid circle with its associated error standard
deviation. The 18-member ensemble consists of three temperature de-
pendence factors (q10 = 1,2,3), three global scale factors
(s = 125,166,208), and two wetland extent maps (A) from
the Global Lakes and Wetlands Database (GLWD; Lehner
and Dölla, 2004) and GLOBCOVER (Bontemps et al.,
2011). The heterotrophic respiration rate (R) is the me-
dian output from the carbon data model framework (CAR-
DAMOM; Bloom et al., 2016), and is not varied across that
ensemble. Figure 6 shows the Southeast US wetland emissions for
each WetCHARTs member, along with the root-mean-square
error (RMSE) of its spatial distribution relative to our opti-
mized posterior estimate on the 0.25◦× 0.3125◦grid. J.-X. Sheng et al.: Methane emissions in the Southeast US using SEAC4RS aircraft observations Con-
sistency in spatial distribution with our optimized estimate
is indicated by a low RMSE. We find that the specification
of wetland extent is the most systematic source of error in
wetland emission estimates; all GLOBCOVER-based mod-
els underestimate wetland emissions, while all GLWD-based
models overestimate emissions. Estimates using q10 = 1 (no
temperature dependence in the CH4 : C respiration ratio) ex-
hibit the lowest RMSE values. The WetCHARTs ensemble
mean used as prior for our inversion has the lowest RMSE,
although this may reflect its influence on the posterior solu-
tion. Our inversion results suggest that the EPA emission inven-
tory has no significant bias on the regional scale for the major
anthropogenic source sectors (livestock, oil and gas, waste,
coal), while the mean of the WetCHARTs wetland ensem-
ble needs to be reduced by 27 % over the inversion domain. These results are supported by independent methane obser-
vations from the NOAA/ESRL surface network and from
the TCCON site in Lamont, Oklahoma. The specification of
wetland areal extent is the dominant source of error in the
WetCHARTs ensemble. Results also indicate that a low tem-
perature dependence for the CH4 : C heterotrophic respira-
tion ratio best explains the spatial variability of the posterior
emissions. Despite regional agreement with the EPA anthro-
pogenic inventory, we still find significant local discrepan-
cies with the EPA inventory for the oil and gas sector, sug-
gesting that methane emission factors are more variable than
assumed in the inventory. www.atmos-chem-phys.net/18/6483/2018/ in the Southeast US using SEAC4RS aircraft observations
Figure 6. Range of wetland emission estimates in the Southeast
US (domain of Fig. 1) for August–September 2013. The figure
shows the spread of the WetCHARTs extended ensemble and com-
pares with the posterior emission estimate from our inversion in
terms of emission total and root-mean-square error (RMSE) on the
0.25◦× 0.3125◦spatial grid. The RMSE of the posterior emission
estimate with itself is zero by definition. WetCHARTs ensemble
members use wetland areal extent data from either the GLOB-
COVER (Bontemps et al., 2011) or GLWD (Lehner and Dölla,
2004) databases, as well as different estimates of temperature sen-
sitivity q10 and global scaling factors s (see Eq. 5 and text). The
posterior wetland emission estimate from our inversion is shown
as black solid diamond with its associated error standard deviation. The mean of the WetCHARTs ensemble used as prior for our inver-
sion is shown as blue solid circle with its associated error standard
deviation. 4
Conclusions We used extensive boundary layer methane observations
from the SEAC4RS aircraft campaign over the Southeast US
in August–September 2013 to optimize methane emissions
in that region with up to 0.25◦× 0.3125◦spatial resolution
and with detailed error characterization. The inversion used
new state-of-the-art inventories as prior information, includ-
ing the gridded version of the EPA (2016) national anthro-
pogenic inventory from Maasakkers et al. (2016) and the
WetCHARTs wetlands extended ensemble from Bloom et al. (2017). The inversion domain over the Southeast US ac-
counts for 45 % of national methane emissions in the EPA
inventory, and for 56 % of wetland emissions over the con-
tiguous US in the mean WetCHARTs estimate. Data availability. SEAC4RS
aircraft
methane
observations
are available at https://www-air.larc.nasa.gov/cgi-bin/ArcView/
seac4rs\T1\textbackslash#BLAKE.DONALD/
(last
access:
1 November 2017). TCCON data (Wunch et al., 2011; Wennberg
et al., 2017) are available at http://tccondata.org (last access:
1 November 2017). The NOAA data (Andrews et al., 2014) are Atmos. Chem. Phys., 18, 6483–6491, 2018 References Alexe, M., Bergamaschi, P., Segers, A., Detmers, R., Butz, A.,
Hasekamp, O., Guerlet, S., Parker, R., Boesch, H., Frankenberg,
C., Scheepmaker, R. A., Dlugokencky, E., Sweeney, C., Wofsy,
S. C., and Kort, E. A.: Inverse modelling of CH4 emissions
for 2010–2011 using different satellite retrieval products from
GOSAT and SCIAMACHY, Atmos. Chem. Phys., 15, 113–133,
https://doi.org/10.5194/acp-15-113-2015, 2015. Fisher, J. A., Jacob, D. J., Travis, K. R., Kim, P. S., Marais, E. A., Chan Miller, C., Yu, K., Zhu, L., Yantosca, R. M., Sul-
prizio, M. P., Mao, J., Wennberg, P. O., Crounse, J. D., Teng,
A. P., Nguyen, T. B., St. Clair, J. M., Cohen, R. C., Romer,
P., Nault, B. A., Wooldridge, P. J., Jimenez, J. L., Campuzano-
Jost, P., Day, D. A., Hu, W., Shepson, P. B., Xiong, F., Blake,
D. R., Goldstein, A. H., Misztal, P. K., Hanisco, T. F., Wolfe,
G. M., Ryerson, T. B., Wisthaler, A., and Mikoviny, T.: Or-
ganic nitrate chemistry and its implications for nitrogen budgets
in an isoprene- and monoterpene-rich atmosphere: constraints
from aircraft (SEAC4RS) and ground-based (SOAS) observa-
tions in the Southeast US, Atmos. Chem. Phys., 16, 5969–5991,
https://doi.org/10.5194/acp-16-5969-2016, 2016. p
g
p
Andrews, A. E., Kofler, J. D., Trudeau, M. E., Williams, J. C., Neff,
D. H., Masarie, K. A., Chao, D. Y., Kitzis, D. R., Novelli, P. C.,
Zhao, C. L., Dlugokencky, E. J., Lang, P. M., Crotwell, M. J.,
Fischer, M. L., Parker, M. J., Lee, J. T., Baumann, D. D., Desai,
A. R., Stanier, C. O., De Wekker, S. F. J., Wolfe, D. E., Munger,
J. W., and Tans, P. P.: CO2, CO, and CH4 measurements from tall
towers in the NOAA Earth System Research Laboratory’s Global
Greenhouse Gas Reference Network: instrumentation, uncer-
tainty analysis, and recommendations for future high-accuracy
greenhouse gas monitoring efforts, Atmos. Meas. Tech., 7, 647–
687, https://doi.org/10.5194/amt-7-647-2014, 2014. G. M., Ryerson, T. B., Wisthaler, A., and Mikoviny, T.: Or-
ganic nitrate chemistry and its implications for nitrogen budgets
in an isoprene- and monoterpene-rich atmosphere: constraints
from aircraft (SEAC4RS) and ground-based (SOAS) observa-
tions in the Southeast US, Atmos. Chem. Phys., 16, 5969–5991,
https://doi.org/10.5194/acp-16-5969-2016, 2016. Heald, C. L., Jacob, D. J., Jones, D. B. A., Palmer, P. I., Logan, J. A.,
Streets,
D. G., Sachse, G. W., Gille, J. C., Hoffman, R. J.-X. Sheng et al.: Methane emissions in the Southeast US using SEAC4RS aircraft observations Brandt, A. R., Heath, G. A., Kort, E. A., O’Sullivan, F.,
Pétron, G., Jordaan, S. M., Tans, P., Wilcox, J., Gop-
stein, A. M., Arent, D., Wofsy, S., Brown, N. J., Bradley, R.,
Stucky, G. D., Eardley, D., and Harriss, R.: Methane Leaks from
North American Natural Gas Systems, Science, 343, 733–735,
https://doi.org/10.1126/science.1247045, 2014. available at https://www.esrl.noaa.gov/gmd/ccgg/obspack/ (last
access: 1 November 2017). Competing interests. The authors declare that they have no conflict
of interest. Competing interests. The authors declare that they have no conflict
of interest. Brasseur, G. P. and Jacob, D. J.: Modeling of Atmospheric Chem-
istry, 1st Edn., Cambridge University Press, 2017. Chan Miller, C., Jacob, D. J., Marais, E. A., Yu, K., Travis, K. R.,
Kim, P. S., Fisher, J. A., Zhu, L., Wolfe, G. M., Hanisco, T. F.,
Keutsch, F. N., Kaiser, J., Min, K.-E., Brown, S. S., Washen-
felder, R. A., González Abad, G., and Chance, K.: Glyoxal yield
from isoprene oxidation and relation to formaldehyde: chemi-
cal mechanism, constraints from SENEX aircraft observations,
and interpretation of OMI satellite data, Atmos. Chem. Phys., 17,
8725–8738, https://doi.org/10.5194/acp-17-8725-2017, 2017. Acknowledgements. This work was funded by the NASA Earth
Science Division. Part of this research was carried out at the
Jet Propulsion Laboratory, California Institute of Technology,
under a contract with NASA. Special thanks to Donald R. Blake
for providing SEAC4RS aircraft methane observations (avail-
able
at
https://www-air.larc.nasa.gov/cgi-bin/ArcView/seac4rs#
BLAKE.DONALD/, last access: 1 November 2017). TCCON
data were obtained from the TCCON Data Archive, hosted
by CaltechData (http://tccondata.org, last access: 1 Novem-
ber 2017). The NOAA data are available from the ObsPack
portal (https://www.esrl.noaa.gov/gmd/ccgg/obspack/, last access:
1 November 2017). EPA: Inventory of US Greenhouse Gas Emissions and Sinks 1990–
2011, US Environmental Protection Agency, Washington DC,
2013. EPA: Inventory of US Greenhouse Gas Emissions and Sinks 1990–
2014, US Environmental Protection Agency, Washington DC,
2014. Edited by: Christoph Gerbig EPA: Inventory of US Greenhouse Gas Emissions and Sinks 1990–
2014,US Environmental Protection Agency, Washington DC,
2016. Reviewed by: two anonymous referees European Commission: Emission Database for Global Atmospheric
Research (EDGAR), release version 4.2, Tech. rep., Joint Re-
search Centre (JRC)/Netherlands Environmental Assessment
Agency (PBL), available at: http://edgar.jrc.ec.europa.eu (last ac-
cess: 1 November 2017), 2011 www.atmos-chem-phys.net/18/6483/2018/ www.atmos-chem-phys.net/18/6483/2018/ 6489 Atmos. Chem. Phys., 18, 6483–6491, 2018 J.-X. Sheng et al.: Methane emissions in the Southeast US using SEAC4RS aircraft observations J.-X. Sheng et al.: Methane emissions in the Southeast US using SEAC4RS aircraft observations 6490 Kim, P. S., Jacob, D. J., Fisher, J. A., Travis, K., Yu, K., Zhu, L.,
Yantosca, R. M., Sulprizio, M. P., Jimenez, J. L., Campuzano-
Jost, P., Froyd, K. D., Liao, J., Hair, J. W., Fenn, M. A., But-
ler, C. F., Wagner, N. L., Gordon, T. D., Welti, A., Wennberg,
P. O., Crounse, J. D., St. Clair, J. M., Teng, A. P., Millet,
D. B., Schwarz, J. P., Markovic, M. Z., and Perring, A. E.:
Sources, seasonality, and trends of southeast US aerosol: an in-
tegrated analysis of surface, aircraft, and satellite observations
with the GEOS-Chem chemical transport model, Atmos. Chem. Phys., 15, 10411–10433, https://doi.org/10.5194/acp-15-10411-
2015, 2015. Qin, D., Plattner, G.-K., Tignor, M., Allen, S. K., Boschung, J.,
Nauels, A., Xia, Y., Bex, V. and Midgley, P. M., Cambridge Uni-
versity Press, Cambridge, UK and New York, NY, USA, 2013. Park, R. J., Jacob, D. J., Field, B. D., Yantosca, R. M.,
and Chin, M.: Natural and transboundary pollution influ-
ences on sulfate-nitrate-ammonium aerosols in the United
States: Implications for policy, J. Geophys. Res., 109, D15204,
https://doi.org/10.1029/2003JD004473, 2004. Prather, M. J., Holmes, C. D., and Hsu, J.: Reactive greenhouse
gas scenarios: Systematic exploration of uncertainties and the
role of atmospheric chemistry, Geophys. Res. Lett., 39, L09803,
https://doi.org/10.1029/2012GL051440, 2012. Lehner, B. and Dölla, P.: Development and validation of a global
database of lakes, reservoirs and wetlands, J. Hydrol., 296, 1–22,
https://doi.org/10.1016/j.jhydrol.2004.03.028, 2004. Rodgers, C. D.: Inverse Methods for Atmospheric Sounding: The-
ory and Practice, World Scientific Publishing, Singapore, 2000. Sheng, J.-X., Jacob, D. J., Maasakkers, J. D., Sulprizio, M. P.,
Zavala-Araiza, D., and Hamburg, S. P.: A high-resolution (0.1◦×
0.1◦) inventory of methane emissions from Canadian and Mex-
ican oil and gas systems, Atmos. Environ., 158, 211–215,
https://doi.org/10.1016/j.atmosenv.2017.02.036, 2017. Lyon, D. R., Zavala-Araiza, D., Alvarez, R. A., Harriss, R., Pala-
cios, V., Lan, X., Talbot, R., Lavoie, T., Shepson, P., Yacov-
itch, T. I., Herndon, S. C., Marchese, A. J., Zimmerle, D.,
Robinson, A. L., and Hamburg, S. P.: Constructing a Spa-
tially Resolved Methane Emission Inventory for the Bar-
nett Shale Region, Environ. Sci. Technol., 49, 8147–8157,
https://doi.org/10.1021/es506359c, 2015. Simpson, I. J., Chenm, T.-Y., Blake, D. R., and Rowland, F. S.:
Implications of the recent fluctuations in the growth rate of
tropospheric methane, Geophys. Res. Lett., 29, 1171–1174,
https://doi.org/10.1029/2001GL014521, 2002. Maasakkers, J. J.-X. Sheng et al.: Methane emissions in the Southeast US using SEAC4RS aircraft observations D., Jacob, D. J., Sulprizio, M. P., Turner, A. J.,
Weitz, M., Wirth, T., Hight, C., DeFigueiredo, M., Desai, M.,
Schmeltz, R., Hockstad, L., Bloom, A. A., Bowman, K. W.,
Jeong, S., and Fischer, M. L.: Gridded National Inventory of US
Methane Emissions, Environ. Sci. Technol., 50, 13123–13133,
https://doi.org/10.1021/acs.est.6b02878, 2016. Toon, O. B., Maring, H., Dibb, J., Ferrare, R., Jacob, D. J.,
Jensen, E. J., Luo, Z. J., Mace, G. G., Pan, L. L., Pfister, L.,
Rosenlof, K. H., Redemann, J., Reid, J. S., Singh, H. B., Thomp-
son, A. M., Yokelson, R., Minnis, P., Chen, G., Jucks, K. W.,
and Pszenny, A.: Planning, implementation, and scientific goals
of the Studies of Emissions and Atmospheric Composition,
Clouds and Climate Coupling by Regional Surveys (SEAC4rs)
field mission, J. Geophys. Res.-Atmos., 121, 2015JD024297,
https://doi.org/10.1002/2015JD024297, 2016. Marais, E. A., Jacob, D. J., Jimenez, J. L., Campuzano-Jost, P.,
Day, D. A., Hu, W., Krechmer, J., Zhu, L., Kim, P. S., Miller,
C. C., Fisher, J. A., Travis, K., Yu, K., Hanisco, T. F., Wolfe,
G. M., Arkinson, H. L., Pye, H. O. T., Froyd, K. D., Liao, J.,
and McNeill, V. F.: Aqueous-phase mechanism for secondary or-
ganic aerosol formation from isoprene: application to the south-
east United States and co-benefit of SO2 emission controls, At-
mos. Chem. Phys., 16, 1603–1618, https://doi.org/10.5194/acp-
16-1603-2016, 2016. p
g
Travis, K. R., Jacob, D. J., Fisher, J. A., Kim, P. S., Marais, E. A.,
Zhu, L., Yu, K., Miller, C. C., Yantosca, R. M., Sulprizio, M. P., Thompson, A. M., Wennberg, P. O., Crounse, J. D., St. Clair,
J. M., Cohen, R. C., Laughner, J. L., Dibb, J. E., Hall, S. R.,
Ullmann, K., Wolfe, G. M., Pollack, I. B., Peischl, J., Neuman, J. A., and Zhou, X.: Why do models overestimate surface ozone in
the Southeast United States?, Atmos. Chem. Phys., 16, 13561–
13577, https://doi.org/10.5194/acp-16-13561-2016, 2016. Melton, J. R., Wania, R., Hodson, E. L., Poulter, B., Ringeval, B.,
Spahni, R., Bohn, T., Avis, C. A., Beerling, D. J., Chen, G.,
Eliseev, A. V., Denisov, S. N., Hopcroft, P. O., Lettenmaier, D. P., Riley, W. J., Singarayer, J. S., Subin, Z. M., Tian, H., Zürcher,
S., Brovkin, V., van Bodegom, P. M., Kleinen, T., Yu, Z. C.,
and Kaplan, J. O.: Present state of global wetland extent and
wetland methane modelling: conclusions from a model inter-
comparison project (WETCHIMP), Biogeosciences, 10, 753–
788, https://doi.org/10.5194/bg-10-753-2013, 2013. Turner, A. References N.,
and Nehrkorn, T.: Comparative inverse analysis of satellite (MO-
PITT) and aircraft (TRACE-P) observations to estimate Asian
sources of carbon monoxide, J. Geophys. Res., 109, D23306,
https://doi.org/10.1029/2004JD005185, 2004. Bloom, A. A., Exbrayat, J.-F., van der Velde, I. R., Feng, L.,
and Williams, M.: The decadal state of the terrestrial carbon
cycle: Global retrievals of terrestrial carbon allocation, pools,
and residence times, P. Natl. Acad. Sci. USA, 113, 1285–1290,
https://doi.org/10.1073/pnas.1515160113, 2016. Henze, D. K., Hakami, A., and Seinfeld, J. H.: Development of
the adjoint of GEOS-Chem, Atmos. Chem. Phys., 7, 2413–2433,
https://doi.org/10.5194/acp-7-2413-2007, 2007. Bloom, A. A., Bowman, K. W., Lee, M., Turner, A. J., Schroeder,
R., Worden, J. R., Weidner, R., McDonald, K. C., and Ja-
cob, D. J.: A global wetland methane emissions and un-
certainty dataset for atmospheric chemical transport models
(WetCHARTs version 1.0), Geosci. Model Dev., 10, 2141–2156,
https://doi.org/10.5194/gmd-10-2141-2017, 2017. Jacob, D. J., Turner, A. J., Maasakkers, J. D., Sheng, J., Sun,
K., Liu, X., Chance, K., Aben, I., McKeever, J., and Franken-
berg, C.: Satellite observations of atmospheric methane and
their value for quantifying methane emissions, Atmos. Chem. Phys., 16, 14371–14396, https://doi.org/10.5194/acp-16-14371-
2016, 2016. Bontemps, S., Defourny, P., Bogaert, E. V., Arino, O., Kalogirou, V.,
and Perez, J R.: GLOBCOVER 2009: Products Description and
Validation Report, Tech. rep., ESA, 2, 2011. www.atmos-chem-phys.net/18/6483/2018/ Atmos. Chem. Phys., 18, 6483–6491, 2018 J.-X. Sheng et al.: Methane emissions in the Southeast US using SEAC4RS aircraft observations J. and Jacob, D. J.: Balancing aggregation and smoothing
errors in inverse models, Atmos. Chem. Phys., 15, 7039–7048,
https://doi.org/10.5194/acp-15-7039-2015, 2015. Turner, A. J., Jacob, D. J., Wecht, K. J., Maasakkers, J. D., Lund-
gren, E., Andrews, A. E., Biraud, S. C., Boesch, H., Bowman, K. W., Deutscher, N. M., Dubey, M. K., Griffith, D. W. T., Hase,
F., Kuze, A., Notholt, J., Ohyama, H., Parker, R., Payne, V. H., Sussmann, R., Sweeney, C., Velazco, V. A., Warneke, T.,
Wennberg, P. O., and Wunch, D.: Estimating global and North
American methane emissions with high spatial resolution us-
ing GOSAT satellite data, Atmos. Chem. Phys., 15, 7049–7069,
https://doi.org/10.5194/acp-15-7049-2015, 2015. Miller, S. M., Wofsy, S. C., Michalak, A. M., Kort, E. A.,
Andrews,
A. E.,
Biraud,
S. C.,
Dlugokencky,
E. J.,
Eluszkiewicz, J., Fischer, M. L., Janssens-Maenhout, G.,
Miller, B. R., Miller, J. B., Montzka, S. A., Nehrkorn, T.,
and Sweeney, C.: Anthropogenic emissions of methane in the
United States, P. Natl. Acad. Sci. USA, 110, 20018–20022,
https://doi.org/10.1073/pnas.1314392110, 2013. Turner, A. J., Frankenberg, C., Wennberg, P. O., and Jacob, D. J.:
Ambiguity in the causes for decadal trends in atmospheric
methane and hydroxyl, P. Natl. Acad. Sci. USA, 114, 5367–5372,
https://doi.org/10.1073/pnas.1616020114, 2017. Myhre, G., Shindell, D., Bréon, F.-M., Collins, W., Fuglestvedt,
J., Huang, J., Koch, D., Lamarque, J.-F., Lee, D., Mendoza,
B., Nakajima, T., Robock, A., Stephens, G., Takemura, T., and
Zhang, H.: Anthropogenic and Natural Radiative Forcing, in:
Climate Change 2013: The Physical Science Basis. Contribution
of Working Group I to the Fifth Assessment Report of the Inter-
governmental Panel on Climate Change, edited by: Stocker, T. F., United Nation: United Nations Framework Convention on Climate
Change, Article 4(1)(a), available at: https://unfccc.int/resource/
docs/convkp/conveng.pdf (last access: November 2017), 1992. www.atmos-chem-phys.net/18/6483/2018/ Atmos. Chem. Phys., 18, 6483–6491, 2018 Atmos. Chem. Phys., 18, 6483–6491, 2018 J.-X. Sheng et al.: Methane emissions in the Southeast US using SEAC4RS aircraft observations 6491 J.-X. Sheng et al.: Methane emissions in the Southeast US using SEAC4RS aircraft observations Wecht, K. J., Jacob, D. J., Sulprizio, M. P., Santoni, G. W., Wofsy,
S. C., Parker, R., Bösch, H., and Worden, J.: Spatially resolv-
ing methane emissions in California: constraints from the Cal-
Nex aircraft campaign and from present (GOSAT, TES) and fu-
ture (TROPOMI, geostationary) satellite observations, Atmos. Chem. Phys., 14, 8173–8184, https://doi.org/10.5194/acp-14-
8173-2014, 2014. Yu, K., Jacob, D. J., Fisher, J. A., Kim, P. S., Marais, E. A., Miller,
C. C., Travis, K. R., Zhu, L., Yantosca, R. M., Sulprizio, M. P.,
Cohen, R. C., Dibb, J. E., Fried, A., Mikoviny, T., Ryerson, T. B., Wennberg, P. O., and Wisthaler, A.: Sensitivity to grid res-
olution in the ability of a chemical transport model to simulate
observed oxidant chemistry under high-isoprene conditions, At-
mos. Chem. Phys., 16, 4369–4378, https://doi.org/10.5194/acp-
16-4369-2016, 2016. Wennberg, P. O., Wunch, D., Roehl, C., Blavier, J. F., Toon, G. C.,
Allen, N., Dowell, P., Teske, K., Martin, C., and Martin, J.:
TCCON
data
from
Lamont,
Oklahoma,
USA,
Release
GGG2014r1,
TCCON
data
archive,
hosted
by
Caltech-
DATA, California Institute of Technology, Pasadena, CA, USA,
https://doi.org/10.14291/tccon.ggg2014.lamont01.R1/1255070,
2017. Zhu, L., Jacob, D. J., Kim, P. S., Fisher, J. A., Yu, K., Travis,
K. R., Mickley, L. J., Yantosca, R. M., Sulprizio, M. P., De
Smedt, I., González Abad, G., Chance, K., Li, C., Ferrare, R.,
Fried, A., Hair, J. W., Hanisco, T. F., Richter, D., Jo Scarino,
A., Walega, J., Weibring, P., and Wolfe, G. M.: Observing at-
mospheric formaldehyde (HCHO) from space: validation and
intercomparison of six retrievals from four satellites (OMI,
GOME2A, GOME2B, OMPS) with SEAC4RS aircraft observa-
tions over the southeast US, Atmos. Chem. Phys., 16, 13477–
13490, https://doi.org/10.5194/acp-16-13477-2016, 2016. Wolf, J., Asrar, G. R., and West, T. O.: Revised methane emis-
sions factors and spatially distributed annual carbon fluxes for
global livestock, Carbon Balance and Management, 12, 16,
https://doi.org/10.1186/s13021-017-0084-y, 2017. Wunch, D., Toon, G. C., Blavier, J.-F. L., Washenfelder, R. A.,
Notholt, J., Connor, B. J., Griffith, D. W. T., Sherlock, V.,
and
Wennberg,
P. O.:
The
Total
Carbon
Column
Ob-
serving Network, Philos. T. R. Soc. A, 369, 2087–2112,
https://doi.org/10.1098/rsta.2010.0240, 2011. Atmos. Chem. Phys., 18, 6483–6491, 2018 www.atmos-chem-phys.net/18/6483/2018/ www.atmos-chem-phys.net/18/6483/2018/ www.atmos-chem-phys.net/18/6483/2018/
|
https://openalex.org/W2901409012
|
https://edoc.unibas.ch/66972/1/20181127161003_5bfd5e4baff58.pdf
|
English
| null |
How does membership in local savings groups influence the determinants of national health insurance demand? A cross-sectional study in Kisumu, Kenya
|
International journal for equity in health
| 2,018
|
cc-by
| 7,035
|
Oraro and Wyss International Journal for Equity in Health (2018) 17:170
https://doi.org/10.1186/s12939-018-0889-7 Oraro and Wyss International Journal for Equity in Health (2018) 17:170
https://doi.org/10.1186/s12939-018-0889-7 Open Access How does membership in local savings
groups influence the determinants of
national health insurance demand? A
cross-sectional study in Kisumu, Kenya Tessa Oraro1,2,3*
and Kaspar Wyss1,2 Tessa Oraro1,2,3*
and Kaspar Wyss1,2 Tessa Oraro1,2,3*
and Kaspar Wyss1,2 Abstract Background: Rotating savings and credit associations (ROSCAs) are highly active in many sub-Saharan African
countries, serving as an important gateway for coping with financial risk. In light of the Kenya’s National Hospital
Insurance Fund’s (NHIF’s) strategy of targeting ROSCAs for membership enrolment, this study sought to estimate
how ROSCA membership influences the determinants of voluntary health insurance enrolment. Methods: A cross-sectional survey of 444 households was carried out in Kisumu City between July and August
2016. A structured questionnaire was administered on health insurance membership, household attributes,
headship characteristics and health-seeking behaviour. We assessed the influence of ROSCA membership on the
associations between NHIF enrolment and the explanatory variables using univariate logistic regression. Results: The study found that education was associated with NHIF demand regardless of ROSCA membership. Both ROSCA and non-ROSCA households with high socioeconomic status showed stronger health insurance
demand compared with poorer households; there was, however, no evidence that the strength of this association
was influenced by ROSCA status (p-value = 0.47). Participants who were self-employed were significantly less likely
to enrol into the NHIF if they did not belong to a ROSCA (interaction test p-value = 0.03). NHIF enrolment was
found to be lower among female-headed households. There was a borderline effect of ROSCA membership on
this association, with a lower odds ratio amongst non-ROSCA members (p-value = 0.09): the low treatment numbers
amongst the insured infers that ROSCA membership may play a role on the association between gender and NHIF
demand. Conclusions: Our findings suggest that ROSCA membership may play a role in increasing health insurance demand
amongst some traditionally under-represented groups such as women and the self-employed. However, the
strategy of targeting ROSCAs to increase national health insurance enrolment may yield exiguous results, given
that ROSCA membership is itself influenced by several non-observable factors – such as time-availability and self-
selection. It is therefore important to anchor outreach to ROSCAs within a broader, multi-pronged approach that
targets households within their social, economic and political realities. Keywords: ROSCAs, Health insurance, Social capital, Kenya * Correspondence: tessa.oraro@swisstph.ch
1Swiss Tropical and Public Health Institute, Basel, Switzerland
2University of Basel, Basel, Switzerland
Full list of author information is available at the end of the article Introduction While different interpretations of its nature
and effect exist, it is generally agreed that social capital in-
fluences the depth of individuals’ engagement with local
and national structures by facilitating trust and reciprocity
within communities. ROSCAs – as local informal group-
ings – seek to empower economic cooperation in the face
of complex resource and capacity constraints. In doing so,
they tap into two key dimensions of social capital: intra-
and inter-community collaborations. When constituting
ROSCAs in low-resource settings, members often recruit
those in whom they have inherent trust – ostensibly those
with similar socioeconomic, professional or ethno-cultural
characteristics. This structure mimics traditional African
obligations to familial and tribal alliances [3], and de-
rives its power from harnessing existing links within
homogenous groups: a concept known as bonding social
capital [30]. Concurrently, in order to make ROSCAs sus-
tainable, members seek to increase the associations’ capital
base and minimise the risk of default [2]. They thus ex-
pand their radius of trust beyond their typical personal,
sociocultural and organisational bases - a phenomenon
known as bridging social capital [30]. In this vein, ROS-
CAs foment economic, social, and cultural links within
local communities [35]. Against the backdrop of increasing ROSCA member-
ship across sub-Saharan Africa, financial risk protection
in health care has remained low. Several governments
in the region such as those in Ghana and Kenya have
sought to promote national health insurance schemes
as a means of minimising the high out-of-pocket costs
associated with ill health [38, 39]. In Kenya – where up
to 28% of slum-dwelling households face catastrophic
health expenditure [5] – the national government has
established
the
National
Hospital
Insurance
Fund
(NHIF) as a means of providing financial risk protec-
tion to its citizens. The NHIF operates like most na-
tional health insurance schemes in the region, with the
government and the formal sector providing the bulk of
financial contributions to the scheme [40]. However,
given the high levels of informal employment in the
Kenyan economy, the depth of insurance coverage has
remained low, with only 13.5% of the population receiv-
ing NHIF coverage [20]. A large proportion of Kenya’s
informally employed households – who constitute 85%
of the working population – are thus left susceptible to
financial shocks in the event of ill health [20, 41]. Introduction social solidarity and health insurance demand across
various sub-Saharan African settings [1, 14]. This grow-
ing body of evidence has led Kenyan policymakers to
target ROSCAs as a means of increasing NHIF demand
amongst the informal sector, in line with their role as ve-
hicles for social solidarity [33]. Consequently, we seek to
understand how ROSCA membership influences the
drivers of voluntary health insurance in Kisumu, Kenya. This paper will be structured as follows: We will first
provide a review of the research on the intersection be-
tween local group membership, social capital, and volun-
tary health insurance membership. We will subsequently
report our methodology and results, and conclude with
an analysis of our findings. Informal economic networks based on geographical,
sociocultural or occupational proximity are highly ac-
tive in many sub-Saharan African countries, with an es-
timated 24% of the population participating in local
savings clubs in 2014 [9]. In Kenya, where 24.6% of the
population is excluded from formal savings and loans
institutions, local rotating savings and credit associa-
tions (ROSCAs) have emerged as important gateways
for coping with financial instability [7]. ROSCAs are
typically created to meet an unmet social and economic
need amongst vulnerable groups such as the working
poor and women [2, 9], with participants pooling their
financial resources to create communal savings funds
as a form of financial risk protection [42]. In this sense,
ROSCAs serve as social networks through which individ-
uals obtain financial security through collective resource
mobilisation. Due to their informal nature, ROSCAs’
success is contingent upon the trust, reciprocity and col-
lective strength shared amongst group members – funda-
mental components of a phenomenon known as social
capital [12, 42]. It is this reliance on social capital that has
driven the use of similar welfare associations as launch-
pads for the national health insurance schemes in Germany
and Japan [6, 36]. y
g
In order to better understand the mechanisms through
which ROSCAs may influence health insurance enrol-
ment, it is first necessary to define the social capital
underpinning the success of ROSCAs. Social capital as a
concept focuses on the utility of relationships and re-
sources in facilitating collaboration towards a common
goal [4, 8, 34]. Page 2 of 9 Oraro and Wyss International Journal for Equity in Health (2018) 17:170 Page 2 of 9 Introduction This
problem
echoes
findings
across
health
insurance
schemes in several sub-Saharan countries, where low
health insurance demand has been attributed to high
levels of labour informality. In spite of the pervasiveness of ROSCAs and their link-
ages to national health insurance enrolment in several
sub-Saharan African settings [24, 27], there remains a pau-
city of research on the influence of ROSCAs on national
health insurance demand. We can, however, draw lessons
from studies carried out on the influence of social capital
as a general concept on voluntary health insurance enrol-
ment. According to existing research, active members of
local mutual assistance groups such as ROSCAs are more
likely to be part of voluntary health insurance schemes As part of efforts to better understand the reasons
behind the low uptake of voluntary health insurance,
several research studies have identified a link between Oraro and Wyss International Journal for Equity in Health (2018) 17:170 Oraro and Wyss International Journal for Equity in Health (2018) 17:170 Page 3 of 9 Page 3 of 9 due to a shared ethos of social cooperation [11, 15, 31]. These findings highlight the importance of common
norms, values and objectives – characteristic features of
social capital – in influencing households’ decision to
enrol into voluntary health insurance. Given this, our
paper will seek to answer the following research question:
How does ROSCA membership influence the determi-
nants of voluntary health insurance enrolment? We will
further seek to understand how existing knowledge on so-
cial capital can help explain the influence of ROSCAs on
the determinants of health insurance demand. Multistage cluster sampling was used to obtain the
study sample, with enumeration maps of the six
sub-locations used to randomly identify existing infor-
mal settlements in Kisumu [19]. In the second stage,
we selected existing water points within informal set-
tlements, and randomly selected starting points lo-
cated near the water points. This strategy was driven
by the use of watering points as navigation tools by
the local communities. We subsequently carried out
systematic sampling, with every nth household inter-
viewed. Approximately 98% of the targeted households
could be interviewed. In order to answer these questions, we can apply several
hypotheses to our study based on the findings of existing
research. Introduction Given the link identified between local informal
group membership and social capital, we anticipate that
non-ROSCA members may have weaker bonding and
bridging social capital compared with their ROSCA mem-
ber counterparts. As a result of this, we expect wealth to
manifest as an overarching determinant of NHIF demand
amongst non-ROSCA members in line with existing na-
tional health financing studies [10, 17, 18, 24, 26, 27]. Additionally, we hypothesise that sociocultural sensitivities
such as household composition; gender of household
head; and marital status will influence households’ likeli-
hood of enrolling into national health insurance [10, 17,
18, 24–27]. We however postulate that these inequities
may be reduced by ROSCA membership due to their high
social capital stock. Data collection
Th The survey was administered as a structured question-
naire using Open Data Kit (ODK) software on handheld
tablets to answer questions on household composition;
household assets; household expenditure and consump-
tion; and health-seeking behaviour. Research assistants
who were fluent in English, Kiswahili and the local dialect,
Luo, and who had knowledge of the local geographical
and sociocultural context were hired. Training was pro-
vided to familiarise research assistants with the question-
naire and data collection using handheld tablets. Household heads were targeted as the primary respon-
dents for this study. In cases where the household head
was not available, their spouse was interviewed. In order to identify the informally employed in our
study area, we used a deductive approach of identifying
households where the household head did not make
mandatory monthly payments into the National Social
Security Fund (NSSF). This approach was taken, as
there is no clearly defined approach for identifying the
informal sector in Kenya. However, the formal sector is
defined distinctly within Kenya’s legal framework, with
sector members legally obligated to make monthly pen-
sion payments to the NSSF [32]. Sampling methodology A cross-sectional household survey was carried out in
Kisumu City between July and August 2016 to identify
the extent to which NHIF enrolment is influenced by
ROSCA membership; household composition and attri-
butes; as well as household head characteristics such as
occupation, age, gender and perceived health status of
household members. Insured households were defined as those where mem-
bers were voluntarily enrolled into the NHIF in the year
preceding the study. Our study aimed to estimate the proportion of the in-
formally employed population with voluntary NHIF
health insurance, and to assess the association between
explanatory variables and this proportion. In order to
achieve these aims, we calculated the sample size based
on the ability to estimate the proportion insured with a
certain precision - in this case, with an expected NHIF
population coverage of 13.8% [29]. Using the Hayes and
Bennett equation and taking into account clustering in
six sub-locations across Kisumu, required sample size
was calculated to be 440 households with the precision
of a 95% confidence interval with a width of 10 to 18%
[16]. The number of households interviewed in each
sub-location was proportional to its demographic size,
which was obtained from the 2009 Kenya Population
and Housing Census [22]. Statistical analysis
h
d The data were analysed using STATA version 14.1 for
Windows (STATA Corporation, College Station, Texas). The main outcome variable was voluntary enrolment
into the NHIF in the previous year. The explanatory var-
iables were divided into five components: household
composition and attributes; household head factors; per-
ceived household health status; and proxies for exposure
to financial risk pooling. In order to measure wealth, we used the asset index
as described within the Kenya Demographic and Health
Survey (DHS) to calculate households’ asset-based wealth
[21]. We collected data on consumer items, dwelling Oraro and Wyss International Journal for Equity in Health (2018) 17:170 Oraro and Wyss International Journal for Equity in Health (2018) 17:170 Page 4 of 9 characteristics such as housing materials, water source,
sleeping arrangements, and other characteristics linked to
wealth status. We subsequently computed the asset index. Households were then classified into one of five socioeco-
nomic groups based on quintiles (Q): Q1 or the poorest
20%; Q2; Q3; Q4; and Q5 or the wealthiest 20% in the
sample. These groups were subsequently clustered into
three groups: Q1 and Q2; Q3 and Q4; and Q5 respect-
ively, to account for the minimal differences amongst the
proximal quintiles. Literacy within the study context was
defined as those above the age of 15 years who could read
and write [37]. The mean perceived health status of each
household was calculated as the average self-reported
health status value for all household members. The final
value of the mean perceived health status was assigned a
value of between 1 and 4, with 1 being “very good” health
status; 2 being “good” health status; 3 being “poor” health
status; and 4 being “very poor” health status. Most household heads in our study population re-
ported themselves as economically active, with 94% of
respondents carrying out casual income-generating ac-
tivities in lieu of formal employment. The most common
economic activities amongst our study population were
small-scale business ventures, with 56% of the popula-
tion engaging in activities such as carpentry, masonry,
and hairdressing. This notwithstanding, our findings re-
vealed that 26% of our study population lived below the
poverty line, which is defined as a household income of
less than $1.90 a day [41]. The household characteristics and health status for the
sample population were broadly similar regardless of
ROSCA membership status. Statistical analysis
h
d Overall, the median house-
hold size was 4, with 77% of all households having chil-
dren below the age of 15 years. The number of households
with elderly members was low across the study popula-
tion, with only 8% of households having members above
the age of 65 years. The study population also showed
similarities in the gender and marital status of household
heads, irrespective of ROSCA membership status. Further,
the study found similarities in the mean household health
status, with 5% of households reporting the presence of
chronic disease. We used multilevel mixed-effects logistic regression
models to estimate the association between health insur-
ance enrolment and the explanatory variables. We took
the clustering of households at the village level into ac-
count using random effects for village. In light of the
low numbers of insured households within our study
population, it was not possible to include multiple covar-
iates simultaneously in the model. Therefore, univariate
models were run to estimate the association of each vari-
able and health insurance enrolment for the ROSCA
and non-ROSCA households respectively. We ran the
regression models for (i) ROSCA and non-ROSCA sep-
arately to easily gain estimates of the univariate odds ra-
tios and 95% confidence intervals; and (ii) on the full
dataset with ROSCA and non-ROSCA households in
order to use interaction terms to directly test whether
the associations between each covariate and NHIF enrol-
ment were affected by ROSCA membership. Each model
included the covariate, ROSCA membership, and the
interaction term as explanatory variables. The study found important distinctions in the economic
characteristics of ROSCA and non-ROSCA households. ROSCA members were more likely to have more than one
earner in the household and higher monthly incomes
compared to non-members. Factors associated with voluntary health insurance
enrolment amongst ROSCA and non-ROSCA member
households Household-level regression results are presented in Table 2
for ROSCA and non-ROSCA member households. Our study found that education and asset-based SES
were associated with NHIF demand, regardless of
ROSCA membership status. Household heads in the
study population were more likely to enrol into the na-
tional health insurance scheme if they were educated to
at least secondary school level (OR = 1.80 (CI:1.05–3.11)
for ROSCA members, and OR = 3.23 (CI:1.47–7.08) for
non-ROSCA members (interaction test = 0.84). Our find-
ings also suggest that SES was associated with health in-
surance demand, although there was no evidence that the
strength of the association was influenced by ROSCA
membership status (p-value = 0.47). Study population characteristics Oraro and Wyss International Journal for Equity in Health (2018) 17:170 Page 5 of 9 Table 1 Descriptive statistics by gender and insurance status
ROSCA member
Non-ROSCA member
Insured
(n = 38)
Uninsured
(n = 125)
Combined
(n = 163)
Insured
(n = 81)
Uninsured
(n = 200)
Combined
(n = 281)
Household characteristics
Median household sizeb
4 (3,5)
4 (3,5)
4 (3,5)
5 (3,6)
4 (3,5)
4 (3,5)
Households with children <=5 years
48%
59%
56%
61%
58%
58%
Households with children <=15 years
75%
76%
76%
84%
78%
79%
Households with elderly > = 60 years
11%
6%
7%
8%
10%
10%
Average socioeconomic status (SES)
scoreaa
3.47 (0.28)
2.95 (0.27)
3.19 (0.27)
3.16 (0.21)
2.69 (0.26)
2.80 (0.25)
% of poor households (Q1-Q2)
26%
42%
37%
29%
50%
45%
% of rich households (Q5)
32%
18%
22%
18%
15%
16%
Annual average household expenditure
(KShs)a
365,060
(154435)
361,195
(734010)
362,309
(624283)
373,513
(571020)
230,524
(208211)
263,859
(333668)
Household head characteristics
Median ageb
39 (29,45)
35 (28,40)
36 (27,40)
34 (27,42)
35 (26,39)
35 (27,40)
Female household heads
21%
29%
26%
13%
34%
29%
Secondary-level education and above
63%
49%
53%
71%
43%
50%
Married household heads
83%
72%
75%
87%
66%
71%
Paid employee
38%
38%
38%
61%
30%
37%
Small-scale business owner
56%
57%
57%
32%
62%
55%
No fixed employment
6%
5%
5%
8%
7%
7%
More than one earner
52%
32%
37%
39%
19%
24%
Profits as remuneration
27%
24%
25%
18%
18%
18%
% living below poverty linec
14%
20%
19%
39%
38%
38%
Household health status
Mean health statusa
2.04 (0.59)
1.96 (0.52)
1.98 (0.54)
2.03 (0.51)
2.04 (0.56)
2.04 (0.55)
Presence of chronic disease
4%
6%
5%
3%
7%
6%
Use of curative services
Outpatient service use in past 30 days
48%
49%
48%
45%
37%
39%
Inpatient service use in past year
36%
25%
28%
16%
22%
21%
Annual per capita health expenditure
(KShs)a
1729 (18607)
739 (3542)
1025 (10516)
569 (2946)
818 (6442)
760 (5821)
aMean (±Standard Deviation) Table 1 Descriptive statistics by gender and insurance status an influencing role on the association between the ex-
planatory variable and NHIF enrolment. Our findings also raise the possibility of a stronger rela-
tionship between gender and national health insurance de-
mand if one does not belong to a ROSCA than if one does
(p-value = 0.09). Study population characteristics Study population characteristics
A total of 444 households were enumerated in the
survey, with ROSCA member households accounting
for 63% of the study population. Approximately 29% of
ROSCA households and 23% of non-ROSCA house-
holds were voluntarily enrolled into the NHIF. While
ROSCA members are actively targeted by the NHIF for
enrolment, our study did not find a statistically signifi-
cant association between ROSCA membership and health
insurance enrolment (OR: 1.33 [0.85–2.09]). Table 1 re-
ports the differences in each test variable according to
insurance status and household ROSCA membership
status. Our analysis showed that the relationship between
NHIF enrolment and some socio-demographic and eco-
nomic variables could be influenced by a household’s
ROSCA status. The study found that household heads
that were self-employed or had no fixed employment
were significantly less likely to enrol into the NHIF if
they did not belong to a ROSCA (p-value = 0.03). Study population characteristics According to our analysis, non-ROSCA
female household heads were less likely to enrol into the
NHIF compared with their male counterparts (OR = 0.29
(CI:0.11–0.79). The odds ratio among the ROSCA mem-
ber group was 0.65 (CI:0.35–1.23). Given the low treat-
ment numbers amongst the insured, the borderline effect
of ROSCA membership on the relationship between gen-
der and health insurance demand infers that it may play In addition to the above associations, our findings sug-
gest that NHIF demand may increase if the household
head was married or received a fixed salary. There was,
however, no evidence that these associations were influ-
enced by ROSCA membership. Discussion This study investigated the influence of informal finan-
cial coping mechanisms on the demand for formal, more Oraro and Wyss International Journal for Equity in Health (2018) 17:170 Page 6 of 9 Table 2 Univariate logistic estimates for probability of purchasing NHIF health insurance at household level
ROSCA members (n = 163)
Non-ROSCA members (n = 281)
Interaction* (n = 444)
Variable description
Odds ratio
CI
Odds ratio
CI
p-value
Household characteristics
Presence of children < 15 years
0.92
0.49–1.72
1.54
0.59–4.05
0.50
Presence of elderly > 60 years
1.93
0.76–4.90
0.74
0.20–2.74
0.15
Socioeconomic status (Reference group: First and second quintiles)
Middle wealth
1.71
0.91–3.23
2.56
1.12–5.88
0.47
Wealthiest quintile
2.81
1.38–5.73
2.08
0.71–6.10
Household head characteristics
Sex of household head (Reference group: Male)
Female
0.65
0.35–1.23
0.29
0.11–0.79
0.09
Age (Reference group: ≤24 years)
25–29 years
0.71
0.27–1.90
1.29
0.35–4.72
0.16
30–45 years
0.92
0.38–2.22
2.16
0.65–7.17
46–59 years
1.92
0.68–5.42
1.79
0.41–7.86
60+ years
1.92
0.59–6.24
0.42
0.04–4.18
Education (Reference group: Primary level or less)
Secondary education or higher
1.80
1.05–3.11
3.23
1.47–7.08
0.84
Employment type (Reference group: Paid employees)
Self-employed
0.95
0.55–1.66
0.25
0.11–0.57
0.03
No fixed employment
1.15
0.35–3.75
0.55
0.14–2.25
Remuneration (Reference group: Fixed salary)
Daily/hourly pay
0.70
0.29–1.71
0.20
0.07–0.53
0.19
Task-based payment
0.52
0.23–1.14
0.15
0.05–0.44
Business profits
0.76
0.33–1.76
0.32
0.11–0.95
Marital status (Reference group: Unmarried)
Married
1.87
0.96–3.64
3.46
1.26–9.51
0.57
Presence of chronic illness in household
0.65
0.17–2.44
0.35
0.43–2.84
0.46
*The p-values assess whether the association between each variable and NHIF membership is significantly different for the ROSCA and non-ROSCA households
The odds in the reference group are multiplied by the odds ratio for each category studies across sub-Saharan Africa that have identified a
positive correlation between health insurance demand
and a household head’s educational or asset-based so-
cioeconomic status [10, 17, 25–27]. secure forms of financial risk protection in health. Be-
fore probing the associations between ROSCA mem-
bership and health insurance enrolment, it is first
necessary to examine the overarching determinants of
NHIF demand within the general population sample. According to our findings, the NHIF has been success-
ful in providing inclusive health insurance coverage re-
gardless of household composition or health status. This may be attributed to the NHIF’s benefit structure,
which enables a household to enrol all its identified de-
pendants under a fixed premium irrespective of their
individual circumstances. Discussion While successful in expand-
ing coverage amongst the above-mentioned groups in
our study area, we found that the NHIF still faces chal-
lenges in increasing health insurance demand amongst
key vulnerable population groups, including those with
low levels of education and limited asset-based wealth. These findings reflect similar national health financing Viewed through the prism of ROSCA membership,
our results inferred that sociocultural disenfranchise-
ment due to female household headship reduced NHIF
demand amongst those not involved in local ROSCAs. However, further analysis revealed that these sociocul-
tural constraints on NHIF enrolment were limited
amongst ROSCA members. With this in mind, it is ne-
cessary to examine how ROSCA membership changes
the terms under which households engage with national
power structures. According to Mladovsky and Mossialos’ social capital
framework, bridging capital is vital in building norms
and rules to facilitate productive behaviour in individuals
[30]. Amongst these norms is the sharing of information Oraro and Wyss International Journal for Equity in Health (2018) 17:170 Page 7 of 9 Page 7 of 9 strategies as collective, as well as effectively identifying
households at the bottom of the pyramid. and resources amongst socially heterogeneous groups,
in order to create a fair market. In the context of ROS-
CAs, regular social and economic interaction mini-
mizes the information and power asymmetry witnessed
in the general population. ROSCAs thus challenge the
power structures that are inherent within our study
area by developing a common identity under which het-
erogeneous groups interact. This implies that ROSCAs
may foster social cohesion amongst its members, osten-
sibly by arming different sociocultural groups with the
information and financial resources to better under-
stand and afford NHIF membership. py
Given the indefinite nature of social capital, care must
be taken in viewing the associations reported in this study
as causal. We must also acknowledge that the structures
of ROSCAs and their relationships with stocks of social
capital will vary considerably. We note that ROSCA mem-
bership may be influenced by several non-observable fac-
tors, such as time-availability and self-selection, which
directly impact one’s ability to participate in ROSCAs. We
further acknowledge that the use of univariate logistic re-
gression in our study limits our ability to control for some
key confounders within our analysis. Discussion We therefore reiter-
ate that the relationship between ROSCA membership
and its associated social capital, NHIF enrolment and the
explanatory variables is associative rather than causative
within the context of our study. Further, the urban setting
of this study limits its external validity in rural areas. Nevertheless, this article provides an insight into the
transmutability of social capital from the local level to na-
tional level. It also provides an insight into how an
in-depth understanding of the social and economic char-
acteristics of target populations can feed into in the design
of national-level schemes. Our study has several design
limitations that may affect the external validity of our
study: the study took place in the transitory period during
which changes to NHIF rates and benefits for the informal
sector were being negotiated. Due to this context, the
study’s applicability may be affected by the new NHIF vol-
untary health insurance regime. p
Based on our findings, however, it is, notable from our
findings that ROSCA membership does not exclusively
negate the influence of sociocultural biases on national
health insurance demand. Cognisant of this, we postu-
late that the social alliances within ROSCAs are ultim-
ately
subservient
to
the
economic
power
held
by
individuals. It is generally accepted that many ROSCAs
exclude those unable to pay their dues during each pay-
ment cycle [13]. This means that the wealthy either
self-select into the ROSCAs, or ultimately exclude their
poorer counterparts due to non-payment of fees [2]. This leads to the pervasive exclusion of the economically
marginalised, thus disempowering actors with weak eco-
nomic agency. The role of potential ostracism in limiting
the bridging social capital in ROSCAs is observed by the
fact that the wealthier business owners in our study
setting, either due to high household head income or
multiple
earners,
were over-represented
in
ROSCA
households. It is for this reason that we found that
small-scale business owners were more likely to be en-
rolled into the NHIF if they had ROSCA membership. Competing interests Competing interests
The authors declare that they have no competing interests. 18. Jehu-Appiah C, Aryeetey G, Spaan E, de Hoop T, Agyepong I, Baltussen R. Equity aspects of the National Health Insurance Scheme in Ghana: who is
enrolling, who is not and why? Soc Sci Med. 2011;72:157–65. https://doi. org/10.1016/j.socscimed.2010.10.025. Abbreviations governance, and competition. Lancet. 2017;390:882–97. https://doi.org/10. 1016/S0140-6736(17)31280-1. governance, and competition. Lancet. 2017;390:882–97. https://doi.org/10. 1016/S0140-6736(17)31280-1. DHS: Demographic and Health Survey; FMS: Free Maternity Services;
HISP: Health Insurance Subsidy Programme for the Poor; NHIF: National
Hospital Insurance Fund; NSSF: National Social Security Fund; ODK: Open
Data Kit; ROSCAs: Rotating Savings and Credit Associations;
SES: Socioeconomic Status; UHC: Universal Health Coverage 7. Central Bank of Kenya, Kenya National Bureau of Statistics, FSD Kenya, 2016. The 2016 FinAccess Household Survey on financial inclusion. FSD Kenya,
Nairobi, Kenya. 8. Coleman, J.S., 1990. The foundations of social theory. 8. Coleman, J.S., 1990. The foundations of social theory. 9. Demirguc-Kunt, A., Klapper, L., Singer, D., Van Oudheusden, P., 2015. The
Global Findex Database 2014: Measuring Financial Inclusion around the
World (Policy Research Working Paper No. 7255). World Bank Group,
Washington, D.C. Received: 15 March 2018 Accepted: 9 November 2018 Received: 15 March 2018 Accepted: 9 November 2018 24. Kimani JK, Ettarh R, Kyobutungi C, Mberu B, Muindi K. Determinants for
participation in a public health insurance program among residents of
urban slums in Nairobi, Kenya: results from a cross-sectional survey. BMC
Health Serv Res. 2012;12(66). https://doi.org/10.1186/1472-6963-12-66. References
1. Adebayo EF, Uthman OA, Wiysonge CS, Stern EA, Lamont KT, Ataguba JE. A
systematic review of factors that affect uptake of community-based health
insurance in low-income and middle-income countries. BMC Health Serv
Res. 2015;15(543). https://doi.org/10.1186/s12913-015-1179-3. 2. Anderson S, Baland J-M, Moene KO. Enforcement in informal saving groups. J Dev Econ. 2009;90:14–23. https://doi.org/10.1016/j.jdeveco.2008.09.003. 3. Biggart, N., 2001. Banking on each other: the situational logic of rotating
savings and credit associations. 4. Bourdieu P. The forms of capital, in: handbook of theory and research for
the sociology of education. Westport, CT: Greenwood; 1986. p. 241–58. 5. Buigut S, Ettarh R, Amendah DD. Catastrophic health expenditure and its
determinants in Kenya slum communities. Int J Equity Health. 2015;14:46. 6. Busse R, Blümel M, Knieps F, Bärnighausen T. Statutory health insurance in
Germany: a health system shaped by 135 years of solidarity, self- Availability of data and materials
Please contact author for data requests. Please contact author for data requests. Conclusion This study examined the influence of ROSCA member-
ship and social capital on voluntary health insurance
enrolment. We posit that socially disadvantaged groups
may accrue strong bridging social capital by participat-
ing in ROSCAs in resource-poor areas. However, we
suggest that the full benefits of social capital in increas-
ing health insurance demand can only be accrued by
taking into account the underlying economic situation
of the target population. We note that the educational
and financial situation of a household serves as a strong
indicator of its capacity to make investment decisions,
including signing up to a voluntary health insurance
scheme. It is therefore important to anchor outreach to
ROSCAs within a broader, multi-pronged approach that
targets households within their social, economic and
political realities. These findings are particularly rele-
vant to countries that may seek to increase voluntary
health insurance coverage through local associations: it
is important to understand the demographic challenges
of these groups and utilise several strategies and target-
ing mechanisms to ensure equitable access to health in-
surance for all citizens. It is important to view the findings of this manuscript
within the context of the NHIF’s broader Universal
Health Coverage (UHC) strategy in Kenya. Since 2015,
the Kenyan Government has provided fully subsidised
NHIF coverage to indigents across the country through
the Health Insurance Subsidy Programme for the Poor
(HISP). This scheme has set the ambitious goal of expand-
ing health insurance coverage to 9 million indigents across
the country by 2020 [28]. Additionally, in 2016, the Ken-
yan Government began offering free maternity services
(FMS) to all Kenyan women through the NHIF (Kenya
National [23]). With these programmes in mind, the
NHIF’s ROSCA targeting strategy would ostensibly be tar-
geted towards a narrow group of informal sector workers
who have some degree of economic agency. With this nu-
anced lens, the NHIF’s strategy of targeting the informal
sector through ROSCAs may be feasible, given their alter-
native means of targeting those at the bottom of the pyra-
mid through premium subsidisation. However, the success
of these strategies will be highly dependent upon the
successful planning and execution of these collaborative Page 8 of 9 Oraro and Wyss International Journal for Equity in Health (2018) 17:170 Authors’ contributions
d
d
d
ll 14. Fadlallah R, El-Jardali F, Hemadi N, Morsi RZ, Abou Samra CA, Ahmad A, Arif K,
Hishi L, Honein-AbouHaidar G, Akl EA. Barriers and facilitators to
implementation, uptake and sustainability of community-based health
insurance schemes in low- and middle-income countries: a systematic review. Int J Equity Health. 2018;17(13). https://doi.org/10.1186/s12939-018-0721-4. TO designed and collected the data for the study. TO completed data
analysis and drafted the manuscript for publication. KW helped draft and
review the manuscript for intellectual content. Both authors read and
approved the final version. 15. Fenenga CJ, Nketiah-Amponsah E, Ogink A, Arhinful DK, Poortinga W,
Hutter I. Social capital and active membership in the Ghana National Health
Insurance Scheme - a mixed method study. Int J Equity Health. 2015;
14(118). https://doi.org/10.1186/s12939-015-0239-y. Publisher’s Note 19. Karanja I. An enumeration and mapping of informal settlements in Kisumu,
Kenya, implemented by their inhabitants. Environ Urban. 2010;22:217–39. https://doi.org/10.1177/0956247809362642. Springer Nature remains neutral with regard to jurisdictional claims in
published maps and institutional affiliations. 20. Kenya National Bureau of Statistics, 2017. Economic Survey 2017. Government of Kenya. Funding Data collection was financed by the Swiss Tropical and Public Health
Institute (Swiss TPH). TO is a recipient of the Swiss Government Excellence
Scholarship (ESKAS), provided by the Swiss Federal Commission for
Scholarships for Foreign Students (FCS). 12. Dupas P, Robinson J. Savings constraints and microenterprise development:
evidence from a field experiment in Kenya. Am Econ J Appl Econ. 2013;5:
163–92. 13. Edoardo, T., 2013. Exploring the black-box of economic informality: social
networks and institutional change among micro and small Enterprises in
Nairobi, Kenya, in: securing livelihoods. Oxford University Press, Oxford
https://doi.org/10.1093/acprof:oso/9780199687015.003.0010. Acknowledgements The authors would like to thank the Kisumu County team for their
cooperation on this project. The author would also like to thank the
Assistant Chiefs of Manyatta A, Manyatta B, Nyalenda, and Obunga for
providing consent for data collection in their areas. The authors
acknowledge Pamoja Trust for availing the Kisumu informal settlement
enumeration maps used for sampling. Finally, the authors also thank Dr. Amanda Ross (Swiss TPH) for statistical support provided. 10. Dixon J, Luginaah I, Mkandawire P. The National Health Insurance Scheme
in Ghana’s upper west region: a gendered perspective of insurance
acquisition in a resource-poor setting. Soc Sci Med. 2014;1982(122):103–12. https://doi.org/10.1016/j.socscimed.2014.10.028. 11. Donfouet HPP, Makaudze E, Mahieu P-A, Malin E. The determinants of the
willingness-to-pay for community-based prepayment scheme in rural
Cameroon. Int J Health Care Finance Econ. 2011;11:209. https://doi.org/10. 1007/s10754-011-9097-3. Consent for publication
Not applicable. Competing interests
The authors declare that they have no competing interests. Ethics approval and consent to participate Ethical approval for this study was obtained through the Institutional Review
Board (IRB) of Strathmore University, Kenya (Ethics Reference No. SU-IRB
0057/16; Date of approval: 15th June 2016). Written consent was acquired
from all respondents. 0057/16; Date of approval: 15th June 2016). Written consent was acquired
from all respondents. 16. Hayes RJ, Bennett S. Simple sample size calculation for cluster-randomized
trials. Int J Epidemiol. 1999;28:319–26. https://doi.org/10.1093/ije/28.2.319. 17. Jehu-Appiah C, Aryeetey G, Agyepong I, Spaan E, Baltussen R. Household
perceptions and their implications for enrolment in the National Health
Insurance Scheme in Ghana. Health Policy Plan. 2012;27:222–33. https://doi. org/10.1093/heapol/czr032. References References
1. Adebayo EF, Uthman OA, Wiysonge CS, Stern EA, Lamont KT, Ataguba JE. A
systematic review of factors that affect uptake of community-based health
insurance in low-income and middle-income countries. BMC Health Serv
Res. 2015;15(543). https://doi.org/10.1186/s12913-015-1179-3. 2. Anderson S, Baland J-M, Moene KO. Enforcement in informal saving groups. J Dev Econ. 2009;90:14–23. https://doi.org/10.1016/j.jdeveco.2008.09.003. 3. Biggart, N., 2001. Banking on each other: the situational logic of rotating
savings and credit associations. 4. Bourdieu P. The forms of capital, in: handbook of theory and research for
the sociology of education. Westport, CT: Greenwood; 1986. p. 241–58. 5. Buigut S, Ettarh R, Amendah DD. Catastrophic health expenditure and its
determinants in Kenya slum communities. Int J Equity Health. 2015;14:46. 6. Busse R, Blümel M, Knieps F, Bärnighausen T. Statutory health insurance in
Germany: a health system shaped by 135 years of solidarity, self- 1. Adebayo EF, Uthman OA, Wiysonge CS, Stern EA, Lamont KT, Ataguba JE. A
systematic review of factors that affect uptake of community-based health
insurance in low-income and middle-income countries. BMC Health Serv
Res. 2015;15(543). https://doi.org/10.1186/s12913-015-1179-3. 25. Kimani JK, Ettarh R, Warren C, Bellows B. Determinants of health insurance
ownership among women in Kenya: evidence from the 2008-09 Kenya
demographic and health survey. Int J Equity Health. 2014;13:27. https://doi. org/10.1186/1475-9276-13-27. Res. 2015;15(543). https://doi.org/10.1186/s12913-015-1179-3. 2. Anderson S, Baland J-M, Moene KO. Enforcement in informal saving groups. J Dev Econ. 2009;90:14–23. https://doi.org/10.1016/j.jdeveco.2008.09.003. 3. Biggart, N., 2001. Banking on each other: the situational logic of rotating
savings and credit associations. 4. Bourdieu P. The forms of capital, in: handbook of theory and research for
the sociology of education. Westport, CT: Greenwood; 1986. p. 241–58. 5. Buigut S, Ettarh R, Amendah DD. Catastrophic health expenditure and its
determinants in Kenya slum communities. Int J Equity Health. 2015;14:46. 6. Busse R, Blümel M, Knieps F, Bärnighausen T. Statutory health insurance in
Germany: a health system shaped by 135 years of solidarity, self- p
g
2. Anderson S, Baland J-M, Moene KO. Enforcement in informal saving groups. J Dev Econ. 2009;90:14–23. https://doi.org/10.1016/j.jdeveco.2008.09.003. 2. Anderson S, Baland J-M, Moene KO. Enforcement in informal saving groups. J Dev Econ. 2009;90:14–23. https://doi.org/10.1016/j.jdeveco.2008.09.003. 26. Macha J, Kuwawenaruwa A, Makawia S, Mtei G, Borghi J. Determinants of
community health fund membership in Tanzania: a mixed methods analysis. BMC Health Serv Res. 2014;14(538). https://doi.org/10.1186/s12913-014-0538-9. 3. Biggart, N., 2001. Banking on each other: the situational logic of rotating
savings and credit associations. Author details
1S i
T
i
l 1Swiss Tropical and Public Health Institute, Basel, Switzerland. 2University of
Basel, Basel, Switzerland. 3Department of Epidemiology and Public Health,
Health Systems Support Unit, Swiss Tropical and Public Health Institute, P.O. Box 4002, Basel, Switzerland. y
21. Kenya National Bureau of Statistics, 2015. Kenya Wealth Index. 21. Kenya National Bureau of Statistics, 2015. Kenya Wealth Index. 22. Kenya National Bureau of Statistics, 2010. The 2009 Kenya Population and
Housing Census. Kenya National Bureau of Statistics, Nairobi, Kenya. 23. Kenya National Treasury, 2017. Budget statement for the fiscal year 2017/
2018. Government of Kenya. References 3. Biggart, N., 2001. Banking on each other: the situational logic of rotating
savings and credit associations. 27. Mebratie AD, Sparrow R, Yilma Z, Alemu G, Bedi AS. Enrollment in Ethiopia’s
community-based health insurance scheme. World Dev. 2015;74:58–76. https://doi.org/10.1016/j.worlddev.2015.04.011. 28. Ministry of Health, 2016. Health Sector Working Group Report: Medium
Term Expenditure Framework for the Period 2017–18 to 2019–20. Government of Kenya, Nairobi. Page 9 of 9 Oraro and Wyss International Journal for Equity in Health (2018) 17:170 29. Ministry of Health, 2013. Kenya Household Health Expenditure and
Utilisation Survey (KHHEUS) 2013 Questionnaire. 29. Ministry of Health, 2013. Kenya Household Health Expenditure and
Utilisation Survey (KHHEUS) 2013 Questionnaire. 30. Mladovsky P, Mossialos E. A conceptual framework for community-based
health insurance in low-income countries: social capital and economic
development. World Dev. 2008;36:590–607. 30. Mladovsky P, Mossialos E. A conceptual framework for community-based
health insurance in low-income countries: social capital and economic
development. World Dev. 2008;36:590–607. 31. Mladovsky P, Soors W, Ndiaye P, Ndiaye A, Criel B. Can social capital help
explain enrolment (or lack thereof) in community-based health insurance? Results of an exploratory mixed methods study from Senegal. Soc Sci Med. 2014;101:18–27. https://doi.org/10.1016/j.socscimed.2013.11.016. 32. National Council for Law Reporting, 1989. The National Social Security Fund
act, in: Laws of Kenya. The Republic of Kenya. 32. National Council for Law Reporting, 1989. The National Social Security Fund
act, in: Laws of Kenya. The Republic of Kenya. 33. National Hospital Insurance Fund, 2017a. Expanding Social Health Insurance
Coverage among the Informal Sector: Strategy 2016–2018. Government of
Kenya, Nairobi, Kenya. 34. Putnam RD. What makes democracy work? Natl Civ Rev. 1993;82:101–7. https://doi.org/10.1002/ncr.4100820204. 34. Putnam RD. What makes democracy wo
https://doi.org/10.1002/ncr.4100820204. 34. Putnam RD. What makes democracy work? Natl Civ Rev. 1993;82:101–7. https://doi.org/10.1002/ncr.4100820204. 34. Putnam RD. What makes democracy wo
https://doi.org/10.1002/ncr.4100820204. 35. Szreter S, Woolcock M. Health by association? Social capital, social theory,
and the political economy of public health. Int J Epidemiol. 2004;33:650–67. https://doi.org/10.1093/ije/dyh013. 35. Szreter S, Woolcock M. Health by association? Social capital, social theory,
and the political economy of public health. Int J Epidemiol. 2004;33:650–67. https://doi.org/10.1093/ije/dyh013. 36. Tatara, K., Okamoto, E., 2009. Japan: Health system review, in: Health
Systems in Transition. pp. 1–164. 36. Tatara, K., Okamoto, E., 2009. Japan: Health system review, in: Health
Systems in Transition. pp. 1–164. 37. UNESCO, 2006. Understandings of literacy, in: education for all global
monitoring report 2006. UNESCO Publishing, Paris. 37. UNESCO, 2006. References Understandings of literacy, in: education for all global
monitoring report 2006. UNESCO Publishing, Paris. 38. United Nations General Assembly, 2012. Global Health and foreign policy. Agenda Item 123. 38. United Nations General Assembly, 2012. Global Health and foreign policy. Agenda Item 123. 39. World Bank, JICA, The Global Fund, African Development Bank, WHO, 2016. Universal health coverage in Africa: A framework for action, in: Joint 2016
TICAD Conference Working Paper. 40. World Health Organization, 2010. The world health report: health systems
financing: the path to universal coverage. World Health Organization,
Geneva. 40. World Health Organization, 2010. The world health report: health systems
financing: the path to universal coverage. World Health Organization,
Geneva. 41. World Health Organization, International Bank for Reconstruction and
Development/World Bank, 2017. Tracking Universal Health Coverage: 2017
Global Monitoring Report. 42. Zollman J. Kenya financial diaries: Shilingi kwa shilingi, the financial lives of
the poor. Nairobi, Kenya: FSD Kenya; 2014. 42. Zollman J. Kenya financial diaries: Shilingi kwa shilingi, the financial lives of
the poor. Nairobi, Kenya: FSD Kenya; 2014.
|
https://openalex.org/W4206209813
|
https://www.vektornaukitech.ru/jour/article/download/178/164
|
Russian
| null |
THE INFLUENCE OF BORON CARBIDE ADDITIVE ON THE STRUCTURE AND HARDNESS OF A NICKEL-BASED COATING
|
Frontier materials & technologies
| 2,021
|
cc-by
| 5,539
|
Влияние добавки карбида бора на структуру
и твердость никелевого покрытия Влияние добавки карбида бора на структуру
и твердость никелевого покрытия
© 2021
Старикова Ульяна Сергеевна*1,4, младший научный сотрудник
Соболева Наталья Николаевна1,2,5, кандидат технических наук, старший научный сотрудник
Макаров Алексей Викторович1,6, член-корреспондент РАН, доктор технических наук,
заведующий отделом материаловедения и лабораторией механических свойств
Харанжевский Евгений Викторович3,7, доктор технических наук, профессор,
заведующий лабораторией
1Институт физики металлов имени М.Н. Михеева Уральского отделения Российской академии наук,
Екатеринбург (Россия)
2Институт машиноведения Уральского отделения Российской академии наук, Екатеринбург (Россия)
3Удмуртский государственный университет, Ижевск (Россия)
*E-mail: ulstar97@mail.ru
4ORCID: https://orcid.org/0000-0001-6960-0619
5ORCID: https://orcid.org/0000-0002-7598-2980
6ORCID: https://orcid.org/0000-0002-2228-0643
7ORCID: https://orcid.org/0000-0002-1525-2169 *E-mail: ulstar97@mail.ru Аннотация: Лазерная наплавка все чаще применяется в различных отраслях машиностроения, так как по
сравнению с традиционными методами создания покрытий она обладает такими преимуществами, как высокие
скорости нагрева и охлаждения, минимальное перемешивание основного и наплавляемого материала. Нанесенные
лазером покрытия обычно характеризуются мелкозернистой структурой и минимальной зоной термического
влияния. Применение покрытий, сформированных из порошков системы Ni–Cr–B–Si, также очень распространено
в промышленности, так как эти покрытия обладают хорошей стойкостью к износу, коррозии, эрозии и т. д. К ука-
занной группе порошков для улучшения свойств наплавляемого покрытия добавляются различные упрочняющие
частицы. В качестве таких частиц могут выступать карбиды бора, обладающие высокой твердостью и термодина-
мической устойчивостью, а также высокими показателями сопротивления изнашиванию. В работе исследовано
влияние добавки 7 масс. % карбида бора В4С на структуру и твердость NiCrBSi покрытия, сформированного ла-
зерной наплавкой из порошка марки ПГ-СР2 на поверхности стали 30ХРА. В работе проведены микроскопиче-
ские исследования структуры NiCrBSi и NiCrBSi–В4С покрытий с использованием растрового электронного мик-
роскопа, приведены результаты рентгеноспектрального микроанализа. Показано, что структура обоих покрытий
в наплавленном состоянии характеризуется однородностью и мелкозернистостью. Выявлено, что образцы
с NiCrBSi и NiCrBSi–В4С покрытиями имеют узкую переходную зону от покрытия к основному металлу. Приве-
дены результаты измерения микротвердости покрытий, свидетельствующие о снижении микротвердости сформи-
рованных лазером никелевых покрытий при добавке карбида бора. р
р
р
р
ва: лазерная наплавка; карбид бора B4C; порошок ПГ-СР2; NiCrBSi покрытие; NiCrBSi–B4C по-
рдость; растровая электронная микроскопия. р
р
р
р
р
р
Ключевые слова: лазерная наплавка; карбид бора B4C; порошок ПГ-СР2; NiCrBSi покрытие
крытие; микротвердость; растровая электронная микроскопия. Благодарности: Работа выполнена в рамках государственных заданий ИФМ УрО РАН по темам
№ 121102900049-1 и № АААА-А18-118020190116-6 и ИМАШ УрО РАН по теме № АААА-А18-118020790147-4. Экспериментальные исследования выполнены на оборудовании ЦКП «Пластометрия» ИМАШ УрО РАН. Статья подготовлена по материалам докладов участников X Международной школы «Физическое материало-
ведение» (ШФМ-2021), Тольятти, 13–17 сентября 2021 года. Влияние добавки карбида бора на структуру
и твердость никелевого покрытия Для цитирования: Старикова У.С., Соболева Н.Н., Макаров А.В., Харанжевский Е.В. Влияние добавки кар-
бида бора на структуру и твердость никелевого покрытия // Frontier Materials & Technologies. 2021. № 4. С. 89–97. DOI: 10.18323/2782-4039-2021-4-89-97. лазерную наплавку для ремонта дорогостоящих изде-
лий, при восстановлении сравнительно небольших де-
фектных участков. лазерную наплавку для ремонта дорогостоящих изде-
лий, при восстановлении сравнительно небольших де-
фектных участков. Frontier Materials & Technologies. 2021. № 4 УДК 621.793:620.187
doi: 10.18323/2782-4039-2021-4-89-97 УДК 621.793:620.187
doi: 10.18323/2782-4039-2021-4-89-97 УДК 621.793:620.187
doi: 10.18323/2782-4039-2021-4-89-97 Влияние добавки карбида бора на структуру
и твердость никелевого покрытия
© 2021
Старикова Ульяна Сергеевна*1,4, младший научный сотрудник
Соболева Наталья Николаевна1,2,5, кандидат технических наук, старший научный сотрудник
Макаров Алексей Викторович1,6, член-корреспондент РАН, доктор технических наук,
заведующий отделом материаловедения и лабораторией механических свойств
Харанжевский Евгений Викторович3,7, доктор технических наук, профессор,
заведующий лабораторией
1Институт физики металлов имени М.Н. Михеева Уральского отделения Российской академии наук,
Екатеринбург (Россия)
2Институт машиноведения Уральского отделения Российской академии наук, Екатеринбург (Россия)
3Удмуртский государственный университет, Ижевск (Россия)
*E-mail: ulstar97@mail.ru
4ORCID: https://orcid.org/0000-0001-6960-0619
5ORCID: https://orcid.org/0000-0002-7598-2980
6ORCID: https://orcid.org/0000-0002-2228-0643
7ORCID: https://orcid.org/0000-0002-1525-2169 МЕТОДИКА ПРОВЕДЕНИЯ ИССЛЕДОВАНИЙ В качестве материала, на котором проводилась лазер-
ная наплавка, выбрана сталь 30ХРА. Для проведения экс-
перимента были подготовлены образцы толщиной 3 мм. NiCrBSi покрытия, сформированные лазерной на-
плавкой, могут быть применены для предотвращения
износа, коррозии, эрозии и окисления поверхности де-
талей при высоких температурах [3–5]. В литературе
отмечается положительное влияние бора в составе
NiCrBSi покрытий в связи с формированием боридов
в структуре покрытий, в результате чего повышается
твердость покрытий, а также их стойкость к окислению
и коррозии при повышенных температурах [6; 7]. В качестве материала покрытия был использован са-
мофлюсующийся порошок системы NiCrBSi марки ПГ-
СР2 (химический состав порошка: Cr – 14,8 масс. %,
B – 2,1 масс. %, Si – 2,9 масс. %, C – 0,48 масс. %,
Fe – 2,6 масс. %, Ni – остальное), а также порошок кар-
бида бора B4C. Размер порошка ПГ-СР2 составлял 40–
160 мкм, а порошка B4C – до 40 мкм. На рис. 1 показа-
ны микрофотографии порошков ПГ-СР2 и B4C, полу-
ченные с использованием растровой электронной мик-
роскопии. При формировании покрытий для улучшения экс-
плуатационных характеристик к хромоникелевым по-
рошкам могут быть добавлены различные упрочняю-
щие частицы, например карбиды WC [8], Cr3C2 [9], TiC
[10], SiC [11], оксиды Al2O3 [12] и др. Карбиды бора
B4C могут быть добавлены в виде упрочняющих частиц
в состав порошков при наплавке, так как они обладают
рядом положительных свойств: высокой твердостью,
термодинамической устойчивостью, высокими показа-
телями износостойкости и др. [13]. Боросодержащие
покрытия применяются для упрочнения деталей ма-
шин, эксплуатирующихся в условиях интенсивного
абразивного изнашивания [14]. Из порошка ПГ-СР2 и смеси порошков ПГ-СР2
и 7 % В4С приготавливалась суспензия, содержащая
1%-ный спиртовой раствор камфоры. Суспензия нано-
силась на поверхность стальной основы, лазерная обра-
ботка проводилась после высыхания связующего. Тол-
щина слоя до оплавления составляла 200 мкм. Лазерная наплавка была проведена при напряжении
на лампах накачки 320 В, энергии одного импульса
6 Дж, частоте импульсов 15 Гц, длительности импульса
15 мс, диаметре луча на поверхности образца 1 мм. Рас-
стояние между дорожками составляло 0,8 мм. В литературе имеются сведения о добавлении карби-
дов бора B4C к NiCrBSi порошкам при формировании
покрытий лазерным излучением. Например, в работе
[15] на сталь марки 12Х18Н10Т были наплавлены
NiCrBSi–B4C (5 и 12 об. %) покрытия методом селектив-
ного лазерного плавления. В структуре покрытий отме-
чается образование карбидов и боридов хрома Cr7C3
и CrB. МЕТОДИКА ПРОВЕДЕНИЯ ИССЛЕДОВАНИЙ Средняя величина микротвердости NiCrBSi по-
крытия достигает значения 485 HV 0,3, в NiCrBSi по-
крытии с 5 % В4С она снижается до 437 HV 0,3, а при
добавке 12 % В4С повышается и достигает значения
519 HV 0,3. Микроструктура покрытий и распределение эле-
ментов по поперечному сечению образцов с покры-
тиями были изучены с применением растрового элек-
тронного микроскопа (РЭМ) TESCAN VEGA II XMU
с вольфрамовым катодом, оборудованного энергодис-
персионным (INCA Energy 450 XT) микроанализато-
ром, позволяющим проводить рентгеноспектральный
микроанализ. Микротвердость по методу восстановленного отпе-
чатка была определена на микротвердомере Shimadzu
HMV-G21DT согласно ГОСТ 9450-76 «Измерение мик-
ротвердости вдавливанием алмазных наконечников»
при нагрузке 0,245 Н, скорости нагружения пирамид-
кой Виккерса 40 мкм/с и выдержке под нагрузкой 15 с
как среднее арифметическое из 10 измерений. Погреш-
ность измерения микротвердости оценивали по величи-
не среднеквадратичного отклонения. В работах [16–18] имеются сведения о влиянии добав-
ки В4С (2, 5 и 10 масс. %) на структуру и твердость хро-
моникелевого покрытия, нанесенного на основу Ti-6Al-4V
лазерной наплавкой. Отмечается, что при добавлении
2 % карбида бора В4С значение микротвердости дости-
гает 1200–1400 HV, в то время как значение микро-
твердости в зоне основного металла достигает 330–
340 HV [16]. При добавлении 5 % В4С значение микро-
твердости составляет 1300 HV [17]. В NiCrBSi–В4С
покрытиях происходит формирование упрочняющих
фаз CrB, TiB2, TiC, что связано с нанесением покрытия
на титановую основу [16; 18]. ВВЕДЕНИЕ Лазерная наплавка – один из популярных и востре-
бованных способов нанесения покрытий на сегодняш-
ний день. Лазерная наплавка применяется в аэрокосми-
ческой, автомобильной, энергетической, горнодобы-
вающей, металлообрабатывающей и других отраслях
промышленности, где используемые машины и меха-
низмы подвергаются воздействию агрессивных сред и
поверхностному износу. В связи с высокой себестоимо-
стью процесса наиболее целесообразно использовать Лазерная наплавка обладает рядом преимуществ по
сравнению с традиционными методами. Так, она харак-
теризуется высокими скоростями нагрева и охлажде-
ния, минимальным перемешиванием основного и на-
плавляемого материала. Это позволяет формировать
покрытия с минимальной зоной термического влияния
и прочной металлургической связью между наплавлен-
ным слоем и основой [1; 2]. При данном методе нанесения 89 Старикова У.С., Соболева Н.Н., Макаров А.В. и др. «Влияние добавки карбида бора на структуру и твердость никелевого…» Старикова У.С., Соболева Н.Н., Макаров А.В. и др. «Влияние добавки карбида бора на структуру и тве покрытий формируются равноосные мелкокристалли-
ческие структуры наплавленного металла [1]. РЕЗУЛЬТАТЫ ИССЛЕДОВАНИЙ На рис. 2 показаны электронные изображения мик-
роструктуры NiCrBSi и NiCrBSi–В4С покрытий, сфор-
мированных лазерной наплавкой. Толщина NiCrBSi
покрытия составляет 140–190 мкм, а толщина NiCrBSi–
В4С покрытия – 120–150 мкм. На рис. 3 и рис. 4 приведены результаты количественного
рентгеноспектрального
микроанализа
NiCrBSi
и NiCrBSi–В4С покрытий соответственно. В никелевом
покрытии (рис. 3) содержание Ni составляет примерно от
30 до 60 масс. %, Cr – от 7 до 12 масс. %, B – до 7 масс. %,
Si – от 2 до 9 масс. %, С – до 5 масс. %, Fe – от 20
до 60 масс. %. В никелевом покрытии с добавлением
карбида бора NiCrBSi–В4С (рис. 4) содержание Ni
составляет примерно от 20 до 60 масс. %, Cr – от 5
до 13 масс. %, B – до 4,5 масс. %, Si – от 2 до 4 масс. %,
С – до 1,2 масс. %, Fe – от 15 до 70 масс. %. Благодаря лазерной наплавке появляется возмож-
ность использовать недорогие и недефицитные марки
сталей и сплавов, при этом создавая износостойкие
слои на поверхности изделий [19]. Однако материал
подложки влияет на свойства формируемого на ней
покрытия. Ранее были упомянуты работы, где в качест-
ве материала основы применялись нержавеющая сталь
марки 12Х18Н10Т [15] и титановый сплав Ti-6Al-4V
[16–18]. В данной работе материалом основы выступает
сталь 30ХРА. Такая сталь может быть применена для
изготовления осей, рычагов, болтов, гаек и др. Цель работы – исследование влияния добавки
7 масс. % В4С на структуру и твердость NiCrBSi по-
крытия, сформированного лазерным излучением на
поверхности стали 30ХРА. На рис. 3 и 4 видно резкое снижение количества Ni
в поперечном сечении образцов с NiCrBSi и NiCrBSi– На рис. 3 и 4 видно резкое снижение количества Ni
в поперечном сечении образцов с NiCrBSi и NiCrBSi– Frontier Materials & Technologies. 2021. № 4 90 Старикова У.С., Соболева Н.Н., Макаров А.В. и др. «Влияние добавки карбида бора на структуру и твердость никелевого…» a
b
Рис. 1. Микрофотографии порошков ПГ-СР2 (a) и В4С (b)
Fig. 1. Microphotographs of the PG-SR2 (a) and В4С (b) powders a b b a Рис. 1. Микрофотографии порошков ПГ-СР2 (a) и В4С (b)
Fig. 1. Microphotographs of the PG-SR2 (a) and В4С (b) powders a
b
Рис. 2. Поперечное сечение покрытий: а – NiCrBSi; b – NiCrBSi–В4С
Fig. 2. Cross-sectional views of coatings: а – NiCrBSi; b – NiCrBSi–В4С b b a Рис. 2. Поперечное сечение покрытий: а – NiCrBSi; b – NiCrBSi–В4С
Fig. 2. РЕЗУЛЬТАТЫ ИССЛЕДОВАНИЙ Cross-sectional views of coatings: а – NiCrBSi; b – NiCrBSi–В4С Frontier Materials & Technologies. 2021. № 4 Frontier Materials & Technologies. 2021. № 4 ОБСУЖДЕНИЕ РЕЗУЛЬТАТОВ Микроструктура (РЭМ) покрытий: a – NiCrBSi; b – NiCrBSi–В4С
Fig. 5. Microstructure (SEM) of coatings: a – NiCrBSi; b – NiCrBSi–В4С Таблица 1. Микротвердость NiCrBSi и NiCrBSi–В4С покрытий
Table 1. Microhardness of NiCrBSi and NiCrBSi–В4С coatings
Покрытие
HV 0,025
NiCrBSi
620±100
NiCrBSi–В4С
530±140 Таблица 1. Микротвердость NiCrBSi и NiCrBSi–В4С покрытий
Table 1. Microhardness of NiCrBSi and NiCrBSi–В4С coatings ние структурных составляющих в покрытии с карбидом
бора (рис. 5 b), очевидно, связано с тем, что мелкие
частицы карбида бора выступают в качестве модифика-
тора при наплавке. Подобный эффект наблюдался в ра-
боте [21], в которой рассмотрена структура металла,
наплавленного порошковой проволокой с алюминиевой
оболочкой. При введении в состав шихты порошковой
проволоки K2TiF6 и K2ZrF6 происходило модифициро-
вание наплавленного металла, вследствие чего он ха-
рактеризовался мелкозернистой структурой и равно-
мерным распределением избыточных фаз, что позволи-
ло повысить износостойкость покрытия. Кроме того,
в работе [22] в результате электрошлаковой сварки
введение наноструктурированных компонентов в виде
порошка карбонитрида титана в сварочную ванну спо-
собствовало измельчению зерна в структуре зоны тер-
мического влияния, а также улучшению свойств метал-
ла шва [22; 23]. ние структурных составляющих в покрытии с карбидом
бора (рис. 5 b), очевидно, связано с тем, что мелкие
частицы карбида бора выступают в качестве модифика-
тора при наплавке. Подобный эффект наблюдался в ра-
боте [21], в которой рассмотрена структура металла,
наплавленного порошковой проволокой с алюминиевой
оболочкой. При введении в состав шихты порошковой
проволоки K2TiF6 и K2ZrF6 происходило модифициро-
вание наплавленного металла, вследствие чего он ха-
рактеризовался мелкозернистой структурой и равно-
мерным распределением избыточных фаз, что позволи-
ло повысить износостойкость покрытия. Кроме того,
в работе [22] в результате электрошлаковой сварки
введение наноструктурированных компонентов в виде
порошка карбонитрида титана в сварочную ванну спо-
собствовало измельчению зерна в структуре зоны тер-
мического влияния, а также улучшению свойств метал-
ла шва [22; 23]. в никелевом покрытии (38 масс. %). Известно, что час-
тицы В4С взаимодействуют с железом. Бор свободно
диффундирует в железе в связи с его малым атомным
радиусом (0,091 нм). Кроме того, при повышении тем-
пературы бор взаимодействует с большинством хими-
ческих элементов с образованием боридов, в то время
как в обычных условиях для бора характерна слабая
химическая активность. Такое взаимодействие способ-
ствует формированию боридов железа FeB и Fe2B
в структуре покрытий в процессе наплавки, что описа-
но в работах [14; 20]. Очевидно, что такое взаимодейст-
вие является причиной повышенного содержания желе-
за в структуре NiCrBSi–В4С покрытия. ОБСУЖДЕНИЕ РЕЗУЛЬТАТОВ В4С покрытиями, наплавленными лазером, на глу-
бине примерно 140 мкм и 150 мкм от поверхности со-
ответственно, и повышение содержания Fe. На микрофотографиях (рис. 2) видно, что покрытия
однородные, характеризуются отсутствием макропор
и включений, в том числе в переходной зоне. На рис. 5 показана микроструктура покрытий
NiCrBSi и NiCrBSi–В4С, полученная растровой элек-
тронной микроскопией. Количественный рентгеноспектральный микроанализ
показал, что исследуемые покрытия по всей толщине
характеризуются достаточно равномерным распределе-
нием элементов в структуре покрытий, что следует из
рис. 3 и рис. 4. При этом следует отметить, что содержа-
ние Fe в покрытии с карбидом бора больше (в среднем
по поперечному сечению покрытия 51 масс. %), чем Среднее значение микротвердости NiCrBSi покры-
тия, сформированного лазерной наплавкой, составляет
620 HV 0,025 (таблица 1). При добавлении 7 масс. % по-
рошка В4С среднее значение микротвердости сформиро-
ванного лазером покрытия снижается до 530 HV 0,025. 91 Старикова У.С., Соболева Н.Н., Макаров А.В. и др. «Влияние добавки карбида бора на структуру и твердость никелевого…» Рис. 3. Распределение элементов в поперечном сечении образцов
с NiCrBSi покрытием в наплавленном состоянии
Fig. 3. Elements distribution in the cross section of samples
with the NiCrBSi clad coating
92
Frontier Materials & Technologies. 2021. №4 Рис. 3. Распределение элементов в поперечном сечении образцов
с NiCrBSi покрытием в наплавленном состоянии
Fig. 3. Elements distribution in the cross section of samples
with the NiCrBSi clad coating
92
Frontier Materials & Technologies. 2021. № 4 Рис. 3. Распределение элементов в поперечном сечении образцов
с NiCrBSi покрытием в наплавленном состоянии
Fig. 3. Elements distribution in the cross section of samples
with the NiCrBSi clad coating Frontier Materials & Technologies. 2021. № 4 92 Старикова У.С., Соболева Н.Н., Макаров А.В. и др. «Влияние добавки карбида бора на структуру и твердость никелевого…» Рис. 4. Распределение элементов в поперечном сечении образцов
с NiCrBSi–В4С покрытием в наплавленном состоянии
Fig. 4. Elements distribution in the cross section of samples
with the NiCrBSi–В4С clad coating Рис. 4. Распределение элементов в поперечном сечении образцов
с NiCrBSi–В4С покрытием в наплавленном состоянии
Fig. 4. Elements distribution in the cross section of samples
with the NiCrBSi–В4С clad coating Frontier Materials & Technologies. 2021. № 4 93 Старикова У.С., Соболева Н.Н., Макаров А.В. и др. «Влияние добавки карбида бора на структуру и твердость никелевого…» a
b
Рис. 5. Микроструктура (РЭМ) покрытий: a – NiCrBSi; b – NiCrBSi–В4С
Fig. 5. Microstructure (SEM) of coatings: a – NiCrBSi; b – NiCrBSi–В4С b a a Рис. 5. Frontier Materials & Technologies. 2021. № 4 СПИСОК ЛИТЕРАТУРЫ 1. Земляков Е., Бабкин К., Корсмик Р., Скляр М., Куз-
нецов М. Перспективы использования технологий
лазерной наплавки для восстановления лопаток
компрессоров газотурбинных деталей // Фотоника. 2016. № 4. С. 10–25. DOI: 10.22184/1993-7296.2016. 13. Li H.J., He Y., Luo P.Y., Fan Y., Yu H., Wang Y.Q.,
He T., Li Z.J., Zhang H.L. Influence of pulse frequency
on corrosion resistance and mechanical properties of Ni-
W/B4C composite coatings // Colloids and Surfaces A:
Physicochemical and Engineering Aspects. 2021. Vol. 629. Article number 127436. DOI: 10.1016/j. colsurfa.2021.127436. 2. Голубовский Е.Н., Паркин А.А., Жаткин С.С. Ре-
монт трещин на секторах лопаток газотурбинного
двигателя методом разделки и лазерной импульсной
наплавки // Известия Самарского научного центра
Российской академии наук. 2020. Т. 22. № 2. С. 107–
112. 14. Кривиженко Д.С., Дробяз Е.А., Зимоглядова Т.А. Особенности структурообразования боросодержа-
щих покрытий, полученных в процессе высокоско-
ростной обработки // Актуальные проблемы в ма-
шиностроении. 2014. № 1. С. 489–492. 3. Соболева Н.Н., Макаров А.В., Малыгина И.Ю. Тех-
нологические аспекты фрикционной обработки по-
крытия ПГ-СР2, сформированного лазерной наплав-
кой // Вектор науки Тольяттинского государствен-
ного университета. 2019. № 3. С. 47–53. DOI: 10.18
323/2073-5073-2019-3-47-53. 15. Golyshev A.A., Filippov A.A. Comparative investiga-
tion of nickel-based metal-ceramic structures with ce-
ramic particles of tungsten and carbides made by the se-
lective laser melting method // Nanoscience and Tech-
nology. 2020. Vol. 11. № 3. Р. 247–257. DOI: 10.1615/
NanoSciTechnolIntJ.2020033784. 4. Гибзун М.С., Макаров А.В., Соболева Н.Н., Малы-
гина И.Ю. Повышение фрикционной теплостойко-
сти хромоникелевого покрытия комбинированной
лазерно-термической обработкой // Master’s journal. 2017. № 1. С. 11–16. 16. Fan X.H., Geng L., Xu B., Li J. Laser Cladding
NiCrBSi+2%B4C Coating on Ti-6Al-4V // Advanced
Materials Research. 2009. Vol. 79-82. Р. 473–476. DOI: 10.4028/www.scientific.net/AMR.79-82.473. 5. Shi B.W., Li T., Wang D., Zhang X.R., Zhang H.C. Investigation on crack behavior of Ni60A alloy coating
produced by coaxial laser cladding // Journal of Materi-
als Science. 2021. Vol. 56. № 23. P. 13323–13336. DOI: 10.1007/s10853-021-06108-5. 17. Meng Q.W., Geng L., Zhang B.Y. Laser cladding of Ni-
base composite coatings onto Ti–6Al–4V substrates
with pre-placed B4C+NiCrBSi powders // Surface and
Coatings Technology. 2006. Vol. 200. № 16-17. P. 4923–4928. DOI: 10.1016/j.surfcoat.2005.04.059. 6. Miguel J.M., Guilemany J.M., Vizcaino S. Tribological
study of NiCrBSi coating obtained by different process-
es // Tribology International. 2003. Vol. 36. № 3. P. 181–187. DOI: 10.1016/S0301-679X(02)00144-5. j
18. Geng L., Meng Q.W., Chen Y.B. ОСНОВНЫЕ РЕЗУЛЬТАТЫ Сформированные лазерным излучением NiCrBSi
и NiCrBSi–B4C покрытия характеризуются отсутствием
макропор, включений, однородной и мелкозернистой
структурой, при этом размер структурных составляю-
щих NiCrBSi–B4C покрытия мельче. 10. Макаров А.В., Соболева Н.Н., Малыгина И.Ю.,
Осинцева А.Л. Формирование композиционного по-
крытия NiCrBSi-TiC с повышенной абразивной из-
носостойкостью методом газопорошковой лазерной
наплавки // Упрочняющие технологии и покрытия. 2013. № 11. С. 38–44. Переходная зона покрытий узкая: у NiCrBSi покры-
тия составляет примерно 10 мкм, а NiCrBSi-B4C покры-
тия – около 4 мкм при общей толщине наплавленных
слоев 140–190 и 120–150 мкм соответственно. Добавление В4С в порошок ПГ-СР2 при лазерной
наплавке приводит к снижению значений средней мик-
ротвердости покрытия (до 530 HV 0,025) по сравнению
со значениями средней микротвердости NiCrBSi по-
крытия (620 HV 0,025), что связано с повышенным со-
держанием железа в NiCrBSi–B4C покрытии. Добавление В4С в порошок ПГ-СР2 при лазерной
наплавке приводит к снижению значений средней мик-
ротвердости покрытия (до 530 HV 0,025) по сравнению
со значениями средней микротвердости NiCrBSi по-
крытия (620 HV 0,025), что связано с повышенным со-
держанием железа в NiCrBSi–B4C покрытии. 11. Fals H.C., Aguiar D., Fanton L., Belém M.J.X., Li-
ma C.R.C. A new approach of abrasive wear perfor-
mance of flame sprayed NiCrSiBFeC/SiC composite
coating // Wear. 2021. Vol. 477. № SI. Article number
203887. DOI: 10.1016/j.wear.2021.203887. 12. Senapati P., Sutar H. Surface erosion behaviour over
NiCrBSi-Al2O3 composite coatings // Materials Re-
search Express. 2020. Vol. 7. № 7. Article number
076512. DOI: 10.1088/2053-1591/aba396. Старикова У.С., Соболева Н.Н., Макаров А.В. и др. «Влияние добавки карбида бора на структуру и твердость никелевого…» 8. Sha J., Chen L.-Y., Liu Y.-T., Yao Z.-J., Lu S.,
Wang Z.-X., Zang Q.-H., Mao S.-H., Zhang L.-C. Phase
transformation-induced improvement in hardness and
high-temperature wear resistance of plasma-sprayed and
remelted NiCrBSi/WC coatings // Metals. 2020. Vol. 10. № 12. Article number 1688. DOI: 10.3390/
met10121688. NiCrBSi покрытия после добавления 5 об. % B4C. В [15]
сделан вывод, что в процессе селективного лазерного
сплавления материал основы (в основном это железо)
смешивается с материалом покрытия. Это приводит
к снижению микротвердости при приближении к пере-
ходной зоне «покрытие – основа». 9. Соболева Н.Н., Николаева Е.П., Макаров А.В., Ма-
лыгина И.Ю. Влияние добавки карбида хрома на
структуру и абразивную износостойкость NiCrBSi
покрытия, сформированного лазерной наплавкой //
Вектор науки Тольяттинского государственного
университета. 2020. № 1. С. 68–76. DOI: 10.18323/
2073-5073-2020-1-68-76. Frontier Materials & Technologies. 2021. № 4 ОБСУЖДЕНИЕ РЕЗУЛЬТАТОВ Резкие изменения количества элементов (снижение
количества Ni и повышение содержания Fe) в попереч-
ном сечении образцов с NiCrBSi и NiCrBSi–В4С покры-
тиями (рис. 3 и рис. 4) свидетельствуют о наличии
в образцах узкой переходной зоны от покрытия к ос-
новному металлу. Переходная зона NiCrBSi покрытия
примерно составляет 10 мкм, а NiCrBSi–B4C покрытия –
примерно 4 мкм. Снижение значения микротвердости покрытия при до-
бавлении 7 масс. % порошка В4С (таблица 1) объясняется
большим содержанием железа в структуре NiCrBSi–В4С
покрытия (51 масс. %), чем в структуре NiCrBSi покры-
тия (38 масс. %). Отмеченный выше результат сниже-
ния микротвердости согласуется с результатом работы
[15], где было показано снижение микротвердости Из рис. 5 следует, что в обоих случаях микрострук-
тура характеризуется однородностью и мелкозернисто-
стью. При этом у покрытия с карбидом бора (рис. 5 b)
размер структурных составляющих мельче. Измельче- Frontier Materials & Technologies. 2021. № 4 94 Старикова У.С., Соболева Н.Н., Макаров А.В. и др. «Влияние добавки карбида бора на структуру и твердость никелевого…» REFERENCES 1. Zemlyakov E., Babkin K., Korsmik R., Sklyar M.,
Kuznetsov M. Prospects of use of laser cladding tech-
nology for restoration of compressor blades of gas tur-
bine engines. Fotonika, 2016, no. 4, pp. 10–25. DOI: 10.22184/1993-7296.2016.58.4.10.22. 14. Krivizhenko D.S., Drobyaz E.A., Zimoglyadova T.A. Structure formation features of boron-containing coat-
ings obtained by highspeed treatment. Aktualnye
problemy v mashinostroenii, 2014, no. 1, pp. 489–492. 2. Golubovskiy E.N., Parkin A.A., Zhatkin S.S. Repair of
cracks on sectors of gas-turbine engine shovels by
method of cutting and laser pulse cladding. Izvestiya
Samarskogo nauchnogo tsentra Rossiyskoy akademii
nauk, 2020, vol. 22, no. 2, pp. 107–112. 15. Golyshev A.A., Filippov A.A. Comparative investiga-
tion of nickel-based metal-ceramic structures with ce-
ramic particles of tungsten and carbides made by the se-
lective laser melting method. Nanoscience and Tech-
nology, 2020, vol. 11, no. 3, pp. 247–257. DOI: 10. 1615/NanoSciTechnolIntJ.2020033784. 3. Soboleva N.N., Makarov A.V., Malygina I.Yu. Techno-
logical aspects of friction treatment of PG-SR2 coating
formed by laser cladding. Vektor nauki Tolyattinskogo
gosudarstvennogo universiteta, 2019, no. 3, pp. 47–53. DOI: 10.18323/2073-5073-2019-3-47-53. 16. Fan X.H., Geng L., Xu B., Li J. Laser Cladding
NiCrBSi+2%B4C Coating on Ti-6Al-4V. Advanced Ma-
terials Research, 2009, vol. 79-82, pp. 473–476. DOI: 10.4028/www.scientific.net/AMR.79-82.473. 4. Gibzun M.S., Makarov A.V., Soboleva N.N., Malygi-
na I.Yu. Increasing frictional thermal resistance
of Ni-Cr coating using combined laser-heat treatment. Master’s journal, 2017, no. 1, pp. 11–16. 17. Meng Q.W., Geng L., Zhang B.Y. Laser cladding of Ni-
base composite coatings onto Ti–6Al–4V substrates
with pre-placed B4C+NiCrBSi powders. Surface and
Coatings Technology, 2006, vol. 200, no. 16-17,
pp. 4923–4928. DOI: 10.1016/j.surfcoat.2005.04.059. pp
5. Shi B.W., Li T., Wang D., Zhang X.R., Zhang H.C. Investigation on crack behavior of Ni60A alloy coating
produced by coaxial laser cladding. Journal of Materi-
als Science, 2021, vol. 56, no. 23, pp. 13323–13336. DOI: 10.1007/s10853-021-06108-5. 18. Geng L., Meng Q.W., Chen Y.B. In-situ Synthesis of
Metal Matrix Composite Coating with Laser Melting-
Solidifying Processes. Composite materials IV. Key En-
gineering Materials, 2006, vol. 313, pp. 139–144. DOI: 10.4028/www.scientific.net/KEM.313.139. 6. Miguel J.M., Guilemany J.M., Vizcaino S. Tribological
study of NiCrBSi coating obtained by different process-
ses. Tribology International, 2003, vol. 36, no. 3,
pp. 181–187. DOI: 10.1016/S0301-679X(02)00144-5. 19. Golubev V.S., Vegera I.I., Chernasheyus O., Chaev-
skiy V.V. Laser treatment of materials with change of
chemical composition of the surface layer. Vestnik
Baranovichskogo
gosudarstvennogo
universiteta. Seriya: Tekhnicheskie nauki, 2019, no. 7, pp. 34–42. 7. Старикова У.С., Соболева Н.Н., Макаров А.В. и др. «Влияние добавки карбида бора на структуру и твердость никелевого…» Старикова У.С., Соболева Н.Н., Макаров А.В. и др. «Влияние добавки карбида бора на структуру и твердость никелевого…» no. 12, article number 1688. DOI: 10.3390/met
10121688. химического состава поверхностного слоя // Вестник
Барановичского государственного университета. Се-
рия: Технические науки. 2019. № 7. С. 34–42. 9. Soboleva N.N., Nikolaeva E.P., Makarov A.V.,
Malygina I.Yu. The influence of chromium carbide ad-
ditive on the structure and abrasive wear resistance of
the NiCrBSi coating formed by laser cladding. Vektor
nauki Tolyattinskogo gosudarstvennogo universiteta,
2020, no. 1, pp. 68–76. DOI: 10.18323/2073-5073-
2020-1-68-76. 20. Дробяз Е.А., Кривиженко Д.С., Поляков И.А., На-
гавкин С.Ю., Иванцивский В.В. Структура и свой-
ства боросодержащих покрытий, наплавленных
электронным лучом, выведенным в воздушную ат-
мосферу // Обработка металлов (технология, обору-
дование, инструменты). 2012. № 4. С. 83–85. ру
)
21. Зусин В.Я. Исследование модифицирования метал-
ла, наплавленного порошковой проволокой с алю-
миниевой оболочкой // Вестник Приазовского госу-
дарственного университета. Серия: Технические
науки. 2011. № 23. С. 180–183. 10. Makarov A.V., Soboleva N.N., Malygina I.Yu.,
Osintseva A.L. The formation of NiCrBSi–TiC compo-
site coating with increased abrasive wear resistance by
gas
powder
laser
cladding. Uprochnyayushchie
tekhnologii i pokrytiya, 2013, no. 11, pp. 38–44. 22. Еремин Е.Н. Применение наночастиц тугоплавких
соединений для повышения качества сварных со-
единений из жаропрочных сплавов // Омский науч-
ный вестник. 2009. № 3. С. 63–67. 11. Fals H.C., Aguiar D., Fanton L., Belém M.J.X., Li-
ma C.R.C. A new approach of abrasive wear perfor-
mance of flame sprayed NiCrSiBFeC/SiC composite
coating. Wear, 2021, vol. 477, no. SI, article number
203887. DOI: 10.1016/j.wear.2021.203887. 23. Зернин Е.А., Кузнецов М.А. Способы модифициро-
вание наплавленного металла наноструктурирован-
ными порошками для увеличения механических
свойств сварных соединений // Современные про-
блемы науки и образования. 2014. № 5. С. 206–212. 12. Senapati P., Sutar H. Surface erosion behaviour over
NiCrBSi-Al2O3 composite coatings. Materials Research
Express, 2020, vol. 7, no. 7, article number 076512. DOI 10 1088/2053 1591/ b 396 13. Li H.J., He Y., Luo P.Y., Fan Y., Yu H., Wang Y.Q.,
He T., Li Z.J., Zhang H.L. Influence of pulse frequency
on corrosion resistance and mechanical properties of Ni-
W/B4C composite coatings. Colloids and Surfaces A:
Physicochemical and Engineering Aspects, 2021,
vol. 629, article number 127436. DOI: 10.1016/j. colsurfa.2021.127436. Frontier Materials & Technologies. 2021. № 4 СПИСОК ЛИТЕРАТУРЫ In-situ Synthesis of
Metal Matrix Composite Coating with Laser Melting-
Solidifying Processes // Composite materials IV. Key
Engineering Materials. 2006. Vol. 313. P. 139–144. DOI: 10.4028/www.scientific.net/KEM.313.139. 7. Niranatlumpong P., Koiprasert H. Phase transformation
of NiCrBSi–WC and NiBSi–WC arc sprayed coatings //
Surface and Coatings Technology. 2011. Vol. 206. № 2-3. P. 440–445. DOI: 10.1016/j.surfcoat.2011.07.057. 19. Голубев В.С., Вегера И.И., Чернашеюс О., Чаевс-
кий В.В. Лазерная обработка материалов с изменением Frontier Materials & Technologies. 2021. № 4 95 The influence of boron carbide additive on the structure
and hardness of a nickel-based coating © 2021
Uliana S. Starikova*1,4, junior researcher
Natalia N. Soboleva1,2,5, PhD (Engineering), senior researcher
Aleksey V. Makarov1,6, Corresponding member of RAS, Doctor of Sciences (Engineering)
Head of Department of Materials Science and the Laboratory of Mechanical Properties,
Evgeny V. Kharanzhevsky3,7, Doctor of Sciences (Engineering), Professor, Head of Laboratory
1M.N. Mikheev Institute of Metal Physics of the Ural Branch of the Russian Academy of Sciences, Yekaterinburg (Russia)
2Institute of Engineering Science of the Ural Branch of the Russian Academy of Sciences, Yekaterinburg (Russia)
3Udmurt State University, Izhevsk (Russia) © 2021
Uliana S. Starikova*1,4, junior researcher
Natalia N. Soboleva1,2,5, PhD (Engineering), senior researcher
Aleksey V. Makarov1,6, Corresponding member of RAS, Doctor of Sciences (Engineering)
Head of Department of Materials Science and the Laboratory of Mechanical Properties,
Evgeny V. Kharanzhevsky3,7, Doctor of Sciences (Engineering), Professor, Head of Laboratory
1M.N. Mikheev Institute of Metal Physics of the Ural Branch of the Russian Academy of Sciences, Yekaterinburg (Russia)
2Institute of Engineering Science of the Ural Branch of the Russian Academy of Sciences, Yekaterinburg (Russia)
3Udmurt State University, Izhevsk (Russia) © 2021
Uliana S. Starikova*1,4, junior researcher
Natalia N. Soboleva1,2,5, PhD (Engineering), senior researcher
Aleksey V. Makarov1,6, Corresponding member of RAS, Doctor of Sciences (Engineering)
Head of Department of Materials Science and the Laboratory of Mechanical Properties,
Evgeny V. Kharanzhevsky3,7, Doctor of Sciences (Engineering), Professor, Head of Laboratory
1M.N. Mikheev Institute of Metal Physics of the Ural Branch of the Russian Academy of Sciences, Yekaterinburg (Russia)
2Institute of Engineering Science of the Ural Branch of the Russian Academy of Sciences, Yekaterinburg (Russia)
3Udmurt State University, Izhevsk (Russia) 4ORCID: https://orcid.org/0000-0001-6960-0619
5ORCID: https://orcid.org/0000-0002-7598-2980
6ORCID: https://orcid.org/0000-0002-2228-0643
7ORCID: https://orcid.org/0000-0002-1525-2169 4ORCID: https://orcid.org/0000-0001-6960-0619
5ORCID: https://orcid.org/0000-0002-7598-2980
6ORCID: https://orcid.org/0000-0002-2228-0643
7ORCID: https://orcid.org/0000-0002-1525-2169 *E-mail: ulstar97@mail.ru *E-mail: ulstar97@mail.ru *E-mail: ulstar97@mail.ru Abstract: Laser cladding is increasingly frequently used in various branches of mechanical engineering since it has
such advantages over traditional methods of depositing coatings as high heating and cooling rates and minimal mixing of
base and melting materials. Laser-clad coatings are usually characterized by a fine-grained structure and a minimal heat-
affected zone. Coatings formed from the Ni–Cr–B–Si powders are also very common in industrial applications, as they
have good resistance to wear, corrosion, erosion, etc. Various strengthening particles can be added to this group of pow-
ders to improve the properties of the deposited coating. Boron carbides can act as such particles since they have high hard-
ness, thermodynamic stability, and wear resistance. REFERENCES Niranatlumpong P., Koiprasert H. Phase transformation
of NiCrBSi–WC and NiBSi–WC arc sprayed coatings. Surface and Coatings Technology, 2011, vol. 206,
no. 2-3, pp. 440–445. DOI: 10.1016/j.surfcoat.2011.07.057. 8. Sha J., Chen L.-Y., Liu Y.-T., Yao Z.-J., Lu S.,
Wang Z.-X., Zang Q.-H., Mao S.-H., Zhang L.-C. Phase
transformation-induced improvement in hardness and
high-temperature wear resistance of plasma-sprayed and
remelted NiCrBSi/WC coatings. Metals, 2020, vol. 10, 20. Drobyaz E.A., Krivizhenko D.S., Polyakov I.A.,
Nagavkin S.Yu., Ivantsivskiy V.V. Structure and
properties of boron-containing coatings, deposited
by non-vacuum electron beam. Obrabotka metallov Frontier Materials & Technologies. 2021. № 4 96 Старикова У.С., Соболева Н.Н., Макаров А.В. и др. «Влияние добавки карбида бора на структуру и твердость никелевого…» Старикова У.С., Соболева Н.Н., Макаров А.В. и др. «Влияние добавки карбида бора на структуру и твердость никелевого…»
(tekhnologiya, oborudovanie, instrumenty), 2012, no. 4,
pp. 83–85. made of heat-resistant alloys. Omskiy nauchnyy vestn
2009, no. 3, pp. 63–67. (tekhnologiya, oborudovanie, instrumenty), 2012, no. 4,
pp. 83–85. made of heat-resistant alloys. Omskiy nauchnyy vestnik,
2009, no. 3, pp. 63–67. pp
21. Zusin V.Ya. Investigation of metal modification, de-
posited with powder wire with aluminums shell. Vestnik Priazovskogo gosudarstvennogo universiteta. Seriya: Tekhnicheskie nauki, 2011, no. 23, pp. 180–
183. 23. Zernin E.A., Kuznetsov M.A. Method of modifying
the weld metal nanostructured powders for increasing
the mechanical properties of welded joints. Sovre-
mennye problemy nauki i obrazovaniya, 2014, no. 5,
pp. 206–212. 22. Eremin E.N. Application of nanoparticles of refractory
compounds for improving the quality of welding joints 22. Eremin E.N. Application of nanoparticles of refractory
compounds for improving the quality of welding joints Frontier Materials & Technologies. 2021. № 4 The influence of boron carbide additive on the structure
and hardness of a nickel-based coating In this regard, the paper investigated the influence of the 7 wt. % of
boron carbide B4C addition on the structure and hardness of the NiCrBSi coating formed by laser cladding of PG-SR2
powder on the surface of 30KhRA steel. Using the scanning electron microscope, the authors carried out microscopic stud-
ies of the structure of NiCrBSi and NiCrBSi–B4C coatings and presented the results of X-ray spectral microanalysis. The study shows that the structures of both coatings in the deposited state are characterized by uniformity and fine-grain
structure. The investigation revealed that the samples with NiCrBSi and NiCrBSi–B4C coatings have a narrow transition
zone from the deposit to the base metal. The paper presents the results of measuring the microhardness of coatings indica-
ting a decrease in the microhardness of laser-clad nickel-based coatings with the boron carbide addition. Keywords: laser cladding; NiCrBSi coating; NiCrBSi–B4C coating; microhardness; scanning electron microscopy. Acknowledgments: The work was carried out within the government assignment to the Institute of Metal Physics of
the UB of RAS on the topics No. 121102900049-1 and No. АААА-А18-118020190116-6 and the Institute of Engineering
Science, RAS (Ural Branch) on the topic No. АААА-А18-118020790147-4. The experimental studies were performed on
the equipment of the “Plastometry” Core Facility Center of the IES UB RAS. y
g
g
g
g
py
Acknowledgments: The work was carried out within the government assignment to the Institute of Metal Physics of
the UB of RAS on the topics No. 121102900049-1 and No. АААА-А18-118020190116-6 and the Institute of Engineering
Science, RAS (Ural Branch) on the topic No. АААА-А18-118020790147-4. The experimental studies were performed on
the equipment of the “Plastometry” Core Facility Center of the IES UB RAS. q p
y
y
The paper was written on the reports of the participants of the X International School of Physical Materials Science
(SPM-2021), Togliatti, September 13–17, 2021. q p
y
y
The paper was written on the reports of the participants of the X International School of Physical Materials Science
(SPM-2021), Togliatti, September 13–17, 2021. For citation: Starikova U.S., Soboleva N.N., Makarov A.V., Kharanzhevsky E.V. The influence of boron carbide addi-
tive on the structure and hardness of a nickel-based coating. Frontier Materials & Technologies, 2021, no. 4, pp. 89–97. DOI: 10.18323/2782-4039-2021-4-89-97. For citation: Starikova U.S., Soboleva N.N., Makarov A.V., Kharanzhevsky E.V. The influence of boron carbide addi-
tive on the structure and hardness of a nickel-based coating. For citation: Starikova U.S., Soboleva N.N., Makarov A.V., Kharanzhevsky E.V. The influence of boron carbide addi-
tive on the structure and hardness of a nickel-based coating. Frontier Materials & Technologies, 2021, no. 4, pp. 89–97.
DOI: 10.18323/2782-4039-2021-4-89-97. The influence of boron carbide additive on the structure
and hardness of a nickel-based coating Frontier Materials & Technologies, 2021, no. 4, pp. 89–97. DOI: 10.18323/2782-4039-2021-4-89-97. 97
|
https://openalex.org/W2620991428
|
https://journals.squ.edu.om/index.php/tjer/article/download/179/191
|
English
| null |
Performance of Cement-Based Patch Repair Materials in Plain and Reinforced Concrete Members
|
The journal of engineering research
| 2,016
|
cc-by
| 5,856
|
Performance of Cement-Based Patch Repair Materials in
Plain and Reinforced Concrete Members
A.H. Al-Saidy*
Department of Civil & Architectural Engineering, College of Engineering, Sultan Qaboos University, Oman A.H. Al-Saidy*
Department of Civil & Architectural Engineering, College of Engineering, Sultan Qaboos University, Oman A.H. Al-Saidy* Received 2 March 2015; Accepted 12 April 2016 Abstract: Structural elements such as beams, slabs, and columns may require strengthening or repair
during their service life. Different repair materials (RMs) are available and it is usually difficult to
choose the best ones, especially when considering the cost of such materials. This paper presents the
results of an experimental investigation of patch RMs on plain concrete prisms as well as on
reinforced concrete beams. Three cement-based RMs available in the market with different
mechanical properties and an ordinary Portland cement (OPC) mix produced in the lab were used in
the study. Damage was induced in prisms/beams and then repaired using different materials. The
experimental work included assessment of the flexural strength of damaged/repaired plain concrete
prisms; slant shear (bond) strength between the concrete and the RM; axial strength of
damaged/repaired plain concrete prisms and bond of the repair materials in damaged/repaired
reinforced concrete beams loaded to failure. The test results showed that all RMs performed well in
restoring the strength of damaged plain concrete. Compatibility of the RMs with substrate concrete
was found to be more important in the behavior than superior mechanical properties of the RMs. No
difference was noted in the behavior between the RMs in repairing reinforced concrete beams at the
tension side. Keywords: Patch repair, Repair materials, Rehabilitation, Retrofitting. The Journal of Engineering Research (TJER), Vol. 13, No. 2 (2016) 160-171 The Journal of Engineering Research (TJER), Vol. 13, No. 2 (2016) 160-171 * Corresponding author’s email: alsaidy@squ.edu.om 1. Introduction Repair and rehabilitation of reinforced concrete
structures are common requirements in modern
construction. Structural elements such as beams,
slabs, and columns may require strengthening
or repair during their service life due to aging,
increased loads, and deterioration caused by the
surrounding harsh environment. Steel corrosion
is a major cause of deterioration which disrupts
the cover zone of reinforced concrete. The
conventional approach to repair a corroded
reinforced concrete member is the use of
patches to repair in the zones where cracking
and spalling take place. Repairs are generally
undertaken by removing the damaged concrete
beyond
the
steel
reinforcement,
and
reinforcements are then cleaned in cases of
minor
corrosion
or
replaced
if
severely
damaged. Finally, a prepackaged cementitious
repair material (RM) is applied. Many
types
of
RMs
are
available
commercially, and the selection for reinstating a
deteriorated concrete structure is usually done
on a relatively ad hoc basis. The repair and
rehabilitation of reinforced concrete structures
can only be successful if the new materials
interact effectively with the parent concrete and
form a durable barrier against the ingress of
carbon dioxide and chlorides. Problems may
arise whenever a dimensionally unstable RM is
placed
against
an
aged
(and,
therefore,
dimensionally stable) substrate concrete. The efficiency and durability of a repaired
system depends on the bond between the
concrete substrate and RM. By increasing the
surface roughness, the surface treatment of
concrete substrate can promote mechanical
interlocking,
which
is
one
of
the
basic
mechanisms of adhesion. The durability of a
bond depends on several factors, each acting
with different degrees of influence. The five
major factors as reported by researchers in the
literature are micro cracking, the absence of a
laitance layer, cleanliness before an overlay
placement,
and
compaction
and
curing
procedures. The first three parameters are
directly related to substrate characteristics,
which can be modified by surface treatment. The treatment of a concrete substrate is
commonly used for cleaning, removing the
laitance layer, and roughening the surface. However, it can induce micro cracking if it is
not well operated with regard to the quality and
strength of concrete (Courard et al. 2006;
Courard et al. 2014; Garbacz et al. 2005; Liu et al. 2014; Qian et al. 2014). Several materials are available in the market
for
the
repair
of
deteriorated
concrete
structures. A.H. Al-Saidy A.H. Al-Saidy A.H. Al-Saidy deformations. Numerous
studies
in
the
literature deal with the evaluation of different
RMs. A great deal of the research has focused
on durability issues associated with RMs such
as chloride permeability, electrical resistivity,
and carbonation (Al-Zahrani et al. 2003; Mangat
and Limbachiya 1995; Ueda et al. 2011). Such
studies have concluded that there is no clear
difference between the cement- and polymer-
based repair mortars with regard to the chloride
permeability. However, carbonation has been
identified as an issue in cement- and polymer-
based repair mortars except in dense mortars
which has been attributed to the addition of
silica fume, fibers, and/or other additives. Other studies have focused on the properties
that affect the stability of RMs such as
shrinkage, thermal properties, and modulus of
elasticity
(Mangat
and
Limbachiya
1997;
Margan 1996). Polymer-based repair mortars
have been found to have lower shrinkage
strains
compared
to
cement-based
repair
mortar. Compatibility of the modulus of the
elasticity of RM with that of the substrate
concrete were found to have a great influence
on load sharing or the contribution of RMs in
resisting the applied load of the repaired
member. RMs with a modulus of elasticity close
to that of the substrate concrete tends to provide
a stable repair under applied loads. Performance of Cement-Based Patch Repair Materials in
Plain and Reinforced Concrete Members
A.H. Al-Saidy*
Department of Civil & Architectural Engineering, College of Engineering, Sultan Qaboos University, Oman א
אא
א
א
א
א
א
א
א
אא
* א
אא
א
א
א
א
א
א
א
אא
* אא
* אW
א
א
אא
א
K
א
אא
אא
א
א
،א
א
אא
K
אא
א
א
א
א
א
א
א
א
אאאK
א
א
א
א
א
א
אאאאא
אK
א
א
א
אאK
אא
א
א
،א
א
א
א
،א
א
א
K
א
אK
א
אא
א
א
א
א
א
א
אK
א
א
א
א
א
אא
א
א
א
אא
א
אא
א
א
א
א
א
אK
א
אא
א
א
א
א
א
א
א
א
K א
אW
א
،אאא
،א
،א
אא א
אW
א
،אאא
،א
،א
אא * Corresponding author’s email: alsaidy@squ.edu.om 3. Specimen Details To evaluate the performance of each RM under
service load, four different tests were carried
out. The bond of the RM to the substrate
concrete was evaluated through the slant shear
test. Behavior of the RM applied to plain
concrete
was
evaluated
using
damaged/repaired prisms tested in flexure and
axial compression. To simulate the behavior of a
real field situation, reinforced concrete beams
were damaged and then patch-repaired using
each of the RMs. The reinforced concrete beams
were tested to failure to observe the bond of the
RM through all loading stages. 2.3 Mechanical Properties a) Compressive
and
tensile
strength
according to American Society for Testing
and Materials (ASTM) standard C109. b) Flexural strength according to ASTM
standard C78. c) Elastic
modulus
according
to
ASTM
standard C469 . This paper presents the results of three
commercially available RMs on the Omani
market and one laboratory-prepared material. The experimental work presented here includes
bond strength; flexural and axial strength of
damaged/repaired prisms; bond and strength
behavior
of
damaged/repaired
reinforced
concrete beams. 2.1 Repair Materials (RM1, RM2, and RM3) 2.1 Repair Materials (RM1, RM2, and RM3)
The three RMs were pre-packed commercial
cement-based materials supplied as ready-to-
use dry powders which require only the
addition of clean water to produce a highly
consistent, lightweight repair mortar suitable
for general purpose concrete and masonry
repairs. Each material is packaged in 20 kg
bags. To get the required mix, water at a
measurement of 15–18% of the powder weight
was added according to the manufacturer's
instructions. However, the difference between
the three materials is mainly in the mechanical
properties (Table 1). 1. Introduction Such RMs are classified, for example,
as
cement,
epoxy
resins,
polymer
latex,
polyester resins, and polyvinyl acetate, with
cement-based and epoxy-based RMs being the
two most widely used. However, manufacturers
are reluctant to provide complete details of their
materials,
which
makes
the
selection
of
appropriate RMs difficult. Potential users are
faced with a wide choice of materials and little
guidance on their properties and long-term
performance (Cabrera and Al-Hasan 1997). The design of patch repair is usually based
on the experience of specialist contractors, and
selection of RMs is made based on their relative
short-term properties such as strength, bond,
and early age plastic shrinkage. Although these
properties indicate the immediate performance
of the repair, they give little information on its
long-term performance with respect to cracking
and composite action with the substrate
concrete
to
carry
loads
and
prevent The performance of the RMs in service under
structural loading has received relatively little 161 Performance of Cement-Based Patch Repair Materials in Plain and Reinforced Concrete Members 2.2 Ordinary Portland Cement (OPC) attention
from
the
research
community,
especially in the case of patch repairs. Some
studies have considered the behavior of
repaired reinforced concrete members under
service loads (Abu-Tair 1997; AlFarabi et al. 2006; Ghassan and Zai-UL-Hasan 1999; Rio et al. 2005). However, there is no agreement between
researches regarding the properties, which
should be considered in the evaluation. This
lack of agreement can have serious effects
where differences in mechanical properties can
lead to short- or long-term performance or
durability problems. This RM was obtained by mixing OPC with
water at a ratio of w/c of 0.25. OPC was chosen
in order to have a consistency similar to that of
the other RMs. 3.1 Slant Shear Specimens The bond strength test (slant shear test) was
done according to ASTM standard C882 using
75 x 150 mm concrete cylinders. Each cylinder
was cut as specified in the standard at 30° by a
concrete cutting machine after curing for 28
days (Fig. 1a). After cutting these cylinders,
each piece was replaced in a cylindrical mold to
cast the RM above it. Three specimens were
used for each RM. Figure 1b shows the
combined specimen after casting the RM. Table 1. Properties of the repair materials after
28 days. 2. Repair Materials Four RMs were used in this study. A brief
description of each of these materials follows. 3.2 Damaged/Repaired Prisms Repair
Material
fc
(Mpa)
ft
(Mpa)
E
(Gpa)
fr
(Mpa)
RM1
54.3
4.4
31.7
7.8
RM2
52.5
4.3
31.1
8.1
RM3
63.7
4.2
34.4
7.8
OPC
51.2
4.2
30.0
7.7
fc = Cube compressive strength
ft = Cylinder tensile strength
E = Modulus of elasticity
fr = Prism flexural strength To test the flexural strength, six 100 x 100 x
500 mm prisms were made from each mix or
RM. Two prisms were undamaged, two were
damaged, and the other two were damaged but
then
repaired
with
the
RMs
(damaged
/repaired). A similar number of prisms were
also made to test the axial strength: two
undamaged,
two
damaged,
and
two
damaged/repaired. In order to produce the
same damage in all specimens, a preformed
mold was used while casting the prisms to Repair
Material
fc
(Mpa)
ft
(Mpa)
E
(Gpa)
fr
(Mpa)
RM1
54.3
4.4
31.7
7.8
RM2
52.5
4.3
31.1
8.1
RM3
63.7
4.2
34.4
7.8
OPC
51.2
4.2
30.0
7.7 162 A.H. Al-Saidy 4. Test Setup All specimens were tested after curing using a
wet Hessian cloth for 35 days (28 days for
concrete substrate + seven days for RM). A
bond strength test (slant shear test) was done
according to the ASTM Standard C882. Each
cylinder of the combined material (concrete +
RM) was placed under the compression
machine, and the failure load was recorded. The
flexural strength of the undamaged, damaged,
and damaged/repaired plain concrete prisms
was measured according to ASTM Standard
C78. Each prism was loaded under four-point
bending; the failure load was recorded and the
location of the failure was observed. A.H. Al-Saidy (a)
(b)
Figure 1. Slant Shear Test: (a) Concrete cylinder after cutting, and (b) Combined cylinder of repair
material and concrete substrate. (a) (a) (b) (b) ( )
Figure 1. Slant Shear Test: (a) Concrete cylinder after cutting, and (b) Combined cylinder of repair
material and concrete substrate. a) Casting of white cement in damaged area. b) Removing the white cement patch. c) Placing the white cement patch in the prism molds before casting. a) Casting of white cement in damaged area. b) Removing the white cement patch. b) Removing the white cement patch. c) Placing the white cement patch in the prism molds before casting. c) Placing the white cement patch in the prism molds before casting. c) Placing the white cement patch in the prism molds before casting Figure 2. Preparation of damage mold to induce uniform damage in prisms. 163 Performance of Cement-Based Patch Repair Materials in Plain and Reinforced Concrete Members create a damaged section measuring 200 mm in
length with a maximum depth of 25 mm within
the middle third of the prism (Fig. 2). To
accomplish this specific type of damage, a
concrete prism was damaged by chipping the
concrete with 200 mm width and a maximum
depth of 25 mm. To simulate the actual damage
in concrete, the damage made by chipping the
concrete was not uniform. After chipping, the
damaged area was patched with white cement
mortar. Once the white cement patch had set, it
was removed from the concrete with all of the
rough surfaces imprinted on the surface of the
patch. The patch was then placed in the prism
mold before casting (Fig. 2(c)). After casting, this
patch was removed from the prism, creating the
required damage. The same procedure was
repeated with the other specimens. In this way,
the damage in all specimens was uniform in
order to make a fair comparison between all
RMs. reinforcement) at the middle of the beam. After
casting, each beam was cured for 28 days using
polythene sheets. Four of the damaged beams
were repaired using the selected RMs and were
cured for an additional seven days prior to
testing. 5.1 Bond Test The specimens were loaded axially in
compression, which caused failure along the
line where the two layers are jointed (Fig. 1). Bond strength was calculated by dividing the
failure load by the interface area (ellipse area). The
average
bond
strength
(using
three
specimens) in N/mm2 were 3.81, 4.97, 5.78, and
4.05
for
RM1,
RM2,
RM3,
and
OPC,
respectively. As can be observed, the maximum
bond strength was 5.78 N/mm2 for RM3,
whereas the minimum bond strength value was
3.81 N/mm2 for RM1. There was not much
difference between RM1 (3.81 N/mm2) and the
control RM OPC (4.05 N/mm2). The mode of
failure was almost the same for RM1, RM2, and
RM3, and failure occurred at the interface
surface between the two materials (Fig. 5(a)). Failure of the OPC specimen was due to
crushing failure (Fig. 5(b)). A summary of the results of the flexural
strength
of
the
undamaged,
damaged,
damaged/repaired prisms is shown in Fig. 6. Here the failure load which caused the prism to Table 3. Properties of concrete substrate of
prisms after 28 days. Prism
fc
(Mpa)
ft
(Mpa)
E
(Gpa)
fr
(Mpa)
Used with
RM1
36.2
2.6
25.3
6.9
Used with
RM2
35.4
3.1
24.5
7.0
Used with
RM3
34.1
2.5
24.1
6.2
Used with
OPC
37.4
2.6
26.0
6.8
fc, ft, E, fr : as defined in Table 1. Prism
fc
(Mpa)
ft
(Mpa)
E
(Gpa)
fr
(Mpa)
Used with
RM1
36.2
2.6
25.3
6.9
Used with
RM2
35.4
3.1
24.5
7.0
Used with
RM3
34.1
2.5
24.1
6.2
Used with
OPC
37.4
2.6
26.0
6.8
fc, ft, E, fr : as defined in Table 1. 3.3 Damaged/Repaired Reinforced Concrete
Beams In
addition
to
flexural
strength,
a
compressive strength test was done on the plain
concrete prisms. Six prisms were tested for each
RM: two undamaged, two damaged, and two
damaged/repaired prisms. The failure load was
recorded for each prism, and the compressive
strength was calculated. Figure 3 shows the test
setup of plain concrete prisms under axial
compression. In order to study the behavior of reinforced
concrete elements, six reinforced concrete beams
were cast in the laboratory (Table 2). The beams
consisted
of
one
undamaged
beam,
one
damaged beam, and four damaged/repaired
beams. The beams were reinforced for tension at
their bottom sides using two bars, each 10 mm
in diameter. Each beam was 2,700 mm long with
a cross-section of 100 x 150 mm and a cover of
25 mm. For the damaged and repaired beams,
the damage was induced during casting with
approximately the same shape and size by the
same procedure described in Section 3.2 for
prism preparation. The damage zone was 400
mm long and 40 mm deep (25 mm cover + 10
mm bar diameter + 5 mm above the The reinforced concrete beams were simply
supported and subjected to four-point bending
(Fig. 4). The beams were supported over 2,400
mm. Linear variable differential transducers
(LVDTs) were used to measure deflection at the
mid-span of the beam. All beams were loaded at
a rate of 0.06 kN/s until the failure load was
reached. Figure 3. Test setup of plain concrete prisms
under axial compression. Table 2. Description of Beams. Beam
Description
C
Control beam without damage
D
Damaged beam
DRM1
Damaged and batch repaired using
RM1
DRM2
Damaged and batch repaired using
RM2
DRM3
Damaged and batch repaired using
RM3
DRM4
Damaged and batch repaired using
OPC Beam
Description
C
Control beam without damage
D
Damaged beam
DRM1
Damaged and batch repaired using
RM1
DRM2
Damaged and batch repaired using
RM2
DRM3
Damaged and batch repaired using
RM3
DRM4
Damaged and batch repaired using
OPC Figure 3. Test setup of plain concrete prisms
under axial compression. 164 A.H. Al-Saidy Figure 4. Test setup of reinforced concrete beams. P
150
950
500
950
150
Spreader beam
100
150
25
2-Ø 10 mm
400 Spreader beam Figure 4. Test setup of reinforced concrete beams. 5. Test Results and Discussion Table 3. These are the properties of the concrete
used in the prisms that later were repaired
using each of the four RMs. It should be noted
that the flexural strength was assessed for an
undamaged concrete prism. Each result was an
average of three specimens. 5.2 Flexural Strength of Damaged/Repaired
Plain Concrete Prisms The mechanical properties of the concrete
substrate which received the RMs are shown in 165 Performance of Cement-Based Patch Repair Materials in Plain and Reinforced Concrete Members Performance of Cement-Based Patch Repair Materials in Plain and Reinforced Concrete Members a)
Bond failure of RM1, RM2 and RM3 b) Crushing failure of OPC
Figure 5. Modes of failure of the specimens observed during Slant Shear Bond test. a)
Bond failure of RM1, RM2 and RM3 b) Crushing failure of OPC b) Crushing failure of OPC a)
Bond failure of RM1, RM2 and RM3 Figure 5. Modes of failure of the specimens observed during Slant Shear Bond test. Figure 6. Flexural strength of prisms. 13.91
14.05
12.43
13.31
8.2
7.42
7.32
8.02
18.18
16.12
16.86
13.85
0
2
4
6
8
10
12
14
16
18
20
RM1
RM2
RM3
OPC
Flexural Failure Load (kN)
Repair Material
Undamaged
Damaged
Repaired
Repaired Material Undamaged
Damaged
Repaired Figure 6. Flexural strength of prisms. break was reported instead of the flexural stress
or modulus of rupture, since the latter might be
misleading in the case of damaged prisms
where the cross-sectional area was reduced as a
result of the damage. It can be seen that the
damage caused large reduction in the flexural
strength by 45% of the undamaged prism. However, all repaired prisms failed at a higher
load than that of the undamaged prisms. An
increase in the failure load of 30%, 14%, 36%,
and 5% were observed in prism repaired using
RM1, RM2, RM3, and OPC, respectively. This
clearly shows that the RMs are very effective break was reported instead of the flexural stress
or modulus of rupture, since the latter might be
misleading in the case of damaged prisms
where the cross-sectional area was reduced as a
result of the damage. It can be seen that the
damage caused large reduction in the flexural
strength by 45% of the undamaged prism. However, all repaired prisms failed at a higher
load than that of the undamaged prisms. An
increase in the failure load of 30%, 14%, 36%,
and 5% were observed in prism repaired using
RM1, RM2, RM3, and OPC, respectively. This
clearly shows that the RMs are very effective when loaded under tension. In all prisms, no
debonding failure was observed. The location of
the cracks and the failure lines of the prisms are
shown in Fig. 7. 5.2 Flexural Strength of Damaged/Repaired
Plain Concrete Prisms RM1 and RM3 were so strong
that the concrete cracked just at the end of the
patch where the substrate concrete was weaker
in tension than the RM (Fig. 8). 5.3 Compressive Strength of Damaged/
Repaired Plain Concrete Prism p
This group of prisms had the same damage
as the ones that were tested under flexure and 166 A.H. Al-Saidy
a) Failure mode of RM2, and OPC
b) Failure mode of RM1 and RM3
Figure 7. Schematic of failure modes in damaged/repaired prisms. (a) RM1 (b) RM2
(c) RM3 (d) OPC
Figure 8. Modes of failure observed during flexural test of prisms for all repaired materials. repaired similarly; however, they were tested
under axial compression. A summary of the
results of the compressive strength of the
undamaged,
damaged,
damaged/repaired
prisms is shown in Fig. 9. As expected, all
damaged prisms failed at a lower load than the
reduced cross-sectional area caused by t
damage. However, all damaged/repair
prisms showed equal or higher failure loa
compared to the undamaged prisms, indicati
that the RM contributed to the axial lo
carrying capacity of the damaged/repair A.H. Al-Saidy a) Failure mode of RM2, and OPC
b) Failure mode of RM1 and RM3
Fi
7 S h
ti
f f il
d
i
d
d/
i
d
i a) Failure mode of RM2, and OPC
b) Failure mode of RM1 and RM3 a) Failure mode of RM2, and OPC b) Failure mode of RM1 and RM3 b) Failure mode of RM1 and RM3 Figure 7. Schematic of failure modes in damaged/repaired prisms. (a) RM1 (b) RM2
(c) RM3 (d) OPC
Figure 8. Modes of failure observed during flexural test of prisms for all repaired materials. (b) RM2 (a) RM1 (a) RM1 (b) RM2 ( )
(d) OPC (c) RM3 (c) RM3 (c) RM3 Figure 8. Modes of failure observed during flexural test of prisms for all repaired materials. reduced cross-sectional area caused by the
damage. However, all damaged/repaired
prisms showed equal or higher failure loads
compared to the undamaged prisms, indicating
that the RM contributed to the axial load
carrying capacity of the damaged/repaired
prisms. repaired similarly; however, they were tested
under axial compression. A summary of the
results of the compressive strength of the
undamaged,
damaged,
damaged/repaired
prisms is shown in Fig. 9. As expected, all
damaged prisms failed at a lower load than the
undamaged and repaired prisms due to the repaired similarly; however, they were tested
under axial compression. A summary of the
results of the compressive strength of the
undamaged,
damaged,
damaged/repaired
prisms is shown in Fig. 9. 5.3 Compressive Strength of Damaged/
Repaired Plain Concrete Prism As expected, all
damaged prisms failed at a lower load than the
undamaged and repaired prisms due to the 167 Performance of Cement-Based Patch Repair Materials in Plain and Reinforced Concrete Members Performance of Cement Based Patch Repair Materials in Plain and Reinforced Concrete Members
Figure 9. Compressive strength of prisms. (a) RM1 (b) RM2
(c) RM3 (d) OPC
Figure 10. Modes of failure observed during compressive test for all repaired materials. 199.5
297.5
191
276.6
123.5
196.8
140.1
189.5
203.2
285.7
201.2
266.6
0
50
100
150
200
250
300
350
RM1
RM2
RM3
OPC
Compressive Failure Load (kN)
Repair Material
Undamaged
Damaged
Repaired
Repaired Material
Compressive Failure Load (kN) 199.5
297.5
191
276.6
123.5
196.8
140.1
189.5
203.2
285.7
201.2
266.6
0
50
100
150
200
250
300
350
RM1
RM2
RM3
OPC
Compressive Failure Load (kN)
Repair Material
Undamaged
Damaged
Repaired
Repaired Material
Compressive Failure Load (kN) Figure 9. Compressive strength of prisms. Figure 9. Compressive strength of prisms. (a) RM1 (b) RM2
(c) RM3 (d) OPC
Figure 10. Modes of failure observed during compressive test for all repaired materials. (b) RM2 (a) RM1 (b) RM2 (a) RM1 ( )
(c) RM3 (d) OPC
essive test for all repaired materials (c) RM3 ( )
( )
Figure 10. Modes of failure observed during compressive test for all repaired materials. In
the
compressive
strength
test,
the
specimens repaired with OPC failed due to
compression without detachment of the repair
material from their locations, while, the other
specimens repaired with RM1, RM2 and RM3
failed due to shear because the RMs detached
from their locations as indicated by the inclined
failure line in Fig. 10. The OPC RM showed
better bond behavior under compression which
was also observed in the slant bond test
discussed in Section 5.1. 5.4 Repaired Reinforced Concrete Beams
Unlike plain concrete, which is very brittle,
reinforced concrete behaves in a ductile manner. Therefore, it was possible to observe the
behavior of the combination of repair + concrete
in all loading stages as encountered in the field. Load versus mid-span deflections are compared
in Fig. 11, showing the undamaged, damaged,
and damaged/repaired beams. From Fig. 11, it
can be observed that all of the repaired beams
resisted an ultimate load slightly higher than
the undamaged beam. All beams failed in a 168 A.H. Al-Saidy A.H. Al-Saidy
Figure 11. Load vs. 5.3 Compressive Strength of Damaged/
Repaired Plain Concrete Prism mid span deflection of undamaged, damaged, and damaged/repaired beams. Figure 12. Repaired beam after failure. ductile manner as in the case of under
reinforced concrete beams. Failure occurred as
the steel yielded, and ultimately the concrete
was crushed in the compression zone. All beams
showed similar behaviors until failure as
expected. However, the objective was to
observe the behavior of the patch RM under all
loading stages, including in the inelastic (post-
yielding) stage. A summary of the results is shown in Table 4. In all beams, the first cracks appeared in the
DRM1, the cracks first appeared in concrete at
3.3 kN whereas in the RM the cracks appeared
at 5.6 kN; the other repaired beams also
followed the same pattern (Table 4). The
variations in crack appearance can be attributed
to the difference in the tensile strength as the
RMs have higher tensile strength than the
concrete substrate. In all beams, the RMs
remained attached to the concrete until the
beam failed (Figs. 12 and 13), and no major
difference was observed among the repairs
Mid span deflections (mm)
Load (kN) Mid span deflections (mm)
Load (kN) Load (kN) Load (kN) Mid span deflections (mm) gure 11. Load vs. mid span deflection of undamaged, damaged, and damaged/repaired beams Figure 12. Repaired beam after failure. DRM1, the cracks first appeared in concrete at
3.3 kN whereas in the RM the cracks appeared
at 5.6 kN; the other repaired beams also
followed the same pattern (Table 4). The
variations in crack appearance can be attributed
to the difference in the tensile strength as the
RMs have higher tensile strength than the
concrete substrate. In all beams, the RMs
remained attached to the concrete until the
beam failed (Figs. 12 and 13), and no major
difference was observed among the repairs
made with different materials. ductile manner as in the case of under
reinforced concrete beams. Failure occurred as
the steel yielded, and ultimately the concrete
was crushed in the compression zone. All beams
showed similar behaviors until failure as
expected. However, the objective was to
observe the behavior of the patch RM under all
loading stages, including in the inelastic (post-
yielding) stage. ductile manner as in the case of under
reinforced concrete beams. Failure occurred as
the steel yielded, and ultimately the concrete
was crushed in the compression zone. All beams
showed similar behaviors until failure as
expected. 5.3 Compressive Strength of Damaged/
Repaired Plain Concrete Prism However, the objective was to
observe the behavior of the patch RM under all
loading stages, including in the inelastic (post-
yielding) stage. A summary of the results is shown in Table 4. In all beams, the first cracks appeared in the
concrete substrate and then in the RM. In beam 169 Performance of Cement-Based Patch Repair Materials in Plain and Reinforced Concrete Members Table 4. Summary of test results of repaired reinforced beams. Beam
Load for first
crack in
concrete (kN)
Load for first
crack in repair
material (kN)
Maximum
Load (kN)
Maximum
Deflection (mm)
C
-
-
19.2
47.0
D
-
-
18.5
45.0
DRM1
3.3
5.6
19.4
41.0
DRM2
4.0
6.2
19.3
44.0
DRM3
4.4
6.7
19.3
35.5
DRM4
5.0
6.9
19.7
43.0
Figure 13. Close up view of the repair material (RM2) at failure. Repair material
Concrete Table 4. Summary of test results of repaired reinforced beams. Beam
Load for first
crack in
concrete (kN)
Load for first
crack in repair
material (kN)
Maximum
Load (kN)
Maximum
Deflection (mm)
C
-
-
19.2
47.0
D
-
-
18.5
45.0
DRM1
3.3
5.6
19.4
41.0
DRM2
4.0
6.2
19.3
44.0
DRM3
4.4
6.7
19.3
35.5
DRM4
5.0
6.9
19.7
43.0 Table 4. Summary of test results of repaired reinforced beams. Table 4. Summary of test results of repaired reinforced beams. Figure 13. Close up view of the repair material (RM2) at failure. Repair material
Concrete Repair material Figure 13. Close up view of the repair material (RM2) at failure. References Abu-Tair AI (1997), Heat cycling of epoxy
mortar repaired beams. ACI Materials Journal
94(2): 129-133. Courard L, Piotrowski T, Garbacz A (2014),
Near-to-surface properties affecting bond
strength in concrete repair. Cement and
Concrete Composites 46: 73–80. Al-Farabi S, Kalimur RM, Al-Gahtani A,
Hameeduddin M (2006), Behaviour of patch
repair of axially loaded reinforced concrete
beams. Cement and Concrete Composites 28:
734–741. p
Garbacz, A., Górka, M., Courard, L., (2005), On
the effect of concrete surface treatment on
adhesion in repair systems. Magazine of
Concrete Research 57(1): 49–60. American Society for Testing and Materials,
ASTM C109. Standard Test Method for
Compressive Strength of Hydraulic Cement
Mortars. Ghassan N, Zai-UL-Hasan C (1999), Reinforced
concrete repairs in beams. Construction and
Building Materials 1: 195-212. Liu YP, Wang FZ, Liu MY, Hu SG (2014), A
microstructural
approach
to
adherence
mechanism of cement and asphalt mortar
(CA mortar) to repair materials. Construction
and Building Materials 66: 125–131. American Society for Testing and Materials,
ASTM C469. Standard Test Method for Static
Modulus of Elasticity and Poisson's Ratio of
Concrete in Compression. American Society for Testing and Materials,
ASTM C78. Standard Test Method for
Flexural Strength of Concrete (Using Simple
Beam with Third-Point Loading). Mangat PS, Limbachiya MK (1995), Repair
material properties which influence long-
term performance of the concrete structure. Construction and Building Materials 11(2): 81-
90. American Society for Testing and Materials,
ASTM C882. Standard Test Method for Bond
Strength of Epoxy-Resin Systems used with
Concrete by Slant Shear. Mangat PS, Limbachiya MK (1997), Repair
materials properties for effective structural
application. Cement and Concrete Research
27(4): 601-617. Al-Zahrani MM, Maslehhuddin M, Ibrahim M
(2003) Mechanical properties and durability
characteristics of polymer- and cement-based
repair materials. Cement and Concrete Research
25: 527-537. ( )
Margan DR (1996), Compatibility of concrete
repair materials and systems. Construction
and Building Materials 10(1): 57-67. Qian J, You C, Wang Q, Wang H, Jia X, (2014),
A method for assessing bond performance of
cement-based repair materials. Construction
and Building Materials 68: 307–313. Cabrera JG, Al-Hasan AS (1997), Performance
properties of the concrete repair materials. Construction and Building Materials 11(5-6):
283-290. g
Rio O, Andrade C, Izquierdo D, Alonso C
(2005), Behavior of patch-repaired concrete
structural elements under increasing static
loads to flexural failure. Journal of Materials in
Civil Engineering 17: 168-177. 6. Conclusion
Compatibility of the mechanical properties
between the RM and the substrate concrete
is very important. RM4 made with OPC
was the nearest in mechanical properties to
that of the substrate concrete and hence
more compatible. The failure mode of RM4
in slant shear and axial compression was the
best among all RMs in terms of adhesion to
the substrate concrete at failure, indicating
the importance of compatibility between the
substrate
concrete
and
the
RMs
in
enhancing the performance of the composite
material. This study presented test results on the behavior
of patch RMs in plain and reinforced concrete. Four RMs were used and each differed in the
mechanical properties of compressive strength,
tensile strength and modulus of elasticity. Based on the test results, it can be concluded
that:
Damage in plain concrete reduced the
flexural and axial capacity considerably.
All RMs performed very well in restoring
the strength lost due to damage and shared
the load with the substrate concrete in
flexure and axial compression of plain
concrete.
In reinforced concrete beams, no difference
was observed in the behavior of the RMs
and all RMs showed good bond till failure. 170 A.H. Al-Saidy A.H. Al-Saidy References Courard L, Schwall D, Garbacz A, Piotrowski T
(2006), Effect of concrete substrate texture on
the adhesion properties of PCC repair
mortar. In: Aguiar JB, Jalali S, Camões A,
Ferreira R.M. (Eds). Proceedings ISPIC 2006
International Symposium Polymers on Concrete
Guimarães, Oficinas Gráficas de Barbosa and
Xavier, Lda, Braga, Portugal 99–110. Ueda H, Tamai Y, Kudu T (2011), Evaluation of
the
durability
of
cement-based
repair
materials. Quarterly Report of Railway
Technical Research Institute, Japan 52(2): 92-96. 171
|
W2800491299.txt
|
https://link.springer.com/content/pdf/10.1007%2F978-3-319-52980-6_3.pdf
|
en
|
Armenia: Transformational Peculiarities of the Soviet and Post-Soviet Higher Education System
|
Palgrave studies in global higher education
| 2,018
|
cc-by
| 7,875
|
CHAPTER 3
Armenia: Transformational Peculiarities
of the Soviet and Post-Soviet Higher
Education System
Susanna Karakhanyan
Introduction
Schools of higher learning were initiated in Armenia as early as the ninth
century and one of the first schools was Tatev Academy. From the thirteenth through the fifteenth century, the fame of some schools spread
beyond the borders of the country. This included the University of
Gladzor, which celebrated its 700th anniversary in 1980 under the aegis
of UNESCO (2000). Historically, schools of higher learning were located
in churches. Strict management rules were applied as early as the thirteenth century to ensure adequate qualifications for teachers and admission of the best students. In medieval Armenia, schools of higher learning
had already begun conferring the scientific degrees of “Archimandrite”
and “Rabbi” upon successful completion of oral and written examinations, and thesis defence (ibid).
Armenian schools of higher learning saw a major expansion between
the eighteenth and nineteenth century to various worldwide locations
S. Karakhanyan (*)
ANQA Accreditation Commission, Yerevan, Armenia
© The Author(s) 2018
J. Huisman et al. (eds.), 25 Years of Transformations of Higher
Education Systems in Post-Soviet Countries, Palgrave Studies in Global
Higher Education, https://doi.org/10.1007/978-3-319-52980-6_3
73
74
S. KARAKHANYAN
such as Venice (Mkhitarian College), India (Calcutta College), Moscow
(Lazarian Seminary), Tiflis (Nersisian School), Echmiadzin (Gevorkian
Seminary), Madras and Rostov (ibid).
In 1920, schools were separated from the church and the whole system
of education became state-owned. As elsewhere in the Soviet republics,
primary, secondary and tertiary education was free, and tertiary education
was elite in both social and intellectual dimensions. Considering that education has always been a central value of Armenian culture, the country
enjoyed a 100 % literacy rate as early as 1960 (Suny 1996, 36).
With the fall of the Soviet Union (SU) in 1991 and the sudden cessation of Soviet standards and rules, the first years of Armenia’s independence were marked by a vacuum in education and culture. The abrupt
absence of a dominating power created social and political confusion, thus
filling the vacuum with the standards of a new and more powerful country
(Terzian 2010) mainly influenced by educational policies from Anglophone
and Anglo-centric systems. In the early 1990s, Armenia made substantial
changes to a centralised and regimented system that evolved with advantages and disadvantages.
The aim of this chapter is to explore the HE landscape in Armenia before
and after the fall of the Soviet regime and the respective transformations
reflected by social needs, economic demands and political goals. In particular, we will look at the Soviet model of institutional diversity in Armenia,
followed by the main drivers of transformation after the fall of the Soviet
regime and the factors that stimulated or impeded institutional differentiation (van Vught 2007). The chapter will culminate in a presentation of the
current institutional landscape and the contextual factors affecting it.
The Soviet Model of Institutional Diversity
in Armenia
Armenia became part of the SU in December 1923. Consequently, at
the start of the twentieth century, the whole concept and ideology of
education radically changed to become permeated with one idea: the
collective self, which became more important than the individual self
(Sarafian 1930). By becoming part of the SU, the higher education
(HE) system in Armenia witnessed a cessation of the ecclesiastical era
and the beginning of horizontal differentiation with the emergence of
professionally oriented schools, new professionally oriented institutes, universities and art schools to serve societal needs (Sarafian 1930). The only
ARMENIA: TRANSFORMATIONAL PECULIARITIES OF THE SOVIET...
75
university from the Soviet era in Armenia, the National University of
Armenia, was renamed Yerevan State University (YSU) in 1922. It was
founded in 1919 by Ministerial decree and was established based on a rich
history of higher learning preserved by the church; it “renewed the ancient
traditions of Armenian scholarship in language and history that during
600 years of foreign occupation had flourished only among the diaspora
abroad” (NAS 2004, 11).
From the 1920s to the fall of the Soviet regime, the system evolved in
line with Communist Party (CP) directives, socialist and communist ideology and the demands of the industrialisation agenda. Higher education institutions (HEIs) in that period were purely public and free of
charge. As early as at the beginning of the 1920s in tandem with the
Cultural Revolution reforms that swept the USSR, technical and agricultural schools as well as workers’ universities gradually evolved into new
professionally oriented institutes (e.g. in 1922 Armenian State Pedagogic
Institute and in 1933 Yerevan Polytechnic Institute).
In 1930, in accordance with the government and on the basis of YSU
faculties, independent professionally oriented institutes were established.
Among these were the State Medical Institute, the Armenian Construction
Institute, and the Yerevan Agriculture Institute. In the 1970s, a major boom
in HE enrolment was recorded (Fig. 3.1). This could be partly explained by
industrialisation policy requiring more educated employees and partly by
the full transition from seven-year education to ten-year secondary education as well as massive provision for compulsory secondary education (with
consideration of vocational education) for all citizens (Chabe 1971).
58.1
54.4
55.7
54.8
87
–8
8
13
19
85
–8
6
13
19
84
19
–8
1
80
19
–7
1
70
# of HEIs
13
–8
5
13
12
19
60
57.5
20.2
10
–6
1
11.1
19
9
–4
1
–3
1
30
19
19
19
–2
0
1
3.4
40
8
19
60
40
20
0
#of students
Fig. 3.1 Number of HEIs with total enrolment figures (in thousands) at the
start of selected academic years (Data collected from different sources: UNESCO,
HE in the USSR, Monographs on HE edited by L. C. Barrows; Khudaverdyan,
K. S, 1960)
76
S. KARAKHANYAN
Based on a strictly centralised model, the HE system was uniform. In
1988 just before the fall of the Soviet regime, there were 13 HEIs in
Soviet Armenia, of which only YSU was qualified as a full university. The
others were 11 professionally oriented institutes and 1 conservatorium.
The schools of higher learning had a combined student body of 55,700
specialising in 103 professions leading to a 5-year diploma specialist qualification (Table 3.1).
During the Soviet regime, Armenia had one of the highest percentages
of HE attendees per capita in comparison to other USSR republics, and
science was a particularly popular field of study (NAS 2004). One major
development during Soviet times that drastically affected HEI research
capacity was the removal of research from HEIs. It was placed under the
Armenian Branch of the Academy of Sciences of the USSR in 1935. As a
result of this separation, the boundaries between professional and research
HEIs gradually blurred, as the only distinction between the HEIs was
whether they comprehensively covered a variety of study fields or only one
Table 3.1
Higher education landscape in Armenia in 1991
Type of HEIs
Full university
Number
of HEIs
1
Professionally
oriented institute
11
Conservatorium
1
Specific example
Description
Yerevan State
University
Higher, postgraduate and
supplementary education in a
wide variety of natural and
sociological fields, science,
technology and culture, as
well as providing
opportunities for scientific
research and studies
HEI conducting specialised
and postgraduate academic
programmes and applied
research in a number of
field-related scientific,
economic and cultural
branches
HEI preparing specialists in
the field of music, providing
qualifications, development
and postgraduate academic
programmes
Yerevan Polytechnic
Institute
Yerevan Komitas State
Conservatorium
ARMENIA: TRANSFORMATIONAL PECULIARITIES OF THE SOVIET...
77
single field. Although the separation strictly served the Soviet agenda of
supporting military-industrial complexes, it also resulted in the establishment of a rich tradition in research activities, particularly in physics, and
ensured strong government support to promote education in science and
engineering in Armenia. The Academy of Sciences became a centre of science and technology research providing support services to the entire SU
(NAS 2004). Just before the fall of the regime, there were about 36
research institutions within and outside the Academy. In 1940, 11 of the
36 belonged to the Armenian Branch of the Academy of Sciences of the
USSR and directly reported to either federal agencies in Moscow or to
local Armenian ministries (Khudaverdyan 1960) (NAS 2004).
All the HEIs were under the auspices of the Ministry of Education and
Science (MoES), with respective line ministries for the Medical Institute
and the Agriculture Institute. They were state funded and followed the
model of curriculum and teaching methods imposed from Moscow. The
HE system was unitary with no differentiation between professional and
academic programmes/qualifications and a strong bias towards the provision of practical knowledge directly linked with industry (Sarafian 1930).
Further, HE system uniformity was spelled out in a centralised and unified approach to HEI governance, since the only directives eligible for
implementation came from a higher level, the Central CP in Moscow, and
were imposed on HEIs without any right to deviate. In fact, the CP considered education too important to delegate to education professionals,
and it was thus the political leaders who designed education policies and
steered the respective developments (Chabe 1971). One of the negative
impacts of such a centralised approach was a decrease in system capacity to
develop and innovate. The isolated system gradually turned the HE leaders in Armenia and other Soviet republics into mere implementers with no
opportunities to reflect on the imperatives coming from Moscow or question approaches related to content and methodology.
Although paralysed in the sphere of social sciences, the technical and
natural sciences proliferated in Armenia. Just before the fall of the SU,
Armenia enjoyed a strong body of professionals advancing research in the
fields of hydro-energy, nuclear energy, radio-electronics, machinery production, precise machine-making, laser technology, biochemistry, microbiology, and light and heavy textile industry. The number of persons per
10,000 enrolled in Armenian HE was 161, compared to the overall average of 177 for the USSR (UNESCO 1990). From every thousand
employed people, 192 and 222 had a higher and secondary professional
education, respectively (UNESCO 2000).
78
S. KARAKHANYAN
Post-Soviet Transformations
After the dissolution of the SU, Armenia faced challenges related to resurrection of its identity and recovery of its economic, cultural and educational values. The country was in major need of transformation at different
levels to ensure its survival and later competitiveness at the international
level. The economic crisis and political tensions were priority issues to
be dealt with, as they were caused by the radical change in the political
system.
Firstly, the country experienced a drastic earthquake in 1988 followed
by political tension with neighbouring Azerbaijan over a historically
Armenian territory, Mountainous Karabagh. This geared the major investments of the Armenian government from development towards a vision of
preserving the national identity and resurrecting rich historical and cultural legacy. Two major trends related to political tension evolved, which
eventually had a major impact on the socio-economic development of the
country. One was an inflow of refugees from Azerbaijan, and the other was
the “brain drain” of Armenian human capacity to countries offering more
opportunities for growth. According to CARIM, major social changes
resulted from hosting refugees from Azerbaijan and other Soviet republics; there were over 420,000 between 1988 and 1991, with 360,000
from Azerbaijan alone (UNHCR 2004; CARIM 2013). The trend is still
persistent with a refugee inflow from Syria. Further, a major outflow of
both citizens and refugees has been registered. The 2004 UNHCR
Statistical Yearbook estimates the number of outflowing people from
Armenia to developed countries to be above 13,000. According to the
State Migration Authority data as of 2013, about 42,800 people left the
country in 2012 without returning. Thus, the two trends have contributed immensely to the existing lag in economic development and consequently reflected in government investment into HE.
While popular during Soviet times, Armenian industry has now declined
and the country has experienced a major blow to the economy. Caused by
political tension, war and blockades, most industries were closed, which led
to rising unemployment and economic paralysis. In addition to economic,
political and social issues, the system had to deal with the legacy of several
decades of a communist regime that was deeply rooted in all aspects of life
and therefore the culture, beliefs and values of Armenia (Kozma and
Polonyi 2004) (Zelvys 2004). Formed throughout the 70 years of the
Soviet regime, exceptionally peculiar culture and values made the transformations to the market economy and democracy complicated, leading to
ARMENIA: TRANSFORMATIONAL PECULIARITIES OF THE SOVIET...
79
distortions in many cases. These factors included Soviet trust vs. Western
responsibility, nihilism and negotiation vs. competition, humanity vs. professionalism, truth vs. rules, faith vs. stimulus, “universalism” vs. individualism, spirituality vs. interests and charity vs. justice (Khrushcheva and
Benvenuti 2002).
In the 2000s, Armenia demonstrated steady economic growth until the
global crisis hit the economy. According to the Index of Economic Freedom
2009, Armenia was ranked the 39th most economically free nation in the
world, and as of 2012 it had made a full transfer to the market economy
(MoES 2014). In 2009, the real GDP rate declined by 14.1 %, followed
by a slow recovery, registering a GDP growth of 2.2 % in 2010 and 4.7 %
in 2011.
So, what were the changes in the HE arena based on these trends? In
tandem with the changes in the political system, changes took place in
HE. After the fall of the communist regime, the liberalisation of the country and a move towards a market-driven economy was the apparent trend.
The first step was related to the resurrection of the Armenian identity and
was registered with curricula emphasising Armenian language, history and
culture. The Armenian language became the dominant language of
instruction followed by English, although Russian was still widely taught
as a foreign language. Because of the political tension, few and fragmented
changes were possible in HE in the early 1990s.
As seen from the figure below, the enrolment of students in the HE
sector remained unchanged up to 1999. Starting with the 1999–2000
academic year, a move from “elite” HE to mass education became apparent. This trend was also conditioned by the appearance of private sector
providers enforced by the Law on Education adopted in 1999. The same
cannot be said about vocational education providers (so-called technicums
or professional colleges/uchilishe). This sector has remained relatively
constant since Soviet times (Fig. 3.2).
Massification of HE is explained by several factors: the demand for a
more qualified workforce in the market, the growing prestige of HE
enrolment over Vocational Education and Training (VET) and the country’s strategic priority of establishing a knowledge economy.
As can be seen from Fig. 3.3, among the most preferred programmes
offered by HEIs predominance is in humanities, education and pedagogy,
economics and management, the agro/food sphere, public health and culture and arts. To a lesser extent, programmes in physical-mathematical and
natural sciences are also registered as preferred. There has been a steady
80
S. KARAKHANYAN
140
120
100
97.8
80
60
40
20
61.7 60.7 65.6
39.4 35.5 36 38.5
27.2 30.1 28.7 31
21.1 24.2 24.3
72.3
77.9
112.2114.4114.6111
105.8
85.1
95.3 90.1
85.9
29.4 31.1 30.5 30.8 30.8 31.1 31.8 29.9 29.6 29.2 29.330.1
0
Middle voca onal educa onal ins tu ons
Higher educa onal ins tu ons
Fig. 3.2 Enrolment in tertiary education institutions in Armenia (Statistical
Yearbook of Armenia: http://www.armstat.am/en/?nid=45&year=2014)
agro food sphere
military education
construction and architecture
chemical technologies and biotechnology
informatics and computer engineering
economics and management
culture and art
public health
education and pedagogy
social sciences
humanities
natural science
physical-mathematical science
0
5000
2013/14
10000
15000
20000
25000
30000
35000
2009/10
Fig. 3.3 Student enrolment per major (BA level) (Statistical Yearbook of
Armenia): http://www.armstat.am/en/?nid=45&year=2014)
rise in information technologies throughout the last 6 years, which is a
good sign of demand on the HE system by the labour market.
One of the consequences of the granted freedom in HE provisions is
the overabundance of professionals in management and economics, law,
ARMENIA: TRANSFORMATIONAL PECULIARITIES OF THE SOVIET...
81
humanities and some spheres of engineering. This adds to unemployment
to a major extent, taking it from 3.8 % in 2008 to 18.7 % in 2009 and
beyond (NSSRA 2014).
According to the Statistical Yearbook of Armenia (2014), only 62.8 %
of HE graduates were actually employed in 2013.
Further in terms of employment, the official demand from the labour
force (for wage earners) as submitted by employers is steadily growing,
while there are fewer and fewer applicants per vacant position. A high
degree of unemployment is apparent and it has marred the economy of the
country for two major reasons. Firstly, HEIs have continued offering traditional study programmes based on the academic standards stipulated by
the government without harmonising the offered qualifications with market needs. Secondly, the market itself was and still is in the process of formation with high reliance on the personal capacities of the workforce to
manage environmental and organisational changes, rather than on the
education and qualifications received (Fig. 3.4).
The period after 1998 can be characterised as a recovery period for the
country in all the dimensions: economic, political and cultural. One sign
was a steady increase in public spending on education until the late 2000s
when the next economic crisis hit in 2009. The education sector suffered
with a drastic cut of about 15 % from the original budget with no indication of recovery (WB 2013). Overall, according to the WB Public
Expenditure Review (WB 2011), public funding for tertiary education was
well below the regional average of 1.3 % and the OECD average of 2.0 %,
taking it down to 0.3 % of GDP and 10–12 % of total education spending,
1689
1800
1514
1600
1400
1000
800
600
1185
1129 1167
1200 1077
855
531
896
792
554
738
561 520
944
816
622
668
738
400
200
0
210 231
102 78 95 115 117 85 60
38 38
1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013
Demand of wage-earners, total
Applicants per vacant posi on
Fig. 3.4 Official labour force demand submitted by employers (Statistical
Yearbook of Armenia: http://www.armstat.am/en/?nid=45&year=2014)
82
S. KARAKHANYAN
respectively. Public funding for research and development (R&D) is as low
as 0.2 % of GDP and the priority allocation is for the NAS, which doubly
limits HEI research capacity. The 2010–2013 state budget allocated for
education and science underwent a steady decline, taking the share for
education and science to 2.4 % of GDP in 2013.
Considering that state allocation amounts to 23 % of the public HEI
budget on average, HEIs generally cover capital expenditures and salaries
from tuition fees. The budget of private HEIs is primarily based on tuition
fees and amounts to 94–100 %. According to the WB Report 2013, the
tuition level in Armenia is considered relatively high if measured against
the GDP and compared with lower-middle income countries on the
OECD list. It is calculated at 7–37 % of GDP per capita PPP, whereas the
maximum is 14 % of GDP per capita PPP in the USA (OECD, Education
at a Glance, 2012). Despite the steady increase in HEI gross enrolment
from 19.6 % in 2001 to 28.6 % in 2008, which compares well with the
countries at a similar economic level in the ECA region (WB report 2013),
equity of access is still a challenge.
HE Landscape in Armenia: Current State of Affairs
The reform of Armenian HE was initiated bottom-up in the early 1990s
by leading scholars and top management at some leading universities. The
first steps taken through pilot projects were the introduction of a two-tier
degree structure, changes to curricula, and student assessment systems. In
some cases, the changes were supported by international projects such as
Tempus, Open Society Institute Assistance Foundation Armenia, the
World Bank and the United States Departments of State and Education.
However, ambiguity with regard to what should be done in what sequence,
as well as how and why, resulted from a lack of clear vision for educational
reforms (Zelvys 2004) and insufficient administrative capacity for change
management, coupled with a lack of MoES guidance. In fact, the trend
was predominant reliance on international consultants instead of building
the capacities of local change owners. This, according to Fullan and Scott
(2009), may have made things worse instead of better as the use of external consultants was not cost-effective and has caused overreliance on
external support at the expense of developing inner capacity (Karakhanyan
et al. 2011).
Further, the newly introduced approaches raised the issue of legitimacy;
this was caused by lack of inquiry into the context in which the diffused
ARMENIA: TRANSFORMATIONAL PECULIARITIES OF THE SOVIET...
83
policy was planted (Karakhanyan et al. 2011). One example is the revision
of the approach to governance, which authorises public HEIs to be governed by a Governing Board (GB) with equal representation by stakeholders, state employees, renowned individuals, faculty members and
students. While democratic in nature, an absent preparatory phase
enabling the meaningful participation of such key stakeholders combined
with negligence of contextualisation later resulted in decision-making
manipulation.
Another trend was the unprecedented decline in the status and prestige
of scholars and researchers conditioned by decades of low payment, overload, insufficient and inadequate resources for teaching and research, and
demoralising management. The once highly prestigious profession lost its
attraction among youth, which resulted in an aging faculty and therefore
non-relevant methods and content delivery. One of the consequences of
this trend is a tendency for highly qualified staff to leave academia for
more lucrative positions so that vacancies are filled by less qualified
individuals.
Bologna Reforms
In the 2000s, geared towards the establishment of an independent country and a democratic society, the Armenian government began to reconsider the whole architecture of the HE system. In 2003, the MoES
developed the Strategy of HE Reforms, which led Armenia to join the
Bologna Process in 2005. The MoES took the initiatives at the major
policy making level, while giving HEIs some autonomy to make institutional and programme level changes. In May 2005 (Bergen Communiqué),
the Development Strategy of Education for 2008–2015 was adopted and put
into practice. The document was revised and reinforced through adoption
of the Law on the Republic of Armenia Education Development Strategy on
June 23, 2011. The main objectives were reflected in the State Program
for Education Development 2011–2015, which sets key objectives and strategic directions for HE revolving around widening access to HE, a national
qualifications framework, enhancement of quality assurance, revising
funding mechanisms, recognition and comparability of degrees, student
mobility, strengthening the ties between HE and the labour market and
ensuring effective governance and financial management of HEIs.
Drawing on the Strategy, the shift to a two-tier degree system (MoES
decree, 2004) has been completed and almost 100 % of students below
84
S. KARAKHANYAN
the doctoral level are enrolled in two-cycle programmes. All HEIs issue
Diploma Supplements and implement the European Credit Transfer and
Accumulation System (ECTS) as of 2008, although with some difficulties.
However, the move towards a two-tier degree has proved to have insufficient legitimacy for many professional sectors across the system (e.g. medicine) and required a necessary revision in 2015–2016 to align with market
needs.
A new actor in the HE sector appeared as a direct result of joining the
Bologna Declaration. A buffer body, the National Center for Professional
Education Quality Assurance (ANQA), was established in 2008 as a new
governmental tool to hold all HEIs accountable for their operations and
outcomes. ANQA policies and procedures are aligned with the European
Standards for Quality Assurance in the European HE Area (ESGs) of the
European Network of Quality Assurance (ENQA) as well as the Guidelines
of Good Practices (GGP) of the International Network for Quality
Assurance Agencies in HE (INQAAHE). At HEI level, with major support from ANQA, internal quality assurance systems have been put in
place and HEIs have completed the first round of institutional self-
assessments. This is a move towards programme level self-assessment.
In 2011, the Armenian government adopted the National Education
Qualifications Framework of the Republic of Armenia (ANQF) consisting of
eight levels, and the responsibilities for the operation and maintenance of the
ANQF are within the jurisdiction of the MoES. Currently, the ANQF is under
revision based on the first round of its implementation and pilot evalution.
Armenia is pursuing a strong internationalisation policy. Currently, it is
a member of the United Nations, the Commonwealth of Independent
States, the Council of Europe and more than 40 other international organisations including OSCE, the World Bank, the International Monetary
Fund, WTO, WHO and UNESCO, to name but a few. Supported by
international missions, the leading HEIs are geared towards internationalisation and have been revamping their approaches to governance, administration and overall programme delivery. Internationalisation is pursued
through development and integration of such dimensions as:
–– Integration of an international dimension into the teaching, learning and research functions of universities by developing and
implementing respective policies and procedures
–– Identification and development of new skills, attitudes and knowledge in students, faculty and staff to promote internationalisation
ARMENIA: TRANSFORMATIONAL PECULIARITIES OF THE SOVIET...
85
–– Promoting scholar and student exchange and technical cooperation of HEIs
–– Development of ethos and culture that values and supports intercultural and international perspectives, initiatives and their quality
assurance
Among the current priorities on the government agenda is the refinement of the ANQF to move to the next step of self-certification, its full
implementation and respective recognition internationally. Also included
are a move from academic standards to a learning outcome approach to
programme development and delivery as well as a revision of funding
mechanisms.
Legal and Regulatory Developments
At the legal framework level, Armenian HE is regulated by the Law on
Education adopted in 1999 and the Law on Higher and Postgraduate
Professional Education adopted in 2004. The two documents clearly state
the vision of HE, which is aimed at international recognition, competitiveness and full integration into the European Higher Education Area
(EHEA). As for regulations related to the research dimension, a new Law
on the National Academy of Sciences was adopted in 2011 to ensure an
autonomous legal status for the NAS. This sort of division actually created
a gulf between HE and research and thus deprives HEIs of the opportunity to strengthen research functions.
The Law on Education (1999) and the Law on HE (2004) define the
overall governance framework for HE in detail, but with ambiguity in
favour of government control (WB 2013). However, due to their status,
some HEIs are also regulated by the Law on State Non-Commercial
Organizations and the Law on State Governing Institutions (SGIs), both
adopted in 2001. The latter was not specifically developed for HEIs and
does not take into account governance, autonomy and academic freedom
guaranteed in education laws, thus contributing even more controversy.
Given the current legal framework, the governance model can be defined
as semi-autonomous (WB 2013).
Recently, a new trend of transitioning to foundation status has been
observed. A foundation is a not-for-profit independent legal entity that
enjoys the following privileges (Hasan 2007):
86
S. KARAKHANYAN
(i) It is an independent legal entity.
(ii) It has a mission (or charter or mandate) to serve defined public (or
national and societal) interests in HE and research.
(iii) As a not-for-profit public interest legal entity, it has favourable tax
treatment on its incomes, assets and trading activities undertaken in
pursuit of its foundation goals.
(iv) It has the autonomy to raise funds and manage its assets in pursuit of
the foundation goals, for which it receives favourable tax treatment
(2007, p.7).
Throughout the last couple of years, six state HEIs have changed their
legal status to foundation, each by individual MoES decree. Thus, public
HEIs in Armenia are now legally differentiated between State NonCommercial Organizations (SNCOs) and foundations.
The same cannot be said for private and intergovernmental HEIs, since
they are less restricted in their operations. Private HEIs, depending on
their status, are regulated by the Law on Education, the Law on HE, the
Law on Enterprises and Entrepreneurial Activity, the Law on Foundations,
the Law on LLCs, the Law on Cooperatives and/or the Law on Joint
Stock Company (JSCs). Intergovernmental HEIs are regulated by the
Law on Education, the Law on HE and the respective legal frameworks of
their counterparts in home or host countries. Table 3.2 summarises the
types of Armenian HEIs by legal status as of 2015.
Funding
Diversification also affected the funding of HEIs. Public HEIs, which
were previously run 100 % on the state budget, now have legal authorisation to charge tuition fees. This has been the case since 1999. By the late
2000s, the average proportion of HEI budget share coming from the
government was only 20 %, with the rest coming predominantly from
student fees. Paradoxically, “public” HEIs in Armenia actually receive very
little state contribution (WB 2013). New funding mechanisms include
competitive innovation funds with a target to promote HEI competitiveness. Funding mechanisms are currently under revision with a diversification goal in mind.
ARMENIA: TRANSFORMATIONAL PECULIARITIES OF THE SOVIET...
87
New Admissions Policy
The next major change was the revision of admissions policy, which moved
from exams given by individual HEIs to a Centralised Admission Exam
(CAE), which is both a school final and a university entrance examination.
From 2005 to 2012, only state HEIs utilised a centralised admission exam
process, which is organised and administered by the MoES’s State
Admission Commission (SAC). In the academic year 2012–2013, private
HEIs were also subject to CAE for the first time, which was a major blow
to the private sector threatening a total closure of some institutions.
Exceptions apply to some public HEIs that still reserved the right to conduct subject specific exams as supplementary to the CAE.
Admissions to foreign-affiliated institutions are governed by individual
HEIs, although these schools may choose to use state exam scores in their
admission decisions.
With the introduction of unified exams, the landscape has transformed
significantly due to intensified competition among HEIs for students. As
a result, weaker public and most private HEIs are now forced to revise
their approaches. The choice is whether or not to merge or revise their
missions and concentrate mostly on life-long learning or further education
courses to become more competitive in the market.
Classification and Ranking of HEIs
With regard to the vertical stratification of HEIs, under the Soviet
approach distinctions in quality were reflected in the privileged status of a
university vs. an institute. Recently, the MoES has come up with new
mechanisms to classify HEIs:
–– Ranking Web of Universities, which provides rankings per country
as well1
–– National classification of HEIs and ranking of programmes, based
on a pilot project conducted by the MoES to classify and then
rank HEI programmes within each classification.
The same trend could be observed implicitly, through the reputational
stratification of HEIs broadly discussed at the society level. Such a stratification mainly places public and intergovernmental HEIs on top as the
most prestigious HEIs. Recently, to move forward on the international
88
S. KARAKHANYAN
visibility agenda, highly prestigious universities have been invited to establish branch campuses in Armenia as exemplified by the MoES’s invitation
of Moscow State University. Despite a history of about 25 years, private
HEIs have not been able to live up to the standards expected by society.
The main indicator for the public at large is the quality of teaching staff,
availability of resources and, most importantly, national and international
recognition of awarded degrees.
Vertical stratification could also be described as promoted by national
accreditation, which serves as an accountability tool for the government and
a tool for financial allocation. According to the ANQA revised procedures
(2015), regardless of HEI legal status, those that fail to obtain accreditation
will be deprived of state funds and will have limits set on tuition fees to be
charged. This will become effective as of 2018. Thus, the new stratification
tool has the potential to substantially change the HE landscape.
Organisational Interrelationships
Last but not least, an aspect of diversity worth elaborating on is organisational interrelationships. Although not very significant in the context of
the developing Armenian system, these alliances create synergies in teaching and learning, research and community outreach activities.
Interrelationships in the context of Armenia are promoted in the following ways:
–– Static—empowered by buffer bodies or international organisations;
–– Dynamic—natural evolutions based on the needs of the HE system and society at large.
With regard to the static, in its quest to develop a quality culture,
ANQA tirelessly invests in the capacity building of different stakeholders.
Capacity building events and peer reviews organised by ANQA actually
create a collaborative culture and establish a firm platform for a quality
education dialogue.
SCS,2 through its grants, promotes research projects that bring together
faculty members from different HEIs and, in some cases, industries. The
same applies to international projects like Tempus, Erasmus+ and
Twinning, which actually bring together HEIs to collectively pursue project objectives. As a natural evolution to this process, for example, the
International Association of Educationalists (IntAE) has been established
to bring together professionals from Armenia, Georgia and Europe.
89
ARMENIA: TRANSFORMATIONAL PECULIARITIES OF THE SOVIET...
The dynamic form of interrelationships revolves around the establishment of industry-university collaborative partnerships, merging HEIs and
partnership agreements. A prominent example of an industry-university
collaborative partnership is Synopsis, which is the Silicon to Software™
partner for innovative companies developing electronic products and software applications. Additionally, it offers courses in microelectronics at the
bachelor and master levels and the degree is awarded with its collaborative
partners which include such IT leaders as YSU, SEUA, and RAU (Slavonic).
Most graduates of this partnership are then hired by Synopsis itself.
A recent trend of merging HEIs is gradually becoming apparent.
Currently, this form of interrelationship has been registered only in the
private sector with some HEIs merging to serve the same purposes.
However, at the government level there are also plans to merge public
HEIs to achieve economies of scale.
Classification of HEIs 2014–2015
The HE system in the 1990s was characterised by only 1 full university, 11
professionally oriented institutes and 1 conservatorium. The new Law on
Education adopted in 1999, however, allowed for a diversity of HE providers to enter the market. Thus, as seen from Fig. 3.5, from 1999 to
140
112.2114.4114.6 111
105.8
120
98
100
80
20
91
61.7 60.7 65.6
60
40
90
97.8
93 88 89
85.1
77.9
92
72.3
87
85
95.3
90
77
74
90.1
68
85.9
65
63
39.4 35.5 36 38.5
15
15
16
16
0
# of students
# of HEIs
Fig. 3.5 Distribution of HEIs and number of students (Statistical Yearbook of
Armenia: http://www.armstat.am/en/?nid=45&year=2014)
90
S. KARAKHANYAN
2008 a new type of HEI mushroomed. The initiators were mainly private
entrepreneurs (e.g. Armenian diaspora to the USA, leading professionals
in different fields like law and economics) and former leaders of public
HEIs. Table 3.2 illustrates the diversity of HE providers as of 2015.
However, having set no boundaries on private initiatives, the system
found itself with an abundance of private providers with quality levels that
are still largely questioned. Starting in 2008 with persistent MoES efforts,
the number of private HEIs was reduced from 98 in 1999 to 63 in 2013
and 57 in 2015. The decline in the number of private HEIs was due to (1)
toughening licensure criteria; (2) university mergers, which were applied
in very few cases; and (3) imposing state unified entrance exams on private
HEIs. The steady decline in enrolments is mainly explained by the decline
in birth rates throughout the 1990s and beginning of the 2000s.
Currently, the HE system in Armenia consists of public, private, intergovernmental and transnational HEIs. At the national level, the executive
authority to elaborate and implement government policies is the MoES,
which tends to its mandate in cooperation with regional and municipal
authorities. State HEIs operate under the responsibility of several ministries but most of them are under the supervision of the MoES. In total
there are 26 state HEIs, of which 16 with 14 branches are under MoES
jurisdiction; 4 HEIs were founded by intergovernmental agreements and
partly funded by the MoES; 2 HEIs are under the Ministry of Defence,
and one HEI is under each of the following ministries/bodies: Police,
Ministry of Emergency Situations and Mother See of Holy Etchimiadzin.3
Of all the public HEIs, YSU has a special distinction as the only Armenian
HEI with its own separate provision in the Law on Higher and Professional
Education and a separate mention in the national budget (Table 3.3).
As demonstrated above, driven by the vision of internationalisation and
the demands of the market economy, the HE landscape is gradually
becoming diverse with three full universities qualified as comprehensive as
well as specialised universities, institutes, academies, a conservatorium and
research institutes to meet a diversity of needs.
Private HEIs are out of the scope of MoES jurisdiction to a considerable extent; however, the government imposes accountability mechanisms.
Intergovernmental institutions are universities established on the basis of
agreement between two countries, for example, the American University
of Armenia, the Russian-Slavonic University of Armenia and the French
University of Armenia. The trend of promoting transnational providers is
escalating with the introduction of academic programmes within the
6
Law on
Foundations,
Law on
Education,
Law on HE
Law on
SNCOs,
Law on
Education,
Law on
HE
Foundation
14
SNCOs
Public
Law on
Foundations and
Agreement
between two
countries, Law on
Education, Law
on HE
2
Law on
Foundations,
Civil Code,
Law on
Education,
Law on HE
2
Intergovernmental Interstate
foundations
Intergovernmental
State
Armenian HEIs by legal status: regulatory landscape
Source: WB Report 2013, modified by author in 2015
No. of
HEIs
Major
governing
laws
Types
Table 3.2
Law on
LLCs; Law
on Coops,
Law on
JSCs, Law
on
Education,
Law on
HE
31
Law on
Foundations,
Law on
Education,
Law on HE
2
For-profit
Not-for-profit
foundations foundations
Local
Law on registering of
Legal Entities and
Records on Separate Units
of Legal Entities,
respective counterpart
institutions and Law on
Private Enterprises, Law
on Education, Law on
HE
9
branches of foreign
universities/transnational
providers
Foreign
Private
ARMENIA: TRANSFORMATIONAL PECULIARITIES OF THE SOVIET...
91
0
1
1
Offering
professional/applied
programmes and
conducting
field-specific
research
Total
44
Total
9
a
PUB Public, PRV Private, IG Intergovernmental, TN Transnational
66 HEIs + 35 RIs
Total
3
1
Offering specialised
and postgraduate
academic programmes
and scientific research
in a number of
field-related scientific,
economic and cultural
branches
Offer higher,
postgraduate and
supplementary
education in a wide
variety of branches of
natural and
sociological fields,
science, technology
and culture, as well as
providing
opportunities for
scientific research and
studies
PUBa PRV IG TN
Institute
Offer applied
programmes in
specific fields (arts,
police), leading to
bachelor, master
and candidate of
Science
qualifications, as
well as conduct
research in the field
Academy
Offers programmes
in the field of
music, leading to
bachelor, master
and candidate of
Science
qualifications
Conservatorium
Total
8
Total
1
Total
35
Public
(outside
NASRA)
1
Offer master
degrees and above
all conduct
scientific research
Research institute
PUB PRV IG TN PUB PRV IG TN PUB PRV IG TN PUB PRV IG TN Public
(within
NASRA)
10
24
2 8
3
6
0 0
4
3
1 0
1
0
0 0
34
Specialised university
Higher education landscape in Armenia in 2015
Full university
Table 3.3
92
S. KARAKHANYAN
ARMENIA: TRANSFORMATIONAL PECULIARITIES OF THE SOVIET...
93
framework of existing HEIs (e.g. the Armenian University of Economics is
hosting a programme from the United Kingdom). This also involves Armenia
hosting worldwide leading university branches (e.g. Moscow State Lomonosov
University), thereby bringing in a more diverse range of HEl providers.
All the HEIs are degree-awarding entities. All the private HEIs accredited by the MoES starting from 1999 under the old accreditation policy
issue state-standard diplomas, which qualifies the graduates as the same
level as those from public HEIs. However, the trust in degrees awarded by
private HEIs is much lower in the labour market and society at large.
Conclusions
For the last century and a half, HE in Armenia has been driven by ideological and political factors and undergone major transformations resulting
in system differentiation at horizontal and to some extent vertical level.
The first major transformation was compelled by the change to a Soviet
and socialist ideology and industrialisation agenda when joining the USSR
in 1923. Driven by the imperatives of the Soviet regime, there was a radical cessation of the ecclesiastic nature leading to the establishment of professionally oriented HEIs outside the church to meet the demands of
society. As shown in Table 3.1, this resulted in the emergence of a full
university as well as professionally oriented institutes and a conservatorium. The system was predominantly characterised by uniformity and was
centrally planned and controlled by the government, with no opportunity
for any other types of HE providers (e.g. private).
With the collapse of the Soviet system, HE in Armenia has undergone
a major transformation at the horizontal level. The institutional landscape
has expanded, not only in numbers but also in types, to include such HEIs
as academies, professionally oriented universities, educational centres
(foundations) and research institutes within National Academy of Sciences
of the Republic of Armenia (NASRA); this is very different to the landscape in 1991 (Table 3.1 vs. 3.2). Further, the system evolved to host
HEIs with diversity in terms of legal status: public, intergovernmental,
transnational and private, which could further be differentiated between
for-profit and not-for-profit providers. The contextual factors affecting the
alteration of the institutional landscape are mainly related to the move to
a market economy, national identity resurrection and the internationalisation agenda promoted by the government.
To achieve international visibility and respond to the changing trends
in HE, the transformations entailed the encouragement of private, inter-
94
S. KARAKHANYAN
governmental and transnational providers. On the other hand, concerned
with the rapidly increasing number of private providers with questionable
quality, the government took steps to regulate the market newcomers by
introducing accountability tools, for example by setting strict regulations
and licensing and accreditation policies. This has led to the closure or
merger of private and poorly performing public providers. Another example is the extension of centralised admission policy to private providers,
which eventually compelled the latter to reorient their missions. Thus, the
stimulators for horizontal differentiation could be summed up as driven
by market demands for modernisation of qualifications and massification
of HE as well as the government agenda for accountability, funding, internationalisation and recognition of qualifications.
Vertical differentiation is increasingly becoming a major concern for the
government, HEIs and stakeholders. HEIs are becoming part of international and national rankings and classifications to enable measurement of
achievements and comparative analysis of those achievements throughout
time and across systems. Among the steps leading to vertical differentiation is the government attempt to invite highly ranked HEIs to establish
branch campuses in Armenia (e.g. Moscow State University).
In sum, the transformations in the Armenian HE system have come in
the form of differentiation at diverse levels. Considering that the drivers
behind this differentiation are predominantly market requirements and
political strategies at the government level leading to international visibility, the HE system is predominantly governed by a balance of national and
global forces leading to convergence. Although it is still premature to
speak about the level of convergence in actual implementation practices,
HEIs in Armenia are becoming more convergent with those at the
European level through such major tools as the NQF and its alignment
with EQF, the independent quality assurance and accountability system,
operationalisation of the credit transfer and accumulation system and a
move towards two-tier education (bachelor and master).
Notes
1. http://www.webometrics.info/en/europe/armenia
2. State Committee of Science, established within the MoES in 2008 with a
mandate to improve the science sector in Armenia. The body mainly promotes research by offering grants.
3. http://studyinarmenia.org/hea
ARMENIA: TRANSFORMATIONAL PECULIARITIES OF THE SOVIET...
95
References
CARIM. 2013. Refugees, Displaced Persons and Asylum Seekers in Armenia.
RSCAS, CARIM East Project, European University Institute, Florence,
Italy.
Chabe, A.M. 1971. Soviet Educational Policies: Their Development,
Administration and Content. Association for Supervision and Curriculum
Development, 525–531 in August, 2015. www.ascd.org/ASCD/pdf/journals/
ed_lead/el_197102_chabe.pdf
Hasan, A. 2007. Independent Legal Status and Universities as Foundations. Paris:
UNESCO, International Institute for Educational Planning.
Karakhanyan, S., K. van Veen, and T. Bergen. 2011. Higher Education Policy
Transfer and Diffusion: The Case of Armenia. Higher Education Policy 24:
53–83.
Khrushcheva, N., and A. Benvenuti. 2002. Lessons of Transition: The Cultural
Contradictions and the Future of Russian Liberalization. New York: World
Policy Institute.
Khudaverdyan, K.S. 1960. Cultural Changes in the Soviet Armenia During Pre-
war Five-Year-Time-Spans. Izvestiya Academii Nauk Armyanskoi SSR 3–17.
Kozma, T., and T. Polonyi. 2004. Understanding Education in European-East
Frames of Interpretation and Comparison. International Journal of Education
Development 24 (5): 467–477.
MoES. 2014. Education for All 2015 National Review Report: Armenia. Paris:
UNESCO.
NAS. 2004. Science and Technology in Armenia: Toward a Knowledge-Based
Economy. Washington, DC: National Academy of Sciences, The National
Academies Press.
NSSRA. 2014. National Statistical Services of the Republic of Armenia. Yearbooks,
March 2015. http://www.armstat.am/en/?nid=45&year=2014
Sarafian, K. 1930. History of Education in Armenia. California: La Verne Leader.
Suny, R. 1996. Armenia, Azerbaijan, and Georgia: Country Studies. Dane
Publishing, Washington D. C.
Terzian, M.S. 2010. Curriculum Reform in Post-Soviet Armenia: Balancing Local
and Global Contexts in Armenian Secondary Schools. Dissertations/107, Loyola
University Chicago.
UNESCO. 1990. Higher Education in the USSR. Bucharest: CEPES.
———. 2000. The EFA 2000 Assessment: Country report, Armenia. Paris.
UNHCR. 2004. UNHCR Statistical Yearbook. UNHCR Refugee Agency,
Geneva, Switzerland.
van Vught, F. 2007. Diversity and Differentiation in Higher Education Systems.
CHET Anniversary Conference, Cape Town.
WB. 2011. Public Expenditure Review. Washington, DC: World Bank.
96
S. KARAKHANYAN
———. 2013. Addressing Governance at the Center of Higher Education Reforms
in Armenia. Washington, DC: World Bank.
Zelvys, R. 2004. Development of Education Policy in Lithuania During the Years
of Transformations. International Journal of Educational Development 24 (5):
559–571.
Susanna Karakhanyan is one of the founders of the National Center for
Professional Education Quality Assurance in Armenia (ANQA) as well as the first
President and current member of the (ANQA) Accreditation Commission. She is
the current President of the International Network for Quality Assurance Agencies
in Higher Education (INQAAHE). Dr. Karakhanyan holds an M.S.Ed. in
Educational Administration from the University of Pennsylvania, USA, and Ph.D.
in Social Sciences from the Radboud University Nijmegen, the Netherlands. Her
research interests evolve around higher education policy development, diffusion
and transfer, HE governance, administration and management in general and its
quality assurance, in particular.
Open Access This chapter is distributed under the terms of the Creative
Commons Attribution 4.0 International License (http://creativecommons.org/
licenses/by/4.0/), which permits use, duplication, adaptation, distribution and
reproduction in any medium or format, as long as you give appropriate credit to
the original author(s) and the source, provide a link to the Creative Commons
license and indicate if changes were made.
The images or other third party material in this chapter are included in the
chapter’s Creative Commons license, unless indicated otherwise in a credit line to
the material. If material is not included in the chapter’s Creative Commons license
and your intended use is not permitted by statutory regulation or exceeds the
permitted use, you will need to obtain permission directly from the copyright
holder.
|
|
https://openalex.org/W2942383839
|
https://hal.archives-ouvertes.fr/hal-02117153/file/lz5024.pdf
|
English
| null |
A simple and versatile microfluidic device for efficient biomacromolecule crystallization and structural analysis by serial crystallography
|
IUCrJ
| 2,019
|
cc-by
| 10,281
|
A simple and versatile microfluidic device for efficient
biomacromolecule crystallization and structural analysis
by serial crystallography Raphaël de Wijn, Oliver Hennig, Jennifer Roche, Sylvain Engilberge, Kevin
Rollet, Pablo Fernandez-Millan, Karl Brillet, Heike Betat, Mario Mörl, Alain
Roussel, et al. To cite this version: Raphaël de Wijn, Oliver Hennig, Jennifer Roche, Sylvain Engilberge, Kevin Rollet, et al.. A simple
and versatile microfluidic device for efficient biomacromolecule crystallization and structural analysis
by serial crystallography. International Union of Crystallography journal, 2019, 6 (3), pp.454-464. 10.1107/S2052252519003622. hal-02117153 A simple and versatile microfluidic device for efficient
biomacromolecule crystallization and structural analysis
by serial crystallography Distributed under a Creative Commons Attribution 4.0 International License HAL Id: hal-02117153
https://hal.science/hal-02117153v1
Submitted on 22 Jan 2020 L’archive ouverte pluridisciplinaire HAL, est
destinée au dépôt et à la diffusion de documents
scientifiques de niveau recherche, publiés ou non,
émanant des établissements d’enseignement et de
recherche français ou étrangers, des laboratoires
publics ou privés. HAL is a multi-disciplinary open access
archive for the deposit and dissemination of sci-
entific research documents, whether they are pub-
lished or not. The documents may come from
teaching and research institutions in France or
abroad, or from public or private research centers. Distributed under a Creative Commons Attribution 4.0 International License research papers A simple and versatile microfluidic device for
efficient biomacromolecule crystallization and
structural analysis by serial crystallography ISSN 2052-2525 ISSN 2052-2525 BIOLOGYjMEDICINE Raphae¨l de Wijn,a Oliver Hennig,b Jennifer Roche,c Sylvain Engilberge,d Kevin
Rollet,a Pablo Fernandez-Millan,a Karl Brillet,a Heike Betat,b Mario Mo¨rl,b Alain
Roussel,c Eric Girard,d Christoph Mueller-Dieckmann,e Gavin C. Fox,f Vincent
Olieric,g Jose´ A. Gavira,h Bernard Lorbera and Claude Sautera* Received 13 December 2018
Accepted 14 March 2019 aArchitecture et Re´activite´ de l’ARN, UPR 9002, CNRS, Institut de Biologie Mole´culaire et Cellulaire (IBMC), Universite´ de
Strasbourg, 15 Rue Rene´ Descartes, 67084 Strasbourg, France, bInstitute for Biochemistry, Leipzig University,
Bruederstrasse 34, 04103 Leipzig, Germany, cArchitecture et Fonction des Macromole´cules Biologiques, UMR 7257
CNRS–Aix Marseille University, 163 Avenue de Luminy, 13288 Marseille, France, dUniversite´ Grenoble Alpes, CEA,
CNRS, IBS, 38000 Grenoble, France, eStructural Biology, European Synchrotron Radiation Facility, 38043 Grenoble,
France, fPROXIMA 2A beamline, Synchrotron SOLEIL, L’Orme des Merisiers, Saint-Aubin, 91192 Gif-sur-Yvette, France,
gPaul Scherrer Institute, Swiss Light Source, Forschungsstrasse 111, 5232 Villigen PSI, Switzerland, and hLaboratorio de
Estudios Cristalogra´ficos, IACT, CSIC–Universidad de Granada, Avenida Las Palmeras 4, 18100 Armilla, Granada, Spain. *Correspondence e-mail: c.sauter@ibmc-cnrs.unistra.fr Edited by Z.-J. Liu, Chinese Academy of
Sciences, China Keywords: macromolecule; crystallization;
counter-diffusion; microfluidics; seeding; ligand
soaking; trace fluorescent labeling; serial
crystallography; room temperature; protein
structure; ChipX3. Determining optimal conditions for the production of well diffracting crystals is
a key step in every biocrystallography project. Here, a microfluidic device is
described that enables the production of crystals by counter-diffusion and their
direct on-chip analysis by serial crystallography at room temperature. Nine ‘non-
model’ and diverse biomacromolecules, including seven soluble proteins, a
membrane protein and an RNA duplex, were crystallized and treated on-chip
with a variety of standard techniques including micro-seeding, crystal soaking
with ligands and crystal detection by fluorescence. Furthermore, the crystal
structures of four proteins and an RNA were determined based on serial data
collected on four synchrotron beamlines, demonstrating the general applic-
ability of this multipurpose chip concept. PDB references: CCA-adding enzyme, 6ibp;
CCA-adding enzyme + CMPcPP, 6q52;
nanobody 02, 6gzp; protease 1, 6q3t; lipase,
6hw1; RNA duplex, 6ibq PDB references: CCA-adding enzyme, 6ibp;
CCA-adding enzyme + CMPcPP, 6q52;
nanobody 02, 6gzp; protease 1, 6q3t; lipase,
6hw1; RNA duplex, 6ibq Supporting information: this article has
supporting information at www.iucrj.org Supporting information: this article has
supporting information at www.iucrj.org 454
https://doi.org/10.1107/S2052252519003622 2.1. Biomacromolecules, biochemicals and chemicals 2.1. Biomacromolecules, biochemicals and chemicals The recombinant proteins used in this work include
protease 1 from Pyrococcus horikoshii (PhP1), the llama
nanobody PorM_02 (Nb02), a lipase from Thermomyces
lanuginosus (Lip; provided by Macrocrystal Oy, Finland), the
CCA-adding
enzyme
from
the
psychrophilic
bacterium
Planococcus halocryophilus (CCA), the TonB-dependent
heme/hemoglobin outer membrane transporter (OMT) ShuA
from the pathogen Shigella dysenteriae (OMT ShuA), the
human mitochondrial aspartyl-tRNA synthetase (hmDRS)
and aspartyl-tRNA synthetase 1 from the bacterium Thermus
thermophilus (ttDRS), which were purified as described
previously (Engilberge et al., 2018; Duhoo et al., 2017; Ernst et
al., 2018; Brillet et al., 2009; Sauter et al., 2015; Zhu et al., 2001). Horse hemoglobin was purchased from Sigma. The nine-base-
pair RNA duplex [r(CGUGAUCG)dC]2 was prepared as
described by Masquida et al. (1999). Stock concentrations and
storage buffers are indicated in Table 1. j
y
g
p
y
To expand the functionality and attractiveness of micro-
chips beyond crystallization and HTP screening, several teams
have explored the possibility of analyzing crystals directly in
their microfluidic environment (Yadav et al., 2005; Ng et al.,
2008; Sauter et al., 2007; Dhouib et al., 2009; Emamzadah et al.,
2009; Hansen et al., 2006; Stojanoff et al., 2011). Various
geometries and materials have been tested and have led to
promising results in terms of data collection, anomalous
phasing or time-resolved applications (Pinker et al., 2013;
Khvostichenko et al., 2014; Perry et al., 2013, 2014). The
difficulty of cryopreserving crystals to protect them from
radiation damage inside chips, owing to the wide flat surfaces
of the device causing vapor condensation and ice formation in
the cryojet, was first perceived as an obstacle. However, the
recent revival of multi-crystal data-collection techniques at
room temperature by the X-ray free-electron laser (XFEL)
community has changed the paradigm and popularized serial
crystallography (Chapman et al., 2011; Stellato et al., 2014;
Ayyer et al., 2015). In this context, microfluidic systems
provide promising solutions to prepare, handle and analyze
crystals at both synchrotron beamlines and XFELs (Heymann
et al., 2014; Sui et al., 2016). To facilitate the detection of CCA crystals by trace fluor-
escent labeling (TFL; Pusey et al., 2015), the protein was
fluorescently labeled with carboxyrhodamine-succinimidyl
ester (Invitrogen, catalog No. C6157) as described by de Wijn
et al. (2018). 1. Introduction Crystallography plays a central role in contemporary biology
because it enables the visualization of the 3D architecture of
biological macromolecules, which provides insights into their
cellular functions and partnerships on the atomic scale (Giege´
& Sauter, 2010; Jaskolski et al., 2014). Over the past two
decades, the advent of structural genomics and associated
high-throughput (HTP) technologies (Vincentelli et al., 2003;
Pusey et al., 2005), together with dramatic improvements in
experimental setups and the computational environment at
synchrotron facilities (Terwilliger et al., 2009; Owen et al.,
2016), have revolutionized the field and led to a torrent of new
crystal structures. This productivity boost is clear from the
number of structures deposited in the Protein Data Bank
(PDB), which recently exceeded 150 000 entries. In spite of such advances, the time-consuming and costly
mapping of reagents and phase space to identify conditions
that yield diffraction-quality crystals from a limited amount of
the macromolecule remains a bottleneck in crystallographic
studies (McPherson & Gavira, 2014; Luft et al., 2014; Giege´,
2017). This process generally involves a trial-and-error
sampling of chemical and physical space by screening hundreds
of different cocktails composed of buffers at different pH 454 IUCrJ (2019). 6, 454–464 454
https://doi.org/10.1107/S2052252519003622 https://doi.org/10.1107/S2052252519003622 research papers the
chip
design,
called
ChipX3,
incorporates
several
improvements in terms of sample injection, reservoir loading
and design to allow low-cost manufacturing by injection
molding. With ChipX3, we demonstrate that crystals can (i) be
easily produced by seeding, (ii) be soaked in situ with ligands
or (iii) be visualized by fluorescence imaging. In addition, the
chip provides a stable platform for crystal storage, handling,
shipment and in situ analysis by serial crystallography. We
illustrate a range of applications for ChipX3 by the crystal-
lization of seven soluble proteins, a membrane protein and an
RNA duplex, as well as the structure determination of five
‘non-model’ macromolecules at room temperature using data
collected on four beamlines at three different synchrotron
sites. This lab-on-a-chip approach simplifies and efficiently
miniaturizes
the
crystallographic
structure-determination
process, from the sample to its 3D structure, in a single device. It offers a user-friendly, cost-effective solution for routine
biocrystallographic investigations at room temperature. values, various crystallants (salts, alcohols and polymers) and
temperature to find at least one appropriate solvent and the
right supersaturation conditions. Raphae¨l de Wijn et al.
Microfluidic device for serial crystallography
455 1. Introduction The miniaturization of crys-
tallization assays in microplates with drop volumes of 0.1–1 ml
(typically containing 1–10 mg of the macromolecule) and
automation of the screening procedure have made this task
considerably more efficient, making it possible to successfully
conduct a project with only a few milligrams of pure sample
(Sauter et al., 2012). the
chip
design,
called
ChipX3,
incorporates
several
improvements in terms of sample injection, reservoir loading
and design to allow low-cost manufacturing by injection
molding. With ChipX3, we demonstrate that crystals can (i) be
easily produced by seeding, (ii) be soaked in situ with ligands
or (iii) be visualized by fluorescence imaging. In addition, the
chip provides a stable platform for crystal storage, handling,
shipment and in situ analysis by serial crystallography. We
illustrate a range of applications for ChipX3 by the crystal-
lization of seven soluble proteins, a membrane protein and an
RNA duplex, as well as the structure determination of five
‘non-model’ macromolecules at room temperature using data
collected on four beamlines at three different synchrotron
sites. This lab-on-a-chip approach simplifies and efficiently
miniaturizes
the
crystallographic
structure-determination
process, from the sample to its 3D structure, in a single device. It offers a user-friendly, cost-effective solution for routine
biocrystallographic investigations at room temperature. (
)
With the introduction of the first microfluidic systems
dedicated to HTP screening 15 years ago, the sample volume
required for a single experiment was reduced by another order
of magnitude, down to a few nanolitres (Hansen et al., 2002;
Zheng et al., 2003). Indeed, microfluidics was immediately
regarded as a major breakthrough, especially for biochemists
dealing with samples that are difficult to purify in large
quantities, such as macromolecules from higher eukaryotes,
large biological assemblies and membrane proteins (Hansen &
Quake, 2003; van der Woerd et al., 2003). However, despite
their potential, microfluidic technologies have not yet been
massively adopted by the global community for crystal growth,
as illustrated by the limited number of PDB entries (only
about 30 as of March 2019) that specifically cite the use of
microfluidic systems. This can be partly explained by the cost
of these microsystems and their associated equipment, but
also by the difficulty in successfully extracting fragile crystals
from the chips or the requirement to reproduce them using
conventional crystallization methods before they can be
subjected to crystallographic analysis. 2. Materials and methods Table 1
Biomolecules and crystallization conditions. Biomolecules and crystallization conditions. Biological source
No. of residues/
molecular mass
(kDa)
Biomolecule
concentration
(mg ml1)
Biomolecule buffer solution
Crystallant solution
CCA-adding
enzyme
Planococcus
halocryophilus
420/48.5
5.5
50 mM Tris–HCl pH 7.5,
200 mM NaCl, 5 mM MgCl2
30%(m/v) PEG 3350, 200 mM sodium formate
pH 6.6
Nanobody 02
Llama
129/14.5
13.8
10 mM HEPES–NaOH pH 7.25,
150 mM NaCl
20%(m/v) PEG 3000, 0.1 M trisodium citrate pH 5.5
Protease 1
Pyrococcus
horikoshii
6 166/111.6
7.4
20 mM Tris–HCl pH 7.5,
10 mM Xo4
3.4 M malonate pH 7.5
Lipase
Thermomyces
lanuginosus
269/29.3
30
25 mM HEPES–NaOH pH 7.5
0.3 M sodium/potasssium phosphate, 50 mM
sodium acetate pH 4.5
Aspartyl-tRNA
synthetase 1
Thermus
thermophilus
2 580/132
19
50 mM Tris–HCl pH 7.2,
1 mM EDTA, 1 mM DTT
10%(m/v) PEG 8000
Mitochondrial
aspartyl-tRNA
synthetase
Homo sapiens
2 630/140
30
50 mM HEPES–NaOH pH 7.5,
150 mM NaCl, 10% glycerol,
1 mM DTT
100 mM Tris–HCl pH 7.0, 40%(m/v) PEG 3350,
0.2 M sodium thiocyanate
OMT ShuA
Shigella
dysenteriae
632/69.5
20
10 mM Tris–HCl pH 8.0,
1.4% -d-octyl-glucoside
0.1 M sodium acetate, 20%(m/v) PEG 400,
15%(m/v) PEG 4000, 10%(m/v) PEG 8000 pH 5.0
RNA duplex
Synthetic
2 9/5.8
10
10 mM sodium cacodylate
pH 6.0, 5 mM MgCl2
2.6 M ammonium sulfate, 50 mM sodium
cacodylate pH 6.0, 5 mM MgSO4, 1 mM
spermine
Hemoglobin
Equus caballus
574/62
20
50 mM potassium
phosphate pH 7.5
3.3 M ammonium sulfate, 50 mM potassium
phosphate pH 7.5 same crystallant solution. This suspension was stored at 277 K
and is referred to as the ‘seed stock’. Protein crystallization
solutions were prepared by mixing 6 ml enzyme solution
(5.5 mg ml1), 1.5 ml seed stock (either the original or diluted
solution) and 1 ml CCA-TFL and were immediately injected
into the chip channels. Crystallization of the ttDRS enzyme in
the ChipX3 was also performed using seeds. The ‘seed stock’
suspension was prepared as described for the CCA enzyme by
crushing crystals grown by vapor diffusion in hanging drops
with a reservoir consisting of 7%(m/v) PEG 8000, 10 mM
MgCl2. ttDRS crystallization solutions were prepared as a
mixture consisting of 6.5 ml enzyme solution (19 mg ml1) and
1.5 ml seed stock, either the original or diluted solution, and
were immediately injected into the chip channels. France and were manufactured by MicroLIQUID, Arraste,
Spain. Table 1
Biomolecules and crystallization conditions. The fluidic layer (thickness 1 mm) was produced in
cyclic olefin copolymer (COC; TOPAS 5013F-04) by injection
molding. Channels and reservoirs were sealed with a second
layer of COC (thickness 100 mm). The bonding process was
carried out at 398 K and a pressure of 500 kPa. The straight
section of the microfluidic channels is 4 cm long with a cross-
section of 80 80 mm, to give a volume of 260 nl. The reser-
voir at their extremity has a volume of 10 ml (Fig. 1). 2.6. X-ray data collection and analysis Diffraction data were collected either (i) on beamline PXII
(Fuchs et al., 2014) equipped with a PILATUS 6M detector or
beamline PXIII (Bingel-Erlenmeyer et al., 2011) equipped
with a MAR CCD or a PILATUS 2M-F detector at the Swiss
Light Source (SLS), Villigen, Switzerland, (ii) on the
PROXIMA-2A (PX2A) beamline (Duran et al., 2013)
equipped with an EIGER X 9M detector at SOLEIL, Saint-
Aubin, France or (iii) on beamline ID30B (McCarthy et al.,
2018) equipped with a PILATUS3 6M detector at the ESRF,
Grenoble, France. 2.5. Crystal soaking with substrate 2.5. Crystal soaking with substrate To soak CCA crystals grown in the ChipX3, the tape
covering the reservoirs was removed and 3 ml of 10 mM
CMPcPP solution was added to selected reservoirs (final
concentration of 3.75 mM) before sealing them again. This
step was performed a week before data collection to ensure
good diffusion along the microfluidic channels and in an
attempt to maximize site occupancy in the crystals. 2.1. Biomacromolecules, biochemicals and chemicals The labeled protein solution was stored at 277 K
and mixed with the protein stock solution just before
preparing crystallization assays as a fraction corresponding to
less than 1% of the total protein stock. This solution will be
referred to as ‘CCA-TFL’. The nonhydrolyzable analog of cytidyl triphosphate (CTP)
that was soaked into the CCA crystals, cytidine-50-[(,)-
methyleno]triphosphate (CMPcPP), was purchased from Jena
Bioscience (catalog No. NU-438). The lanthanide complex Tb-
Xo4 (commercial name Crystallophore) used to crystallize
PhP1 was synthesized as described by Engilberge et al. (2017). 2.2. ChipX3 manufacturing In this report, we describe a versatile and low-cost micro-
fluidic chip for crystal production and characterization. This
chip was initially designed to miniaturize and facilitate the
identification of crystal-growth conditions using the counter-
diffusion method and its efficient self-optimizing process
(Dhouib et al., 2009; Pinker et al., 2013). The latest version of ChipX3 devices were designed at IBMC, Strasbourg, France
in collaboration with Synchrotron SOLEIL, Saint-Aubin, Raphae¨l de Wijn et al. Microfluidic device for serial crystallography 455 IUCrJ (2019). 6, 454–464 2.3. Sample loading and crystallization 2.3. Sample loading and crystallization Crystallization experiments in the ChipX3 were set up in
three steps with a conventional 10 ml micropipet (Gilson) and
regular tips (StarLab). Firstly, 4–6 ml of macromolecule solu-
tion was injected into the sample inlet connecting all channels
to fill the entire arborescence up to the reservoirs. Secondly,
1 ml of paraffin oil (Fluka) was injected into the sample inlet to
isolate the channels from each other and the inlet was sealed
with CrystalClear tape (Hampton Research) to prevent
evaporation and solution movements. The third and last step
consisted of filling the reservoirs with 5 ml crystallization
solution before sealing them with CrystalClear tape. The
solutions used to set up the chips are listed in Table 1. All
experiments were incubated at 293 K, except for the RNA
duplex, which was crystallized at 310 K. 456
Raphae¨l de Wijn et al.
Microfluidic device for serial crystallography research papers research papers 3.1. ChipX3 design and setup ChipX3 was designed to perform counter-diffusion (CD)
experiments and take advantage of convection-free conditions
(a prerequisite of CD) in channels of small cross-section
(width 80 mm, depth 80 mm) to enable the creation of crys-
tallant concentration gradients by pure diffusion [Fig. 1(a)]. The channels, with a length of 4 cm, allow a broad screening
of supersaturation states, as does conventional CD in micro-
capillaries (Garcı´a-Ruiz et al., 2001; Ota´lora et al., 2009). p
(
)
The geometry of the sample inlet was adapted to fit stan-
dard P2/P10 micropipet tips for chip loading using standard
laboratory materials. No extra equipment (such as a pump) is
required. The standard micropipet is used to inject the solu-
tion into the fluidic system. The branching channel config-
uration [Fig. 1(b)] allows the simultaneous loading of the eight
channels in a single manipulation, thus limiting the loading
time and solution dead volumes. Note that in the case of
membrane-protein samples containing a detergent (such as
ShuA in this work), solutions enter and fill the channels
spontaneously owing to capillary action and the native wett-
ability of the COC material. Labels embossed along the
channels facilitate crystal location and
grid mapping on synchrotron beamlines
[Fig. 1(c)]. Partial data sets were individually processed with XDS
(Kabsch, 2010). When their number did not exceed ten, they
were manually merged with XSCALE to find the best
combination and determine the appropriate resolution range. In the case of the PhP1 enzyme, ccCluster (Santoni et al., 2017)
was used to determine the best partial data sets to merge
among the 35 available. For all remaining steps, the PHENIX
package was used (Adams et al., 2010). Phases were deter-
mined by molecular replacement using the following struc-
tures: PDB entries 1miv (Li et al., 2002) for CCA, 5lmw
(Duhoo et al., 2017) for Nb02, 1g2i (Du et al., 2000) for PhP1, Figure 1
ChipX3 setup. (a) Schematic view of the chip, which has the dimensions of a microscope slide (75
25 mm) and eight channels with a straight segment of 4 cm and a cross-section of 80 80 mm. Close-
up views are shown of (b) the inlet for the biomacromolecule solution, (c) the channels and labels,
and (d) the end of the channel and the crystallant reservoir. research papers research papers All serial data collections were performed at room
temperature (T = 293–298 K) on crystals inside ChipX3, owing
to the reduced scattering background of the chip (Pinker et al.,
2013). In most cases a dedicated 3D-printed holder mounted
on a standard goniometer (see Fig. 5 and Supplementary Fig. S2) was used for data collections. To collect the widest possible
rotation angle for each crystal in ChipX3, the channel
containing the crystal was aligned with the rotation axis of the
goniometer. Crystal alignment was performed either by stan-
dard low-dose grid screening at SLS and SOLEIL, or by a one-
click procedure at ESRF as described by McCarthy et al. (2018). To avoid collisions with the surrounding equipment
(beamstop and collimator), we typically collected 30 rota-
tions per crystal or crystal sector between goniometer posi-
tions 30 and +30 (where 0 corresponding to the channels
being perpendicular to the X-ray beam). Two data-collection
strategies were used to obtain complete data: either merging
several partial data sets (sweeps) from the same crystal (one
orientation and a wide rotation range) or merging several data
sets from different crystals (several orientations and a smaller
rotation range per crystal). Table 2 provides details of data
collection and processing. 4gwl (P. K. Shukla, M. Sinha, J. Mukherjee, M. N. Gupta, P. Kaur, S. Sharma & T. P. Singh, unpublished work) for Lip and
485d (Masquida et al., 1999) for the RNA. The latter crystals
(space group H3) presented translational pseudo-symmetry
owing to the intrinsic symmetry of the duplex and merohedral
twinning (twin fraction 0.21–0.39). Hence, the structure was
refined using the twin law h, k h, l. All structures were
built and refined with Coot and phenix.refine (Emsley &
Cowtan, 2004; Adams et al., 2010). 4gwl (P. K. Shukla, M. Sinha, J. Mukherjee, M. N. Gupta, P. Kaur, S. Sharma & T. P. Singh, unpublished work) for Lip and
485d (Masquida et al., 1999) for the RNA. The latter crystals
(space group H3) presented translational pseudo-symmetry
owing to the intrinsic symmetry of the duplex and merohedral
twinning (twin fraction 0.21–0.39). Hence, the structure was
refined using the twin law h, k h, l. All structures were
built and refined with Coot and phenix.refine (Emsley &
Cowtan, 2004; Adams et al., 2010). 2.4. Crystallization by seeding The condition producing the best CCA crystals (de Wijn et
al., 2018) was found using the microseed matrix screening
(MMS) method described by D’Arcy et al. (2007, 2014). Small
crystals grown by the hanging-drop method using a reservoir
consisting of 1 M diammonium hydrogen phosphate, 100 mM
sodium acetate pH 4.5 (condition E8 from the commercial
screen JCSG++ from Jena Biosciences) were recovered,
vigorously resuspended, vortexed and diluted in 50 ml of the 456
Raphae¨l de Wijn et al. Microfluidic device for serial crystallography IUCrJ (2019). 6, 454–464 Raphae¨l de Wijn et al.
Microfluidic device for serial crystallography research papers Values in parentheses correspond to the high resolution range. p
p
g
g
CCA-adding enzyme
CCA-adding enzyme
+ CMPcPP
Nanobody 02
Protease 1
Lipase
RNA duplex
X-ray beamline
PXIII, SLS
PXII, SLS
PX2A, SOLEIL
PXIII, SLS
ID30B, ESRF
PXIII, SLS
Wavelength (A˚ )
1.000
1.000
0.826
1.240
0.976
0.826
Temperature (K)
293
293
293
293
293
293
Detector
PILATUS 2M-F
PILATUS 6M
EIGER
PILATUS 2M-F
PILATUS3 6M
MAR CCD
Crystal-to-detector distance (mm)
300
400
154
150/200
502
200
Crystals collected
6
14
9
1/11
14
3
Crystals selected
5
5
1
8
2
3
Rotation range per image ()
0.1
0.2
0.1
0.2
0.1
2–3
No. of images selected
1000
540
500
1300
600
80
Total rotation range ()
100
108
50
260
60
155
Exposure time per image (s)
0.1
0.1
0.1
0.1
0.02
1–2
Space group
P43212
P43212
P43212
P41212
P61
R3
a, c (A˚ )
71.5, 293.8
71.4, 293.6
66.7, 91.8
125.6, 133.9
142.6, 80.7
40.0, 69.1
Solvent content (%)
68.3
67.8
65.0
74.0
68.6
54.7
Mean mosaicity ()
0.04
0.04
0.07
0.04
0.03
0.15
Resolution range (A˚ )
46–2.54
(2.60–2.54)
48–2.30
(2.40–2.30)
50–2.10
(2.18–2.10)
50–2.15
(2.21–2.15)
49.06–2.50
(2.60–2.50)
23–1.55
(1.59–1.55)
Total No. of reflections
176105 (9374)
232642 (32937)
45307 (4574)
1095436 (85346)
102820 (11312)
21681 (605)
No. of unique reflections
23922 (1598)
34862 (4066)
12281 (1196)
57690 (4522)
31982 (3668)
5485 (304)
Completeness (%)
90.6 (84.6)
99.5 (100.0)
97.2 (98.3)
98.5 (99.6)
98.5 (98.9)
91.5 (69.7)
Multiplicity
7.5 (6.0)
6.7 (8.1)
3.7 (3.8)
19.0 (18.9)
3.2 (3.1)
3.9 (2.0)
hI/(I)i
8.1 (1.3)
6.9 (0.7)
11.3 (1.8)
12.0 (1.4)
6.3 (0.8)
6.1 (1.8)
Rmeas (%)
18.9 (126.0)
18.0 (231.2)
7.5 (84.7)
17.4 (206.4)
8.6 (86.8)
17.9 (45.6)
CC1/2 (%)
98.7 (55.0)
98.7 (46.9)
99.7 (73.5)
99.7 (69.4)
99.4 (49.4)
98.8 (75.5)
B factor from Wilson plot (A˚ 2)
57.4
60.6
45.2
50.8
63.3
23.6
Reflections in working/test sets
23583/1180
34840/3405
11053/1228
57659/5758
31516/1573
5484/382
Final Rwork/Rfree (%)
18.8/21.4
20.0/22.9
16.9/21.1
16.2/18.4
17.2/19.9
19.2/22.3
No. research papers Typical counter-
diffusion patterns can be observed along the concentration
gradient, with microcrystalline material close to the reservoirs
where supersaturation is maximal and larger crystals towards chip with eight different conditions takes less than 5 min even
for untrained experimenters, as attested by numerous assays
performed in the five laboratories involved in this work and by
the many participants of crystallization workshops [FEBS
courses in 2014–2018 in Nove´ Hrady, Czech Republic; Inter-
national School of Biological Crystallization (ISBC) 2015–
2017 in Granada, Spain]. concept using real cases beyond classical model proteins such
as lysozyme or thaumatin. We report here on eight proteins of
different sizes and sources (from bacteria to human) and an
RNA oligomer (Table 1) crystallized in ChipX3. Crystal-
lization conditions were adapted from those initially used in
vapor diffusion or batch crystallization: while the biomacro-
molecule concentration was kept unchanged, the crystallant
concentration was increased by a factor of 1.5–2, as recom-
mended by Ota´lora et al. (2009). Representative examples
ranging from small microcrystals to large crystals filling a
portion of the channel are shown in Fig. 2. Typical counter-
diffusion patterns can be observed along the concentration
gradient, with microcrystalline material close to the reservoirs
where supersaturation is maximal and larger crystals towards 3.1. ChipX3 design and setup Once the channels have been filled
with the macromolecule solution and
the inlet closed with tape, crystallant
solutions are deposited in the reservoirs
[Fig. 1(d)]. The setup is fully compatible
with viscous solutions such as the PEG
mixtures used in CD screens (Gonza´lez-
Ramı´rez et al., 2017). The funnel-like
channel shape has been optimized to
facilitate the contact between the crys-
tallization and macromolecule solutions
and to avoid trapping air bubbles, which
could prevent the diffusion process. 1 ml
low-gelling temperature agarose solution
at 1%(w/v) can optionally be deposited
in the funnel prior to the crystallization
cocktail to constitute a physical buffer
at the entry to the channels that stabi-
lizes the diffusion interface. Fi
1 Figure 1 Figure 1 Figure 1
ChipX3 setup. (a) Schematic view of the chip, which has the dimensions of a microscope slide (75
25 mm) and eight channels with a straight segment of 4 cm and a cross-section of 80 80 mm. Close-
up views are shown of (b) the inlet for the biomacromolecule solution, (c) the channels and labels,
and (d) the end of the channel and the crystallant reservoir. The loading procedure of ChipX3 is
fast and straightforward. Setting up a Raphae¨l de Wijn et al. Microfluidic device for serial crystallography
4 457 IUCrJ (2019). 6, 454–464 458
Raphae¨l de Wijn et al.
Microfluidic device for serial crystallography research papers of non-H atoms
Total
2998
3028
970
4017
4446
390
Protein
2989
2989
947
3921
4404
342
Solvent
9
10
23
96
47
43
Ligand
0
29
0
0
33
0
Ion
0
0
0
0
2
5
R.m.s.d., bonds (A˚ )
0.009
0.010
0.008
0.012
0.004
0.004
R.m.s.d., angles ()
1.23
1.22
0.897
1.43
1.08
0.680
Average B factors (A˚ 2)
Overall
60.1
62.6
53.8
57.1
83.8
17.8
Biomolecule
60.1
60.1
53.8
57.1
82.9
17.6
Solvent
52.7
55.5
50.9
56.1
62.6
15.1
Ramachandran plot regions (%)
Most favored
98.1
97.2
95.8
98.4
96.6
Allowed
1.9
2.8
4.2
1.6
3.2
PDB code
6ibp
6q52
6gzp
6q3t
6hw1
6ibq
PDB code at 100 K†
6qy6
6qxn
5lmw
6hf6
4zgb
485d
R.m.s. distance (A˚ 2)/Vc (%)†
0.79/5.0
0.83/3.8
0.79/6.6
0.47/4.3
1.0/3.9
0.40/2.7
† Structures solved at room temperature (this work) are compared with equivalent structures determined at cryogenic temperatures (100 K). R.m.s. distances are calculated taking into
account all biomolecule atoms and Vc quantifies the increase in the unit-cell volume (Vc) at room temperature. this work) are compared with equivalent structures determined at cryogenic temperatures (100 K). R.m.s. distances are calculated taking into
uantifies the increase in the unit-cell volume (Vc) at room temperature. † Structures solved at room temperature (this work) are compared with equivalent structures determined at cryogenic temperatures (100 K). account all biomolecule atoms and Vc quantifies the increase in the unit-cell volume (Vc) at room temperature. † Structures solved at room temperature (this work) are compared with equivalent structures determined at cryogenic temperatures (100 K). R.m.s. distances are calculated taking into
account all biomolecule atoms and Vc quantifies the increase in the unit-cell volume (Vc) at room temperature. concept using real cases beyond classical model proteins such
as lysozyme or thaumatin. We report here on eight proteins of
different sizes and sources (from bacteria to human) and an
RNA oligomer (Table 1) crystallized in ChipX3. Crystal-
lization conditions were adapted from those initially used in
vapor diffusion or batch crystallization: while the biomacro-
molecule concentration was kept unchanged, the crystallant
concentration was increased by a factor of 1.5–2, as recom-
mended by Ota´lora et al. (2009). Representative examples
ranging from small microcrystals to large crystals filling a
portion of the channel are shown in Fig. 2. 3.3. Advanced crystallogenesis strategies In addition to providing an efficient screening of super-
saturation conditions, the CD process has other practical
benefits, including the possibility of diffusing anomalous
scatterers into pregrown crystals for phasing, or cryopro-
tecting with compounds such as glycerol (Gavira et al., 2002;
Ng et al., 2003). In a previous study, we demonstrated the
feasibility of on-chip SAD phasing at room temperature using
crystals soaked by CD with a lanthanide complex (Pinker et
al., 2013). Along the same lines, CMPcPP, a nonhydrolyzable
analog of CTP, which is a substrate of CCA-adding enzymes,
was added to the reservoirs once CCA crystals had grown and
one week before the synchrotron session. The resulting X-ray
structures confirmed that the crystals were derivatized by
smooth diffusion without any sign of damage (Figs. 4 and 6). Microseeding can also be used together with CD crystal-
lization (Bergfors, 2003; Gavira et al., 2011) to bypass the
nucleation step and promote rapid crystal growth. Hence,
CCA and ttDRS crystals were grown by a combination of CD
and seeding. Microseeds were added to the protein solution
just before it was injected into the chips and the first crystals
appeared in the channels after a few days. Seeding proved
to be an effective way to trigger rapid and abundant crystal
production, which is of particular interest for serial analysis. We also used a new nucleant called crystallophore or Xo4
(Engilberge et al., 2018) in the case of the protein PhP1, for
which the crystallization conditions (Table 1) were determined
only in the presence of this terbium complex. Tb-Xo4 was
added to the protein solution before filling the channels. It
triggered the nucleation and the growth of large PhP1 crystals,
which completely filled the available volume. An added value
for macromolecules crystallized in the presence of Xo4 is the
strong luminescence when illuminated by UV light (see Fig. 3). research papers research papers in Martiel et al., 2018), which hardly affects data processing
and quality. During data collection, the chip is oriented with its
thickest layer facing the direct beam and the thinnest face in Martiel et al., 2018), which hardly affects data processing
and quality. During data collection, the chip is oriented with its
thickest layer facing the direct beam and the thinnest face the other extremity of the channels (see Supplementary
Fig. S1). in Martiel et al., 2018), which hardly affects data processing
and quality. During data collection, the chip is oriented with its
thickest layer facing the direct beam and the thinnest face
ae¨l de Wijn et al
Microfluidic device for serial crystallography
459
Figure 2
Examples of crystals obtained in ChipX3. Crystals were grown as
described in Table 1. (a) CCA, (b) PhP1, (c, d) Nb02 before (c) and after
(d) data collection, with the X-ray beam footprint, (e) lipase, (f) ttDRS,
(g) hmDRS, (h) OMT ShuA, (i) oligo RNA duplex and (j) hemoglobin. The scale bar is 0.1 mm in length. Crystals appeared after a few hours or days and could be
visualized under polarized light. To facilitate the detection of
small crystals, we exploited different fluorescence approaches
such as classical UV excitation (Meyer et al., 2015), the
fluorescent lanthanide compound Tb-Xo4 developed by
Engilberge et al. (2017) and trace fluorescent labeling (TFL) as
developed by Pusey et al. (2015). All three approaches were
compatible with ChipX3, but the Tb-Xo4 molecule and TFL
gave a much brighter signal (Fig. 3). Fluorescence has the
advantage of rapidly localizing samples in the channels and
may be used in the future to automate and speed up serial
analysis. 3.2. Crystallization in ChipX3 After a prototyping phase of small batches made by
hot embossing (Pinker et al., 2013), a 3D mold was machined
with the new ChipX3 specifications to produce a larger
batch by injection molding. This enabled validation of the 458
Raphae¨l de Wijn et al. Microfluidic device for serial crystallography IUCrJ (2019). 6, 454–464 research papers (Table 2, Figs. 5 and 6). Hemoglobin crystals also yielded
complete data to 2.8 A˚ resolution (data not shown), whereas
the microcrystals of the aspartyl-tRNA synthetases and OMT
ShuA only diffracted to low resolution and could not be used
for structure determination without further optimization. behind the crystal to minimize the attenuation of the diffrac-
tion signal (Fig. 5). Labels embossed along the channels
enable the easy localization of crystals before analysis, with a
view to future automation of the procedure on synchrotron
beamlines. The chip can also be positioned in the beam using a
plate gripper, as illustrated in Supplementary Fig. S3. To avoid
intervention from beamline staff to mount/unmount the
gripper, we developed a light chip holder that can be directly
attached to a standard goniometer. This chip holder is
manufactured by 3D printing (Fig. 5, Supplementary Fig. S2)
and integrates a standard metal base (B5, SPINE-style;
MiTeGen) that is in contact with the goniometer magnet. The
holder can be used with any flat device of microscope-slide
dimensions on synchrotron beamlines and laboratory-based
instruments. The 3D description file for printing this device is
provided as supporting information. The data collections were carried out on series of crystals
and their parameters were adapted for crystal size and sensi-
tivity to radiation damage. When collecting several paths from
the same crystal signs of radiation damage could clearly be
seen [see Fig. 2(d)], accompanied by the formation of gas
bubbles as described by Meents et al. (2010) and by the
deterioration of data-collection statistics (data not shown). Note that performing in situ analysis, i.e. without direct
handling of the crystals, is a guarantee that their genuine
diffraction properties have been preserved. Comparative
tests on thaumatin or lipase crystals in ChipX3 sent by
regular postal mail or carried to the synchrotron by
experimenters did not show significant differences (results
not shown), indicating that the chip is a stable and robust
container for crystal storage and transport. To illustrate the general applicability of on-chip serial
crystallography at room temperature, we present the results of
structure determination in the 1.5–2.5 A˚ resolution range of
four proteins (CCA, Lip, Nb02 and PhP1) and an RNA g
p
Final crystal structures were obtained
either from a single large crystal and
two wedges (Nb02) or from combining
partial data sets from several individual
crystals (RNA, CCA, Lip and PhP1). research papers In
the latter case, the use of ccCluster
considerably facilitated the choice of
partial data sets to be merged. The
comparison of these structures with
equivalent structures solved at cryo-
genic temperatures only showed small
differences (see the r.m.s. distances in
Table 2), although the unit-cell volumes
were significantly larger (2.7–6.6%) at
room
temperature
than
at
100 K
because of crystal shrinkage occurring
during cryocooling. Figure 3
Crystal detection in ChipX3 by fluorescence. (a, b) CCA crystals grown as described in Table 1 with
0.6% CCA-TFL; (c, d) PhP1 crystals grown as described in Table 1 with 10 mM Tb-Xo4. (a, c)
Crystals illuminated with white light. (b) Crystal illuminated with a 520 nm light source and image
taken with a low-pass filter at 550 nm (LP550); inset, structure of carboxyrhodamine-succinimidyl
ester. (d) Crystal illuminated with a 280–380 nm UV source; inset, structure of Tb-Xo4. The scale
bar is 0.1 mm in length. The high sensitivity and low back-
ground of the latest hybrid pixel detec-
tors
(HPDs)
compared
with
CCD
detectors (Pinker et al., 2013), and the
very short analysis time (seconds) of the
largest wedge of reciprocal space from
single crystals are crucial to outrun
radiation damage for room-temperature
data collection. The analysis in shutter-
less mode also limits systematic errors in
crystal orientation and thus improves
the data quality. For example, the
highest apparent mosaicity of the RNA
crystals (see Table 2), which were
analyzed at an early stage of this work
with a MAR CCD detector, is a direct
symptom of the data-collection strate-
gies used before the advent of HPDs. In
the future, the widespread integration
of HPD technology at synchrotron sites g
Crystal detection in ChipX3 by fluorescence. (a, b) CCA crystals grown as described in Table 1 with
0.6% CCA-TFL; (c, d) PhP1 crystals grown as described in Table 1 with 10 mM Tb-Xo4. (a, c)
Crystals illuminated with white light. (b) Crystal illuminated with a 520 nm light source and image
taken with a low-pass filter at 550 nm (LP550); inset, structure of carboxyrhodamine-succinimidyl
ester. (d) Crystal illuminated with a 280–380 nm UV source; inset, structure of Tb-Xo4. The scale
bar is 0.1 mm in length. g
Crystal detection in ChipX3 by fluorescence. (a, b) CCA crystals grown as described in Table 1 with
0.6% CCA-TFL; (c, d) PhP1 crystals grown as described in Table 1 with 10 mM Tb-Xo4. 3.4. Serial crystal analysis inside ChipX3 The ChipX3 was designed for in situ characterization. Its
overall thickness was optimized to give a good compromise
between material rigidity and X-ray absorption/scattering
(Pinker et al., 2013). The COC material produces a char-
acteristic diffuse scattering ring [Fig. 5(c)] in the resolution
range 4–6 A˚ (see also Fig. 4 in Dhouib et al., 2009 and Fig. 10.4 Figure 2
Examples of crystals obtained in ChipX3. Crystals were grown as
described in Table 1. (a) CCA, (b) PhP1, (c, d) Nb02 before (c) and after
(d) data collection, with the X-ray beam footprint, (e) lipase, (f) ttDRS,
(g) hmDRS, (h) OMT ShuA, (i) oligo RNA duplex and (j) hemoglobin. The scale bar is 0.1 mm in length. Figure 2
Examples of crystals obtained in ChipX3. Crystals were grown as
described in Table 1. (a) CCA, (b) PhP1, (c, d) Nb02 before (c) and after
(d) data collection, with the X-ray beam footprint, (e) lipase, (f) ttDRS,
(g) hmDRS, (h) OMT ShuA, (i) oligo RNA duplex and (j) hemoglobin. The scale bar is 0.1 mm in length. Raphae¨l de Wijn et al. Microfluidic device for serial crystallography 459 IUCrJ (2019). 6, 454–464 research papers (a, c)
Crystals illuminated with white light. (b) Crystal illuminated with a 520 nm light source and image
taken with a low-pass filter at 550 nm (LP550); inset, structure of carboxyrhodamine-succinimidyl
ester. (d) Crystal illuminated with a 280–380 nm UV source; inset, structure of Tb-Xo4. The scale
bar is 0.1 mm in length. 460
Raphae¨l de Wijn et al. Microfluidic device for serial crystallography IUCrJ (2019). 6, 454–464 research papers Figure 5
Diffraction analysis in ChipX3. (a) ChipX3 on its holder. (b) ChipX3 on
beamline PXIII at the SLS synchrotron. (c) Example of a diffraction
pattern of the CCA adding-enzyme in ChipX3 at room temperature
(exposure 0.1 s, rotation 0.2). and on laboratory-based X-ray sources will undoubtedly
facilitate the development of serial crystallography. The concept of serial crystallography was introduced with
XFEL sources and their extremely intense X-ray pulses that
destroy the sample upon signal emission (a process called
‘diffraction before destruction’). As a consequence, large
numbers (thousands) of micro/nanocrystals are necessary to
obtain a complete data set from series of individual still
images. The serial approach has been extended to room-
temperature data collection using synchrotron radiation. However, with a lower beam intensity (compared with
XFELs) crystals can be used to collect more than a single
image and up to several degrees of rotation. With very stable
crystals (see Nb02 in Table 2), a single crystal may even be
sufficient to collect complete data with the help of high
symmetry and rapid analysis using HPDs. More generally, the
number of crystals that are required for structure determina-
tion will depend on their size, their symmetry and their
sensitivity to radiation damage. Most of our structures were
derived from rather small series of 6–14 crystals and the
combination of best data sets (Table 2). With highly sensitive
samples such as membrane proteins, in situ room-temperature
serial crystallography can still be carried out successfully using
several hundred crystals (Huang et al., 2015). In this context,
ChipX3 provides a convenient means to produce batches of
crystals distributed along chip channels and, in the future,
automatic crystal detection and characterization should Figure 5
Diffraction analysis in ChipX3. (a) ChipX3 on its holder. (b) ChipX3 on
beamline PXIII at the SLS synchrotron. (c) Example of a diffraction
pattern of the CCA adding-enzyme in ChipX3 at room temperature
(exposure 0.1 s, rotation 0.2). Figure 4
Crystals before and after soaking in ChipX3. research papers Images of CCA crystals
grown as described in Table 1 with 0.6% CCA-TFL. (a, b) Before soaking. (c, d) Images taken six days after soaking with CMPcPP at a final
concentration of 3.75 mM. (a, c) White-light illumination. (b, d) Images
taken with a 520 nm light source and a low-pass filter at 550 nm (LP550). The scale bar is 0.1 mm in length. contribute to speeding up data collection and popularizing this
kind of serial RT analysis. 4. Conclusion
Microfluidics has demonstrated its value in terms of minia-
turization for macromolecular crystallization experiments and
HTP screening. With ChipX3, we propose a versatile tool that
integrates all of the steps of a crystallographic study on a
IUC J (2019) 6 454 464
R
h ¨l d
Wij
l
Mi
fl idi
d
i
f
i l
ll
h
461
Figure 4
Crystals before and after soaking in ChipX3. Images of CCA crystals
grown as described in Table 1 with 0.6% CCA-TFL. (a, b) Before soaking. (c, d) Images taken six days after soaking with CMPcPP at a final
concentration of 3.75 mM. (a, c) White-light illumination. (b, d) Images
taken with a 520 nm light source and a low-pass filter at 550 nm (LP550). The scale bar is 0.1 mm in length. Figure 5
Diffraction analysis in ChipX3. (a) ChipX3 on its holder. (b) ChipX3 on
beamline PXIII at the SLS synchrotron. (c) Example of a diffraction
pattern of the CCA adding-enzyme in ChipX3 at room temperature
(exposure 0.1 s, rotation 0.2). g
Diffraction analysis in ChipX3. (a) ChipX3 on its holder. (b) ChipX3 on
beamline PXIII at the SLS synchrotron. (c) Example of a diffraction
pattern of the CCA adding-enzyme in ChipX3 at room temperature
(exposure 0.1 s, rotation 0.2). contribute to speeding up data collection and popularizing this
kind of serial RT analysis. Raphae¨l de Wijn et al.
Microfluidic device for serial crystallography
46 4. Conclusion g
Crystals before and after soaking in ChipX3. Images of CCA crystals
grown as described in Table 1 with 0.6% CCA-TFL. (a, b) Before soaking. (c, d) Images taken six days after soaking with CMPcPP at a final
concentration of 3.75 mM. (a, c) White-light illumination. (b, d) Images
taken with a 520 nm light source and a low-pass filter at 550 nm (LP550). The scale bar is 0.1 mm in length. Crystals before and after soaking in ChipX3. Images of CCA crystals
grown as described in Table 1 with 0.6% CCA-TFL. (a, b) Before soaking. (c, d) Images taken six days after soaking with CMPcPP at a final
concentration of 3.75 mM. (a, c) White-light illumination. (b, d) Images
taken with a 520 nm light source and a low-pass filter at 550 nm (LP550). The scale bar is 0.1 mm in length. Microfluidics has demonstrated its value in terms of minia-
turization for macromolecular crystallization experiments and
HTP screening. With ChipX3, we propose a versatile tool that
integrates all of the steps of a crystallographic study on a Raphae¨l de Wijn et al. Microfluidic device for serial crystallography
461 Raphae¨l de Wijn et al. Microfluidic device for serial crystallography 461 IUCrJ (2019). 6, 454–464 research papers Figure 6
Electron-density maps and structures of target macromolecules. (a) CCA-adding enzyme with the
positive density from the ligand, (b) nanobody, (c) protease 1, (d) lipase, (e) RNA duplex. Insets:
schematic representations of the whole macromolecules. This figure was prepared using PyMOL
(v1.8.6; Schro¨dinger) with 2Fo Fc electron-density maps (in blue) contoured at 1.2 and the
difference map (in green) contoured at 4. physiological conditions (Martin-Garcia
et al., 2016; Johansson et al., 2017). physiological conditions (Martin-Garcia
et al., 2016; Johansson et al., 2017). physiological conditions (Martin-Garcia
et al., 2016; Johansson et al., 2017). Acknowledgements The
authors
thank
the
following
synchrotron facilities and associated
scientists for beamtime allocation to the
project and for assistance during data
collection: beamlines X06DA (PXIII)
and X10SA (PXII) at the Swiss Light
Source,
Villigen,
Switzerland,
PROXIMA
2A
at
the
SOLEIL
synchrotron, Saint-Aubin, France and
A. McCarthy for support on ID30B at
the European Synchrotron Radiation
Facility, Grenoble, France. They also
acknowledge F. Riobe´, O. Maury from
Laboratoire de Chimie at ENS-Lyon
and
Polyvalan
(Lyon,
France)
for
providing Tb-Xo4, and V. Vanel and the
team of the Shadok fablab (Strasbourg,
France) for assistance with 3D printers,
as well as the participants and organi-
zers of crystallization schools at Nove´
Hrady in the Czech Republic (FEBS
courses 2014–2018) and at Granada in
Spain (ISBC 2013–2015) for their active
participation in operational tests of
ChipX3 and their technical feedback. gu e 6
Electron-density maps and structures of target macromolecules. (a) CCA-adding enzyme with the
positive density from the ligand, (b) nanobody, (c) protease 1, (d) lipase, (e) RNA duplex. Insets:
schematic representations of the whole macromolecules. This figure was prepared using PyMOL
(v1.8.6; Schro¨dinger) with 2Fo Fc electron-density maps (in blue) contoured at 1.2 and the
difference map (in green) contoured at 4. Adams, P. D., Afonine, P. V., Bunko´czi, G., Chen, V. B., Davis, I. W.,
Echols, N., Headd, J. J., Hung, L.-W., Kapral, G. J., Grosse-
Kunstleve, R. W., McCoy, A. J., Moriarty, N. W., Oeffner, R., Read,
R. J., Richardson, D. C., Richardson, J. S., Terwilliger, T. C. &
Zwart, P. H. (2010). Acta Cryst. D66, 213–221.
Ayyer, K., Geloni, G., Kocharyan, V., Saldin, E., Serkez, S., Yefanov,
O. & Zagorodnov, I. (2015). Struct. Dyn. 2, 041702.
Bergfors, T. (2003). J. Struct. Biol. 142, 66–76.
Bingel-Erlenmeyer, R., Olieric, V., Grimshaw, J. P. A., Gabadinho, J.,
Wang, X., Ebner, S. G., Isenegger, A., Schneider, R., Schneider, J., Funding information The following funding is acknowledged:
Agence Nationale de la Recherche
(contract
No. ANR-11-LABX-
0057_MITOCROSS to Claude Sauter, Bernard Lorber;
contract No. ANR-10-LABX-0036_NETRN to Claude Sauter,
Bernard Lorber; contract No. ANR-13-BS07-0007-01 to Eric
Girard, Sylvain Engilberge); Ministe`re des Affaires Etran-
ge`res (contract No. PROCOPE Hubert Curien to Claude
Sauter,
Mario
Mo¨rl);
Deutsche
Forschungsgemeinschaft
(contract No. Mo 634/10-1 to Mario Mo¨rl, Heike Betat);
Universite´ de Strasbourg [grant No. Initiative d’excellence
(IDEX) to Claude Sauter, Raphae¨l de Wijn]; Centre National
de
la
Recherche
Scientifique
(grant
No. MRCT-
2012_PTI_UPR9002 to Claude Sauter). p
g
y
ured at 1.2 and the single device with the size of a microscope slide. The same chip
serves to produce crystals by counter-diffusion (including
seeding techniques), to soak them with ligands (for substrate
catalysis, ligand screening in fragment-based drug design or
phasing purposes) and to perform their diffraction analysis by
in situ serial crystallography. The latter step, which is carried
out on-chip at room temperature, no longer requires any
crystal handling: neither fishing, nor mounting nor cryo-
cooling. This guarantees the preservation of the intrinsic
crystal quality, with the chip being a safe means of sample
storage and transportation. ChipX3 is easy to use with stan-
dard laboratory equipment for sample loading and crystal
observation, making it cost-effective, with minimal training or
expertise required. We show the general applicability of this
lab-on-chip concept with several case studies. Sample fluor-
escent labeling, as exemplified in this work, may be exploited
to detect and center individual crystals in the X-ray beam and
to perform their characterization fully automatically. Such
microfluidic devices show great promise in the future in the
combination
of
serial
analysis
pipelines
developed
at
advanced X-ray sources (XFELs and synchrotrons) for
routine structure determination at temperatures close to research papers Martin-Garcia, J. M., Conrad, C. E., Coe, J., Roy-Chowdhury, S. &
F
P (2016) A h Bi
h
Bi
h
602 32 47 D’Arcy, A., Bergfors, T., Cowan-Jacob, S. W. & Marsh, M. (2014). Acta Cryst. F70, 1117–1126. Martin-Garcia, J. M., Conrad, C. E., Coe, J., Roy-Chowdhury, S. &
Fromme, P. (2016). Arch. Biochem. Biophys. 602, 32–47. D’Arcy, A., Villard, F. & Marsh, M. (2007). Acta Cryst. D63, 550–554. Masquida, B., Sauter, C. & Westhof, E. (1999). RNA, 5, 1384–1395. Dhouib, K., Khan Malek, C., Pfleging, W., Gauthier-Manuel, B.,
Duffait, R., Thuillier, G., Ferrigno, R., Jacquamet, L., Ohana, J.,
Ferrer, J.-L., The´obald-Dietrich, A., Giege´, R., Lorber, B. & Sauter,
C. (2009). Lab Chip, 9, 1412–1421. McCarthy, A. A., Barrett, R., Beteva, A., Caserotto, H., Dobias, F.,
Felisaz, F., Giraud, T., Guijarro, M., Janocha, R., Khadrouche, A.,
Lentini, M., Leonard, G. A., Lopez Marrero, M., Malbet-Monaco,
S., McSweeney, S., Nurizzo, D., Papp, G., Rossi, C., Sinoir, J., Sorez,
C., Surr, J., Svensson, O., Zander, U., Cipriani, F., Theveneau, P. &
Mueller-Dieckmann, C. (2018). J. Synchrotron Rad. 25, 1249–1260. McPherson, A. & Gavira, J. A. (2014). Acta Cryst. F70, 2–20. Du, X., Choi, I. G., Kim, R., Wang, W., Jancarik, J., Yokota, H. & Kim,
S.-H. (2000). Proc. Natl Acad. Sci. USA, 97, 14079–14084. Duhoo, Y., Roche, J., Trinh, T. T. N., Desmyter, A., Gaubert, A.,
Kellenberger, C., Cambillau, C., Roussel, A. & Leone, P. (2017). Acta Cryst. F73, 286–293. Meents, A., Gutmann, S., Wagner, A. & Schulze-Briese, C. (2010). Proc. Natl Acad. Sci. USA, 107, 1094–1099. Duran, D., Couster, S. L., Desjardins, K., Delmotte, A., Fox, G.,
Meijers, R., Moreno, T., Savko, M. & Shepard, W. (2013). J. Phys. Conf. Ser. 425, 012005. Meyer, A., Betzel, C. & Pusey, M. (2015). Acta Cryst. F71, 121–131. Ng, J. D., Clark, P. J., Stevens, R. C. & Kuhn, P. (2008). Acta Cryst. D64, 189–197. f
Emamzadah, S., Petty, T. J., De Almeida, V., Nishimura, T., Joly, J.,
Ferrer, J.-L. & Halazonetis, T. D. (2009). Acta Cryst. D65, 913–920. Emamzadah, S., Petty, T. J., De Almeida, V., Nishimura, T., Joly, J.,
Ferrer, J.-L. & Halazonetis, T. D. (2009). Acta Cryst. D65, 913–920. Emsley, P. & Cowtan, K. (2004). Acta Cryst. D60, 2126–2132. Ng, J. D., Gavira, J. A. & Garcı´a-Ruiz, J. M. (2003). J. Struct. Biol. 142,
218–231. (
)
y
Emsley, P. & Cowtan, K. (2004). Acta Cryst. D60, 2126–2132. Glettig, W., Pradervand, C., Panepucci, E. H., Tomizaki, T., Wang,
M. & Schulze-Briese, C. (2011). Cryst. Growth Des. 11, 916–923.
Brillet, K., Meksem, A., Thompson, A. & Cobessi, D. (2009). Acta
Cryst. F65, 402–405.
Chapman, H. N., Fromme, P., Barty, A., White, T. A., Kirian, R. A.,
Aquila, A., Hunter, M. S., Schulz, J., DePonte, D. P., Weierstall, U.,
Doak, R. B., Maia, F. R. N. C., Martin, A. V., Schlichting, I., Lomb,
L., Coppola, N., Shoeman, R. L., Epp, S. W., Hartmann, R., Rolles,
D., Rudenko, A., Foucar, L., Kimmel, N., Weidenspointner, G.,
Holl, P., Liang, M., Barthelmess, M., Caleman, C., Boutet, S.,
Bogan, M. J., Krzywinski, J., Bostedt, C., Bajt, S., Gumprecht, L.,
Rudek, B., Erk, B., Schmidt, C., Ho¨mke, A., Reich, C., Pietschner,
D., Stru¨der, L., Hauser, G., Gorke, H., Ullrich, J., Herrmann, S.,
Schaller, G., Schopper, F., Soltau, H., Ku¨hnel, K.-U., Messer-
schmidt, M., Bozek, J. D., Hau-Riege, S. P., Frank, M., Hampton,
C. Y., Sierra, R. G., Starodub, D., Williams, G. J., Hajdu, J.,
Timneanu, N., Seibert, M. M., Andreasson, J., Rocker, A., Jo¨nsson,
O., Svenda, M., Stern, S., Nass, K., Andritschke, R., Schro¨ter, C.-D.,
Krasniqi, F., Bott, M., Schmidt, K. E., Wang, X., Grotjohann, I.,
Holton, J. M., Barends, T. R. M., Neutze, R., Marchesini, S.,
Fromme, R., Schorb, S., Rupp, D., Adolph, M., Gorkhover, T.,
Andersson, I., Hirsemann, H., Potdevin, G., Graafsma, H., Nilsson,
B. & Spence, J. C. H. (2011). Nature (London), 470, 73–77. research papers Ota´lora, F., Gavira, J. A., Ng, J. D. & Garcı´a-Ruiz, J. M. (2009). Prog. Biophys. Mol. Biol. 101, 26–37. Engilberge, S., Riobe´, F., Di Pietro, S., Lassalle, L., Coquelle, N.,
Arnaud, C.-A., Pitrat, D., Mulatier, J.-C., Madern, D., Breyton, C.,
Maury, O. & Girard, E. (2017). Chem. Sci. 8, 5909–5917. Owen, R. L., Juanhuix, J. & Fuchs, M. (2016). Arch. Biochem. Biophys. 602, 21–31. Engilberge, S., Riobe´, F., Wagner, T., Di Pietro, S., Breyton, C.,
Franzetti, B., Shima, S., Girard, E., Dumont, E. & Maury, O. (2018). Chem. Eur. J. 24, 9739–9746. Perry, S. L., Guha, S., Pawate, A. S., Bhaskarla, A., Agarwal, V., Nair,
S. K. & Kenis, P. J. A. (2013). Lab Chip, 13, 3183. Perry, S. L., Guha, S., Pawate, A. S., Henning, R., Kosheleva, I., Srajer,
V., Kenis, P. J. A. & Ren, Z. (2014). J. Appl. Cryst. 47, 1975–1982. Ernst, F. G. M., Erber, L., Sammler, J., Ju¨hling, F., Betat, H. & Mo¨rl,
M. (2018). RNA Biol. 15, 144–155. Fuchs, M. R., Pradervand, C., Thominet, V., Schneider, R., Panepucci,
E., Grunder, M., Gabadinho, J., Dworkowski, F. S. N., Tomizaki, T.,
Schneider, J., Mayer, A., Curtin, A., Olieric, V., Frommherz, U.,
Kotrle, G., Welte, J., Wang, X., Maag, S., Schulze-Briese, C. & Wang,
M. (2014). J. Synchrotron Rad. 21, 340–351. Pinker, F., Brun, M., Morin, P., Deman, A.-L., Chateaux, J.-F., Olie´ric,
V., Stirnimann, C., Lorber, B., Terrier, N., Ferrigno, R. & Sauter, C. (2013). Cryst. Growth Des. 13, 3333–3340. Kotrle, G., Welte, J., Wang, X., Maag, S., Schulze-Briese, C. & Wang,
M. (2014). J. Synchrotron Rad. 21, 340–351. Pusey, M., Barcena, J., Morris, M., Singhal, A., Yuan, Q. & Ng, J. (2015). Acta Cryst. F71, 806–814. Garcı´a-Ruiz, J. M., Ota´lora, F., Novella, M. L., Gavira, J. A., Sauter,
C. & Vidal, O. (2001). J. Cryst. Growth, 232, 149–155. Pusey, M. L., Liu, Z.-J., Tempel, W., Praissman, J., Lin, D., Wang,
B.-C., Gavira, J. A. & Ng, J. D. (2005). Prog. Biophys. Mol. Biol. 88,
359–386. Gavira, J. A., Hernandez-Hernandez, M. A., Gonzalez-Ramirez,
L. A., Briggs, R. A., Kolek, S. A. & Shaw Stewart, P. D. (2011). Cryst. Growth Des. 11, 2122–2126. Santoni, G., Zander, U., Mueller-Dieckmann, C., Leonard, G. &
Popov, A. (2017). J. Appl. Cryst. 50, 1844–1851. Sauter, C., Dhouib, K. & Lorber, B. (2007). Cryst. Growth Des. 7,
2247–2250. Gavira, J. A., Toh, D., Lope´z-Jaramillo, J., Garcı´a-Ruı´z, J. M. & Ng,
J. D. (2002). research papers Glettig, W., Pradervand, C., Panepucci, E. H., Tomizaki, T., Wang,
M. & Schulze-Briese, C. (2011). Cryst. Growth Des. 11, 916–923. Brillet, K., Meksem, A., Thompson, A. & Cobessi, D. (2009). Acta
Cryst. F65, 402–405. Chapman, H. N., Fromme, P., Barty, A., White, T. A., Kirian, R. A.,
Aquila, A., Hunter, M. S., Schulz, J., DePonte, D. P., Weierstall, U.,
Doak, R. B., Maia, F. R. N. C., Martin, A. V., Schlichting, I., Lomb,
L., Coppola, N., Shoeman, R. L., Epp, S. W., Hartmann, R., Rolles,
D., Rudenko, A., Foucar, L., Kimmel, N., Weidenspointner, G.,
Holl, P., Liang, M., Barthelmess, M., Caleman, C., Boutet, S.,
Bogan, M. J., Krzywinski, J., Bostedt, C., Bajt, S., Gumprecht, L.,
Rudek, B., Erk, B., Schmidt, C., Ho¨mke, A., Reich, C., Pietschner,
D., Stru¨der, L., Hauser, G., Gorke, H., Ullrich, J., Herrmann, S.,
Schaller, G., Schopper, F., Soltau, H., Ku¨hnel, K.-U., Messer-
schmidt, M., Bozek, J. D., Hau-Riege, S. P., Frank, M., Hampton,
C. Y., Sierra, R. G., Starodub, D., Williams, G. J., Hajdu, J.,
Timneanu, N., Seibert, M. M., Andreasson, J., Rocker, A., Jo¨nsson,
O., Svenda, M., Stern, S., Nass, K., Andritschke, R., Schro¨ter, C.-D.,
Krasniqi, F., Bott, M., Schmidt, K. E., Wang, X., Grotjohann, I.,
Holton, J. M., Barends, T. R. M., Neutze, R., Marchesini, S.,
Fromme, R., Schorb, S., Rupp, D., Adolph, M., Gorkhover, T.,
Andersson, I., Hirsemann, H., Potdevin, G., Graafsma, H., Nilsson,
B. & Spence, J. C. H. (2011). Nature (London), 470, 73–77. Hansen, C. & Quake, S. R. (2003). Curr. Opin. Struct. Biol. 13, 538–
544. Glettig, W., Pradervand, C., Panepucci, E. H., Tomizaki, T., Wang,
M. & Schulze-Briese, C. (2011). Cryst. Growth Des. 11, 916–923. Brillet, K., Meksem, A., Thompson, A. & Cobessi, D. (2009). Acta
Cryst. F65, 402–405. Chapman, H. N., Fromme, P., Barty, A., White, T. A., Kirian, R. A.,
Aquila, A., Hunter, M. S., Schulz, J., DePonte, D. P., Weierstall, U.,
Doak, R. B., Maia, F. R. N. C., Martin, A. V., Schlichting, I., Lomb,
L., Coppola, N., Shoeman, R. L., Epp, S. W., Hartmann, R., Rolles,
D., Rudenko, A., Foucar, L., Kimmel, N., Weidenspointner, G.,
Holl, P., Liang, M., Barthelmess, M., Caleman, C., Boutet, S.,
Bogan, M. research papers J., Krzywinski, J., Bostedt, C., Bajt, S., Gumprecht, L.,
Rudek, B., Erk, B., Schmidt, C., Ho¨mke, A., Reich, C., Pietschner,
D., Stru¨der, L., Hauser, G., Gorke, H., Ullrich, J., Herrmann, S.,
Schaller, G., Schopper, F., Soltau, H., Ku¨hnel, K.-U., Messer-
schmidt, M., Bozek, J. D., Hau-Riege, S. P., Frank, M., Hampton,
C. Y., Sierra, R. G., Starodub, D., Williams, G. J., Hajdu, J.,
Timneanu, N., Seibert, M. M., Andreasson, J., Rocker, A., Jo¨nsson,
O., Svenda, M., Stern, S., Nass, K., Andritschke, R., Schro¨ter, C.-D.,
Krasniqi, F., Bott, M., Schmidt, K. E., Wang, X., Grotjohann, I.,
Holton, J. M., Barends, T. R. M., Neutze, R., Marchesini, S.,
Fromme, R., Schorb, S., Rupp, D., Adolph, M., Gorkhover, T.,
Andersson, I., Hirsemann, H., Potdevin, G., Graafsma, H., Nilsson,
B. & Spence, J. C. H. (2011). Nature (London), 470, 73–77. Hansen, C. L., Classen, S., Berger, J. M. & Quake, S. R. (2006). J. Am. Chem. Soc. 128, 3142–3143. Hansen, C. L., Skordalakes, E., Berger, J. M. & Quake, S. R. (2002). Proc. Natl Acad. Sci. USA, 99, 16531–16536. Heymann, M., Opthalage, A., Wierman, J. L., Akella, S., Szebenyi,
D. M. E., Gruner, S. M. & Fraden, S. (2014). IUCrJ, 1, 349–360. D. M. E., Gruner, S. M. & Fraden, S. (2014). IUCrJ, 1 (
)
Huang, C.-Y., Olieric, V., Ma, P., Panepucci, E., Diederichs, K., Wang,
M. & Caffrey, M. (2015). Acta Cryst. D71, 1238–1256. y
(
)
y
Jaskolski, M., Dauter, Z. & Wlodawer, A. (2014). FEBS J. 281, 3985–
4009. Johansson, L. C., Stauch, B., Ishchenko, A. & Cherezov, V. (2017). Trends Biochem. Sci. 42, 749–762. Kabsch, W. (2010). Acta Cryst. D66, 125–132. Khvostichenko, D. S., Schieferstein, J. M., Pawate, A. S., Laible, P. D. & Kenis, P. J. A. (2014). Cryst. Growth Des. 14, 4886–4890. ,
,
,
,
,
,
& Kenis, P. J. A. (2014). Cryst. Growth Des. 14, 4886–4890. Li, F., Xiong, Y., Wang, J., Cho, H. D., Tomita, K., Weiner, A. M. &
Steitz, T. A. (2002). Cell, 111, 815–824. (
)
Luft, J. R., Newman, J. & Snell, E. H. (2014). Acta Cryst. F70, 835–
853. Martiel, I., Olieric, V., Caffrey, M. & Wang, M. (2018). Protein
Crystallography: Challenges and Practical Solutions, edited by K. Beis & G. Evans, pp. 1–27. Cambridge: Royal Society of Chemistry. Crystallography: Challenges and Practical Solutions, edited by K. Beis & G. Evans, pp. 1–27. Cambridge: Royal Society of Chemistry. References 462
Raphae¨l de Wijn et al. Microfluidic device for serial crystallography IUCrJ (2019). 6, 454–464 research papers 464
Raphae¨l de Wijn et al.
Microfluidic device for serial crystallography Zhu, D.-W., Lorber, B., Sauter, C., Ng, J. D., Be´nas, P., Le Grimellec,
C. & Giege´, R. (2001). Acta Cryst. D57, 552–558. research papers Acta Cryst. D58, 1147–1154. Giege´, R. (2017). IUCrJ, 4, 340–349. Sauter, C., Lorber, B., Gaudry, A., Karim, L., Schwenzer, H., Wien, F.,
Roblin, P., Florentz, C. & Sissler, M. (2015). Sci. Rep. 5, 17332. Giege´, R. & Sauter, C. (2010). HFSP J. 4, 109–121. Sauter, C., Lorber, B., McPherson, A. & Giege´, R. (2012). International Tables for Crystallography, Vol. F, edited by E. Arnold, D. M. Himmel & M. G. Rossmann, pp. 99–121. Chester:
International Union of Crystallography. Gonza´lez-Ramı´rez, L. A., Ruiz-Martı´nez, C. R., Estremera-Andu´jar,
R. A., Nieves-Marrero, C. A., Garcı´a-Caballero, A., Gavira, J. A.,
Lo´pez-Garriga, J. & Garcı´a-Ruiz, J. M. (2017). Cryst. Growth Des. 17, 6780–6786. Raphae¨l de Wijn et al. Microfluidic device for serial crystallography 463 IUCrJ (2019). 6, 454–464 research papers Vincentelli, R., Bignon, C., Gruez, A., Canaan, S., Sulzenbacher, G.,
Tegoni, M., Campanacci, V. & Cambillau, C. (2003). Acc. Chem. Res. 36, 165–172. Stellato,
F.,
Oberthu¨r,
D.,
Liang,
M.,
Bean,
R.,
Gati,
C.,
Yefanov, O., Barty, A., Burkhardt, A., Fischer, P., Galli, L.,
Kirian, R. A., Meyer, J., Panneerselvam, S., Yoon, C. H.,
Chervinskii,
F.,
Speller,
E.,
White,
T. A.,
Betzel,
C.,
Meents,
A. &
Chapman,
H. N. (2014). IUCrJ,
1,
204–
212. Wijn, R. de, Hennig, O., Ernst, F. G. M., Lorber, B., Betat, H., Mo¨rl,
M. & Sauter, C. (2018). Acta Cryst. F74, 747–753. Woerd, M. van der, Ferree, D. & Pusey, M. (2003). J. Struct. Biol. 142,
180–187. Stojanoff, V., Jakoncic, J., Oren, D. A., Nagarajan, V., Navarro
Poulsen, J.-C., Adams-Cioaba, M. A., Bergfors, T. & Sommer,
M. O. A. (2011). Acta Cryst. F67, 971–975. Yadav, M. K., Gerdts, C. J., Sanishvili, R., Smith, W. W., Roach, L. S.,
Ismagilov, R. F., Kuhn, P. & Stevens, R. C. (2005). J. Appl. Cryst. 38,
900–905. (
)
Sui, S., Wang, Y., Kolewe, K. W., Srajer, V., Henning, R., Schiffman,
J. D., Dimitrakopoulos, C. & Perry, S. L. (2016). Lab Chip, 16, 3082–
3096. Zheng, B., Roach, L. S. & Ismagilov, R. F. (2003). J. Am. Chem. Soc. 125, 11170–11171. Zhu, D.-W., Lorber, B., Sauter, C., Ng, J. D., Be´nas, P., Le Grimellec,
C. & Giege´, R. (2001). Acta Cryst. D57, 552–558. Terwilliger, T. C., Stuart, D. & Yokoyama, S. (2009). Annu. Rev. Biophys. 38, 371–383. IUCrJ (2019). 6, 454–464
|
https://openalex.org/W3209018864
|
https://backend.orbit.dtu.dk/ws/files/274837521/cjfas_2021_0178.pdf
|
English
| null |
Potential for codends with shortened lastridge ropes to replace mandated selection devices in demersal trawl fisheries
|
Canadian journal of fisheries and aquatic sciences
| 2,022
|
cc-by
| 17,497
|
Citation (APA):
Sistiaga, M., Brinkhof, J., Herrmann, B., Larsen, R. B., Grimaldo, E., Cerbule, K., Brinkhof, I., & Jørgensen, T.
(2022). Potential for codends with shortened lastridge ropes to replace mandated selection devices in demersal
trawl fisheries. Canadian Journal of Fisheries and Aquatic Sciences, 79(5), 834-849.
https://doi.org/10.1139/cjfas-2021-0178 General rights
Copyright and moral rights for the publications made accessible in the public portal are retained by the authors and/or other copyright
owners and it is a condition of accessing publications that users recognise and abide by the legal requirements associated with these rights.
Users may download and print one copy of any publication from the public portal for the purpose of private study or research.
You may not further distribute the material or use it for any profit-making activity or commercial gain
You may freely distribute the URL identifying the publication in the public portal
If you believe that this document breaches copyright please contact us providing details, and we will remove access to the work immediately
and investigate your claim. Downloaded from orbit.dtu.dk on: Oct 24, 2024 Downloaded from orbit.dtu.dk on: Oct 24, 2024 Downloaded from orbit.dtu.dk on: Oct 24, 2024 General rights
Copyright and moral rights for the publications made accessible in the public portal are retained by the authors and/or other copyright
owners and it is a condition of accessing publications that users recognise and abide by the legal requirements associated with these rights. General rights
Copyright and moral rights for the publications made accessible in the public portal are retained by the authors and/or other copyright
owners and it is a condition of accessing publications that users recognise and abide by the legal requirements associated with these rights.
Users may download and print one copy of any publication from the public portal for the purpose of private study or research.
Y
f
h
di
ib
h
i l
i f
fi
ki
i i
i l
i Can. J. Fish. Aquat. Sci. Downloaded from cdnsciencepub.com by Danmarks Tekniske Informationscenter on 04/28/22
For personal use only. Abstract: In many trawl fisheries, codend size selectivity is supplemented by adding selection devices to the gear. In the
Barents Sea gadoid fishery, combining diamond mesh codends with sorting grids is compulsory. However, the use of grids
increases the costs and complexity of the gear, causing discontent among fishermen and prompting researchers to seek al-
ternative solutions. Lastridge ropes are ropes attached to the selvedges of the codend. In this study, we tested the effect of
shortening the lastridge ropes of two diamond mesh codends with different mesh sizes on the size selectivity of cod (Gadus
morhua), haddock (Melanogramus aeglefinnus), and redfish (Sebastes spp.). Shortening the lastridge ropes by 15% increased the
mesh opening during the fishing process, which significantly improved the size-selective properties of the codends. Fur-
ther, the L50 values were always higher for the codends in the short lastridge configuration. Therefore, codends with short-
ened lastridge ropes may be a simpler alternative to sorting grids in this fishery, and they may be applicable to many other
fisheries in which additional selection devices are used. Résumé : Dans de nombreuses pêches au chalut, la sélectivité selon la taille des culs de chalut est rehaussée en ajoutant
aux engins des dispositifs de sélection d’appoint. Dans la pêche aux gadidés de la mer de Barents, il est obligatoire de
combiner des grilles séparatrices aux culs de chalut à maille en losange. L’utilisation de grilles accroît les coûts et la com-
plexité des engins, ce qui cause du mécontentement chez les pêcheurs et incite les chercheurs à explorer d’autres solutions. Les ralingues sont des cordes attachées aux lisières du cul de chalut. Nous avons vérifié l’effet du raccourcissement des
ralingues de deux culs de chalut à maille en losange à mailles de différentes dimensions sur la sélectivité de la taille des
morues (Gadus morhua), des aiglefins (Melanogramus aeglefinnus) et des sébastes (Sebastes spp.). Un raccourcissement de 15 %
des ralingues accroît l’ouverture des mailles durant le processus de pêche, ce qui améliore significativement les propriétés
de sélection selon la taille des culs de chalut. En outre, les valeurs L50 sont toujours plus grandes pour les configurations de
cul de chalut intégrant des ralingues raccourcies. ARTICLE ARTICLE Can. J. Fish. Aquat. Sci. Downloaded from cdnsciencepub.com by Danmarks Tekniske Informationscenter on 04/28/22
For personal use only. Les culs de chalut dotés de ralingues raccourcies pourraient donc constit-
uer une solution de rechange plus simple aux grilles séparatrices dans cette pêche et pourraient éventuellement être
employés dans de nombreuses autres pêches dans lesquelles des dispositifs de sélection d’appoint sont utilisés. [Traduit par
la Rédaction] Can. J. Fish. Aquat. Sci. Downloaded from cdnsciencepub
For perso 2007). The meshes at these stages will generally maintain their dia-
mond shape and fish will most likely not be able to deform the net-
ting and escape. However, when the codend is at the surface with
low or no tension, the meshes can be both wide open (up to 90
degrees) and slack, which could give fish trying to escape the chance
to distort the mesh shape to fit their cross-sectional shape and
escape through it (Herrmann et al. 2016). Introduction Can. J. Fish. Aquat. Sci. Downloaded fro Diamond mesh codends are the most widespread and simplest
size-selection device used in demersal trawls, and in some fish-
eries, size selectivity relies solely on the selective properties of
this type of codend (Cheng et al. 2019). However, diamond mesh
codends can pose challenges and yield varying or unsatisfactory
results (Robertson and Stewart 1988; Sala et al. 2008; Wienbeck
2011). For example, as the catch in the codend builds up during
towing and haul-back, tension increases and the longitudinal
forces in the mesh bars close the codend meshes, affecting selectiv-
ity (Robertson and Stewart 1988; Herrmann 2005a, 2005b; Herrmann
and O’Neill 2005; Herrmann et al. 2007; O’Neill and Herrmann An obvious management approach to solving trawl selectivity
issues would be to modify the codend. However, in many fish-
eries the approach adopted has been to insert additional devices
into the gear, such as square mesh panels (Graham et al. 2003;
Herrmann et al. 2015; Cuende et al. 2020) or sorting grids (Sistiaga
et al. 2008; Brinkhof et al. 2020), to supplement codend size Potential for codends with shortened lastridge ropes to replace mandated selection
devices in demersal trawl fisheries Sistiaga, Manu; Brinkhof, Jesse; Herrmann, Bent; Larsen, Roger B.; Grimaldo, Eduardo; Cerbule,
Kristine; Brinkhof, Ilmar; Jørgensen, Terje Published in:
Canadian Journal of Fisheries and Aquatic Sciences Document Version
Publisher's PDF, also known as Version of record Link back to DTU Orbit Citation (APA):
Sistiaga, M., Brinkhof, J., Herrmann, B., Larsen, R. B., Grimaldo, E., Cerbule, K., Brinkhof, I., & Jørgensen, T. (2022). Potential for codends with shortened lastridge ropes to replace mandated selection devices in demersal
trawl fisheries. Canadian Journal of Fisheries and Aquatic Sciences, 79(5), 834-849. https://doi.org/10.1139/cjfas-2021-0178 Can. J. Fish. Aquat. Sci. 79: 834–849 (2022) dx.doi.org/10.1139/cjfas-2021-0178 Potential for codends with shortened lastridge ropes to replace
mandated selection devices in demersal trawl fisheries Manu Sistiaga, Jesse Brinkhof, Bent Herrmann, Roger B. Larsen, Eduardo Grimaldo, Kristine Cerbule,
Ilmar Brinkhof, and Terje Jørgensen Can. J. Fish. Aquat. Sci. Downloaded from cdnsciencepub.com by Danmarks Tekniske Informationscenter on 04/28/22
For personal use only. 1. Do shortened lastridge ropes modify the selection properties
of diamond mesh codends for cod, haddock, and redfish? If
so, then to what extent? (
)
The fishing industry would like to remove the mandatory use
of grids from the regulations because they are expensive, heavy,
and can substantially influence water flow in the extension piece
and codend (Grimaldo et al. 2016). Reduced water flow in the aft
part of the trawl can lead to fish accumulation, which can result
in section breakage (Sistiaga et al. 2016) and failure of catch limit-
ers and catch sensors (Grimaldo et al. 2014). In addition, the three
grid systems may not be equally efficient, and their performance
can vary substantially depending on factors such as catch den-
sities and whether the section is constructed of two or four pan-
els (Sistiaga et al. 2016; Brinkhof et al. 2020). 2. Is it possible to explain the selectivity results obtained for
cod, haddock, and redfish by their species-specific character-
istics and potential changes in the codend meshes generated
by shortening the lastridge ropes? 3. Can shortened lastridge codends provide the desired catch
patterns for cod, haddock, and redfish so that they could
replace the grid system required in the Barents Sea demersal
trawl fishery? The mandatory use of selection grids in the Barents Sea demer-
sal trawl fishery has been questioned since it was made compul-
sory in 1997 (Jørgensen et al. 2006). Simple codend modifications
or additional devices such as exit windows or square mesh sec-
tions have been tested as potential alternatives (Jørgensen et al. 2006; Grimaldo et al. 2008, 2018). Although some of the sorting
devices have shown selection properties similar to those of the
sorting grids, issues related to how to mount the devices and how
to objectively monitor and control their use have prevented their
implementation. Another approach that does not require addi-
tional devices and is relatively simple to implement and control
is to attach short lastridge ropes in the codend. Lastridge ropes
are ropes attached to the selvedges of the codend, and they are
normally slightly shorter than the codend netting (e.g., typically
0%–5% in the Barents Sea). This type of rope is normally used in
fisheries like the Barents Sea demersal fishery where the catches
can be large (e.g., >10 tonnes). Study area, experimental design, and data collection Study area, experimental design, and data collection Study area, experimental design, and data collection
Experimental fishing was conducted onboard the research ves-
sel Helmer Hanssen (63.9 m long, 4080 HP (3000 kW)) from 8 to
16 January 2021 in the southern part of the Barents Sea (71°
2206500N–72°0803000N, 25°4809200E–30°1304400E). The experimental
fishing was conducted using an Alfredo 5 twin-body trawl (trouser-
trawl; Grimaldo et al. 2007) combined with a set of Injector Scor-
pion trawl doors, each weighing 3100 kg and with an area of 8 m2. The trawl doors were connected to the sweeps with 3 m long back-
straps followed by 7 m long connector wires. The sweeps were
2 m 30 m long and divided by a Ø53 cm steel bobbin in the mid-
dle to protect them from excessive abrasion. The sweeps were con-
nected to a 48 m long ground gear, which consisted of a 14 m long
chain (Ø19 mm) with four equally spaced bobbins (Ø53 cm) on each
side with a rock-hopper gear in the middle. The rock-hopper gear
was 21 m long and equipped with Ø53 cm discs.The headline in the
trawl was 38 m long, and it was equipped with 170 floats (8 inch;
1 inch = 2.5 cm). The trawl net itself was a modified 155 mm, two-
panel Alfredo 5 twin-body trawl. A vertical panel (# 80 mm) was
inserted in the front part of the trawl body to divide it into two
equal sections. At the end of the vertical net, the trawl body was
split into two equal 23.3 m long tapered funnels (Fig. 1a). Each fun-
nel was followed by a 14.1 m long extension piece, which took the
place of the grid section that is compulsory in the commercial fish-
ery. The codends were mounted directly onto the extension pieces
and consisted of two panels made of single braided polyethylene
hotmelt twine (Ø8 mm). Each codend was 12 m long and 60 free
meshes in circumference. The two codends had different mesh
sizes: 128.23 6 3.97 mm and 137.08 6 2.28 mm. These two mesh
sizes represent the minimum mesh size used by the fleet in the
fishery (130 mm) and a codend with approximately 1 cm bigger
meshes. The selvedges of the codends were strengthened with
Ø32 mm (Polyethylene, Danline) thick lastridge ropes. Can. J. Fish. Aquat. Sci. Downloaded from cdnsciencepub.com by Danmarks Tekniske Informationscenter on 04/28/22
For personal use only. When the catch builds up, most of
the load is carried by these ropes rather than by the netting in
the codend. By shortening the lastridge ropes further, they would
bear the load of the catch to a greater extent than in a typical
codend. Consequently, the tension in the codend netting would
remain low as the catch accumulates, resulting in more open
meshes during fishing, which should improve the selective prop-
erties of the codend (Isaksen and Valdemarsen 1990; Lök et al. 1997; Ingolfsson and Brinkhof 2020). Published by Canadian Science Publishing Received 23 June 2021. Accepted 25 October 2021. M. Sistiaga.* Institute of Marine Research, Postbox 1870 Nordnes, N-5817 Bergen, Norway; Norwegian University of Science and Technology, Otto Nielsens
veg 10, N-7491 Trondheim, Norway. I. Brinkhof. The Arctic University of Norway, UiT, Breivika, N-9037 Tromsø, Norway. J. Brinkhof,* R.B. Larsen, and I. Brinkhof. The Arctic University of Norway, UiT, Breivika, N-9037 Tromsø, Norway. B. Herrmann.* The Arctic University of Norway, UiT, Breivika, N-9037 Tromsø, Norway; SINTEF Ocean, Brattørkaia 17C, N-7010 Trondheim, Norway; DTU
Aqua, Technical University of Denmark, Willemoesvej 2, DK-9850 Hirtshals, Denmark. E. Grimaldo and K. Cerbule. The Arctic University of Norway, UiT, Breivika, N-9037 Tromsø, Norway; SINTEF Ocean, Brattørkaia 17C, N-7010 Trondheim,
Norway J. Brinkhof,* R.B. Larsen, and I. Brinkhof. The Arctic University of Norway, UiT, Breivika, N-9037 Tromsø, Norway. B. Herrmann.* The Arctic University of Norway, UiT, Breivika, N-9037 Tromsø, Norway; SINTEF Ocean, Brattørkaia 17C, N-7010 Trondheim, Norway; DTU
Aqua, Technical University of Denmark, Willemoesvej 2, DK-9850 Hirtshals, Denmark. E. Grimaldo and K. Cerbule. The Arctic University of Norway, UiT, Breivika, N-9037 Tromsø, Norway; SINTEF Ocean, Brattørkaia 17C, N-7010 Trondheim, T. Jørgensen. Institute of Marine Research, Postbox 1870 Nordnes, N-5817 Bergen, Norway. T. Jørgensen. Institute of Marine Research, Postbox 1870 Nordnes, N-5817 Bergen, Norway. T. Jørgensen. Institute of Marine Research, Postbox 1870 Nordnes, N-581 q
p
© 2021 The Author(s). This work is licensed under a Creative Commons Attribution 4.0 International License (CC BY 4.0), which permits unrestricted use,
distribution, and reproduction in any medium, provided the original author(s) and source are credited. q
p
© 2021 The Author(s). This work is licensed under a Creative Commons Attribution 4.0 International License (CC BY 4.0), which permits unrestricted use,
distribution, and reproduction in any medium, provided the original author(s) and source are credited. Published at www.cdnsciencepub.com/cjfas on 28 October 2021. Can. J. Fish. Aquat. Sci. 79: 834–849 (2022) dx.doi.org/10.1139/cjfas-2021-0178 835 Sistiaga et al. selectivity. One such fishery is the Barents Sea gadoid trawl fish-
ery, which is one of the most important demersal fisheries in the
world (Bergstad et al. 1987; Olsen et al. 2010). In this fishery, the
diamond mesh codend is supplemented by a rigid sorting grid,
which became compulsory in 1997 due to unsatisfactory size
selection of the diamond mesh codend alone (Larsen and Isaksen
1993). Received 23 June 2021. Accepted 25 October 2021. The current compulsory size-selection gear is a dual system
composed of a sorting grid with a minimum bar spacing of
55 mm and a subsequent diamond mesh codend with a mini-
mum mesh size of 130 mm (Norwegian Directorate of Fisheries
2017). Fishermen can choose among three different sorting grid
systems (Sort-X, Sort-V, and Flexigrid) that have been developed
over time since the first trials were conducted in the early 1990s
(Larsen and Isaksen 1993; Grimaldo et al. 2016). selectivity. One such fishery is the Barents Sea gadoid trawl fish-
ery, which is one of the most important demersal fisheries in the
world (Bergstad et al. 1987; Olsen et al. 2010). In this fishery, the
diamond mesh codend is supplemented by a rigid sorting grid,
which became compulsory in 1997 due to unsatisfactory size
selection of the diamond mesh codend alone (Larsen and Isaksen
1993). The current compulsory size-selection gear is a dual system
composed of a sorting grid with a minimum bar spacing of
55 mm and a subsequent diamond mesh codend with a mini-
mum mesh size of 130 mm (Norwegian Directorate of Fisheries
2017). Fishermen can choose among three different sorting grid
systems (Sort-X, Sort-V, and Flexigrid) that have been developed
over time since the first trials were conducted in the early 1990s
(Larsen and Isaksen 1993; Grimaldo et al. 2016). ropes in the codend is limited (Isaksen and Valdemarsen 1990). Thus, the objectives of this study were to investigate the effect of
shortening the lastridge ropes on codends with different mesh
sizes and to evaluate how the changes affect the selectivity and
catch patterns of cod, haddock, and redfish in the Barents Sea
demersal trawl fishery. Considering the MLS and exploitation
pattern desired by fishermen for the different species involved,
we also investigated whether codends with shortened lastridge
ropes could realistically replace the grid system required in the
fishery today. Specifically, the research was designed to answer
the following research questions: ropes in the codend is limited (Isaksen and Valdemarsen 1990). Thus, the objectives of this study were to investigate the effect of
shortening the lastridge ropes on codends with different mesh
sizes and to evaluate how the changes affect the selectivity and
catch patterns of cod, haddock, and redfish in the Barents Sea
demersal trawl fishery. Received 23 June 2021. Accepted 25 October 2021. Considering the MLS and exploitation
pattern desired by fishermen for the different species involved,
we also investigated whether codends with shortened lastridge
ropes could realistically replace the grid system required in the
fishery today. Specifically, the research was designed to answer
the following research questions: Study area, experimental design, and data collection Illustration showing the twin-body trawl (a) and the configuration of the covers (CC) over the codends (C) (b). [Colour online.] m cdnsciencepub.com by Danmarks Tekniske Informationscenter on 04/28/22
For personal use only. Can. J. Fish. Aquat. Sci. Downloaded from cdnsciencepub.com by Danmarks Tekniske Informationscenter on 04/28/22
For personal use only. Can. J. Fish. Aquat. Sci. Downloaded from cdnsciencepub.com by Danmarks Tekniske Informationscenter on 04/28/22
For personal use only. Can. J. Fish. Aquat. Sci. Downloaded from cdnsciencepub.com by Danmarks Te
For personal use only. Can. J. Fish. Aquat. Sci. Downloaded from cdnsciencepub.com
For personal us rcodend (l, vcodend) for the codend size selection, where vcodend is
a vector consisting of the parameters in the model. The purpose
of the analysis was to estimate the values of the parameters in
vcodend that maximized the likelihood for the experimental data
(averaged over hauls) to be obtained. For this purpose, the follow-
ing expression was minimized, which corresponds to maximizing
the likelihood for obtaining the observed experimental data: piece of chain on the top, side, and bottom part of the codend,
respectively (Fig. 1b). Further, each of the covers had 12 kites
attached to the cover around the bulk of the catch in the codend. The covers had a nominal mesh size of 50 mm and were strength-
ened with an outer layer of large-meshed netting in the aft part. Can. J. Fish. Aquat. Sci. Downloaded fro The performance of the trawl was monitored continuously with a
set of trawl door sensors, a trawl height sensor, and a catch volume
sensor. During the trials, the catch from each compartment was kept
in separate holding bins. The length of all cod, haddock, and redfish
above 20 cm was measured to the nearest centimetre rounded down. Can. J. Fish. Aquat. Sci. Dow ð1Þ
X
m
j¼1
X
l
nClj ln rcodend l; vcodend
ð
Þ
½
þ nCClj
ð1Þ ln 1:0 rcodend l; vcodend
ð
Þ
½
g Published by Canadian Science Publishing Study area, experimental design, and data collection During the
first part of the experimental period, the two codends were tested
with a regular lastridge rope configuration (no shortening),
whereas in the second part of the experimental period the lastridge
ropes in the last 6 m of both codends were shortened by 15%. Experimental fishing was conducted onboard the research ves-
sel Helmer Hanssen (63.9 m long, 4080 HP (3000 kW)) from 8 to
16 January 2021 in the southern part of the Barents Sea (71°
2206500N–72°0803000N, 25°4809200E–30°1304400E). The experimental
fishing was conducted using an Alfredo 5 twin-body trawl (trouser-
trawl; Grimaldo et al. 2007) combined with a set of Injector Scor-
pion trawl doors, each weighing 3100 kg and with an area of 8 m2. Can. J. Fish. Aquat. Sci. Downloaded from cdnsciencepu
For perso Cod (Gadus morhua) and haddock (Melanogrammus aeglefinus) are
the main target species in the Barents Sea demersal trawl fishery,
and redfish (Sebastes spp.) are among the main bycatch species. Large cod and haddock often acquire a higher price per kilogram
than smaller individuals, and fishermen generally aim to maxi-
mize the revenue from their limited quotas. Therefore, fisher-
men in this area often are interested in only catching cod and
haddock well above (approximately 5 cm) the minimum legal
size (MLS), which is 44 cm for cod and 40 cm for haddock. The
MLS for redfish is 32 cm. These three fish have substantial mor-
phological (Sistiaga et al. 2011; Herrmann et al. 2012) and behav-
ioral differences (Engås and Godø 1989; Larsen et al. 2016). Thus,
the effects of gear modifications on size-selection properties and
catch patterns vary among them. The entire length of the codends was covered with small-
meshed covers that caught fish escapees. To ensure that the cov-
ers stayed clear of the codend netting, the front part of each of
the covers was equipped with six floats, three kites, and a 12 kg Although earlier studies have documented the performance of
codends with shortened lastridge ropes compared to other gear
(Lök et al. 1997; Ingolfsson and Brinkhof 2020), research docu-
menting the potential gains of applying shortened lastridge Can. J. Fish. Aquat. Sci. Vol. 79, 2022 836 Fig. 1. Illustration showing the twin-body trawl (a) and the configuration of the covers (CC) over the codends (C) (b). [Colour online.] Fig. 1. Sistiaga et al. (Herrmann et al. 2009), to estimate the size-selective potential for
the diamond mesh codends used during the experimental fishing. Application of FISHSELECT to simulate size selectivity through
codend meshes for a species requires (i) a morphological model
describing the cross-sections of importance for size selection of the
species and (ii) a model describing how and to what extent the fish
cross-sections can be squeezed when trying to pass through a mesh. The FISHSELECT models necessary to study cod, haddock, and red-
fish size selectivity in diamond mesh codends for the Barents Sea
demersal trawl fishery were already available from studies con-
ducted by Sistiaga et al. (2011) and Herrmann et al. (2012). Based on
these FISHSELECT models, we simulated the size selection in stiff di-
amond meshes (mesh shape cannot be deformed by fish trying to
escape through it) with a mesh size identical to the two codends
applied in the experimental fishing. Mesh opening angles between
10 and 90 degrees, in 10-degree increments, were tested to establish
the potential size selection in the codend and its dependency on the
mesh opening angle. In addition, we simulated the potential size
selection for slack meshes (meshes can potentially be fully deformed
by the effort of the fish while trying to escape) of the same mesh
size. For each simulated size-selection data set obtained in this way,
we fitted a logit selection model to obtain a size-selection curve. It is
likely that fish will have multiple chances to attempt to escape, espe-
cially in the catch accumulation zone (Herrmann 2005a). If unsuc-
cessful in a prior attempt, it is likely that decisive attempts will not
be represented by the average mesh size but instead by meshes bi-
ased to some extent towards the maximum mesh size available in
the codend. To account for this scenario in the simulations, we con-
sidered mean mesh sizes of 128 and 137 mm as well as mesh size +
2 times the standard deviations as an estimate for maximum mesh
size for each of the codends (i.e.,134 and 142 mm, respectively). Evaluating the ability of a model to describe the data sufficiently well
was based on estimating the corresponding p value, which expresses
the likelihood of obtaining at least as big a discrepancy between the
fitted model and the observed experimental data by coincidence. Exploitation pattern indicators for the codends Finally, based on the bootstrap population, Efron 95% percen-
tile confidence limits were obtained for Dr(l) as described above. To investigate how the different codend configurations affected the
capture pattern for each species separately and address research ques-
tion 3, we estimated the value of three exploitation pattern indicators,
nP, nP+, and nDiscard (discard ratio). These indicators are often used
in fishing gear size selectivity studies to supplement assessment solely
based on selectivity curves (Santos et al. 2016; Sala et al. 2017; Cheng
et al. 2019; Kalogirou et al. 2019; Melli et al. 2020). To estimate these
exploitation pattern indicators, we first applied the predicted size-
selection curves for each codend to the population of each species
entering the fishing gear, which was estimated from the population
entering the gear summed over all codends during the experimental
fishing. The population size structure nPopl for each individual cur
199
com
mo
O
cie
ma
the
sel
me
sel
pin
tak
tio
gro
boo
inc
for
rep
ana
str
lyz
run
con
on
201
Est
T
qu
ene
me
T
an
ð2Þ
wh
95%
the
tiv
bo
pro
ð3Þ
F
tile
Can. J. Fish. Aquat. Sci. Downloaded from cdnsciencepub.com by Danmarks Tekniske Informationscenter on 04/28/22
For personal use only. (
)
Once the specific size-selection model was identified for each spe-
cies and codend configuration, bootstrapping was applied to esti-
mate the confidence limits for the average size selection. We used
the software tool SELNET (Herrmann et al. 2012) for the size-
selection analysis, and the double bootstrap method was imple-
mented in the tool to obtain the confidence limits for the size-
selection curve and the corresponding parameters. This bootstrap-
ping approach is identical to the one described in Millar (1993) and
takes into consideration both within-haul and between-haul varia-
tion. The hauls for each codend configuration were treated as a
group of hauls. To account for between-haul variation, an outer
bootstrap resample with replacement from the group of hauls was
included in the procedure. Within each resampled haul, the data
for each length class were bootstrapped in an inner bootstrap with
replacement to account for within-haul variation. For each species
analyzed, 1000 bootstrap repetitions were conducted. Each boot-
strap run resulted in a set of data that was pooled and then ana-
lyzed using the identified selection model. Thus, each bootstrap
run resulted in an average selection curve. The Efron percentile 95%
confidence limits for the average selection curve were obtained based
on the same 1000 bootstrap repetitions (Efron 1982; Herrmann et al. 2012). Can. J. Fish. Aquat. Sci. Downloaded from cdnsciencepub.com by
For personal use o (
p
y)
We also investigated whether the experimental size-selection data
for cod, haddock, and redfish obtained for the different codends in
the sea trials could be understood based on the FISHSELECT simula-
tions. Therefore, we evaluated whether the experimental size-
selection curves based on the data collected during the sea trials
could be replicated by simulating scenarios assuming different
combinations of mesh states (i.e., mesh sizes and opening angles). We considered stiff diamond meshes for both the mean mesh sizes
and the mean mesh sizes + 2 times the standard deviation for open-
ing angles between 10 and 90 degrees. We also considered slack
meshes for all four mesh sizes. We then identified the combination
of varying mesh openness and state that was best able to reproduce
the experimental size-selection curves obtained during the experi-
mental fishing for each species for each codend separately. Estimation of difference in size selectivity between codends Modeling and estimation of the size selection in the codends Modeling and estimation of the size selection in the codends
To identify potential selectivity differences between the different
codends tested, it was first necessary to estimate the size-selection
properties of each of the different codends tested individually. The
data for each species were analyzed separately using the method
described here. The experimental design (Fig. 1) applied to test the
codends enabled us to analyze the catch data as binominal data. The numbers of individuals per length class, retained either
by the codend cover or by the codend itself, were used to esti-
mate the size selection in the codend (i.e., length-dependent
retention probability). The size selectivity between hauls for
the same codend is expected to vary (Fryer 1991). However, we
were interested in the size selection averaged over hauls because
it would provide information about the average consequences
for the size-selection process when using the codend in the
fishery. We tested different parametric models of the form To identify potential selectivity differences between the different
codends tested, it was first necessary to estimate the size-selection
properties of each of the different codends tested individually. The
data for each species were analyzed separately using the method
described here. The experimental design (Fig. 1) applied to test the
codends enabled us to analyze the catch data as binominal data. where nClj and nCClj are the numbers of fish in the codend and
codend cover for length class l in haul j, respectively. The outer summation in eq. 1 comprises the hauls j (from 1 to
m) conducted with the specific codend, and the inner summation
is over the length classes l in the data. Four different models were chosen as basic candidates to describe
rcodend (l, vcodend) for each codend and species individually: Logit, Pro-
bit, Gompertz, and Richard.The first three models are fully described
by the selection parameters L50 (length of fish with 50% probability
of being retained) and SR (difference in length between fish with 75%
and 25% probability of being retained, respectively), whereas the
Richard model requires an additional parameter (d) that describes
the asymmetry of the curve.The formulas for the four selection mod-
els and additional information can be found in Lomeli (2019). 837 Sistiaga et al. Published by Canadian Science Publishing Estimation of difference in size selectivity between codends The analysis presented in this subsection was linked to research
question 1, which aimed at discerning if, and to what extent, short-
ened lastridge ropes modify the selection properties of diamond
mesh codends. Can. J. Fish. Aquat. Sci. Downloaded from cdnsc
F The difference in size selectivity Dr(l) between two codends x
and y was estimated by ð2Þ
Dr lð Þ ¼ ry lð Þ rx lð Þ To conduct this analysis, we used the selection curves, with CIs
and retention lengths, obtained from the analysis of the sea trial
data and the simulated retention data for different mesh openness
and different mesh states from FISHSELECT. We estimated the con-
tributions needed from the different retention data to obtain com-
bined selection curves that best fitted the experimentally obtained
data. This procedure is identical to the one applied by Herrmann
et al. (2013, 2016) and Cuende et al. (2020), who provide detailed in-
formation on the technical aspects of the method. where x and y represent the different codends, respectively. The
95% confidence intervals (CIs) for Dr(l) were obtained based on
the two bootstrap population results for rx(l) and ry(l), respec-
tively. As they were obtained independently of each other, a new
bootstrap population of results for Dr(l) was created using the
procedure described in Larsen et al. (2018): ð3Þ
Dr lð Þi ¼ ry lð Þi rx lð Þi
i 2 1 . . . 1000
½
ð3Þ Sistiaga et al. Therefore, for the fitted model to be a candidate to model the size-
selection data, this p value should not be <0.05 (Wileman et al. 1996). In case of a poor fit statistic (p < 0.05), the residuals were inspected to
determine whether the poor result was due to structural problems
when modeling the experimental data with the different selection
curves or if it was due to overdispersion in the data (Wileman et al. 1996). The best model among the four considered was selected by
comparing their Akaike information criterion (AIC) values. The
model with the lowest AIC value was selected (Akaike 1974). Understanding codend size selection based on fish
morphology and mesh geometry For cod and haddock, we esti-
mated the indicators for the current MLS (44 and 40 cm, respectively)
and for an MLS of 50 cm for cod and 45 cm for haddock, which repre-
sents the scenario in which fishermen are interested in catching fish
only well above the MLS. Ideally, nP and nDiscard should be low
(close to 0), while nP+ should be high (close to 100). The indicators
were estimated for the different codends by Overview of sea trials We conducted 31 hauls during the experimental period, 6 of them
with the 128 and 137 mm codends in the standard configuration
(without shortened lastridge ropes) and 25 with the same codends in
the shortened lastridge configuration. In total, we measured 12938
cod,12162 haddock, and 3119 redfish during the trials (Table 1). Effect of shortening lastridge ropes on size selectivity g
g
p
y
The L50 values estimated for the two codends in the shortened
lastridge rope configuration were always higher than the equiva-
lent in the standard configuration (Table 2). A comparison of the
selectivity curves and the corresponding delta plots obtained for
cod, haddock, and redfish with the codends in the standard con-
figuration and the shortened lastridge rope configuration showed
that in general, shortening the lastridge ropes decreased the reten-
tion probability for the smaller length classes (Fig. 4). For the
128 mm codend, shortening the lastridge ropes resulted in no
significant decrease in the retention probability of cod, a slight
but significant decrease for some length classes of haddock, and a
more considerable and significant effect on redfish (Figs. 4b, 4f, 4j). For the 137 mm codend, on the other hand, shortening the las-
tridge ropes led to a more pronounced reduction over a larger
range of length classes for all three species (Figs. 4d, 4h, 4l). For this
codend, the effect was largest for redfish and similar for cod and
haddock. The probability of retaining fish above the MLS and the discard
ratio are two important indicators to consider when comparing
the performance of different gear, as the former is a measure of
the efficiency of the gear and the latter is a measure of the under-
sized fish caught with respect to the number of fish above the
MLS caught. We used these two indicators to compare the per-
formance of the four codend configurations tested in the present
study with that of the Sort-V and Flexigrid grid sections com-
bined with a 130 mm codend. Comparison with the gear currently used in the fishery p
g
y
y
To complete the answer to research question 3 and assess the
performance of the four codend configurations tested in this
study relative to the gear currently used in the Barents Sea, we
first estimated the exploitation pattern indicators for the Sort-V
and Flexigrid grid systems combined with a diamond mesh
codend for cod, haddock, and redfish. We then compared these
results to those obtained in the present study for the four codend
configurations tested. The selectivity data used to estimate the
indicators for cod and haddock with a Sort-V grid combined with
a diamond mesh codend were obtained from Sistiaga et al. (2010),
whereas the data for the Flexigrid and codend system for these
two species were obtained from Brinkhof et al. (2020). Note that
the codend used together with the Sort-V grid in Sistiaga et al. (2010) had a mesh size of 135 mm, which was the minimum mesh
size in the codend at the time. The selectivity data used for red-
fish were presented in Herrmann et al. (2013). As the exploitation
pattern indicators depend on the fish population in the area at
the time the trials are conducted (nPopl), the nPopl used to esti-
mate the indicators for cod, haddock, and redfish with the grid
systems was the same as that used to estimate the indicators for
the four codend configurations tested in the present study. Can. J. Fish. Aquat. Sci. Downloaded from cdnsciencepub.com by Danmarks Tekniske Informationscenter on 04/28/22
For personal use only. The size selectivity analysis results showed primarily that the
models used to represent the data for all four codend configura-
tions tested for cod, haddock, and redfish were adequate. The
Richard model was found to describe the data best in most cases. In all cases, the p value for the model with the lowest AIC value
among the models considered was >0.05, which indicates that
the difference between the experimental points and the model
used in every case could be coincidental (Table 2). This result was
corroborated by the selectivity curves, which fitted the experi-
mental data well in every case (Fig. 2). ð4Þ
nP ¼ 100
X
l<MLS rcodend l; vcodend
ð
Þ nPopl
X
l<MLS nPopl
ð
Þ
nPþ ¼ 100
X
l>MLS rcodend l; vcodend
ð
Þ nPopl
X
l>MLS nPopl
ð
Þ
nDiscard ¼ 100
X
l<MLS rcodend l; vcodend
ð
Þ nPopl
X
l rcodend l; vcodend
ð
Þ nPopl
ð4Þ
nP ¼ 100
X
l<MLS rcodend l; vcodend
ð
Þ nPopl
X
l<MLS nPopl
ð
Þ
nPþ ¼ 100
X
l>MLS rcodend l; vcodend
ð
Þ nPopl
X
l>MLS nPopl
ð
Þ
nDiscard ¼ 100
X
l<MLS rcodend l; vcodend
ð
Þ nPopl
X
l rcodend l; vcodend
ð
Þ nPopl
ð4Þ
nP ¼ 100
X
l<MLS rcodend l; vcodend
ð
Þ nPopl
X
l<MLS nPopl
ð
Þ
nPþ ¼ 100
X
l>MLS rcodend l; vcodend
ð
Þ nPopl
X
l>MLS nPopl
ð
Þ
nDiscard ¼ 100
X
l<MLS rcodend l; vcodend
ð
Þ nPopl
X
l rcodend l; vcodend
ð
Þ nPopl
ð4Þ Published by Canadian Science Publishing Can. J. Fish. Aquat. Sci. Downloaded from cdnsciencepub.com by Danmarks Te
For personal use only. All indicators (nP, nP+, and nDiscard) were estimated with
uncertainties for each codend using the bootstrap set for rcodend
(l, vcodend) and nPopl. Specifically, based on Herrmann et al. (2018), the bootstrap set for estimating indicator values was
obtained based on each bootstrap repetition result in which rcodend
(l, vcodend) and nPopl were applied simultaneously in eq. 4. Finally,
based on the resulting bootstrap set, 95% CIs were obtained for
each of the indicators. All analyses of the exploitation pattern indi-
cators were conducted using SELNET (Herrmann et al. 2012). Can. J. Fish. Aquat. Sci. Downloaded from cdnsciencepub
For perso Understanding codend size selection based on fish
morphology and mesh geometry The objective of the analysis in this subsection was to answer
research question 2 in the study; that is, investigate if it is possible to
explain the selectivity results obtained for cod, haddock, and redfish
by their morphological characteristics and the potential changes in
the codend meshes generated by shortening the lastridge ropes. Herein, we applied the FISHSELECT methodology, which is a
framework of methods, tools, and software developed to determine
if a fish can penetrate a certain mesh shape and size in fishing gear 838 Can. J. Fish. Aquat. Sci. Vol. 79, 2022 Can. J. Fish. Aquat. Sci. Vol. 79, 2022 species was obtained based on the data for all hauls from all codend
designs by summing catches in the codend and cover. Uncertainties
in populations were obtained by double bootstrapping following the
approach described in Melli et al. (2020). We then estimated the per-
centage of individuals retained for individuals below (nP) and above
(nP+) a specified MLS, respectively, for each codend. We also estimated
nDiscard, which is a measure of the number of undersized fish rela-
tive to the number of fish in the haul. For cod and haddock, we esti-
mated the indicators for the current MLS (44 and 40 cm, respectively)
and for an MLS of 50 cm for cod and 45 cm for haddock, which repre-
sents the scenario in which fishermen are interested in catching fish
only well above the MLS. Ideally, nP and nDiscard should be low
(close to 0), while nP+ should be high (close to 100). The indicators
were estimated for the different codends by species was obtained based on the data for all hauls from all codend
designs by summing catches in the codend and cover. Uncertainties
in populations were obtained by double bootstrapping following the
approach described in Melli et al. (2020). We then estimated the per-
centage of individuals retained for individuals below (nP) and above
(nP+) a specified MLS, respectively, for each codend. We also estimated
nDiscard, which is a measure of the number of undersized fish rela-
tive to the number of fish in the haul. Effect of increasing mesh size on size selectivity Effect of increasing mesh size on size selectivity For cod, haddock, and redfish, the L50 values estimated for the
128 mm codend with both the standard and the shortened las-
tridge configuration were always lower than those for the
137 mm codend with the same configuration (Table 2). A compari-
son of the selectivity curves and the corresponding delta plots
between the 128 and 137 mm codends in the standard configura-
tion also illustrate the difference between the codends for all
three species (Fig. 3). When the curves were compared for the
codends in the standard configuration, the differences observed
were significant for a few length classes that included fish above
and below the MLS for haddock but only for fish above the MLS
for cod and redfish (Figs. 3b, 3f, 3j). However, when the codends
were compared in the shortened lastridge configuration, the dif-
ferences between the codends increased substantially for all
three species. Not only was the difference larger, but it was also
significant for a larger number of length classes. For all three spe-
cies, the 128 mm codend with shortened lastridge ropes captured
significantly more fish of length classes both above and below
the MLS, although the number of length classes that differed
between the codends was substantially larger for cod and had-
dock than for redfish (Figs. 3d, 3h, 3l). Size selectivity results m cdnsciencepub.com by Danmarks Tekniske Informationscenter on 04/28/22
For personal use only. ulation of the experimental selectivity curves an The simulation results showed that for all four codend configura-
tions and the three species included in the study, the experimental
selectivity curves could be well explained by a combination of 839 Sistiaga et al. Table 1. Overview of the hauls conducted during the experimental sea trials and the numbers (n) of cod, haddock, and redfish retained in the
codend (C) and codend cover (CC) in each haul. Haul
No. Can. J. Fish. Aquat. Sci. Downloaded from cdnsciencepub.com by Danmarks Te
For personal use only. Can. J. Fish. Aquat. Sci. Downloaded from cdnsciencepub
For perso Table 2. Selection model, selectivity parameters, and fit statistics for the four codend configurations tested and the three species sampled
during the sea trials Table 2. Selection model, selectivity parameters, and fit statistics for the four codend configurations teste
during the sea trials. Table 2. Selection model, selectivity parameters, and fit statistics for the four codend configurations tested and the three species sampled
during the sea trials. Table 2. Selection model, selectivity parameters, and fit statistics for the four codend configurations tested and the three species sampled
during the sea trials. Species
Lastrige ropes
Mesh
size (mm)
Model
L50 (cm)
SR (cm)
d
Deviance
df
p
Cod
Standard
128
Logit
41.20 (38.06–43.42)
8.75 (5.58–13.04)
—
34.00
79
>0.999
137
Richard
44.29 (41.25–47.07)
12.28 (8.43–16.61)
0.19 (0.10–0.42)
39.17
82
>0.999
Shortened
128
Probit
41.79 (39.47–43.78)
9.63 (7.95–11.24)
—
53.14
92
>0.999
137
Logit
49.14 (48.21–49.92)
6.13 (5.33–7.00)
—
37.89
88
>0.999
Haddock
Standard
128
Probit
39.20 (38.50–39.86)
7.14 (5.96–8.33)
—
21.71
52
0.993
137
Richard
41.07 (39.61–42.32)
6.75 (4.95–7.86)
0.63 (0.30–1.60)
30.30
49
0.984
Shortened
128
Richard
40.53 (39.77–41.17)
6.75 (6.16–7.33)
0.67 (0.44–1.13)
50.01
54
0.629
137
Richard
45.12 (44.51–45.72)
6.31 (5.79–6.87)
0.62 (0.40–1.01)
38.81
58
0.975
Redfish
Standard
128
Richard
32.77 (31.38–34.93)
6.38 (4.21–8.51)
0.60 (0.13–1.37)
37.09
35
0.373
137
Richard
35.15 (32.54–38.61)
9.05 (6.04–12.50)
0.13 (0.10–0.40)
8.47
37
1.000
Shortened
128
Richard
38.57 (37.17–39.64)
7.60 (5.55–10.06)
0.19 (0.10–0.44)
41.60
39
0.355
137
Richard
42.47 (41.37–43.46)
6.52 (4.81–8.46)
0.35 (0.10–1.01)
42.46
40
0.366
Note: L50 is length of fish with 50% probability of being retained, SR is the difference in length between fish with 75% and 25% probability of being retained, and d
represents the asymmetry parameter in the Richard model (Lomeli 2019). STD is the standard configuration (nonshortened lastridge ropes), and SL is the codend
with shortened lastridge ropes. Values in parentheses represent 95% confidence intervals. Can. J. Fish. Aquat. Sci. Downloaded Note: L50 is length of fish with 50% probability of being retained, SR is the difference in length between fish with 75% and 25% probability of being retained, and d
represents the asymmetry parameter in the Richard model (Lomeli 2019). STD is the standard configuration (nonshortened lastridge ropes), and SL is the codend
with shortened lastridge ropes. Values in parentheses represent 95% confidence intervals. Published by Canadian Science Publishing ulation of the experimental selectivity curves an Duration
tow (min)
Depth
(m)
Gear
Cod
Haddock
Redfish
128 mm
137 mm
128 mm
137 mm
128 mm
137 mm
nC
nCC
nC
nCC
nC
nCC
nC
nCC
nC
nCC
nC
nCC
1
188
315.73
STD
90
10
86
12
131
66
105
75
17
99
12
95
2
149
328.07
STD
116
5
117
16
110
51
90
87
11
94
7
84
3
173
311.19
STD
168
6
207
8
176
72
226
129
31
123
94
94
4
151
287.28
STD
68
3
97
6
115
94
187
114
29
6
24
12
5
124
345.00
STD
54
4
53
0
49
27
56
19
230
101
238
156
6
150
324.05
STD
158
8
209
18
127
61
192
110
81
97
71
169
7
130
254.28
SL
247
17
300
25
145
55
118
95
13
9
17
17
8
120
294.23
SL
108
2
131
8
73
39
78
43
11
7
18
4
9
145
255.55
SL
266
15
297
28
174
117
138
147
16
6
12
8
10
129
315.95
SL
291
3
319
16
148
67
139
108
14
7
24
16
11
120
237.82
SL
190
14
170
15
149
38
118
72
11
10
23
27
12
129
305.34
SL
246
14
391
39
131
65
158
121
20
2
20
9
13
122
321.79
SL
49
2
71
9
43
7
32
31
4
2
7
2
14
120
298.94
SL
108
6
121
12
83
64
91
57
13
2
9
5
15
120
261.87
SL
234
11
293
30
200
141
183
236
13
12
8
10
16
122
311.80
SL
218
3
307
27
158
103
164
206
21
6
5
13
17
121
308.68
SL
167
8
200
11
147
122
122
154
16
7
9
7
18
124
272.86
SL
192
7
259
27
102
48
125
127
11
11
16
13
19
120
312.52
SL
183
11
223
14
133
96
100
128
14
9
25
9
20
120
279.67
SL
222
10
264
34
141
72
153
171
17
6
11
17
21
128
301.11
SL
190
10
226
16
133
101
120
177
10
5
6
15
22
121
282.78
SL
119
3
174
25
90
49
79
126
16
14
14
10
23
125
298.49
SL
199
4
211
9
105
52
89
69
11
6
6
7
24
137
278.92
SL
109
7
146
1
70
38
69
65
8
2
10
6
25
121
299.40
SL
120
2
138
12
111
35
78
47
12
4
6
7
26
123
280.12
SL
162
6
212
16
117
76
117
117
15
13
9
14
27
126
273.58
SL
227
4
283
0
121
62
90
77
19
12
10
20
28
121
261.98
SL
393
10
495
31
128
68
134
111
19
4
7
22
29
125
298.08
SL
199
5
217
14
84
44
70
72
14
16
13
24
30
147
252.07
SL
198
2
243
18
37
20
54
57
9
27
8
19
31
130
266.45
SL
226
2
236
14
12
14
12
20
7
6
3
8
Note: STD is the standard configuration (nonshortened lastridge ropes), and SL is the codend with shortened lastridge ropes. ulation of the experimental selectivity curves an Table 1. Overview of the hauls conducted during the experimental sea trials and the numbers (n) of cod, haddock, and redfish retained in the
codend (C) and codend cover (CC) in each haul. Can. J. Fish. Aquat. Sci. Downloaded from cdnsciencepub.com by Danmarks Te
For personal use only. Finally, the simulation of the
results obtained experimentally for redfish showed that compared
to cod and haddock, redfish potentially have greater ability to
utilize meshes with higher opening angles or slack meshes that
are deformable upon escape. Shortening the lastridge ropes likely
allowed redfish to make use of meshes with higher opening angles
and especially slack meshes. Finally, the simulations estimated that
when the 128 and 137 mm codends were fished in the shortened Published by Canadian Science Publishing Can. J. Fish. Aquat. Sci. Downloaded from cdnsciencepub.com by Danmarks Te
For personal use only. utilize more open meshes or slack meshes to escape when the
codends with short lastridge ropes were used (Table 3). This result
indicates that with this configuration the longitudinal forces in the
codend meshes were lower, providing greater availability of the
more open meshes and slack meshes. contributions from different mesh sizes and opening angles. In
every case, the simulated selectivity curve was within the CIs of
the experimental selectivity curves (Fig. 5). Further, the potential
contributions of the different meshes and mesh openings showed
that in general, cod, haddock, and especially redfish were able to 840 Can. J. Fish. Aquat. Sci. Vol. 79, 2022 Fig. 2. Length-dependent retention probabilities for cod, haddock, and redfish with the four codend configurations tested during the
trials. STD is the standard configuration (nonshortened lastridge ropes), and SL is the codend with shortened lastridge ropes. In each plot,
the circles represent the experimental observations, the solid curve represents the models fitted to the data, and the dashed curves
represent the 95% CIs. The grey line represents the population fished by the gear (codend + cover). The dashed vertical grey lines show the
minimum legal size (MLS) for cod (44 cm), haddock (40 cm), and redfish (32 cm). Can. J. Fish. Aquat. Sci. Downloaded from lastridge configuration, 46.95% and 62.77%, respectively, of the
redfish that escaped through the codend meshes may have done
so through the largest meshes in the slack state available in the
codend (Table 3). lastridge configuration, 46.95% and 62.77%, respectively, of the
redfish that escaped through the codend meshes may have done
so through the largest meshes in the slack state available in the
codend (Table 3). The simulations showed that when the 128 mm codend was
employed, cod may have escaped through similar opening angles
and mesh sizes independent of which gear configuration was used. For the 137 mm mesh codend, however, cod may have been able to
use more of the larger meshes available and meshes with 10°–20°
higher opening angles when the shortened lastridge configuration
was used in the codend (Table 3). The simulation results showed a
similar pattern for haddock. However, the meshes with opening
angles of 40°–50° may have been more important for haddock than
for cod, whereas meshes with opening angles of 20°–30° showed
higher relevance for cod (Table 3). Exploitation pattern indicators for the four codend
configurations tested Exploitation pattern indicators depend on the fish population
in the fishing area at the time of the trials. Therefore, to conduct
a fair comparison between the different codends tested, the indi-
cators for the four codend configurations tested during the trials
were estimated based on the length-frequencies of the fish popu-
lation encountered during the whole trial period (Fig. 6). For cod, the catch pattern indicators showed that the probabil-
ity of catching fish under the MLS of 44 cm and the discard ratio
decreased when we increased the mesh size from 128 to 137 mm,
but the decrease was only statistically significant for the short-
ened lastridge rope configuration. When comparing the two gear 841 Sistiaga et al. Fig. 3. Comparison of the 128 mm (black) and 137 mm (grey) codends tested in both the standard (STD) and the short lastridge (SL)
configurations. Delta plots of the comparisons are also shown. The dashed curves represent the 95% CIs in each case. The dashed vertical
grey lines show the MLS for cod (44 cm), haddock (40 cm), and redfish (32 cm). grey lines show the MLS for cod (44 cm), haddock (40 cm), and redfish (32 cm). m cdnsciencepub.com by Danmarks Tekniske Informationscenter on 04/28/22
For personal use only. Can. J. Fish. Aquat. Sci. Downloaded from cdnsciencepub
For person Can. J. Fish. Aquat. Sci. Downloaded fro configurations for the 128 mm codend, the gear change did not have
a significant effect on either parameter. However, for the 137 mm
codend, shortening the lastridge ropes significantly decreased the
probability of capturing cod < 44 cm and the discard ratio, and the
probability of retaining cod > 44 cm decreased from 97.4% to 94.1%. Increasing the MLS to 50 cm increased the probability of retaining
cod both below and above the MLS. The discard ratio increased sig-
nificantly for all four configurations (Table 4). example, the discard ratio for the 128 mm codend in the standard
configuration increased from 8.7% to 28.1% when the MLS was
increased from 40 to 45 cm. example, the discard ratio for the 128 mm codend in the standard
configuration increased from 8.7% to 28.1% when the MLS was
increased from 40 to 45 cm. For redfish, the probability of catching fish below or above MLS did
not change significantly when the codend mesh size increased from
128 to 137 mm in either configuration. Exploitation pattern indicators for the four codend
configurations tested However, when we compared
the two configurations with 128 or 137 mm codends, the probability
of catching redfish below and above the MLS was substantially lower
in the shortened lastridge configuration, and the reduction was stat-
istically significant for the probability of catching redfish below the
MLS for the 137 mm codend. The discard ratio did not differ signifi-
cantly among any of the four codend configurations tested (Table 4). For haddock, as for cod, increasing mesh size in the standard
codend configuration had no significant effect on any of the pa-
rameters estimated. The discard ratio only decreased from 8.7%
to 6.7%, and although the probability of retaining fish < 40 cm
decreased from 17.9% to 12.8%, the reduction was not statistically
significant. However, all three indicators differed significantly when
the codends were compared in the shortened lastridge configura-
tion. As for cod, increasing the MLS from 40 to 45 cm significantly
increased the retention probability for haddock above and below
the MLS and the discard ratio for all four configurations tested. For Published by Canadian Science Publishing Published by Canadian Science Publishing Comparison of the exploitation pattern indicators of the
four codend configurations tested with those of the gear
currently used in the fishery Regardless of the MLS considered, the retention probability for
cod with the four codend configurations tested was significantly
higher than that for the two grid configurations (Fig. 7). The dis-
card ratio for cod was <1% for all six configurations when the
MLS was 44 cm and <5% when the MLS was increased to 50 cm. Although the discard ratio differences were not large, they were
significant among all codend configurations tested except the m cdnsciencepub.com by Danmarks Tekniske Informationscenter on 04/28/22
For personal use only. Can. J. Fish. Aquat. Sci. Downloaded from codend system were low, but the retention probability for fish
above the MLS was also low and under 30% (Table 5). codend system were low, but the retention probability for fish
above the MLS was also low and under 30% (Table 5). were estimated based on the length-frequencies of the fish popula-
tion encountered during the whole trial period (Fig. 6). The results
for the grid systems showed that the probability of retaining fish
under the MLS was low (<5% for cod, <1% for haddock, and <1% for
redfish). Increasing the MLS to 50 cm for cod and 45 cm for haddock
increased the probability of catching undersized cod to approxi-
mately 15% and <3% for haddock. These increases were significant
in both cases. The results also showed that while the retention
probabilities for cod over MLS (nP+) with the Sort-V grid and Flexi-
grid were over 87% and 83%, respectively, regardless of the MLS
used, for haddock the retention probability with the grid systems
could be as low as 24% and not higher than 47% (Table 5). With
increasing MLS, the discard ratio increased by approximately 1%
for cod for both grids and approximately 4% for haddock with the
Sort-V grid and 130 mm codend, and the increase was statistically
significant in both cases (Table 5). The retention probability for
undersized redfish and the discard ratio with the Sort-V grid and Considering the current MLS for cod in the Barents Sea, the
retention probability of commercial fish for all four codend con-
figurations tested in this study was >94% in all cases, whereas
retention probability was 87% and 83% for the Sort-V grid and
Flexigrid systems, respectively. Comparison of the exploitation pattern indicators of the
four codend configurations tested with those of the gear
currently used in the fishery The exploitation pattern indicators for the Sort-V and Flexigrid
grid systems combined with a diamond mesh codend (Table 5) Published by Canadian Science Publishing 842 Can. J. Fish. Aquat. Sci. Vol. 79, 2022 Fig. 4. Comparison of the 128 and 137 mm codends in the standard (STD) configurations (black) and the configuration with shortened lastridge
ropes (SL; grey). Delta plots of the comparisons are also shown. The dashed curves represent the 95% CIs in each case. The dashed vertical grey
lines show the MLS for cod (44 cm), haddock (40 cm), and redfish (32 cm). were estimated based on the length-frequencies of the fish popula-
tion encountered during the whole trial period (Fig. 6). The results
for the grid systems showed that the probability of retaining fish
under the MLS was low (<5% for cod, <1% for haddock, and <1% for
redfish). Increasing the MLS to 50 cm for cod and 45 cm for haddock
increased the probability of catching undersized cod to approxi-
mately 15% and <3% for haddock. These increases were significant
in both cases. The results also showed that while the retention
probabilities for cod over MLS (nP+) with the Sort-V grid and Flexi-
grid were over 87% and 83%, respectively, regardless of the MLS
used, for haddock the retention probability with the grid systems
could be as low as 24% and not higher than 47% (Table 5). With
increasing MLS, the discard ratio increased by approximately 1%
for cod for both grids and approximately 4% for haddock with the
Sort-V grid and 130 mm codend, and the increase was statistically
significant in both cases (Table 5). The retention probability for
undersized redfish and the discard ratio with the Sort-V grid and
codend system were low, but the retention probability for fish
above the MLS was also low and under 30% (Table 5). Considering the current MLS for cod in the Barents Sea, the
retention probability of commercial fish for all four codend con-
figurations tested in this study was >94% in all cases, whereas
retention probability was 87% and 83% for the Sort-V grid and
Flexigrid systems, respectively. If the MLS was increased to 50 cm
for cod, the retention probability for all four codend configura-
tions tested would be >96%, whereas it would be 90% and 86% for
the Sort-V grid and Flexigrid systems, respectively (Tables 4–5). Published by Canadian Science Publishing Comparison of the exploitation pattern indicators of the
four codend configurations tested with those of the gear
currently used in the fishery The blue line to the right in
each plot shows the selection curve for a slack mesh of 134 mm (a, c, e, g, i, and k) and 142 mm (b, d, f, h, j, and l) in each case. The dashed
vertical grey lines show the MLS for cod (44 cm), haddock (40 cm), and redfish (32 cm). [Colour online.] 137 mm codend with shortened lastridge ropes and the Sort-V
grid and Flexigrid systems (Tables 4–5; Fig. 7). At the MLS of 40 cm, the retention probability for haddock for
the four codend configurations tested varied between 70% and 91%,
whereas the values were 36% to 24% for the Sort-V and Flexigrid sys-
tems, respectively (Tables 4–5). The difference between the four
codend configurations and the grids was significant (Fig. 7). Increas-
ing the MLS to 45 cm increased the retention probability of had-
dock in all cases, with estimated values of 85%–97% for the four
codend configurations tested, 47% for the Sort-V system, and 31% for
the Flexigrid system (Tables 4–5). The difference between all four
codends and the two grid systems was still statistically significant
(Fig. 7). However, the discard ratio was significantly higher for the
test codends than for the two grid systems in every case, regardless of
the MLS considered. At the MLS of 40 cm, the discard ratio for the test
codends never exceeded 9%, but increasing the MLS to 45 cm resulted
in a 28% discard ratio for the 128 mm codend in the standard configu-
ration (Tables 4–5; Fig.7). The retention probability for redfish >32 cm was significantly
higher for the 128 mm codend and the 137 mm codend in the
standard configuration compared to the Sort-V system, because
the lower confidence limit in these three cases does not overlap
with the upper confidence limit for the Sort-V system. The dis-
card ratio was substantially lower with the Sort-V grid than with
vertical grey lines show the MLS for cod (44 cm), haddock (40 cm), and redfish (32 cm). [Colour online.]
q
p
y
For personal use only. 137 mm codend with shortened lastridge ropes and the Sort-V
grid and Flexigrid systems (Tables 4–5; Fig. 7). (Fig. 7). However, the discard ratio was significantly higher for the
test codends than for the two grid systems in every case, regardless of
the MLS considered. Published by Canadian Science Publishing Comparison of the exploitation pattern indicators of the
four codend configurations tested with those of the gear
currently used in the fishery If the MLS was increased to 50 cm
for cod, the retention probability for all four codend configura-
tions tested would be >96%, whereas it would be 90% and 86% for
the Sort-V grid and Flexigrid systems, respectively (Tables 4–5). Regardless of the MLS considered, the retention probability for
cod with the four codend configurations tested was significantly
higher than that for the two grid configurations (Fig. 7). The dis-
card ratio for cod was <1% for all six configurations when the
MLS was 44 cm and <5% when the MLS was increased to 50 cm. Although the discard ratio differences were not large, they were
significant among all codend configurations tested except the Considering the current MLS for cod in the Barents Sea, the
retention probability of commercial fish for all four codend con-
figurations tested in this study was >94% in all cases, whereas
retention probability was 87% and 83% for the Sort-V grid and
Flexigrid systems, respectively. If the MLS was increased to 50 cm
for cod, the retention probability for all four codend configura-
tions tested would be >96%, whereas it would be 90% and 86% for
the Sort-V grid and Flexigrid systems, respectively (Tables 4–5). Regardless of the MLS considered, the retention probability for
cod with the four codend configurations tested was significantly
higher than that for the two grid configurations (Fig. 7). The dis-
card ratio for cod was <1% for all six configurations when the
MLS was 44 cm and <5% when the MLS was increased to 50 cm. Although the discard ratio differences were not large, they were
significant among all codend configurations tested except the 843 Sistiaga et al. Fig. 5. Experimental (black) and simulated (purple) size-selection curves for the four codend configurations tested during the trials. Dashed curves (black) show the 95% CIs. STD is the standard configuration (nonshortened lastridge ropes), and SL is the codend with
shortened lastridge ropes. The red curves show selection curves simulated in FISHSELECT for meshes of 128 mm (a, c, e, g, i, and k) and
137 mm (b, d, f, h, j, and l) with opening angles of 20°, 50°, and 80°. The blue curves show selection curves simulated in FISHSELECT for
meshes of 134 mm (a, c, e, g, i, and k) and 142 mm (b, d, f, h, j, and l) with opening angles of 20°, 50°, and 80°. Comparison of the exploitation pattern indicators of the
four codend configurations tested with those of the gear
currently used in the fishery At the MLS of 40 cm, the discard ratio for the test
codends never exceeded 9%, but increasing the MLS to 45 cm resulted
in a 28% discard ratio for the 128 mm codend in the standard configu-
ration (Tables 4–5; Fig.7). (Fig. 7). However, the discard ratio was significantly higher for the
test codends than for the two grid systems in every case, regardless of
the MLS considered. At the MLS of 40 cm, the discard ratio for the test
codends never exceeded 9%, but increasing the MLS to 45 cm resulted
in a 28% discard ratio for the 128 mm codend in the standard configu-
ration (Tables 4–5; Fig.7). g
g
y
(
g
)
At the MLS of 40 cm, the retention probability for haddock for
the four codend configurations tested varied between 70% and 91%,
whereas the values were 36% to 24% for the Sort-V and Flexigrid sys-
tems, respectively (Tables 4–5). The difference between the four
codend configurations and the grids was significant (Fig. 7). Increas-
ing the MLS to 45 cm increased the retention probability of had-
dock in all cases, with estimated values of 85%–97% for the four
codend configurations tested, 47% for the Sort-V system, and 31% for
the Flexigrid system (Tables 4–5). The difference between all four
codends and the two grid systems was still statistically significant The retention probability for redfish >32 cm was significantly
higher for the 128 mm codend and the 137 mm codend in the
standard configuration compared to the Sort-V system, because
the lower confidence limit in these three cases does not overlap
with the upper confidence limit for the Sort-V system. The dis-
card ratio was substantially lower with the Sort-V grid than with ed in rep
Mesh
size (m
137
137
137
137
137
137
137
137
137
142
142
142
142
142
142
142
142
142
142 opening angles (OAs), and mesh states considered as being potentially involved in r
d redfish. ope
red
12
co
ST
—
—
4.
25
23
2.
1.
—
—
—
—
5.
20
9.
6.
—
—
—
k
—
oden Comparison of the exploitation pattern indicators of the
four codend configurations tested with those of the gear
currently used in the fishery Redfish
128 mm
codend
137 mm
codend
128 mm
codend
STD
SL
STD
SL
STD
SL
Mesh
size (mm)
OA
Mesh
size (mm)
OA
Mes
size
—
—
137
10°
—
—
128
10°
—
—
137
—
—
137
20°
0.45
—
128
20°
—
—
137
4.636
2.420
137
30°
0.028
0.928
128
30°
3.25
—
137
25.78
20.773
137
40°
18.79
3.203
128
40°
9.704
9.828
137
23.95
28.100
137
50°
26.73
24.952
128
50°
18.91
0.001
137
2.089
1.184
137
60°
1.589
12.049
128
60°
5E-04
0.006
137
1.092
2.207
137
70°
0.112
0.648
128
70°
0.004
0.002
137
—
—
137
80°
—
0.373
128
80°
1.072
0.003
137
—
—
137
90°
—
0.085
128
90°
0.009
0.000
137
—
—
142
10°
—
—
134
10°
—
—
142
—
—
142
20°
0.109
—
134
20°
—
—
142
5.32
4.671
142
30°
13.93
3.921
134
30°
3.676
—
142
20.86
19.950
142
40°
31.81
24.213
134
40°
5.827
6.097
142
9.345
14.186
142
50°
4.9
24.454
134
50°
1.747
0.002
142
6.924
6.507
142
60°
0.513
1.013
134
60°
0.002
0.007
142
—
—
142
70°
1.042
0.170
134
70°
26.93
3.646
142
—
—
142
80°
—
0.830
134
80°
26.87
23.173
142
—
—
142
90°
—
3.162
134
90°
1.989
10.279
142
k
—
—
142
Slack
—
—
134
Slack
—
46.950
142
odend with shortened lastridge ropes. 845 Sistiaga et al. Fig. 6. Size distribution of cod, haddock, and redfish populations
encountered during the experimental trials. The vertical dashed
grey lines show the MLS for each species. The black dashed lines
show the 95% confidence intervals for the variation in the
populations encountered for each species. Fig. 6. Size distribution of cod, haddock, and redfish populations
encountered during the experimental trials. The vertical dashed
grey lines show the MLS for each species. The black dashed lines
show the 95% confidence intervals for the variation in the
populations encountered for each species. compared to the standard configuration. The effect on the size-
selection properties of the diamond mesh in the shortened las-
tridge configuration was clear for both codends, but it was more
pronounced for the 137 mm codend. Can. J. Fish. Aquat. Sci. Downloaded from cdnsciencepub.com by Danmarks Tekniske Informationscenter on 04/28/22
For personal use only. In recent years, the use of exploitation pattern indicators has
gained popularity in size selectivity studies (Santos et al. 2016;
Sala et al. 2017; Cheng et al. 2019; Kalogirou et al. 2019; Melli et al. 2020) because they provide a good picture of how the gear per-
forms with respect to the management objectives and alternative
catch pattern objectives in the fishery. Considering the MLS for
cod, haddock, and redfish, the estimated indicator values showed
that the tested codend configurations performed quite differ-
ently. While the 137 mm codend with shortened lastridge ropes
retained <5% of undersized fish of all three species and >94% of
the cod above MLS, it resulted in a loss of 30% and 60% of com-
mercial haddock and redfish, respectively. On the other hand,
reducing the mesh size to 128 mm for the same codend configura-
tion reduced the loss of commercial haddock and redfish to 13%
and 40%, respectively, but the catch of undersized cod with this
codend configuration increased to over 30%. The indicator results
obtained with the 137 mm shortened lastridge codend fit with
the goals of the fleet of keeping haddock and cod larger than 45
and 50 cm, respectively, whereas using the 128 mm codend cap-
tured lower value haddock and cod. However, the shortened las-
tridge 137 mm codend that caught <5% of cod below the MLS also
caught over 25% of fish below 50 cm. Overall, these indicator
results illustrate the challenge of multispecies fisheries and the
difficulty of finding optimal gear solutions that provide satisfac-
tory and efficient results for different species simultaneously. Our results also show that a change of 5 or 6 cm in the legal or
desired minimum size of a certain species can notably change
the performance of the gear with respect to this new potential
goal. However, we must stress that the indicators depend on the
specific population the gear encounters for each species during
the trials and that selectivity estimates can provide a more gen-
eral picture of the selective performance of the gear tested. Can. J. Fish. Aquat. Sci. Downloaded from cdnsciencepub.com
For personal us all codend configurations except for the 137 mm codend with
shortened lastridge ropes. However, the difference was not statis-
tically significant in any of the cases (Fig. 7). Can. J. Fish. Aquat. Sci. Downloaded fro Published by Canadian Science Publishing Comparison of the exploitation pattern indicators of the
four codend configurations tested with those of the gear
currently used in the fishery As the hanging ratio was the
same for both codends, this difference likely was due to the stiff
netting material used, which could have reduced the effect of the
shortened lastridge ropes for the smaller mesh size. The simulation carried out using the existing FISHSELECT mod-
els for cod, haddock, and redfish (Sistiaga et al. 2011; Herrmann
et al. 2012) showed that it is indeed possible to explain the selectiv-
ity results obtained for these three species and the four diamond
mesh codend configurations tested in our study. The results indi-
cate that when using shortened lastridge codends, the availability
of meshes with high opening angles is larger and all three species
investigated are able to escape through these meshes. The largest
contributions were for mesh opening angles of 40°–60° for cod and
haddock and 80°–90° for redfish. It is unclear why the largest con-
tribution to size selectivity for redfish changed from nearly square
meshes when using the standard configuration to slack meshes
when using the short lastridge configuration. Redfish is a robust
fish that tries so hard to squeeze itself through meshes that it often
gets stuck (Isaksen and Valdemarsen 1986; ICES 2012). However,
considering the stiffness of the material used in the codends (single
braided polyethylene hotmelt twine, Ø8 mm), it is difficult to
understand how the meshes could be slack enough to deform and
allow redfish to pass through them. The experimental design and
data analysis in this study do not allow us to provide a clear expla-
nation for the observed redfish selectivity results other than those
already discussed. Discussion In this study, we compared catch patterns for two diamond
mesh codends with different mesh sizes in a standard and a short-
ened lastridge configuration. The results demonstrated that both
increasing mesh size from 128 to 137 mm and shortening the las-
tridge ropes for both codends, so that they were 15% shorter than
the stretched codend netting, can change the selection properties
of the codend for cod, haddock, and redfish (Figs. 3–4). The effect
of mesh size was a consequence of the fact that physically larger
fish are able to penetrate larger meshes. That both codends had
the same number of free meshes around but different mesh size
could also have contributed to the differences observed due to
differences in codend circumference, which have earlier been demon-
strated potentially affect codend size selectivity (Sala and Lucchetti
2011). The selectivity changes caused by shortening the lastridge
ropes occurred because this modification removes the tension
from the netting generated by the accumulation of fish inside the
codend, which results in slacker and more open meshes (Herrmann
2005a, 2005b). The effect on selectivity of increasing mesh size was
more pronounced for the codends in the shortened lastridge con-
figuration than in the standard configuration. Because shortening
lastridge ropes contributes to more slack meshes with higher
opening angles in the codend, we expected a larger effect of
changing codend mesh size on size selection with this configuration Compared to the mandatory sorting grid and codend gear used
in the Barents Sea gadoid fishery, all codend configurations
tested in this study retained significantly more commercial-sized
cod and had a discard ratio that was only marginally larger. The
pattern was similar when the minimum size was for cod was set
at 50 cm, but in this case the retention of commercial cod was
substantially larger and the discard ratio was always <5%. From 846 Can. J. Fish. Aquat. Sci. Vol. 79, 2022 Table 4. Exploitation pattern indicator values for the four codend configurations tested and the
three species sampled during the sea trials. Can. J. Fish. Aquat. Sci. Downloaded from cdnsciencepub.com by Danmarks Tekniske Informationscenter on 04/28/22
For personal use only. For redfish, the differen-
ces between the Sort-V grid system and the tested codends were
similar to but not as clear as those for cod and haddock, so it is
more difficult to draw a conclusion about the extent to which the
fishery would benefit from removing the grids and using any of
the different types of codends tested in this study. this perspective, the diamond mesh codends, and especially the
Table 5. Exploitation pattern indicator values obtained for two grid
and codend gear configurations used in the fishery today. Indicator
Sort-V + codend
Flexigrid + codend
Cod
nP– 44 cm (%)
4.6 (2.8–7.4)
3.7 (1.7–7.0)
nP+ 44 cm (%)
87.2 (84.5–89.4)
83.1 (79.0–86.2)
nDiscard (%)
0.1 (0.1–0.2)
0.1 (0.1–0.2)
nP– 50 cm (%)
16.0 (12.5–20.4)
14.9 (9.7–20.6)
nP+ 50 cm (%)
90.4 (87.9–92.4)
86.1 (82.1–89.1)
nDiscard (%)
1.3 (1.0–1.7)
1.3 (0.9–1.7)
Haddock
nP– 40 cm (%)
0.5 (0.2–0.9)
0.2 (0.1–0.5)
nP+ 40 cm (%)
35.9 (31.3–40.0)
23.7 (20.3–27.7)
nDiscard (%)
0.7 (0.3–1.2)
0.4 (0.1–0.9)
nP– 45 cm (%)
2.2 (1.5–3.0)
0.9 (0.4–1.6)
nP+ 45 cm (%)
46.9 (41.4–51.7)
31.5 (27.2–36.2)
nDiscard (%)
4.5 (3.2–6.1)
2.7 (1.1–4.7)
Redfish
nP– 32 cm (%)
0.1 (0.0–2.6)
—
nP+ 32 cm (%)
29.1 (20.0–45.0)
—
nDiscard (%)
0.2 (0.0–4.7)
—
Note: The minimum mesh size in the codend, which was 135 mm in 2010, is
now 130 mm. The selectivity data for the estimation of the indicators are based
on the data presented in Sistiaga et al. (2010), Herrmann et al. (2012), and
Brinkhof et al. (2020). The populations used for all three species are those
shown in Figure 6. Indicator values for cod are shown for MLS of 44 and 50 cm;
indicator values for haddock are shown for MLS of 40 and 45 cm; and indicator
values for redfish are shown for MLS of 32 cm. Indicator
Sort-V + codend
Flexigrid + codend Can. J. Fish. Aquat. Sci. Downloaded fr In general, the indicator results obtained and our comparison
of the performance of the compulsory grid systems used in the
Barents Sea today with the codends tested in our study showed
that in many cases shortened lastridge codends can provide a bet-
ter catch pattern than the grid system for the species of interest. Published by Canadian Science Publishing Can. J. Fish. Aquat. Sci. Downloaded from cdnsciencepub.com by Danmarks Tekniske Informationscenter on 04/28/22
For personal use only. Note: Indicator values for cod are shown for MLS of 44 and 50 cm; indicator values for haddock are shown fo
MLS of 40 and 45 cm; and indicator values for redfish are shown for MLS of 32 cm. Table 5. Exploitation pattern indicator values obtained for two grid
and codend gear configurations used in the fishery today. Table 5. Exploitation pattern indicator values obtained for two grid
and codend gear configurations used in the fishery today. were similar to those for cod, although for this species the differ-
ences between the grid systems and the codends tested in the
present study were more pronounced. It is clear from the results
that removing the grids from the fishery would significantly
increase the retention of haddock over the current MLS and had-
dock above 45 cm. However, the discard ratio for the codends
tested was much larger than for the grid and codend configura-
tions. For three of the four codends tested when the MLS was set
at 45 cm, 25% of the catch would be below this size. Only the
137 mm codend with shortened lastridge ropes was able to keep
the discard ratio for haddock below 15%. For redfish, the differen-
ces between the Sort-V grid system and the tested codends were
similar to but not as clear as those for cod and haddock, so it is
more difficult to draw a conclusion about the extent to which the
fishery would benefit from removing the grids and using any of
the different types of codends tested in this study. were similar to those for cod, although for this species the differ-
ences between the grid systems and the codends tested in the
present study were more pronounced. It is clear from the results
that removing the grids from the fishery would significantly
increase the retention of haddock over the current MLS and had-
dock above 45 cm. However, the discard ratio for the codends
tested was much larger than for the grid and codend configura-
tions. For three of the four codends tested when the MLS was set
at 45 cm, 25% of the catch would be below this size. Only the
137 mm codend with shortened lastridge ropes was able to keep
the discard ratio for haddock below 15%. Discussion Indicator
Standard
Short lastridges
128 mm
137 mm
128 mm
137 mm
Cod
nP– 44 cm (%)
35.5 (22.1–50.3)
30.1 (18.1–41.8)
33.2 (22.7–44.0)
4.4 (2.5–7.3)
nP+ 44 cm (%)
98.3 (97.0–99.2)
97.4 (95.5–98.7)
98.1 (97.6–98.6)
94.1 (93.0–95.0)
nDiscard (%)
0.9 (0.5–1.4)
0.8 (0.4–1.1)
0.8 (0.5–1.2)
0.1 (0.1–0.2)
nP– 50 cm (%)
66.2 (57.8–74.1)
54.3 (41.4–65.6)
62.9 (55.1–70.4)
26.3 (20.2–33.0)
nP+ 50 cm (%)
99.1 (98.0–99.8)
98.8 (97.6–99.7)
99.1 (98.8–99.4)
96.8 (96.1–97.5)
nDiscard (%)
4.8 (4.1–5.8)
4.0 (3.0–4.9)
4.6 (3.8–5.4)
2.0 (1.5–2.5)
Haddock
nP– 40 cm (%)
17.9 (14.1–22.0)
12.8 (6.6–18.0)
14.2 (11.7–17.4)
4.4 (3.2–5.7)
nP+ 40 cm (%)
90.6 (88.1–92.9)
86.2 (82.0–91.2)
87.8 (85.7–89.8)
70.0 (67.1–72.9)
nDiscard (%)
8.7 (6.7–10.5)
6.7 (3.6–9.1)
7.2 (5.9–8.8)
2.9 (2.1–3.9)
nP– 45 cm (%)
37.5 (33.9–41.7)
30.3 (23.7–37.0)
32.6 (29.3–36.7)
13.9 (12.1–15.9)
nP+ 45 cm (%)
96.8 (94.9–98.3)
94.8 (92.0–98.3)
95.5 (94.5–96.5)
83.9 (81.4–86.2)
nDiscard (%)
28.1 (25.2–31.1)
24.4 (19.9–28.7)
25.6 (22.6–28.5)
14.3 (12.4–16.4)
Redfish
nP– 32 cm (%)
13.8 (5.7–22.9)
16.8 (5.9–25.7)
8.1 (3.5–14.1)
2.3 (0.3–5.4)
nP+ 32 cm (%)
85.4 (72.0–94.4)
75.4 (56.7–89.8)
60.4 (50.6–75.6)
41.3 (30.2–57.9)
nDiscard (%)
8.6 (3.7–14.1)
11.5 (4.4–17.7)
7.3 (2.9–12.1)
3.2 (0.4–7.1)
Note: Indicator values for cod are shown for MLS of 44 and 50 cm; indicator values for haddock are shown for
MLS of 40 and 45 cm; and indicator values for redfish are shown for MLS of 32 cm. Table 4. Exploitation pattern indicator values for the four codend configurations tested and the
three species sampled during the sea trials. Can. J. Fish. Aquat. Sci. Downloaded from cdnsciencepub.com by Danmarks Tekniske Informationscenter on 04/28/22
For personal use only. Values for the indicators nP+ (individuals above MLS in %) and discard ratio (%) for cod for MLS of 44 and 50 cm, for haddock for
MLS of 40 and 45 cm, and for redfish for 32 cm (Tables 4–5) for the four codends tested in this study and the two grid systems used today
in the Barents Sea bottom trawl gadoid fishery. STD is the standard configuration (nonshortened lastridge ropes), and SL is the codend
with shortened lastridge ropes. The abscissa labels are consistent for each column. Fig. 7. Values for the indicators nP+ (individuals above MLS in %) and discard ratio (%) for cod for MLS of 44 and 50 cm, for haddock for
MLS of 40 and 45 cm, and for redfish for 32 cm (Tables 4–5) for the four codends tested in this study and the two grid systems used today
in the Barents Sea bottom trawl gadoid fishery. STD is the standard configuration (nonshortened lastridge ropes), and SL is the codend
with shortened lastridge ropes. The abscissa labels are consistent for each column. gear, species like cod are often more dependent on additional stim-
uli to attempt escape (Tschernij and Suuronen 2002; Grimaldo
et al. 2018). Decompression experienced during haul back can be
an additional escape stimulus (Madsen et al. 2008; Grimaldo et al. 2009, 2014), but it creates additional risk of injury and potentially
reduced survival for the escapees (Breen et al. 2007). Earlier studies
reported that contrary to the selectivity of codends, grid selectivity
is a more mechanical size-selection process that takes place at the
fishing depth (Grimaldo et al. 2009). This argument is often used by
the management authorities in the Barents Sea to maintain the
grid and codend configuration that is compulsory in the area today. Whether the properties of codends with shortened lastridge ropes
are different from ordinary codends in this respect is unknown
and should be investigated, as the availability of more open meshes
in the codend may stimulate fish to escape earlier in the capture
process. MLS of 40 and 45 cm, and for redfish for 32 cm (Tables 4 5) for the four codends tested in this study and the two grid systems used today
in the Barents Sea bottom trawl gadoid fishery. STD is the standard configuration (nonshortened lastridge ropes), and SL is the codend
with shortened lastridge ropes. Can. J. Fish. Aquat. Sci. Downloaded from cdnsciencepub.com by Danmarks Tekniske Informationscenter on 04/28/22
For personal use only. Particularly for cod, and to a large extent for haddock, the
137 mm codend with shortened lastridge ropes resulted in a sig-
nificantly higher retention of fish above the MLS with an insignif-
icant or small increase in the discard ratio compared to the
compulsory grid system. Therefore, in terms of size selection, our
results show that a codend with shortened lastridge ropes is an
alternative to the grid and codend gear currently required in the
Barents Sea demersal trawl fishery. Note: The minimum mesh size in the codend, which was 135 mm in 2010, is
now 130 mm. The selectivity data for the estimation of the indicators are based
on the data presented in Sistiaga et al. (2010), Herrmann et al. (2012), and
Brinkhof et al. (2020). The populations used for all three species are those
shown in Figure 6. Indicator values for cod are shown for MLS of 44 and 50 cm;
indicator values for haddock are shown for MLS of 40 and 45 cm; and indicator
values for redfish are shown for MLS of 32 cm. Note: The minimum mesh size in the codend, which was 135 mm in 2010, is
now 130 mm. The selectivity data for the estimation of the indicators are based
on the data presented in Sistiaga et al. (2010), Herrmann et al. (2012), and
Brinkhof et al. (2020). The populations used for all three species are those
shown in Figure 6. Indicator values for cod are shown for MLS of 44 and 50 cm;
indicator values for haddock are shown for MLS of 40 and 45 cm; and indicator
values for redfish are shown for MLS of 32 cm. Despite the positive selectivity results obtained with the codends
in the shortened lastridge configuration and their maneuverability
and encouraging performance compared to the grid systems, other
aspects need to be considered. For example, it is important to
understand how and when fish escape through the selection de-
vice. Selectivity through codend meshes is highly dependent on
fish behavior, meaning that fish must actively swim through the
meshes to escape. While species such as haddock are active in the this perspective, the diamond mesh codends, and especially the
137 mm codend, with shortened lastridge ropes resulted in more
satisfactory selection than the grid and codend configurations
used in the fishery today. The patterns observed for haddock 847 Sistiaga et al. Fig. 7. Published by Canadian Science Publishing Can. J. Fish. Aquat. Sci. Downloaded from cdnsciencepub.com by Danmarks Tekniske Informationscenter on 04/28/22
For personal use only. The abscissa labels are consistent for each column. Can. J. Fish. Aquat. Sci. Downloaded from cdnsciencepub.com by Danmarks Tekniske Informationscenter on 04/28/22
For personal use only. gear, species like cod are often more dependent on additional stim-
uli to attempt escape (Tschernij and Suuronen 2002; Grimaldo
et al. 2018). Decompression experienced during haul back can be
an additional escape stimulus (Madsen et al. 2008; Grimaldo et al. 2009, 2014), but it creates additional risk of injury and potentially
reduced survival for the escapees (Breen et al. 2007). Earlier studies
reported that contrary to the selectivity of codends, grid selectivity
is a more mechanical size-selection process that takes place at the gear, species like cod are often more dependent on additional stim-
uli to attempt escape (Tschernij and Suuronen 2002; Grimaldo
et al. 2018). Decompression experienced during haul back can be
an additional escape stimulus (Madsen et al. 2008; Grimaldo et al. 2009, 2014), but it creates additional risk of injury and potentially
reduced survival for the escapees (Breen et al. 2007). Earlier studies
reported that contrary to the selectivity of codends, grid selectivity
is a more mechanical size-selection process that takes place at the fishing depth (Grimaldo et al. 2009). This argument is often used by
the management authorities in the Barents Sea to maintain the
grid and codend configuration that is compulsory in the area today. Whether the properties of codends with shortened lastridge ropes
are different from ordinary codends in this respect is unknown
and should be investigated, as the availability of more open meshes
in the codend may stimulate fish to escape earlier in the capture
process. fishing depth (Grimaldo et al. 2009). This argument is often used by
the management authorities in the Barents Sea to maintain the
grid and codend configuration that is compulsory in the area today. Whether the properties of codends with shortened lastridge ropes
are different from ordinary codends in this respect is unknown
and should be investigated, as the availability of more open meshes
in the codend may stimulate fish to escape earlier in the capture
process. Published by Canadian Science Publishing 848 Can. J. Fish. Aquat. Sci. Vol. 79, 2022 Can. J. Fish. Aquat. Sci. Vol. 79, 2022 Netting meshes can change their physical properties over
time, which can lead to that selectivity devices constructed of
netting meshes lose their selection properties. Can. J. Fish. Aquat. Sci. Downloaded from cdnsciencepub.com by Danmarks Tekniske Informationscenter on 04/28/22
For personal use only. Square-meshed
panels (e.g., the BACOMA codend (Herrmann et al. 2015; Madsen
et al. 2015), codends with lateral exit windows (Grimaldo et al. 2008, 2009), and T90 codends (ICES 2011; Madsen et al. 2015;
Cheng et al. 2020)) have good selection properties for cod and
haddock. However, deformation of the meshes and loss of stiff-
ness over time may change the selection properties of these types
of codends. Likewise, codends with lastridge ropes can poten-
tially lose their properties over time. Ropes, especially twisted
ropes, stretch with use, and this property depends on rope con-
struction and material (McKenna et al. 2004). If ropes increase in
length, the effect of shortened lastridge ropes would be reduced
over time and the meshes in the codend would close. If ropes
stretch, the crew may have to adjust them repeatedly to avoid los-
ing the selective properties of the gear and comply with regula-
tions. A potential solution to avoid stretching is the use of
Dyneema ropes, which in principle stretch little (<3.5%; Thomas
and Lekshmi 2017). However, Dyneema ropes have little load
absorption due to their limited stretchability. Thus, material
selection is a key to designing appropriate lastridge ropes, and
further research of the quality and performance over time of dif-
ferent types of lastridge ropes is necessary. Engås, A., and Godø, O.R. 1989. The effect of different sweep lengths on the
length composition of bottom-sampling trawl catches. J. Cons. Int. Explor. 45: 263–268. doi:10.1093/icesjms/45.3.263. Fryer, R.J. 1991. A model of between-haul variation in selectivity. ICES J. Mar. Sci. 48: 281–290. doi:10.1093/icesjms/48.3.281. /
/
Graham, N., Kynoch, R., and Fryer, R. 2003. Square mesh panels in demersal
trawls: further data relating haddock and whiting selectivity to panel
position. Fish. Res. 62: 361–375. doi:10.1016/S0165-7836(02)00279-5. p
/
(
)
Grimaldo, E., Larsen, R.B., and Holst, R. 2007. Exit windows as an alternative
selective system for the Barents Sea demersal fishery for cod and had-
dock. Fish. Res. 85: 295–305. doi:10.1016/j.fishres.2007.03.005. Grimaldo, E., Sistiaga, M., and Larsen, R.B. 2008. Evaluation of codends with
sorting grids, exit windows, and diamond meshes: size selection and fish
behaviour. Fish. Res. 91: 271–280. doi:10.1016/j.fishres.2007.12.003. Grimaldo, E., Larsen, R.B., Sistiaga, M., Madsen, N., and Breen, M. 2009. Se-
lectivity and escape percentages during three phases of the towing pro-
cess for codends fitted with different selection systems. Fish. Res. 95 (2–3):
198–205. doi:10.1016/j.fishres.2008.08.019. /j
Grimaldo, E., Sistiaga, M., and Larsen, R.B. Can. J. Fish. Aquat. Sci. Downloaded from cdnsciencepub.com by Danmarks Tekniske Informationscenter on 04/28/22
For personal use only. 2014. Development of catch con-
trol devices in the Barents Sea cod fishery. Fish. Res. 155: 122–126. doi:10.1016/j.fishres.2014.02.035. /j
Grimaldo, E., Sistiaga, M., Herrmann, B., and Larsen, R.B. 2016. Trawl selec-
tivity in the Barents Sea demersal fishery. In Fisheries and aquaculture in
the modern world. Edited by H. Mikkola. IntechOpen. doi:10.5772/63019. Grimaldo, E., Sistiaga, M., Herrmann, B., Larsen, R.B., Brinkhof, J., and Tatone, I. 2018. Improving release efficiency of cod (Gadus morhua) and haddock
(Melanogrammus aeglefinus) in the Barents Sea demersal trawl fishery by
stimulating escape behaviour. Can. J. Fish. Aquat. Sci. 75(3): 402–416. doi:10.1139/cjfas-2017-0002. / j
Herrmann, B. 2005a. Effect of catch size and shape on the selectivity of dia-
mond mesh cod-ends: I Model development. Fish. Res. 71: 1–13. doi:10.1016/j. fishres.2004.08.024. Considering the results obtained in earlier trials (Isaksen and
Valdemarsen 1990; Lök et al. 1997; Ingolfsson and Brinkhof 2020)
and the results from our study, we conclude that codends with
shortened lastridge ropes are satisfactory selection devices that
could be used in the Barents Sea gadoid fishery and other fish-
eries to replace or supplement other sorting devices. However,
selection during the capture process and the properties and per-
formance of different types of lastridge ropes over time require
further investigation. Herrmann, B. 2005b. Effect of catch size and shape on the selectivity of dia-
mond mesh cod-ends: II Theoretical study of haddock selection. Fish. Res. 71: 15–26. doi:10.1016/j.fishres.2004.08.021. Herrmann, B., and O’Neill, F.G. 2005. Theoretical study of the between-haul
variation of haddock selectivity in a diamond mesh cod-end. Fish. Res. 74: 243–252. doi:10.1016/j.fishres.2005.01.022. Can. J. Fish. Aquat. Sci. Downloaded from cdnsciencepub.com by Da
For personal use only /j
Herrmann, B., Priour, D., and Krag, L.A. 2007. Simulation-based study of the
combined effect on cod-end size selection for round fish of turning mesh
90 degrees and of reducing the number of meshes in the circumference. combined effect on cod-end size selection for round fish
90 degrees and of reducing the number of meshes in th
Fish. Res. 84: 222–232. doi:10.1016/j.fishres.2006.10.020. 90 degrees and of reducing the number of meshes in th
Fish. Res. 84: 222–232. doi:10.1016/j.fishres.2006.10.020. Fish. Res. 84: 222–232. doi:10.1016/j.fishres.2006.10.020. Can. J. Fish. Aquat. Sci. Downloaded from cdnsciencepub
For perso Herrmann, B., Krag, L.A., Frandsen, R.P., Madsen, N., Lundgren, B., and Stæhr, K.J. 2009. Prediction of selectivity from morphological conditions: methodology
and a case study on cod (Gadus morhua). Fish. Res. 97: 59–71. doi:10.1016/j. fishres.2009.01.002. Acknowledgements We express our gratitude to the crew of the R/V Helmer Hanssen and
to John Terje Eilertsen, Ivan Tatone, Clemens Knittel, Gaute Ringvall,
Andreas Eilefsen, and Ronan Gombau for their help during the
experimental trials at sea. We also thank the Norwegian Directorate
of Fisheries and the Norwegian Fisheries and Aquaculture Research
Fund (project number 901633) for funding the project. Herrmann, B., Sistiaga, M., Nielsen, K.N., and Larsen, R.B. 2012. Understand-
ing the size selectivity of redfish (Sebastes spp.) in North Atlantic trawl
codends. J. Northw. Atl. Fish. Sci. 44: 1–13. doi:10.2960/J.v44.m680. Herrmann, B., Sistiaga, M., Larsen, R.B., Nielsen, K.N., and Grimaldo, E. 2013. Understanding sorting grid and codend size selectivity of Greenland halibut
(Reinhardtius hippoglossoides). Fish. Res. 146: 59–73. doi:10.1016/j.fishres.2013.04.004. Herrmann, B., Wienbeck, H., Karlsen, J.D., Stepputtis, D., Dahm, E., and
Moderhak, W. 2015. Understanding the release efficiency of Atlantic cod (Gadus
morhua) from trawls with a square mesh panel: effects of panel area, panel posi-
tion, and stimulation of escape response. ICES J. Mar. Sci. 72: 686–696. doi:10.1093/icesjms/fsu124. Can. J. Fish. Aquat. Sci. Downloade Published by Canadian Science Publishing References Akaike, H. 1974. A new look at the statistical model identification. IEEE
Trans. Autom. Control, 19: 716–723. doi:10.1109/TAC.1974.1100705. Herrmann, B., Krag, L.A., Feekings, J., and Noack, T. 2016. Understanding
and predicting size selection in diamond mesh codends for Danish sein-
ing: a study based on sea trials and computer simulations. Mar. Coast. Fish. 8: 277–291. doi:10.1080/19425120.2016.1161682. Bergstad, O.A., Jørgensen, T., and Dragesund, O. 1987. Life history and ecol-
ogy of the gadoid resources of the Barents Sea. Fish. Res. 5: 119–161. doi:10.1016/0165-7836(87)90037-3. Breen, M., Huse, I., Ingolfsson, O.A., Madsen, N., and Soldal, A.V. 2007. Sur-
vival: An assessment of mortality in fish escaping from trawl codends
and its use in fisheries management. EU Contract Q5RS-2002-01603 Final
Report. Herrmann, B., Krag, L.A., and Krafft, B.A. 2018. Size selection of Antarctic
krill (Euphausia superba) in a commercial codend and trawl body. Fish. Res. 207: 49–54. doi:10.1016/j.fishres.2018.05.028. /
ICES. 2011. Report of the Study Group on Turned 90³ Codend Selectivity, fo-
cusing on Baltic Cod Selectivity (SGTCOD), 4–6 May 2011, IMR, Reykjavik,
Iceland. ICES. CM 2011/SSGESST:08. Brinkhof, J., Larsen, R.B., Herrmann, B., and Sistiaga, M. 2020. Size selectiv-
ity and catch efficiency of bottom trawl with a double sorting grid and
diamond mesh codend in the North-east Atlantic gadoid fishery. Fish. Res. 231: 105647. doi:10.1016/j.fishres.2020.105647. /
ICES. 2012. Report of the ICES-FAO Working Group on Fishing Technology
and Fish Behaviour (WGFTFB). ICES CM 2012/SSGESST:07. Cheng, Z., Einarsson, H.A., Bayse, S., Herrmann, B., and Winger, P. 2019. Com-
paring size selectivity of traditional and knotless diamond-mesh codends in
the Iceland redfish (Sebastes spp.) fishery. Fish. Res. 216: 138–144. doi:10.1016/j. fishres.2019.04.009. Ingolfsson, O.A., and Brinkhof, J. 2020. Relative size selectivity of a four-
panel codend with short lastridge ropes compared to a flexigrid with a
regular codend in the Barents Sea gadoid trawl fishery. Fish. Res. 232:
105724. doi:10.1016/j.fishres.2020.105724. Cheng, Z., Winger, P.D., Bayse, S.M., Kebede, G.E., DeLouche, H., Einarsson, H.A.,
et al. 2020. Out with the old and in with the new: T90 codends improve size
selectivity in the Canadian redfish (Sebastes mentella) trawl fishery. Can. J. Fish. Aquat. Sci. 77(10): 1711–1720. doi:10.1139/cjfas-2020-0063. Isaksen, B., and Valdemarsen, J.W. 1986. Selectivity experiments with square
mesh codends in bottom trawl. ICES C. M. 1986/B:28. Isaksen, B., and Valdemarsen, J.W. 1990. Codend with short lastridge ropes
to improve size selectivity in fish trawls. ICES CM 1990/. B46:8. to improve size selectivity in fish trawls. ICES CM 1990/. References dnsciencepub.com by Danmarks Tekniske Informationscenter on 04/28/22
For personal use only Can. J. Fish. Aquat. Sci. Downloaded from cdnsciencepub.com by Danmarks Tekniske Informationscenter on 04/28/22
For personal use only. / j
Santos, J., Herrmann, B., Mieske, B., Stepputtis, D., Krumme, U., and Nilsson, H. 2016. Reducing flatfish by-catches in roundfish fisheries. Fish. Res. 184: 64–73. doi:10.1016/j.fishres.2015.08.025. nloaded from cdnsciencepub.com by Danmarks Tekniske Informationscenter on 04/28/2
For personal use only. Lomeli, M.J.M. 2019. Bycatch Reduction in Eastern North Pacific Trawl Fish-
eries. A dissertation for the degree of Doctor Philosophiae, The Arctic
University of Norway, Faculty of Biosciences, Fisheries and Economy,
Norwegian College of Fishery Science, Tromsø, Norway. Sistiaga, M., Grimaldo, E., and Larsen, R.B. 2008. Size selectivity patterns in
the north-east Arctic cod and haddock fishery with sorting grids of 55,
60, 70 and 80mm. Fish. Res. 93: 195–203. doi:10.1016/j.fishres.2008.04.014. Sistiaga, M., Herrmann, B., Grimaldo, E., and Larsen, R.B. 2010. Assessment of dual
selection in grid based selectivity systems. Fish. Res. 105: 187–199. doi:10.1016/j. fishres.2010.05.006. Madsen, N., Skeide, R., Breen, M., Krag, L.A., Huse, I., and Soldal, A.V. 2008. Selectivity in trawl codend during haul-back operation — an overlooked
phenomenon. Fish. Res. 91 (2–3): 168–174. doi:10.1016/j.fishres.2007.11.016. p
(
)
/j
Madsen, N., Hansen, K., and Madsen, N.A.H. 2015. Behavior of different trawl
codends concepts. Ocean Eng. 108: 571–577. doi:10.1016/j.oceaneng.2015.08.047. Sistiaga, M., Herrmann, B., Nielsen, K.N., and Larsen, R.B. 2011. Understand-
ing limits to cod and haddock separation using size selectivity in a multi-
species trawl fishery: an application of FISHSELECT. Can. J. Fish. Aquat. Sci. 68(5): 927–940. doi:10.1139/f2011-017. McKenna, H.A., Hearle, J.W.S., and O’Hear, N. 2004. Handbook of fibre rope
technology. Woodhead Publishing, Cambridge, England. Sistiaga, M., Brinkhof, J., Herrmann, B., Grimaldo, E., Langård, L., and Lilleng, D. 2016. Size selective performance of two flexible sorting grid designs in the
Northeast Arctic cod (Gadus morhua) and haddock (Melanogrammus aeglefinus)
fishery. Fish. Res. 183: 340–351. doi:10.1016/j.fishres.2016.06.022. Melli, V., Herrmann, B., Karlsen, J.D., Feekings, J.P., and Krag, L.A. 2020. Pre-
dicting optimal combinations of bycatch reduction devices in trawl
gears: a meta-analytical approach. Fish Fish. 21(2): 252–268. doi:10.1111/
faf.12428. Millar, R.B. 1993. Incorporation of between-haul variation using bootstrap-
ping and nonparametric estimation of selection curves. Fish. Bull. 91:
564–572. /
Thomas, S., and Lekshmi, M. 2017. Recent trends in fishing gear materials. In Recent trends in harvest and post-harvest technologies in fisheries. Edited by C.N. Ravishankar, M.M. Prasad, A.K. Mohanty, V. Published by Canadian Science Publishing References B Cuende, E., Arregi, L., Herrmann, B., Sistiaga, M., and Aboitiz, X. 2020. Predic-
tion of square mesh panel and codend size selectivity of blue whiting based
on fish morphology. ICES J. Mar. Sci. 77: 2857–2869. doi:10.1093/icesjms/
fsaa156. Jørgensen, T., Ingolfsson, I.A., Graham, N., and Isaksen, B. 2006. Size selection of
cod by rigid grids — Is anything gained compared to diamond mesh codends
only? Fish. Res. 79: 337–348. doi:10.1016/j.fishres.2006.01.017. y
/j
Kalogirou, S., Pihl, L., Maravelias, C.D., Herrmann, B., Smith, C.J., Papadopoulou, N.,
et al. 2019. Shrimp trap selectivity in a Mediterranean small-scale-fishery. Fish. Res. 211: 131–140. doi:10.1016/j.fishres.2018.11.006. Efron, B. 1982. The jackknife, the bootstrap and other resampling plans. SIAM Monograph No. 38, CBSM-NSF. doi:10.1137/1.9781611970319. Published by Canadian Science Publishing Sistiaga et al. 849 Larsen, R.B., and Isaksen, B. 1993. Size selectivity of rigid sorting grids in
bottom trawls for Atlantic cod (Gadus morhua) and haddock (Melanogrammus
aeglefinus). ICES Mar. Sci. Symp. 196: 178–182. Robertson, J.H.B., and Stewart, P.A.M. 1988. A comparison of size selection
of haddock and whiting by square and diamond mesh codends. J. Cons. CIEM. 44: 148–161. doi:10.1093/icesjms/44.2.148. g fi
)
y
p
Larsen, R.B., Herrmann, B., Sistiaga, M., Grimaldo, E., Tatone, I., and Onandia, I. 2016. Size selection of redfish (Sebastes spp.) in a double grid system: quantify-
ing escapement through individual grids and comparison to former grid trials. Fish. Res. 183: 385–395. doi:10.1016/j.fishres.2016.07.013. /
j
/
Sala, A., and Lucchetti, A. 2011. Effect of mesh size and codend circumfer-
ence on selectivity in the Mediterranean demersal trawl fisheries. Fish. Res. 110: 252–258. doi:10.1016/j.fishres.2011.04.012. Sala, A., Lucchetti, A., Piccinetti, C., and Ferretti, M. 2008. Size selection by dia-
mond and square-mesh codends in multi-species Mediterranean demersal
trawl fisheries. Fish. Res. 93: 8–21. doi:10.1016/j.fishres.2008.02.003. Fish. Res. 183: 385–395. doi:10.1016/j.fishres.2016.07.013. Larsen, R.B., Herrmann, B., Sistiaga, M., Brcic, J., Brinkhof, J., and Tatone, I. 2018. Could green artificial light reduce bycatch during Barents Sea deep-water
shrimp trawling? Fish. Res. 204: 441–447. doi:10.1016/j.fishres.2018.03.023. Sala, A., Brcic, J., Herrmann, B., Lucchetti, A., and Virgili, M. 2017. Assessment of
size selectivity in hydraulic clam dredge fisheries. Can. J. Fish. Aquat. Sci. 74(3):
339–348. doi:10.1139/cjfas-2015-0199. Lök, A., Tokaç, A., Tosunoˇglu, Z., Metin, C., and Ferro, R.S.T. 1997. The effects
of different codend design on bottom trawl selectivity in Turkish fish-
eries of the Aegean Sea. Fish. Res. 32: 149–156. doi:10.1016/S0165-7836(97)
00048-9. References Murugadas, Thomas, S., and Lekshmi, M. 2017. Recent trends in fishing gear materials. In Recent trends in harvest and post-harvest technologies in fisheries. Edited by C.N. Ravishankar, M.M. Prasad, A.K. Mohanty, V. Murugadas,
N. ManjuLekshmi, R. Anandan, et al. Central Institute of Fisheries Technol-
ogy, Kochi, India. Can. J. Fish. Aquat. Sci. Downloaded from cdnsciencepub.com by Danmarks Te
For personal use only. Edited by C.N. Ravishankar, M.M. Prasad, A.K. Mohanty, V. Murugadas,
N. ManjuLekshmi, R. Anandan, et al. Central Institute of Fisheries Technol-
ogy, Kochi, India. Norwegian Directorate of Fisheries. 2017. J-181-2017: Forskrift om bruk av
sorteringsristsystem i fiske med stormasket trål. [In Norwegian.] Avail-
able from https://www.fiskeridir.no/Yrkesfiske/Regelverk-og-reguleringer/
J-meldinger/Gjeldende-J-meldinger/J-181-2017. Tschernij, V., and Suuronen, P. 2002. Improving trawl selectivity in the Bal-
tic. Nordic Council of Ministers, Copenhagen, Denmark. TemaNord 2002. No. 512. Olsen, E., Aanes, S., Mehl, S., Holst, J.C., Aglen, A., and Gjøsæter, H. 2010. Cod, had-
dock, saithe, and capelin in the Barents Sea and adjacent waters: a review of
the biological value of the area. ICES J. Mar. Sci. 67: 87–101. doi:10.1093/icesjms/
fsp229. Wienbeck, H., Herrmann, B., Moderhak, W., and Stepputtis, D. 2011. Effect of net-
ting direction and number of meshes around on size selection in the codend for
Baltic cod (Gadus morhua). Fish. Res. 109: 80–88. doi:10.1016/j.fishres.2011.01.019. O’Neill, F.G., and Herrmann, B. 2007. PRESEMO — a predictive model of
codend selectivity — a tool for fisheries managers. ICES J. Mar. Sci. 64:
1558–1568. doi:10.1093/icesjms/fsm101. Wileman, D., Ferro, R.S.T., Fonteyne, R., and Millar, R.B. (Editors). 1996. Man-
ual of methods of measuring the selectivity of towed fishing gears. ICES
Cooperative Research report No. 215. doi:10.17895/ices.pub.4628.
|
https://openalex.org/W4392569719
|
https://jurnal.jomparnd.com/index.php/jp/article/download/999/823
|
Indonesian
| null |
Perkembangan Perguruan Tinggi Islam Di Indonesia
|
Jurnal Pendidikan dan Ilmu Sosial
| 2,023
|
cc-by-sa
| 5,956
|
Vol. 3. No. 1 Januari 2024
e-ISSN: 2809-7998 p-ISSN: 2809-8005 Vol. 3. No. 1 Januari 2024
e-ISSN: 2809-7998 p-ISSN: 2809-8005 Vol. 3. No. 1 Januari 2024
e-ISSN: 2809-7998 p-ISSN: 2809-8005 JUPEIS: Jurnal Pendidikan dan Ilmu Sosial
https://jurnal.jomparnd.com/index.php/jp JUPEIS: Jurnal Pendidikan dan Ilmu Sosial
https://jurnal.jomparnd.com/index.php/jp Perkembangan Perguruan Tinggi Islam Di Indonesia Rahmawati Ramli1*, Bahaking Rama2
1,2UIN Alauddin Makasar
Email: rahmawatiramli505@gmail.com 1* Abstrak Tujuan penulisan artikel ini adalah: (1) mengetahui sejarah perguruan tinggi islam di Indonesia,
(2) mengetahui perkembangan perguruan tinggi islam di Indonesia, (3) mengetahui tantangan dan
peluang perguruan tinggi islam di Indonesia. Penulisan makalah ini menggunakan metode Penelitian ini
menggunakan penelitian pustaka, dengan melakukan kajian pustaka yaitu analisis terhadap berbagai
artikel, jurnal dan buku yang terkait dengan tema. Selanjutnya, pengumpulan data penelitian dilakukan
dengan menggunakan tekhnik dokumentasi yaitu dengan upaya mencari data-data dengan menganalisis
dokumen-dokumen terkait perihal yang peneliti tulis. Dari hasil makalah tersebut maka dapar dirtarik
kesimpulan: Konversi alih status dari IAIN ke UIN memang membutuhkan persyaratan yang
cukup banyak, sehingga perubahan itu tidak dapat dilakukan dengan mudah begitu saja. Tidak
cukup persyaratan administrasi saja, akan tetapi juga persyaratan substantif. Saat ini terdapat 23
UIN dari total 58 PTKIN atau sekitar 40 % dari total perguruan tinggi Islam yang ada. Model
integrasi keilmuan di UIN Syarif Hidayatullah ialah reintegrasi ilmu, UIN Sunan Ampel Surabaya
konsep integrated twin towers (ITT), UINSumatera Utara mengusung Konsep Wahdatul Ulu, UIN
Sunan Kalijaga konsep Integrasi Interkoneksi. Kurikulum UIN tidak menganggap bahwa masing-
masing keilmuan terpisah melainkan satu kesatuan, yang bersumber dari Al-Quran dan Hadis. Untuk itulah yang membuatnya begitu berbeda dengan pendidikan tinggi lainnya. Keywords: Indonesia, Perkembangan, Perguruan Tinggi Islam PENDAHULUAN dengan berbagai kebijakan pemerintah yang
tampak
belum
memberikan
dukungan
sepenuhnya terhadap lembaga pendidikan
Islam. Meski demikian, satu hal yang tidak
bisa dipungkiri bahwa Pendidikan Islam
dengan semua lembaga pendidikannya telah
mewarnai
perjalanan
sejarah
bangsa
Indonesia (Rofi, 2016). Pemikir pendidikan
Islam di Indonesia selalu mencari berbagai
cara untuk membangun sistem pendidikan
Islam
yang
lengkap,
sistematis
dan
transparan, mulai pesantren yang sederhana
sampai tingkat perguruan tinggi (Rukiati dan
Henawati, 2006). Institut Agama Islam
Negeri (IAIN) yang sekarang menyebar
hampir
di
seluruh
nusantara
bukan
merupakan bentuk kelembagaan yang final Perjalanan
sejarah
pendidikan
di
Indonesia menjadi diskursus yang tidak ada
akhirnya. Dalam dimensi sejarah, terdapat tiga
bagian yang selalu bergerak bagai roda
kehidupan, yaitu masa lalu, masa kini dan
masa
yang
akan
datang. Sejarah
Perkembangan pendidikan Islam lahir seiring
dengan
masuknya
Islam
ke
Indonesia,
meskipun pada mulanya hanya dalam bentuk
yang
sangat
sederhana. Perkembangan
pendidikan Islam secara historis tidak pernah
sunyi dari persoalan dan rintangan yang
dihadapinya. Pada
masa
sebelum
kemerdekaan berhadapan dengan tekanan dan
intimidasi pemerintah kolonial Belanda dan
Jepang. Pada masa kemerdekaan berhadapan 42 Vol. 3. No. 1 Januari 2024
e-ISSN: 2809-7998 p-ISSN: 2809-8005 Vol. 3. No. 1 Januari 2024
e-ISSN: 2809-7998 p-ISSN: 2809-8005 JUPEIS: Jurnal Pendidikan dan Ilmu Sosial
https://jurnal.jomparnd.com/index.php/jp dalam perkembangan kelembagaan Perguruan
Tinggi Islam di Indonesia. Seperti tercatat
dalam sejarah, nama Perguruan Tinggi Islam
(PTI) di Indonesia terus berubah sebagai
upaya
merespon
perkembangan
ilmu
pengetahuan dan teknologi serta menjawab
kebutuhan masyarakat dan sekaligus juga
sebagai obyek tarik menarik antara berbagai
kekuatan atau kelompok dalam masyarakat. gagal) karena perguruan tinggi Islam yang
didirikan tidak bertahan lama, kecuali sekolah
tinggi yang dibentuk oleh masyumi. Setelah
Indonesia merdeka, lahirlah Perguruan Tinggi
Agama
Islam
Negeri
(PTAIN)
yang
kemudian
berkembang
menjadi
Institut
Agama Islam Negeri (IAIN). Saat ini PTAIN
terdiri atas 3 jenis yakni: Institut Agama
Islam Negeri (IAIN), Sekolah Tinggi Agama
Islam Negeri (STAIN), dan Universitas Islam
Negeri (UIN) (Syahri, 2022). Perjalanan panjang perguruan tinggi
Islam di Indonesia hingga sekarang, dapat
dikategorikan
tiga
periodesasi. Pertama,
periode awal sejak kedatangan Islam yang
ditandai dengan pendidikan Islam yang
terkonsentrasi di pesantren, dayah, surau atau
masjid. Kedua, periode ketika pendidikan
Islam telah dimasuki oleh ide-ide pembaruan
pemikiran Islam pada awal abad ke-20. Periode ini ditandai dengan lahirnya madrasah
yang telah memasukkan pelajaran “umum”
kedalam
program
kurikulumnya. Ketiga,
periode lahirnya perguruan tinggi Islam negeri
dan pendidikan Islam telah terintegrasi ke
dalam system pendidikan nasional. Hal
tersebut menunjukkan bahwa pendidikan
Islam semakin memperhatikan dinamikanya
sejak Indonesia merdeka. PENDAHULUAN Lahirnya perguruan
tinggi Islam inilah yang kemudian melahirkan
sejumlah terobosan yang luar biasa, karena
lembaga
pendidikan
tinggi
Islam
ini
melahirkan sejumlah ilmuan Islam modern di
kemudian hari (Nurkholis, 2019). Sampai saat sekarang ini konsentrasi
kelimuan di IAIN adalah pengembangan
ilmu-ilmu agama. Menyikapi globalisasi
dengan tuntutan yang semakin berkembang
serta cita-cita untuk mengitegrasikan ilmu
yang tergolong perennial knowledge dan ilmu
yang tergolong ecquired knowledge, maka
kemumuncul
ide
atau
gagasan
untuk
mengembangkan
lagi
IAIN
menjadi
universitas merupakan sebuah keniscayaan
untuk menjaga dan merawat ekssitentesinya
dalam
percaturan
global
maupun
internasional. Ide ini akhirnya melahirkan
Universitas Islam Negeri (UIN). Sejarah
perkembangan dan pembaharuan perguruan
Tinggi Islam menjadi sebuah kajian yang
menarik untuk ditelusuri oleh karena itu
dalam makalah ini kita akan menggali
tentaang bagaimana sejarah perkembangan
perguruan tinngi islam yang ada di Indonesia. Berdasarkan uraian pada latar belakang
diatas, maka masalah pokok yang menjadi
pembahasan
dalam
makalah
ini
dapat
dirumuskan
sebagai
berikut:
Bagaimana
sejarah
Perguruan
Tingggi
Islam
di
Indonesia?,
Bagaimana
perkembangan
Perguruan Tinggi
Islam di
Indonesia?, Sebenarnya ide pendirian perguruan
tinggi Islam sudah muncul sebelum Indonesia
merdeka. Namun diantara sekian banyak ide
untuk untuk mendirikan perguruan tinggi
Islam pada masa penjajahan bisa dikatakan
belum berhasil (jika tidak mau mengatakan 43 Vol. 3. No. 1 Januari 2024
e-ISSN: 2809-7998 p-ISSN: 2809-8005 Vol. 3. No. 1 Januari 2024
e-ISSN: 2809-7998 p-ISSN: 2809-8005 JUPEIS: Jurnal Pendidikan dan Ilmu Sosial
https://jurnal.jomparnd.com/index.php/jp mengalami kemajuan. Segala upaya dan
bentuk diskrimisasi terus dilakukan oleh
pemerintah kolonial. Misalnya, pada masa
pemerintahan Belanda, umat Islam (rakyat
Indonesia) mendapat diskriminasi, baik dalam
ranah agama, ras, politik, pendidikan dan
maupun ekonomi. Kondisi tersebut telah
memperlambat kemajuan umat Islam di
Indonesia. Bagaimana Tantangan dan Peluang Perguruan
Tinggi Islam?. Bagaimana Tantangan dan Peluang Perguruan
Tinggi Islam?. METODE PENELITIAN Penelitian ini menggunakan penelitian
studi pustaka, dengan melakukan kajian
pustaka yaitu analisis terhadap berbagai
artikel, jurnal dan buku yang terkait dengan
tema. Selanjutnya,
pengumpulan
data
penelitian dilakukan dengan menggunakan
tekhnik dokumentasi yaitu dengan upaya
mencari
data-data
dengan
menganalisis
dokumen-dokumen
terkait
perihal
yang
peneliti tulis. Hasrat umat Islam untuk mendirikan
pendidikan tinggi sudah dirintis sejak zaman
kolonial Belanda. M. Natsir menulis dalam
Capita Selekta, bahwa keinginan untuk
mendirikan pendidikan tinggi Islam itu telah
muncul di hati umat Islam. M. Natsir
menyebutkan, bahwa Dr. Satiman telah
menulis artikel dalam Pedoman Masyarakat
nomor 15 membentangkan cita-cita beliau
yang mulia akan mendirikan satu sekolah
tinggi Islam itu akan berpusat di tiga tempat,
yakni Jakarta, Solo dan Surabaya. Di Jakarta
akan diadakan sekolah tinggi sebagai bagian
atas Sekolah Menengah Muhammadiyah
(AMS) yang bersifat Westerch (kebaratan). Di Solo akan diadakan sekolah tinggi untuk
mendidik mubalighin. Di Surabaya akan
diadakan sekolah tinggi yang akan menerima
orang-orang pesantren. Kendatipun yang
diungkapkan ini masih dalam bentuk ide,
akan tetapi semangat untuk mendirikan
perguruan tinggi Islam itu telah muncul pada
tahun 1930-an (Amiruddin, 2006). Secara
historis, upaya mendirikan perguruan tinggi
Islam sebagai wahana pengembangan studi
keislaman dalam konteks keindonesiaan yang
pendiriannya dilatarbelakangi oleh banyak hal
atas berbagai macam pertimbangan. Pertama
adalah faktor intern yaitu di Indonesia telah Sejarah Perguruan Tinggi di Indonesia Perguruan tinggi islam adalah salah satu
stratum pendidikan islam yang berada pada
level tertinggi. Eksistensi pendidikan islam
dalam
kancah
pendidikan
nasional
di
indonesia memiliki urgensi yang sangat besar,
utamanya
sebagai
pilar
bagi
bangunan
pendidikan
islam
secara
keseluruhan. Perguruan Tinggi Islam memiliki misi sebagai
center of excellent untuk menghasilkan para
sarjana yang sujana serta manfaat pendidikan
bagi stalkholder. Untuk mewujudkan tujuan
tersebut perguruan tinggi islam merumuskan
visi, misi, strategi dan program kerja yang
terencana, terfokus, dan berkesinambungan
yang dipergunakan civitas akademika sebagai
pedoman untuk mencapai tujuan. Indonesia
sebelum kemerdekaan berada dalam situasi
sulit, di mana Indonesia di bawah pengaruh
kolonial, baik Belanda maupun Jepang. Kondisi tersebut membuat umat Islam tidak
mengalami
kemajuan,
sebab
pemerintah
kolonial tidak ingin umat Islam di Indonesia 44 Vol. 3. No. 1 Januari 2024
e-ISSN: 2809-7998 p-ISSN: 2809-8005 JUPEIS: Jurnal Pendidikan dan Ilmu Sosial
https://jurnal.jomparnd.com/index.php/jp Vol. 3. No. 1 Januari 2024
e-ISSN: 2809-7998 p-ISSN: 2809-8005 Islamiyah al-Hukumiyah yang berada di dua
kota, yaitu Yogyakarta dan Jakarta. Islamiyah al-Hukumiyah yang berada di dua
kota, yaitu Yogyakarta dan Jakarta. berdiri perguruan tinggi umum, antara lain
Sekolah Tinggi Teknik di Bandung 1920,
Sekolah Tinggi Hukum di Jakarta pada tahun
1920, dan Sekolah Tinggi Kedokteran di
Jakarta pada tahun 1927. Kedua adalah faktor
ekstern
yaitu
respon
atas
kebutuhan
masyarakat untuk merealisasikan kehidupan
beragama di tanah air dan masuknya pengaruh
ide-ide
pembaruan
pemikiran
Islam
ke
Indonesia
(Afrizal,
2022). Kehadiran
perguruan tinggi Islam di tengah masyarakat
pada dasarnya merupakan perwujudan dan
suatu cita-cita yang telah lama terkandung di
hati sanubari umat Islam Indonesia. JUPEIS: Jurnal Pendidikan dan Ilmu Sosial
https://jurnal.jomparnd.com/index.php/jp dimana awal sejarahnya, STI Menjadi UII,
Fakultas Agama UII Menjadi PTAIN, PTAIN
dan ADIA Menjadi IAIN, kemudian IAIN
Menjadi UIN (Amiruddin, 2017). dimana awal sejarahnya, STI Menjadi UII,
Fakultas Agama UII Menjadi PTAIN, PTAIN
dan ADIA Menjadi IAIN, kemudian IAIN
Menjadi UIN (Amiruddin, 2017). Penjelasan di atas juga ditunjukkan
dalam ayat al-Qu’ran yang berkaitan dengan
perubahan atau transformasi baik secara
individu mapun lembaga atau organisasi. Dalam al-Qur’an dijelaskan, bahwa semangat
perubahan dan revolusi termasuk transformasi
sehingga
dapat
kita
temukan
pijakan
epistemologisnya dari beberapa ayat tentang
para nabi dan rasul. Dalam al-Qur’an ayat
218, Surah al-Baqarah disebutkan pentingnya
berhijrah (transformasi) (Tim Kemanag RI,
2016). dimana awal sejarahnya, STI Menjadi UII,
Fakultas Agama UII Menjadi PTAIN, PTAIN
dan ADIA Menjadi IAIN, kemudian IAIN
Menjadi UIN (Amiruddin, 2017). Pembaharuan Perguruan Tinggi Islam
sekarang ini, yang berupa UIN tersebut bukan
merupakan
perkembangan
sesaat
dan
mendadak,
tetapi
merupakan
pergulatan
panjang umat Islam bukan hanya di Indonesia
tapi juga di belahan dunia lain, bahkan
menjadi kecenderungan umum dunia kali ini. Seperti dimaklumi, abad ke-21 dikenal
sebagai abad PostModernime yang dicirikan
dengan perlunya penyatuan kembali antara
ilmu dan agama. Dikhotomi ilmu dan agama
atau dikhotomi ilmu umum dan ilmu agama
dipandang
tidak
lagi
relevan. Adapun
perkembangan peguruan tinggi islam di
indonesia terdapat beberapa fase ataupun
tingkatan perguruan islam sebelum menjadi
UIN antara lain yaitu : ِلُإ وُرِجالإولدلهََ إآل وُرِجالإنلُالذَآ إآلالُنلمَآ إنِِإَّلاَِّنِإ
رُ ِإَِآرلإوِفلإ لإلماََللإللُمْلَلإرُ ِإآل رُ َإ لكََلهللُِ َِّهإ
“Sesungguhnya orang-orang yang beriman,
orang-orang yang berhijrah dan berhijrah
dan berjihad di jalan Allah, mereka itu
mengharapkan rahmat Allah, dan Allah maha
pengampun lagi maha penyayang.” ِلُإ وُرِجالإولدلهََ إآل وُرِجالإنلُالذَآ إآلالُنلمَآ إنِِإَّلاَِّنِإ
رُ ِإَِآرلإوِفلإ لإلماََللإللُمْلَلإرُ ِإآل رُ َإ لكََلهللُِ َِّهإ
“Sesungguhnya orang-orang yang beriman,
orang-orang yang berhijrah dan berhijrah
dan berjihad di jalan Allah, mereka itu
mengharapkan rahmat Allah, dan Allah maha
pengampun lagi maha penyayang.” “Sesungguhnya orang-orang yang beriman,
orang-orang yang berhijrah dan berhijrah
dan berjihad di jalan Allah, mereka itu
mengharapkan rahmat Allah, dan Allah maha
pengampun lagi maha penyayang.” Berdasarkan pemaparan di atas, penulis
berasumsi bahwa transformasi dalam lembaga
pendidikan Islam sangat diperlukan pada
zaman ini untuk menumbuhkan karakter
Islami, mencetak lulusan lembaga pendidikan
Islam yang mempunyai multi kecakapan, yaitu
kecakapan umum dan kecapakan Islam. Sehingga lembaga pendidikan Islam nantinya
dapat dijadikan tolak ukur bagi kemajuan
suatu bangsa. Perkembangan Perguruan Tinggi Islam di
Indonesia Saat ini sudah ada banyak usaha yang
dilakukan dalam rangka membenahi lembaga
pendidikan tinggi Islam agar mampu eksis
dan survive di tengah tantangan zaman. Salah
satu usaha yang dimaksud adalah adanya tren
trasnformasi lembaga pendidikan tinggi Islam
dalam hal ini STAIN dan IAIN menjadi
Universitas Islam Negeri atau UIN. Dengan
demikian, lembaga pendidikan Islam yang
mulanya hanya dapat membuka fakultas
agama, maka setelah menjadi UIN lembaga
tersebutdapat membukan fakultas umum. Dengan demikian, akhirnya minat masyarakat
akan kembali tinggi apalagi tarif biaya di UIN
jauh lebih rendah dibandingkan di lembaga
pendidikan
umum
(Wahyono,
2014). Perubahan tersebut merupakan perubahan
kearah yang lebih maju lagi. Hal ini selaras
dengan pengertian dari perubahan itu sendiri. Adapun
pengertian
perubahan
menurut
Kamus Besar Bahasa Indonesia adalah hal
(keadaan) yang berubah atau suatu peralihan
(Darminta, 2005). Perubahan merupakan
suatu keharusan, karena perubahan adalah
esensi dan juga efek dari kemajuan. Menjadi
maju berarti harus mau berpindah posisi
semakin kedepan dari posisi semula. Jika
tidak
mau
berubah
sesuai
dengan
perkembangan tentu lembaga tersebut akan
tertinggal. Sehingga lembaga pendidikan
Islam harus mampu melakukan filter atas
perubahan-perubahan,
agar
mampu
memperhitungkan yang lebih baik untuk
pengembangan lembaga pendidikan Islam di
Indonesia nantinya Sejarah pendidikan tinggi islam dimulai
dengan lahirnya Sekolah Tinggi Islam di
tahun 1940 sebagai hasil pertemuan beberapa
guru muslim di padang. Pada tahun 1945
(sebulan setelah proklamasi kemerdekaan RI)
di tingkat nasional berdiri sekolah tinggi
islam, atas inisiatif dari Moh. Hatta sebagai
ketua dan Moh. Natsir sebagai sekretaris dan
dipimpin oleh Prof. Kahar Muzakir. Pada
tahun 1946, sekolah ini pindah ke Yogyakarta
mengikuti perpindahan ibu kota negara. Berdiri pula Akademi Dakwah Islam (ADIA)
di jakarta berdasarkan peraturan no 1 tahun
1957 (Mudzhar, 2020). Melalui Peraturan No. 34 Tahun 1950, fakultas agama Universitas
Islam Indonesia di Yogyakarta diubah menjadi
Perguruan Tinggi Agama Islam Negeri
(PTAIN). Hal
ini
disebabkan
cakupan
pengetahuan agama Islam yang demikian luas
tidak mencukup untuk diajarkan dalam satu
fakultas. Pada tahun 1960, PTAIN dengan
ADIA Jakarta disatukan. Sejak tanggal 9 Mei
1960 namanya menjadi IAIN al-Jami’ah al- 45 Vol. 3. No. 1 Januari 2024
e-ISSN: 2809-7998 p-ISSN: 2809-8005 JUPEIS: Jurnal Pendidikan dan Ilmu Sosial
https://jurnal.jomparnd.com/index.php/jp JUPEIS: Jurnal Pendidikan dan Ilmu Sosial
https://jurnal.jomparnd.com/index.php/jp Dapat dipahami bahwa Sejarah
perkembangan Pendidikan Tinggi Agama
Islam adalah sejarah Ekspansi perguruan
Tinggi Agama Islam, dalam artian upaya
pengembangan perguruan tinggi Agama Islam
dalam mewjudkan perguruan tinggi agama
Islam lebih berkualitas dan bermutu serta
mampu bersaing dengan PTU yang ada di
Indonesia maupun dunia Islam lainnya pada
masa mendatang. Ekspansi perguruan tinggi
agama Islam terjadi pada aspek kelembagaan, 1. STI MIAI ini pun pada tahun 1943
diubah namanya menjadi Masyumi (Majelis
Syuro Muslimin Indonesia, berarti Majelis
Permusyawaratan
Umat
Islam
Indonesia).Masyumi
yang
merupakan
penjelmaan baru dari MIAI itu merupakan
federasi
dari empat organisasi-organisasi
Islam yang oleh Jepang diizinkan hidup terus
(sebagai organisasi sosial dan dakwah),yakni:
Nahdlatul Ulama (NU), didirikan di Surabaya,
Muhammadiyah (didirikan di Yogyakarta),
Persatuan
Oemat
Islam
Indonesia
(POI),didirikan di Majalengka, Persatuan
Umat islam di Indonesia (PUII) didirikan di
Sukabumi. Dengan
bantuan
pemerintah
Jepang, STI ini akhirnya dapat dibuka secara
resmi pada tanggal 27 Rajab 1364 H, di saat
peringatan Isra’ Mi’raj Nabi Muhammad saw,
bertepatan dengan 8 Juli 1945 di Jakarta. Upacara peresmiannya diselenggarakan di
gedung Kantor Imigrasi Pusat Gondangdia
Jakarta. Pada mulanya, STI didirikan untuk
melatih ulama ulama yang berpendidikan baik secara meluas dan mendalam, dan telah
memperoleh standar pengetahuan umum yang
memadai seperti dituntut oleh masyarakat
sekarang ini. Hal ini sesuai dengan tujuan
didirikannya
STI
yang
pada
dasarnya
merupakan kebutuhan umat Islam Indonesia
akan
adanya
Perguruan
Tinggi
yang
memberikan pelajaran dan pendidikan tinggi
tentang ilmu-ilmu agama Islam dan ilmu agar
menjadi peyiaran agama dan memberikan
pengaruh Islam di Indonesia. Demikian STI
terus berjalan di Yogyakarta pada tahun 1946
dan tahun 1947 dengan agak kurang lancar
karena waktu itu bangsa Indonesia sedang
berjuang
untuk
mempertahankan
kemerdekaan,
apalagi
hampir
warga
akademika STI ikut memanggul senjata
melawan penjajah yang hendak kembali. Sekalipun
ada
gangguan
sehingga
perkuliahan tidak lancar, namun pada akhir
tahun
1946/1947
masih
sempat
menyelenggarakan dua kali ujian untuk
tingkat pendahuluan. Pihak pemerintah dalam
keadaan seperti itu masih memberikan
perhatian cukup baik bagi STI. Terbukti pada
tahun
1947
Kementrian
Agama
RI
memberikan bantuan uang sebesar Rp.60.000
(enam puluh ribu rupiah). Rencana pelajaran
yang ada waktu itu sama seperti ketika di
Jakarta. 2. UII 1. STI Sejarah Pendidikan Islam di Indonesia
yang tertua, dapat disebutkan bahwa Sekolah
Tinggi Islam (STI) merupakan cikal bakal
pendidikan Islam di Indonesia pada masa
masa
selanjutnya. Dalam
usaha
pengembangan lembaga pendidikan Islam
sudah berawal sebelum Indonesia merdeka. Dalam perang dunia II, ternyata Jepang
berhasil merebut Indonesia dari tangan
Belanda
pada
tahun
1942. Sementara
Indonesia terseret dalam kancah peperangan
tersebut. Penguasa Jepang (Dai Nippon)
dengan segera mengeluarkan larangan bagi
kegiatan-kegiatan
pergerakan
nasional
Indonesia. Bahkan
banyak
pemimpin
Indonesia yang direkrut untuk membantu 46 Vol. 3. No. 1 Januari 2024
e-ISSN: 2809-7998 p-ISSN: 2809-8005 JUPEIS: Jurnal Pendidikan dan Ilmu Sosial
https://jurnal.jomparnd.com/index.php/jp JUPEIS: Jurnal Pendidikan dan Ilmu Sosial
https://jurnal.jomparnd.com/index.php/jp administrasi
pemerintahan
Jepang
(pemerintahan Jepang di Indonesia), antara
lain adalah KHA Kahar Muzakkir dan KH. Imam Zarkasyi. Gerakan perjuangan umat
Islam Indonesia tak terlepas dari langkah
Jepang untuk menutup gerakan kebangsaan. Partai-partai politik Islam dibubarkan kecuali
MIAI. MIAI
merupakan
tempat
bermusyawarah untuk kepentingan agama
Islam dan dimaksudkan juga untuk tempat
berkenalan, saling bertemu dan mejalin
persahabatan agar dengan demikian dapat
terwujud persatuan lahir dan batin di antara
para alim ulama dan pemimpin Islam seluruh
Indonesia. MIAI ini pun pada tahun 1943
diubah namanya menjadi Masyumi (Majelis
Syuro Muslimin Indonesia, berarti Majelis
Permusyawaratan
Umat
Islam
Indonesia).Masyumi
yang
merupakan
penjelmaan baru dari MIAI itu merupakan
federasi
dari empat organisasi-organisasi
Islam yang oleh Jepang diizinkan hidup terus
(sebagai organisasi sosial dan dakwah),yakni:
Nahdlatul Ulama (NU), didirikan di Surabaya,
Muhammadiyah (didirikan di Yogyakarta),
Persatuan
Oemat
Islam
Indonesia
(POI),didirikan di Majalengka, Persatuan
Umat islam di Indonesia (PUII) didirikan di
Sukabumi. Dengan
bantuan
pemerintah
Jepang, STI ini akhirnya dapat dibuka secara
resmi pada tanggal 27 Rajab 1364 H, di saat
peringatan Isra’ Mi’raj Nabi Muhammad saw,
bertepatan dengan 8 Juli 1945 di Jakarta. Upacara peresmiannya diselenggarakan di
gedung Kantor Imigrasi Pusat Gondangdia
Jakarta. Pada mulanya, STI didirikan untuk
melatih ulama-ulama yang berpendidikan baik
yaitu orang yang telah mempelajari Islam administrasi
pemerintahan
Jepang
(pemerintahan Jepang di Indonesia), antara
lain adalah KHA Kahar Muzakkir dan KH. Imam Zarkasyi. Gerakan perjuangan umat
Islam Indonesia tak terlepas dari langkah
Jepang untuk menutup gerakan kebangsaan. Partai-partai politik Islam dibubarkan kecuali
MIAI. MIAI
merupakan
tempat
bermusyawarah untuk kepentingan agama
Islam dan dimaksudkan juga untuk tempat
berkenalan, saling bertemu dan mejalin
persahabatan agar dengan demikian dapat
terwujud persatuan lahir dan batin di antara
para alim ulama dan pemimpin Islam seluruh
Indonesia. 2. UII Pada bulan November 1947, STI
dikembangkan menjadi Universitas Islam
Indonesia (UII) mengasuh empat fakultas
yaitu
agama,
hukum,
pendidikan
dan
ekonomi yang dibuka secara resmi pada
tanggal 10 Maret 1948 bertepatan dengan 27
Rajab 1367 H. Perkembangan berikutnya 47 Vol. 3. No. 1 Januari 2024
e-ISSN: 2809-7998 p-ISSN: 2809-8005 Vol. 3. No. 1 Januari 2024
e-ISSN: 2809-7998 p-ISSN: 2809-8005 JUPEIS: Jurnal Pendidikan dan Ilmu Sosial
https://jurnal.jomparnd.com/index.php/jp adalah Fakultas Agama Universitas Islam
Indonesia dinegerikan menjadi
Peguruan
Tinggi
Islam
Negeri
(PTAIN). Panitia
perbaikan STI menjadi UII adalah KHR. Fatchurrahman Kafrawi, KH.Faried Ma’roef,
KH. Malikus Suparto, Sulaiman, Mr. R. Sunandjo,
Drs. A.Sigit,
KHA. Kahar
Muzakkir, Ustaz Sulaiman, Ustaz Husein
Jahja, dan Kartosudarmo. Untuk memulai
secara resmi mengubah STI menjadi UII pada
bulan Maret
1948
diadakanlah upacara
pembukaan pendahuluan yakni pembukaan
kelas pendahuluan di Yogyakarta. Setelah
tingkat pendahuluan resmi dibuka maka
berarti segala sesuatu yang diperlukan telah
dipersiapkan
untuk
segera
meresmikan
perubahan STI menjadi UII. Acara penting
yang disajikan pada upacara pembukaan UII
adalah
pembacaan
Keputusan
Dewan
Pengurus tentang berdirinya UII, juga pidato
oleh KHA. Kahar Muzakir, dan Dr. Mr. Kusumah Atmaja. Dengan demikian UII
berpusat pada Yogyakarta sekarang adalah UII
yang merupakan kelanjutan dan pengganti dari
STI yang dibuka pertama kali di Jakarta
tanggal 27 Rajab 1361 H atau 8 Juli 1945. Dengan perubahan STI menjadi UII ini, tujuan
yang semula dimaksudkan untuk memberikan
pendidikan yang baik bagi para calon ulama,
akhirnya bergeser titik beratnya pada fakultas-
fakultas non-agama yang bersifat sekuler
meskipun tetap berlandaskan agama dan
semangat keagamaan. Perubahan orientasi ini
bukan
tanpa
alasan. Kehadiran
sebuah
perguruan
tinggi
swasta
non-Islam
di
yogyakarta yang didirikan pada bulan Maret
1948 dan berkembang menjadi Universitas
Gajah mada (UGM) sejak tanggal 19 Desember 1949 memperoleh corak nasional. bukan tidak mungkin UGM ini menjadi
saingan yang berat bagi UII. Perubahan
orientasi dalam UII ini berarti UII harus
berkompetisi dengan universitas lain yang
secara ideologis berbeda. UII bercorak
keislaman,
sedangkan
UGM
bercorak
nasionalis. Pembukaan
fakultas-fakultas
sekuler
merupakan
suatu
keharusan
(kebutuhan) zaman spesialisasi ilmu menjadi
tak terhindarkan. Ilmu-ilmu dalam Islam jelas
tidak hanya menyangkut ilmu-ilmu agama
tetapi termasuk ilmu-ilmu umum. UII jelas
dengan
penuh
kesadaran
melakukan
perubahan-perubahan ini. Perubahan itu tidak
keluar dari tujuan pengembangan ilmu yang
hendak dicapai STI, yaitu memadukan ilmu
agama
dan
ilmu
umum
dalam
suatu
universitas yang mencakup ilmu-ilmu Islam. 2. UII Sampai tahun 1950, UII merupakan satu-
satunya perguruan tinggi yang dimiliki oleh
umat Islam Indonesia dan cikal bakal lahirnya
perguruan tinggi Islam di Indonesia pada
tahun-tahun
berikutnya. Dampak
ikutan
kehadiran UII ini secara kelembagaan mulai
menggema pada tahun 1950-1960 yang
ditandai dengan berdirinya PTIS (perguruan
Tinggi Swasta) di berbagai kota. 3. PTAIN/ADIA PTAIN ini diresmikan pada pada
tanggal 26 September 1951 dihadiri oleh
Menteri Agama RI A. Wahid Hasyim. Penyelenggaraan PTAIN selanjutnya diatur
dengan Peraturan Bersama Menteri Agama
dan Menteri PP&K tertanggal 21 Oktober
1951 yang ditandatangani oleh A. Wahid
Hasyim dan Mr. Wongsonegoro. Sebagai
upaya peningkatan kualitas sumber daya 48 Vol. 3. No. 1 Januari 2024
e-ISSN: 2809-7998 p-ISSN: 2809-8005 JUPEIS: Jurnal Pendidikan dan Ilmu Sosial
https://jurnal.jomparnd.com/index.php/jp disebutkan bahwa: “Terwujudnya warga
masyarakat
yang
beriman,
bertaqwa,
berahklak mulia, dan memiliki rasa tanggung
jawab kemasyarakatan, serta mempunyai
kemampuan akademik dan/atau profesional
yang mampu menerapkan, mengembangkan
dan/atau menciptakan ilmu agama Islam dan
ilmu
lain
yang
terkait,
dan
menyebarluaskannya
dengan
menghargai
nilai kemanusiaan untuk meningkatkan harkat
dan kehidupan masyarakat, dan memperkaya
kebudayaan nasional dan umat manusia”. Melalui Visi Akademik ini jelas bahwa IAIN
merupakan PTI yang konsentrasi pada ilmu
agama tetapi sudah mulai memberi perhatian
pada ilmu umum yang dalam statuta tersebut
disebut dengan kalimat “ilmu lain yang
terkait”. Kecenderungan ini bukan monopoli
IAIN Sunan Kalijaga tetapi juga menjadi visi
akademik sejumlah IAIN lainnya. 5. UIN manusia, kemudian juga didirikan Akademi
Dinas Ilmu Agama (ADIA) di Jakarta pada
tanggal 1 Juni 1957, dengan visi: “Guna
mendidik dan mempersiapkan pegawai negeri
yang akan mencapai ijazah pendidikan semi-
akademi dan akademi untuk dijadikan ahli-
didik agama pada sekolah-sekolah lanjutan,
baik umum, maupun kejuruan dan agama”. 4. IAIN Perkembangan PTI tidak berhenti di
situ, setelah memperhatikan situasi sosial
kemasyarakatan waktu itu, PTAIN dan ADIA
akhirnya dilebur menjadi satu lembaga PTI
dengan nama Institut Agama Islam Negeri
(IAIN) yang diresmikan pada tanggal 24
Agustus 1960 di Yogyakarta. Ada beberapa
hal yang menarik untuk dicermati dari
perjalanan sejarah Visi Akademik PTI di atas. Pada proses awal, terutama ketika STI, maka
Visi Akademik mencakup dua hal: ilmu dan
agama (science and religion) atau ilmu agama
dan ilmu umum (religious and seculer
sciences), walau dalam praktiknya, ilmu
agama lebih dominan karena berorientasi ke
Fakultas Ushuluddin Universitas Al-Azhar. Namun keadaan ini mulai beranjak berubah
memasuki akhir tahun 1990-an atau awal
tahun 2000-an. Pada masa ini diperkenalkan
istilah “IAIN with wider-mandate”, atau
“IAIN dengan mandat yang diperluas”. Sejak
itu, IAIN yang semula berkonsentrasi pada
ilmu agama mulai merambah bidang-bidang
studi yang selama ini dikenal dengan ilmu
umum. Lahirnya Jurusan Tadris di Fakultas
Tarbiyah menjadi contoh amat jelas dalam hal
ini. Visi dan orientasi akademik yang baru ini
bisa dilihat misalnya, pada Status IAIN Sunan
Kalijaga
tahun
2001. Dalam
pasal
3 5. UIN Perubahan relatif fundamental mulai
muncul dengan lahirnya Universitas Islam
Negeri (UIN). UIN ini merupakan hasil
transformasi IAIN yang secara kelembagaan
berupa Institut menjadi Universitas. Sebagai
sebuah lembaga Universitas, UIN tidak lagi
seperti IAIN. Universitas tersebut tidak hanya
menekuni ilmu agama tetapi juga ilmu umum. Prodi-Prodi ilmu agama berada di bawah
naungan Departemen Agama, sementara
prodi-prodi umum berada di bawah naungan
Departemen
Pendidikan
Nasional. Perkembangan ini menarik untuk dicermati,
karena memberi gambaran semakin jelas
pentingnya “pendidikan satu atap” yang
sudah
didengungkan
sejak
lama
oleh
sejumlah
pakar
pendidikan
di 49 Vol. 3. No. 1 Januari 2024
e-ISSN: 2809-7998 p-ISSN: 2809-8005 JUPEIS: Jurnal Pendidikan dan Ilmu Sosial
https://jurnal.jomparnd.com/index.php/jp Indonesia.Apabila Pemerintah sudah berhasil
menjadikan “satu atap” Peradilan Agama
dengan Peradilan Umum (juga peradilan
lainnya), maka tinggal diteruskan ke arah
“pendidikan satu atap”. Sehingga dapat
terhindar
dari
kebingungan
birokrasi
sebagaimana
dialami
selama
ini. Perkembangan terakhir dari PTI yang berupa
UIN tersebut bukan merupakan perkembangan
sesaat dan tiba-tiba, tetapi merupakan hasil
dari pergulatan panjang umat Islam di
Indonesia. Di tinjau dari segi sejarah
sosialnya gagasan transformasi IAIN menjadi
UIN pertama kali dikemukakan oleh Rektor
IAIN Jakarta periode 1973-1984, yakni
Prof. Harun Nasution. alasan Harun
Nasution ingin mengembangkan IAIN
menjadi UIN dikemukakan dalam sebuah
wawancara dengan Republika, pada Kamis,
28 Desember 1995. Saat itu ia sudah menjadi
Direktur Program Pascasarjana. “Kita merasa
yang diperlukan umat di zaman sekarang
ini bukan hanya sarjana yang mengetahui
ilmu agama saja, tapi juga ilmu umum. Harus diakui tidak banyak orang yang
bisa menguasai keduanya secara
mumpuni. Hanya orang-orang jenius saja
yang bisa melakukannya”. Berangkat dari
kebutuhan itu, Harun berpendapat, IAIN
perlu ditransformasikan menjadi universitas,
sehingga dapat membuka jurusan-jurusan
umum. Harapannya tentu saja mampu
mencetak sarjana yang memiliki
kompetensi agama namun tidak asing
dengan pengetahuan umum. Hal itu bagi
Harun bukan mustahi. Sejarah mencatat
seorang Ibnu Rusyd dan Ibnu Sina yang
selain ahli filsafat, syariah, juga seorang dokter yang masyhur. “Kalau pada masa
lampau mereka bisa menghasilkan tokoh
seperti itu, kenapa kita tidak mampu
menghasilkannya. Inilah dasar pendirian
kita sehingga ada keinginan untuk
mengubah IAIN menjadi UIN,” tegas Harun. Kepastian pendirian UIN kian menunjukkan
kejelasan setelah diadakannya Sarasehan di
IAIN Jakarta pada 22 Oktober 1994. Sarasehan itu dihadiri oleh beberapa
narasumber ternama, baik dari kalangan
internal IAIN Jakarta dan kalangan luar. Dari
Kalangan internal yakni Quraish Shihab(saat
itu menjabat Rektor), Harun Nasution,
Zakiah Drajat. 5. UIN Di dalam SKB tersebut dinyatakan
bahwa secara kelembagaan, keuangan,
kepegawaian dan pembukaan program
studi agama IAIN berada di bawah
tanggung jawab Departemen Agama,
sedangkan dari segi pengembangan
akademik khususnya bidang studi
umum, berada di bawah tanggung
jawab Departemen Pendidikan Nasional. Dengan SKB tersebut, maka tahapan
perubahan IAIN menjadi UIN
selanjutnya dilakukan melalui
keputusan
Presiden
Republik
Indonesia. Perubahan itu terjadi pada masa Menteri
Agama yang dijabat oleh Prof. Said Agil
Husein al-Munawar, dan Menteri
Pendidikan Nasional ketika itu adalah Abdul
Malik Fadjar, dan disahkan oleh Presiden
Megawati Soekarno Putri. Khusus untuk
Abdul Malik Fadjar menjadi Mendiknas
yang sangat pro aktif dalam realisasi
perubahan IAIN menjadi UIN, ia
sangat mendukunga adanya UIN sebagai
basis integrasi keilmuan. Departemen Pendidikan Nasional,
ataukah tetap berada di bawah pengelolaan
Departemen Agama. Upaya tarik menarik ini
akhirnya
dapat
diselesaikan
melalui
kompromi dalam bentuk dikeluarkannya
Surat Keputusan Bersama antara Menteri
Agama RI dengan Menteri Pendidikan
Nasional. Di dalam SKB tersebut dinyatakan
bahwa secara kelembagaan, keuangan,
kepegawaian dan pembukaan program
studi agama IAIN berada di bawah
tanggung jawab Departemen Agama,
sedangkan dari segi pengembangan
akademik khususnya bidang studi
umum, berada di bawah tanggung
jawab Departemen Pendidikan Nasional. Dengan SKB tersebut, maka tahapan
perubahan IAIN menjadi UIN
selanjutnya dilakukan melalui
keputusan
Presiden
Republik
Indonesia. Perubahan itu terjadi pada masa Menteri
Agama yang dijabat oleh Prof. Said Agil
Husein al-Munawar, dan Menteri
Pendidikan Nasional ketika itu adalah Abdul
Malik Fadjar, dan disahkan oleh Presiden
Megawati Soekarno Putri. Khusus untuk
Abdul Malik Fadjar menjadi Mendiknas
yang sangat pro aktif dalam realisasi
perubahan IAIN menjadi UIN, ia
sangat mendukunga adanya UIN sebagai
basis integrasi keilmuan. Integrasi Keilmuan Perguruan Tinggi 5. UIN Dari Luar yakni mewakili
Departemen Agama ialah Atho
Mudzhar, Perguruan tinggi Umum seperri
Asri Rasyad dari Yayasan Rumah
SakitIslam Indonesia.Hanna Djumhana
Bustaman dan Laode M Kamaluddin dari
Universitas Indonesia(UI), dan Ahmad
Baiquni dari badan Pengkajian dan
Penerapan Teknologi (BPPT). Sarasehan itu
kemudian membuahkan hasil berupa
Departemen
Agama
menetapkan
IAIN
Jakarta
sebagai
pilot
project
dan
mengharapkan agar upaya konversi
IAIN menjadi UIN di awali dengan
studi
kelayakan
beberapa
hal
yakni
pengembangan
kelembagaan,
pengembangan
ketenagaan,
pengembangankurikulum, pengembangan
perpustakaan,
pengembangan sarana
dan pengembangankurikulum, pengembangan
perpustakaan, pengembangan sarana dan
prasarana, dan penyusunan RIP UIN.Proses
ini semua terjadi pada masakepemimpinan
Rektor IAIN M. Quraish Shihab (1992-
1998).Cita-cita tersebut akhirnya terealisasi
pada 20 Mei 2002 pada masa periode 50 Vol. 3. No. 1 Januari 2024
e-ISSN: 2809-7998 p-ISSN: 2809-8005 Vol. 3. No. 1 Januari 2024
e-ISSN: 2809-7998 p-ISSN: 2809-8005 JUPEIS: Jurnal Pendidikan dan Ilmu Sosial
https://jurnal.jomparnd.com/index.php/jp kepemimpinan Rektor Azyumardi Azra(1998-
2006). Setelah berganti nama, infrastruktur
segera dibangun dan arah pengembangan
diperjelas yakni menjadi universitas riset dan
universitas kelas dunia. Jika bisa lebih
diringkas lagi maka jalan perubahan IAIN
menjadi UIN itu ada tiga tapap, yakni: (1)
tahap Perintisan dan penjajagan yang
dilakukan pada masa Harun Nasution sebagai
rektor, (2) Tahap Pelanjutan dan pematangan
Konsep, tahap ini terjadi pada masa Quraish
Shihab sebagai rektor, pada tahap ini
disusun proposal perubahan IAIN
menjadi UIN yang dikomandoi oleh
Azyumardi Azra sebagai pembantu Bidang
Akademik. (3) Tahap pematangan
gagasan dan implementasi. Tahap ini
terjadi pada masa Azyumardi Azra sebagai
rektor (Lubis, 2021). Proposal yang telah
disusun kemudian dipresentasikan di
hadapan senat, dan dilakukan juga
upaya untuk terus beraudiensi baik
kepada DPR, Sekretariat negara,
Departemen
pendayagunaan
Aparatur
Negara. Namun pada masa itu Sejarah
membutktikan bahwa sempat terjadi
ganjalan yakni aturan yang termuat dalam
Peraturan Pemerintah Nomor 60 tahun
1999 entang perguruan tinggi yang
berada dibawah departemen non-kependidikan
(Diknas), hanya dapat menyelenggarakan
pendidikan tinggisampai dengan bentuk
institut. Sedangkan untuk perguruan tinggi
setingkat universitas harusberada di bawah
naungan Diknas. Keadaan ini menimbulkan
pro kontra, tentang apakahdenganperubahan
IAIN
j di UIN
b
l l Departemen Pendidikan Nasional,
ataukah tetap berada di bawah pengelolaan
Departemen Agama. Upaya tarik menarik ini
akhirnya
dapat
diselesaikan
melalui
kompromi dalam bentuk dikeluarkannya
Surat Keputusan Bersama antara Menteri
Agama RI dengan Menteri Pendidikan
Nasional. Integrasi Keilmuan Perguruan Tinggi
Islam Integrasi Keilmuan Perguruan Tinggi
Islam Perubahan dari satu bentuk lembaga
ke bentuk lembaga lain tentu
melahirkan karekteristik yang berbeda beda. Di bawah ini akan dikemukakan karekteristik
UIN: 51 Vol. 3. No. 1 Januari 2024
e-ISSN: 2809-7998 p-ISSN: 2809-8005 JUPEIS: Jurnal Pendidikan dan Ilmu Sosial
https://jurnal.jomparnd.com/index.php/jp a. UIN tidak mengajarkan keilmuan agama,
akan tetapi juga menawarkan ilmu umum. Ada hal yang menarik dicermati dari
perjalanan sejarah visi akademik PTI diatas
pada proses awal, terutama ketika STI,
maka Visi Akademik mencakup dua hal :
ilmu dan agama (science and religion) atau
ilmu agama dan ilmu umum (religius and
seculer sciences), walau dalam praktiknya,
ilmu
agama
lebih
dominan
karena
beriorentasi
ke
Fakultas
Ushuluddin
Universitas Al-Azhar, setelah STI menjadi
UII, maka ada sedikit ergeseran, jika
semula penekanannya pada ilmu agama
kemudian berubah penekanannya pada
ilmu umum. Setelah itu, terutama setelah
berdirinya
PTAIN
dan
ADIA
dan
kemudian
menjadi
IAIN,
Maka
visi
Akademik
lebih
terfokus
pada
pengembangan ilmu agama. Bahkan secara
tegas disebutkan bahwa agama tersebut
diarahkhkan seperti yang terdapat pada
Universitas Al-Azhar, Kairo Mesir. Dan
sejak itu pula pula kemudian dikenal bahwa
PTI menghkhusukan kajiannya pada ilmu
agama (dalam arti sempit) dan sekaligus
menjadi bagian atau dibawah naungan
Departemen Agama. Sedangkan Perguruan
Tinggi Umum misalnya UGM berada
dibawah naungan Departemen Pendidikan. Dengan kata lain, dikhotomi ilmu agama
dan ilmu umum terefleksi juga dengan
adanya
Departemen
Agama
dan
Departemen
Pendidikan,
dan
inilah
fenomena umum di hampir seluruh dunia
islam sejak masa penjajahan Barat. Sebagai
sebuah lembaga Universitas, UIN tidak lagi
i
di d l
b k
h menekuni ilmu agama tetapi juga ilmu
umum. Prodi-prodi ilmu agama berada di
bawah naungan Departemen Pendidikann
Naional (Darda, 2015). b. Integrasi
keilmuan
sebagai
respon
terhadap paradigma dikotomis, dikotomi
pendidikan agama dan umum di Indonesia
yang terjadi pada masa sebelum dan awal
kemerdekaan,
berakibat
kurangnya
perhatian
lembaga
pendidikan
islam
terhadap ilmu umum terutama sains dan
teknologi. Akibatnya,
umat
islam
terbelakang dalam penguasaan ilmu-ilmu
tersebut. Hal ini mendorong lahirnya
upaya menciptakan sistem pendidikan
islam yang tidak dikhotomistik, sebab
hanya mengajarkan keilmuan gama saja. Universitas
mengharuskan
untuk
mempelajari keilmuan umum. b. Integrasi
keilmuan
sebagai
respon
terhadap paradigma dikotomis, dikotomi
pendidikan agama dan umum di Indonesia
yang terjadi pada masa sebelum dan awal
kemerdekaan,
berakibat
kurangnya
perhatian
lembaga
pendidikan
islam
terhadap ilmu umum terutama sains dan
teknologi. Akibatnya,
umat
islam
terbelakang dalam penguasaan ilmu-ilmu
tersebut. Hal ini mendorong lahirnya
upaya menciptakan sistem pendidikan
islam yang tidak dikhotomistik, sebab
hanya mengajarkan keilmuan gama saja. KESIMPULAN Konversi alih status dari IAIN ke UIN
memang membutuhkan persyaratan yang
cukup banyak, sehingga perubahan itu
tidak dapat dilakukan dengan mudah
begitu saja. Tidak cukup persyaratan
administrasi saja, akan tetapi juga
persyaratan substantif.Saat ini terdapat 23
UIN dari total 58 PTKIN atau sekitar 40 %
dari total perguruan tinggi Islam yang ada. 4. Perkembangan Sarana dan Prasarana Saat ini Universitas Islam Negeri
sudah memiliki sarana dan prasarana yang
sama seperti dimiliki oleh Universitas
umum. Namun letak perbedaanya sarana
dan prasarana tersebut tidak semua
dimiliki oleh UIN di Indonesia, ada
beberapa sarana dan prasarana yang
baru dimiliki beberapa UIN saja, seperti:
Rumah sakit (UIN Syarif Hidayatullah),
Planetarium
dan
Observatorium
(UIN
WalisongoSemarang), Perpustakaan
Digital,
Asrama
mahasiswa
berupa
Rusunawa. 2. Perkembangan Kurikulum Dalam
kurikulumnya
sudah
memberlakukan integrasi ilmu sebagai basis
kurikulumnya. Tentunya, UIN menjadi
harapan semua orang, khususnya umat
Islam dan bangsa Indonesia pada umumnya. UIN tidak hanya fokus pada ilmu
keislamanan saja, akan tetapi ilmu
pengetahuan lainnya di luar ilmu
keislamanan. Sebab, tujuan UIN dilahirkan
untuk mengintegrasikan ilmu yang selama
ini ilmu dipandang dalam paradigma
dikotomi, sehingga lulusan dari PTKIN
tidak mampu menjawab kebutuhan
masyarakat. Jika lulusan PTKIN terbatas
geraknya, maka lulusan tersebut tidak
marketable, artinya tidak dapat berkiprah
dalam semua sektor dunia usaha. Karena itu,
UIN adalah solusi untuk menyahuti
dinamika keilmuan dan tuntutan dunia
kerja. UIN merupakan model PTKIN yang
ideal di era globalisasi, di mana UIN
mengembangkan multi dispilin ilmu dalam
konsep integrasi ilmu. 1. Perkembangan Jumlah Sejak tahun 2002 (berdirinya Pertama
kali UIN) hingga kini Perkembanga UIN
mengalami kemajuan yang sangat pesat, Dari
segi kuanitas, saat ini terdapat 23 UIN dari
total 58 PTKIN atau sekitar 40% dari total
perguruan tinggi Islam yang ada jumalh
tersebut termasuk yang cukup banyak. JUPEIS: Jurnal Pendidikan dan Ilmu Sosial
https://jurnal.jomparnd.com/index.php/jp Kemudian
Saat
ini
pembinaan
teknis
penyelenggaraan program pendidikan tinggi
ilmu agama Islam dilakukan oleh menteri
yang menyelenggarakan urusan pemerintah
di bidang agama, dan oembinaan teknis
program pendidikan ilmu lain dilakukan
oleh menteri yang menyelenggarkan urusan
pemerintahan di bidang lain. Universitas
Islam
menyelenggarakan
pendidikan
akademik dan dapat menyelenggarakan
pendidikan vokasi dalam berbagai umpul
ilmu pengetahuan dan /atau teknologi dan
jika memenuhi syarat dapat menyelenggarkan
pendidikan profesi sesuai dengan ketentuan
perundang-undangan. Integrasi Keilmuan Perguruan Tinggi
Islam Universitas
mengharuskan
untuk
mempelajari keilmuan umum. b. Tidak hanya dominan pada orientasi
dakwah, tetapi juga merespon masyarakat
semakin
kompleks. Karekteristik
Keilmuan yang ada di UIN pada dasarnya
tidak
terlepas
dari
masalah
yang
dihadapi oleh lembaga sebelumnya yakni
IAIN
yakni
kurikulum
IAIN belum
mampu merespon perkembangan IPTEK
dan perubahan masyarakat yang semakin
kompleks. Selama ini kerap bahwa
lulusan IAIN sebagai juru dakwah dan
tidak memiliki kemampuan dalam
bidang pengetahuan umun dan teknologi,
sehingga
dengan
stigma
seperti
itu
menghambat laju pertumbuhan karir
dan pekerjaan. c. 52 Vol. 3. No. 1 Januari 2024
e-ISSN: 2809-7998 p-ISSN: 2809-8005 Vol. 3. No. 1 Januari 2024
e-ISSN: 2809-7998 p-ISSN: 2809-8005 Vol. 3. No. 1 Januari 2024
e-ISSN: 2809-7998 p-ISSN: 2809-8005 JUPEIS: Jurnal Pendidikan dan Ilmu Sosial
https://jurnal.jomparnd.com/index.php/jp UCAPAN TERIMAKASIH Terimakasih kepada semua pihak yang
telah membantu sehingga penelitian ini bisa
terlaksana. Tim Kemanag RI. (2016). Qur’an Kemenag. Badan Litbang dan Diklat Lajnah
Pentashihan Mushaf al-Qur’an. 3. Perkembangan Kelembagaan Secara kelembagaan UIN merupakan
perguruan tinggi di Lingkungan
Kementerian Agama yang berada dan di
bawah tanggung jawab menteri agama. 53 Vol. 3. No. 1 Januari 2024
e-ISSN: 2809-7998 p-ISSN: 2809-8005 Vol. 3. No. 1 Januari 2024
e-ISSN: 2809-7998 p-ISSN: 2809-8005 JUPEIS: Jurnal Pendidikan dan Ilmu Sosial
https://jurnal.jomparnd.com/index.php/jp Model integrasi keilmuan di UIN Syarif
Hidayatullah ialah reintegrasi ilmu, UIN
Sunan Ampel Surabaya konsep integrated
twin
towers(ITT),
UINSumatera
Utara
mengusung Konsep Wahdatul Ulu, UIN
Sunan Kalijaga konsep Integrasi Interkoneksi. Kurikulum UIN tidak menganggap bahwa
masing-masing keilmuan terpisah melainkan
satu kesatuan, yang bersumber dari Al-
Quran dan Hadis. Untuk itulah yang
membuatnya begitu berbeda dengan
pendidikan tinggi lainnya. Mudzhar, M. A. (2000). Kedudukan UIN
sebagai
Perguruan
Tinggi. dalam
Problem dan Prospek IAIN: Anotologi
Pendidikan
Tinggi
Islam,
ed. Komaruddin Hidayat&Hendro Prastyo. Jakarta:
Dirjen
Pembinaan
Kelembagaan Agama Islam Depag RI. Rezky Nurhidayah. (2019). Perkembangan
Perguruan Tinggi di Indonesia, (Jurnal
Pendidikan Islam, Vol.2, No.1. Rofi, S. (2016). Sejarah Pendidikan Islam di
Indonesia. Deepublish. Rukiati,
Henawati. (2006). Sejarah
Pendidikan Islam di Indonesia. Pustaka
Setia. Syahri, S. (2022). Pembaharuan Pendidikan
Islam Perspektif Fazlur Rahman Dalam
Perguruan
Tinggi
Agama
Islam. Berkala Ilmiah Pendidikan, 2(1),
1-11. DAFTAR PUSTAKA Abrori, M. S., & Nurkholis, M. (2019). Islamisasi Ilmu Pengetahuan Menurut
Pandangan Syed Muhammad Naquib
Al-Attas Dan Implikasinya Terhadap
Pengembangan
PAI
Di
Perguruan
Tinggi
Umum. Al-I'tibar:
Jurnal
Pendidikan Islam, 6(1), 09-18. Wahyono, A. (2014). Kebijakan Pendidikan
Islam: Hibridasi Lembaga Pendidikan
Tinggi dalam Jurnal Pendidikan Islam. Afrizal, A. (2022). Perkembangan Perguruan
Tinggi Islam Di Indonesia (Kasus
STAIN, IAIN, UIN Dan Perguruan
Tinggi
Islam). Berkala
Ilmiah
Pendidikan, 2(1), 18-32. Amiruddin, A. (2017). Dinamika Lembaga
Pendidikan
Tinggi
Islam
Di
Indonesia. MIQOT: Jurnal Ilmu-Ilmu
Keislaman, 41(1). Darda, A. (2015). Integrasi ilmu dan agama:
Perkembangan
konseptual
di
Indonesia. At-Ta'dib, 10(1). Darminta, P. (2005). Kamus Besar Bahasa
Indonesia. Yogyakarta: Tiara Wacana
Yogya. Lubis, R. R. (2021). Universitas Islam Negeri
(Studi Historisitas, Perkembangan dan
Model
Integrasi
Keilmuan). Hikmah, 18(2), 150-167. 54
|
https://openalex.org/W2068292550
|
https://journals.plos.org/plosone/article/file?id=10.1371/journal.pone.0047884&type=printable
|
English
| null |
Usp14 Deficiency Increases Tau Phosphorylation without Altering Tau Degradation or Causing Tau-Dependent Deficits
|
PloS one
| 2,012
|
cc-by
| 8,325
|
Usp14 Deficiency Increases Tau Phosphorylation without
Altering Tau Degradation or Causing Tau-Dependent
Deficits Youngnam N. Jin1, Ping-Chung Chen1¤a, Jennifer A. Watson1, Brandon J. Walters1¤b, Scott E. Phillips1,
Karen Green2, Robert Schmidt2, Julie A. Wilson1, Gail V. Johnson3, Erik D. Roberson4, Lynn E. Dobrunz1,
Scott M. Wilson1* 1 Department of Neurobiology, Civitan International Research Center, University of Alabama at Birmingham, Birmingham, Alabama, United States of America, 2 Division of
Neuropathology, Department of Pathology and Immunology, Washington University School of Medicine, Saint Louis, Missouri, United States of America, 3 Department of
Anesthesiology, University of Rochester, Rochester, New York, United States of America, 4 Department of Neurology, University of Alabama at Birmingham, Birmingham,
Alabama, United States of America Abstract Competing Interests: The authors have declared that no competing interests exist. Competing Interests: The authors have declared that no competing interests exist. * E-mail: wilson@nrc.uab.edu ¤a Current address: Department of Structural Biology, St. Jude Children’s Research Hospital, Memphis, Tennessee, United States of America
¤b Current address: Department of Developmental Biology, St. Jude Children’s Research Hospital, Memphis, Tennessee, United States of America antagonize substrate degradation. For example, by removing
ubiquitin side chains prior to commitment of substrates to
degradation, proteasomal deubiquitinating enzymes can prevent
substrates from being degraded by the proteasome [9,10]. Usp14
is one of the proteasomal deubiquitinating enzymes that can
remove ubiquitin side chains on proteins bound to the proteasome
and, as such, plays an important role in ubiquitin recycling and
substrate stability [11]. Citation: Jin YN, Chen P-C, Watson JA, Walters BJ, Phillips SE, et al. (2012) Usp14 Deficiency Increases Tau Phosphorylation without Altering Tau Degradation or
Causing Tau-Dependent Deficits. PLoS ONE 7(10): e47884. doi:10.1371/journal.pone.0047884 Abstract Regulated protein degradation by the proteasome plays an essential role in the enhancement and suppression of signaling
pathways in the nervous system. Proteasome-associated factors are pivotal in ensuring appropriate protein degradation,
and we have previously demonstrated that alterations in one of these factors, the proteasomal deubiquitinating enzyme
ubiquitin-specific protease 14 (Usp14), can lead to proteasome dysfunction and neurological disease. Recent studies in cell
culture have shown that Usp14 can also stabilize the expression of over-expressed, disease-associated proteins such as tau
and ataxin-3. Using Usp14-deficient axJ mice, we investigated if loss of Usp14 results in decreased levels of endogenous tau
and ataxin-3 in the nervous system of mice. Although loss of Usp14 did not alter the overall neuronal levels of tau and
ataxin-3, we found increased levels of phosphorylated tau that correlated with the onset of axonal varicosities in the Usp14-
deficient mice. These changes in tau phosphorylation were accompanied by increased levels of activated phospho-Akt,
phosphorylated MAPKs, and inactivated phospho-GSK3b. However, genetic ablation of tau did not alter any of the
neurological deficits in the Usp14-deficient mice, demonstrating that increased levels of phosphorylated tau do not
necessarily lead to neurological disease. Due to the widespread activation of intracellular signaling pathways induced by the
loss of Usp14, a better understanding of the cellular pathways regulated by the proteasome is required before effective
proteasomal-based therapies can be used to treat chronic neurological diseases. Citation: Jin YN, Chen P-C, Watson JA, Walters BJ, Phillips SE, et al. (2012) Usp14 Deficiency Increases Tau Phosphorylation without Altering Tau Degradation or
Causing Tau-Dependent Deficits. PLoS ONE 7(10): e47884. doi:10.1371/journal.pone.0047884 itor: Mel B. Feany, Brigham and Women’s Hospital, Harvard Medical School, United States of America Received August 7, 2012; Accepted September 24, 2012; Published October 29, 2012 Copyright: 2012 Jin et al. This is an open-access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted
use, distribution, and reproduction in any medium, provided the original author and source are credited. Funding: NIH/NINDS Grants NS047533 and NS074456 (S.M.W), and NS075487 (E.D.R.). The funders had no role in study design, d
decision to publish, or preparation of the manuscript. ants NS047533 and NS074456 (S.M.W), and NS075487 (E.D.R.). The funders had no role in study design, data collection and analysis
eparation of the manuscript. Funding: NIH/NINDS Grants NS047533 and NS074456 (S.M.W), and NS075487 (E.D.R.). The
decision to publish, or preparation of the manuscript. October 2012 | Volume 7 | Issue 10 | e47884 Editor: Mel B. Feany, Brigham and Women’s Hospital, Harvard Medical School, United States of America Quantitation of Purkinje cell axonal swellings j
g
Four to five 5 to 6-week-old wt, axJ, tauKO and axJtauKO mice
were euthanized by CO2 asphyxiation. Brains were sagittally cut
and were immersed in methacarne to fix overnight at 4uC. Samples were paraffin embedded and sliced medially in 10 mm
sections with a Microme HM 355S. Twenty medial sections from
each brain were taken, and three sections from each sample were
randomly chosen to quantitate. Chosen sections were deparaffi-
nized and rehydrated. Sections were blocked in a 10 mM PBS
solution containing 1% BSA, 0.2% non-fat dry milk, 0.1% Triton
X-100 for 30 min, and were then incubated with a 1:500 dilution
of a rabbit calbindin polyclonal antibody (Swant, Bellinzona,
Switzerland) for 1 h at room temperature. Samples were washed 3
times with 10 mM PBS for 5 min. The sections were then treated
with Alexa-Fluor goat-anti-rabbit 568 at a 1:500 dilution and
DAPI at a 1:1000 dilution in the dark for 1 h at room
temperature. Sections were washed 3 times with 10 mM PBS for
5 min. Samples were mounted with 50% glycerol in 10 mM PBS
and stored at 220uC. An Olympus BX-51 upright microscope was
utilized to take images at 106of the medial deep cerebellar nuclei. Axonal swellings greater than 2 mm were counted and normalized
to the area of the cerebellar nuclei region with the MicroBright-
Field Stereo Investigator software (Williston, VT). Body mass analysis Body weights were collected from 4 and 8-week-old wt, axJ,
tauKO and axJtauKO mice. Weights were determined for 4 animals
per genotype. Values represent the average body mass 6 SE. Effect of Usp14 Deficiency on Tau Levels in Mice Effect of Usp14 Deficiency on Tau Levels in Mice light/dark cycle in ventilated cages and maintained on Harlan
Teklad 7904 breeders’ diet. Homozygous Usp14axJ mice, which
we refer to as axJ mice, were generated by intercrossing axJ/+
siblings. Tau deficient mice (tauKO) were obtained from Jackson
Laboratories [24]. The axJ mice were crossed to tauKO mice to
generate mice homozygous for the axJ and tau mutations (axJtauKO). Mice were euthanized by CO2 asphyxiation. in an increase in the turnover of ubiquitinated proteins in vitro
[14,15]. Loss of Usp14 in cell culture models resulted in a decrease
in the steady-state levels of aggregate-prone proteins such as tau
and ataxin-3 [14], indicating that Usp14 may serve as a
therapeutic
target
for
the
treatment
of
neurodegenerative
proteinopathies [14]. The association of elevated ubiquitin conjugates with intracel-
lular deposits of aggregate-prone proteins suggests a causative link
between proteasome dysfunction and disease [16]. For example,
tau accumulation can lead to the production of abnormally
modified oligomeric tau and neurofibrillary tangles (NFTs) that
may contribute to neurological disease in patients with Alzheimer’s
disease (AD) [17]. Since reduced proteasomal activity and
increased tau accumulation has been observed in the brains of
AD patients [18], enhancing the activity of the proteasome may
offer a potential therapeutic avenue for the treatment of diseases
associated with the accumulation of aggregate-prone proteins. Animals Wild
type
(wt)
C57BL/6J
and
Usp14axJ
mice
(Jackson
Laboratories, Bar Harbor, ME) have been maintained in our
breeding colony at the University of Alabama at Birmingham,
which is fully accredited by the Association for Assessment and
Accreditation of Laboratory Animal Care International (Animal
Protocol number 110909471). Mice were housed with a 12 hr Ethics Statement The experiments described in this manuscript were approved by
the University of Alabama at Birmingham Institutional Animal
Care and Use Committee (IACUC). All research complied with
the United States Animal Welfare Act and other Federal statutes
and regulations relating to animals and experiments involving
animals, and adhered to principles stated in the Guide for the Care
and Use of Laboratory Animals, United States National Research
Council. All reasonable efforts were made to minimize suffering of
animals. Electron Microscopy Animals were anesthetized with ketamine/xylazine and per-
fused with 25 mL of heparinized saline followed by 25 mL of
modified Karnovsky’s fixative containing 3% glutaraldehyde and
1% paraformaldehyde in sodium cacodylate buffer, pH 7.4. Fixation was continued overnight at 4uC in the same fixative,
and the following day the cerebellum was dissected, cleaned of
extraneous tissue, and rinsed in sodium cacodylate buffer. Tissue
was post-fixed in phosphate cacodylate-buffered 21% OsO4 for
1 h, dehydrated in graded ethanol with a final dehydration in
propyleneoxide, and embedded in EMbed-812 (Electron Micros-
copy Sciences, Hatfield, PA). One-micron-thick plastic sections
were examined by light microscopy after staining with toluidine
blue. Ultrathin sections (90 nm thick) were cut onto formvar-
coated slot grids. Sections were post-stained with uranyl acetate
and Venable’s lead citrate and viewed with a 1200EX electron
microscope (JEOL, Tokyo, Japan). Digital images were acquired
using the AMT Advantage HR camera (Advanced Microscopy
Techniques, Danvers, MA). Proteasome inhibition has been shown to alter the expression of
several mitogen-activated signaling pathways [19,20]. By selec-
tively targeting protein kinases for degradation, the UPS is able to
directly control intracellular signaling events. For example, ASK-
1, the mitogen-activated protein kinase kinase kinase, is ubiqui-
tinated and targeted to the proteasome by the E3 ligase CHIP
[21]. In addition, proteasome inhibition has also been shown to
induce c-Jun NH2-terminal kinase (JNK) and extracellular signal-
regulated kinases (ERKs) [19,22]. While proteasome inhibition has
been shown to have both pro- and anti-apoptotic effects, the result
of enhanced proteasome function on intracellular signaling
pathways is currently unknown [22,23]. In this study, we investigated if loss of Usp14 in vivo affects the
levels of aggregate-prone proteins in the central nervous system. Although loss of Usp14 did not alter the steady-state levels of
either endogenous tau or ataxin-3 in mice, we did observe a
significant increase in the levels of phosphorylated tau and
increased activation of Akt, JNK and ERK. Genetic ablation of
tau did not change the disease course in the axJ mice, indicating
that abnormal tau phosphorylation was not responsible for causing
the neurological deficits in the axJ mice. Our results therefore
suggest that the expression of the aggregate-prone proteins tau and
ataxin-3 are not controlled by Usp14 in neurons and that the loss
of Usp14 activates several different neuronal signaling pathways
that may contribute to the synaptic transmission defects observed
in the axJ mice. Introduction The ubiquitin proteasome system (UPS) functions to control
intracellular protein abundance [1] and is tightly regulated to
facilitate the removal of damaged proteins that can cause disease
[2–4]. Cellular proteins are marked for degradation by tagging
them with a 76 amino acid protein called ubiquitin [5], enabling
them to productively interact with the proteasome, a multi-subunit
protease which catalyzes their breakdown into small peptides. Changes in the targeting and destruction of ubiquitinated proteins
are observed in many chronic neurological diseases, reinforcing
the importance of regulated proteolysis in neuronal viability and
function [6,7]. Our previous studies demonstrated that the loss of Usp14
expression specifically in the nervous system causes axJ mice to
display a resting tremor, gait ataxia, motor-endplate disease and
cerebellar pathology due to impaired neuronal development
[12,13]. It is currently unknown if Usp14 is dispensable in the
adult nervous system. Although the axJ mice have dysfunctional
proteasomes, there is no accumulation of ubiquitinated proteins or
increase in apoptotic cells in the central nervous system [13]. This
observation is consistent with the finding that loss of Usp14 results Following the binding of ubiquitinated substrates to the
proteasome, ubiquitin side chains can be disassembled and/or
modified by further ubiquitination [8]. The proteasomal factors
responsible for these activities are believed to either enhance or October 2012 | Volume 7 | Issue 10 | e47884 PLOS ONE | www.plosone.org 1 October 2012 | Volume 7 | Issue 10 | e47884 repeated a minimum of 10 times at 0.1 Hz. The averaged paired-
pulse ratio (fEPSP2/fEPSP1) was calculated after recording. repeated a minimum of 10 times at 0.1 Hz. The averaged paired-
pulse ratio (fEPSP2/fEPSP1) was calculated after recording. Field excitatory postsynaptic potentials (fEPSPs) were recorded
in response to extracellular stimulation of Schaffer collateral axons
by a bipolar tungsten microelectrode (FHC, Bowdoinham, ME)
placed in the stratum radiatum of the CA1 region. Recording
pipettes filled with external recording solution were placed in the
same region of the CA1 stratum radiatum. Stimulation was
generated from a Grass S48 stimulator (Grass Tech, West
Warwick, RI) and applied with a BSI-2 biphasic stimulus isolator
(BAK Electronics, Mount Airy, MD). Electrophysiology Hippocampal slices were held in a submersion recording
chamber perfused with external recording solution composed of
(in mM): 120 NaCl, 3.5 KCl, 2.5 CaCl2, 1.3 MgCl2, 1.25
NaH2PO4, 26 NaHCO3, and 10 glucose [25]. The solution was October 2012 | Volume 7 | Issue 10 | e47884 PLOS ONE | www.plosone.org 2 Effect of Usp14 Deficiency on Tau Levels in Mice Figure 1. Steady-state levels of aggregate-prone tau and ataxin-3 proteins in the cerebellum of wt and Usp14-deficient axJ mice. (A)
Representative immunoblot of ataxin-3, tau, a- and b-tubulin, and Usp14 from the cerebellum of two different 4-week-old wt and axJ mice. The tau5
antibody was used to determine total tau levels, and the PHF-1 antibody was used to identify phosphorylated tau. (B) Quantitation of total and PHF-
1-reactive tau. n = 4 to 5 animals per genotype (C) Immunohistochemistry of PHF-1-reactive tau in the cerebellum of wt and axJ mice. Cryosections
from each genotype were stained with the PHF-1 antibody and visualized using DAB. The boxed area in a and b is enlarged in c and d, respectively. Arrows indicate localized regions of intense staining with PHF-1 in Purkinje cell dendrites. ML, molecular layer; PCL, Purkinje cell layer; and GC, granule
cell layer. Scale bars 50 mm. doi:10 1371/journal pone 0047884 g001 Figure 1. Steady-state levels of aggregate-prone tau and ataxin-3 proteins in the cerebellum of wt and Usp14-deficient axJ mice. (A)
Representative immunoblot of ataxin-3, tau, a- and b-tubulin, and Usp14 from the cerebellum of two different 4-week-old wt and axJ mice. The tau5
antibody was used to determine total tau levels, and the PHF-1 antibody was used to identify phosphorylated tau. (B) Quantitation of total and PHF-
1-reactive tau. n = 4 to 5 animals per genotype (C) Immunohistochemistry of PHF-1-reactive tau in the cerebellum of wt and axJ mice. Cryosections
from each genotype were stained with the PHF-1 antibody and visualized using DAB. The boxed area in a and b is enlarged in c and d, respectively. Arrows indicate localized regions of intense staining with PHF-1 in Purkinje cell dendrites. ML, molecular layer; PCL, Purkinje cell layer; and GC, granule
cell layer. Scale bars 50 mm. doi:10.1371/journal.pone.0047884.g001 y
m
doi:10.1371/journal.pone.0047884.g001 bubbled
with
95%
O2
and
5%
CO2,
pH 7.35–7.45. All
experiments were performed at 25uC. Tau and ataxin-3 expression in Usp14-deficient mice Tau and ataxin-3 expression in Usp14-deficient mice Tau and ataxin 3 expression in Usp14 deficient mice
We have recently reported that Usp14 is required for the stable
expression of neurological disease-related proteins, such as tau and
ataxin-3, in mouse embryonic fibroblasts [14]. To determine if
Usp14 also stabilizes these proteins in vivo, we used immunoblot
analysis to analyze the expression of endogenous tau and ataxin-3
in wt and axJ mice. In contrast to what was observed in vitro [14],
we found no differences in the levels of ataxin-3 in cerebellar
extracts from 4-week-old wt and axJ mice (Fig. 1a), indicating that
Usp14 was not required for the neuronal stability of ataxin-3. Although immunoblot analysis demonstrated that the overall levels
of tau were also similar in the wt and axJ mice (Fig. 1a and b), the
absence of Usp14 did result in a shift in the migration pattern of
tau in the axJ mice (Fig. 1a). To determine if the slower migrating
form of tau found in the cerebellar extracts from the axJ mice
represents increased levels of phospho-tau, we performed immu-
noblot analysis on cerebellar extracts from wt and axJ mice using
the phospho-specific tau antibody PHF-1 [26]. There was a
significant increase in the expression of PHF-1-reactive tau in the
axJ mice as compared to controls (Fig. 1a and 1b), demonstrating
that, while loss of Usp14 does not affect the steady-state levels of
tau in neurons, loss of Usp14 alters the phosphorylation state of
tau in the axJ mice. Slower migrating forms of tau have also been detected in the
brains of AD patients, and these forms of tau are believed to be
due to hyperphosphorylated tau, which may aggregate to form
neurofibrillary tangles [17,27]. Mislocalization of tau to the
somatodendritic compartment of neurons instead of the axon is
one of the earliest aspects of tau pathology [28]. To determine if
the increased levels of phospho-tau observed in the axJ mice were
associated with aberrant localization of tau, we stained cerebellar
sections from wt and axJ mice with the PHF-1 antibody (Fig. 1c). The cerebellar sections from control mice revealed widespread
PHF-1 staining in the molecular and granule cell layers (Fig. 1c). In contrast, the axJ mice exhibited strong PHF-1 staining in the
Purkinje cell dendrites, cell body, and axonal swellings near the
deep cerebellar nuclei and showed reduced staining in the
molecular layer. Diaminobenzidine (DAB) was used for color development and was
followed with hematoxylin counterstain. Diaminobenzidine (DAB) was used for color development and was
followed with hematoxylin counterstain. Diaminobenzidine (DAB) was used for color development and was
followed with hematoxylin counterstain. Immunoblotting Proteins were resolved on 8% Tris-glycine gels and transferred
onto PVDF membranes. The Usp14 R138 antisera [12], ataxin-3
(gift from Dr. Hank Paulson, University of Michigan), PHF-1
antibody (gift from Dr. Peter Davis, Albert Einstein College of
Medicine), tau1 (dephosphorylated amino acids 189–207), tau5
(amino
acids
210–230),
12E8
(pSer262/pSer356),
AT180
(pThr231), AT100 (pThr212/pSer214) (gifts from Dr. Gail
Johnson, Rochester University), anti-b-tubulin, and anti-a-tubulin
(Developmental Hybridoma Core, Iowa City, IA), GSK3b, Akt,
Cdk5, p35, MEK1/2, JNK1/2 and ERK1/2 (Cell Signaling,
Danvers, MA) antibodies were diluted in PBS containing 5% non-
fat milk. Primary antibodies were detected using an anti-mouse or
anti-rabbit HRP-conjugated antibody (Southern Biotechnology
Associates, Birmingham, AL) and Luminol reagents (Pierce,
Rockford, IL). Our previous studies demonstrated the presence of proximal
Purkinje cell axonal swellings in the axJ mice [12]. To further
examine the distribution of the axonal swellings in the axJ mice, we
examined the distal Purkinje cell axons for any sign of disease. Consistent
with
our
previous
findings
[12],
we
observed
widespread Purkinje cell axonal swellings near the deep cerebellar
nuclei in the axJ mice (Fig. 2a). Since increased tau phosphory-
lation correlates with decreased microtubule binding and disorga-
nized microtubules [29], we used electron microscopy to compare
the Purkinje cell axonal ultrastructure in the wt and axJ mice. The
Purkinje cell axons from wt mice were myelinated and showed no
gross structural alterations (Fig. 2b). Consistent with our immu-
nohistochemical findings, large axonal swellings were observed
along the Purkinje cell axons of the axJ mice. These swollen axons
appeared to be unmyelinated and contained numerous mitochon-
dria and randomly arranged microtubules (Fig. 2b). These changes
in axonal structure did not correlate with any significant changes
in the levels of a- or b-tubulin (Fig. 1a). Therefore, while there PHF-1 Staining Paraffin-embedded sections of mouse brains (7 mm) were
deparaffinized and followed with antigen retrieval by citrate
buffer (10 mM, pH 6). The endogenous peroxidase activity was
blocked with 3% hydrogen peroxide. Sections were blocked with
3% goat serum for 30 min and were incubated with the PHF-1
antibody (1:500) overnight at 4uC. The PHF-1 antibody recog-
nizes tau phosphorylated at Ser 396/404. Sections were then
incubated with biotinylated goat secondary antibody for 30 min,
and then incubated with avidin-biotin peroxidase complex (Vector
Laboratories, Burlingame, CA) for 60 min at room temperature. Paired-pulse stimulation at different intervals (50, 100, 150, 200,
and 500 ms) was applied in a pseudo-random sequence and PLOS ONE | www.plosone.org October 2012 | Volume 7 | Issue 10 | e47884 3 Effect of Usp14 Deficiency on Tau Levels in Mice Figure 2. Structural analysis of the cerebellum from wt and axJ
mice. (A) Indirect immunofluorescence of cerebellum from wt and axJ
mice using calbindin (red), b-tubulin (green) and DAPI (blue). (B)
Electron micrographs of cerebellar sections from wt and axJ mice. Black
line indicates the outline of a Purkinje cell axon. Magnification at
2,500X. n = 3 to 4 animals per genotype. doi:10.1371/journal.pone.0047884.g002 Tau and ataxin-3 expression in Usp14-deficient mice Focal staining with PHF-1 was also noted in
the dendrites of the axJ Purkinje cells (Fig. 1c). Increased PHF-1
staining was also detected in the cortex and hippocampus of the
axJ mice (data not shown). Therefore, in addition to increasing the
levels of phosphorylated tau in the neurons of the axJ mice, loss of
Usp14 also leads to a redistribution of PHF-1-reactive tau in the
Purkinje cells of the axJ mice. Figure 2. Structural analysis of the cerebellum from wt and axJ
mice. (A) Indirect immunofluorescence of cerebellum from wt and axJ
mice using calbindin (red), b-tubulin (green) and DAPI (blue). (B)
Electron micrographs of cerebellar sections from wt and axJ mice. Black
line indicates the outline of a Purkinje cell axon. Magnification at
2,500X. n = 3 to 4 animals per genotype. doi:10.1371/journal.pone.0047884.g002 Quantitation of immunoblots Blots were scanned using a Hewlett Packard Scanjet 3970 and
quantitated using UN-SCAN-IT software (Orem, UT). Each value
represents the average and standard error from 4 blots using at
least three different animals of each genotype. PLOS ONE | www.plosone.org October 2012 | Volume 7 | Issue 10 | e47884 4 Effect of Usp14 Deficiency on Tau Levels in Mice Figure 3. Immunoblot analysis of tau levels in specific brain regions. (A) Cortical and (B) hippocampal extracts were prepared from two
different 3 and 6-week-old wt and axJ mice. The PHF-1 antibody was used to determine levels of phosphorylated tau, the tau5 antibody was used to
detect total tau, and Gapdh was used as a loading control. n = 4 to 5 animals per genotype. doi:10.1371/journal.pone.0047884.g003 Figure 3. Immunoblot analysis of tau levels in specific brain regions. (A) Cortical and (B) hippocampal extracts were prepared from two
different 3 and 6-week-old wt and axJ mice. The PHF-1 antibody was used to determine levels of phosphorylated tau, the tau5 antibody was used to
detect total tau, and Gapdh was used as a loading control. n = 4 to 5 animals per genotype. doi:10.1371/journal.pone.0047884.g003 Figure 3. Immunoblot analysis of tau levels in specific brain regions. (A) Cortical and (B) hippocampal extracts were prepared from two
different 3 and 6-week-old wt and axJ mice. The PHF-1 antibody was used to determine levels of phosphorylated tau, the tau5 antibody was used to
detect total tau, and Gapdh was used as a loading control. n = 4 to 5 animals per genotype. doi:10.1371/journal.pone.0047884.g003 Figure 3. Immunoblot analysis of tau levels in specific brain regions. (A) Cortical and (B) hippocampal extracts were prepared from two
different 3 and 6-week-old wt and axJ mice. The PHF-1 antibody was used to determine levels of phosphorylated tau, the tau5 antibody was used to
detect total tau, and Gapdh was used as a loading control. n = 4 to 5 animals per genotype. doi:10.1371/journal.pone.0047884.g003 tau [26,29]. Using these antibodies, we examined the levels of
phospho-tau epitopes in hippocampal extracts from 3 and 6-week-
old wt and axJ mice (Fig. 4). At 3 weeks of age, a significant
increase in AT100 (pSer212 and pThr214) reactivity was observed
in the axJ mice as compared to controls. Quantitation of immunoblots By 6 weeks of age, there
was a significant increase in reactivity with the AT100, AT180
(pThr231),
and
12E8
(pSer262
and
pSer356)
phospho-tau
epitopes and a significant decrease in reactivity with the non-
phosphorylated tau1 epitope in the axJ mice as compared to
controls. These results demonstrate that there are widespread
changes in the phosphorylation state of tau in the Usp14-deficient
axJ mice, and the time frame of these changes suggests that tau
hyperphosphorylation may contribute to the axonal pathology
found in the axJ mice. were no major changes in the steady-state levels of these
microtubule proteins or tau, loss of Usp14 did result in aberrant
localization of phosphorylated tau and microtubule disorganiza-
tion in the axJ mice. axJ mice have elevated levels of phosphorylated tau in
the cortex and hippocampus NFTs are found in several regions of the brains of AD patients,
including the hippocampus and entorhinal and frontal cortexes
[30]. To determine if frontal cortex and hippocampal brain
regions of the axJ mice have increased levels of phosphorylated
tau, we investigated PHF-1-reactive tau levels by immunoblot
analysis. At both 3 and 6 weeks of age, elevated levels of PHF-1-
reactive tau were observed in cortical and hippocampal extracts
from the axJ mice compared to controls (Fig. 3), demonstrating
that
loss
of Usp14 results
in widespread
changes
in tau
phosphorylation in the axJ mice. Dysregulation of GSK3b and stress kinases in axJ mice
Several kinases, including Cdk5, GSK3b, ERK and JNK, have
been implicated in the hyperphosphorylation of tau [23,31–34],
and these increases in tau phosphorylation are thought to
accelerate NFT formation [35,36]. To determine if the levels of
these
kinases
are
altered
in
the
axJ
mice,
we
examined
hippocampal protein extracts from 6-week-old wt and axJ mice Discussion The UPS regulates numerous cellular pathways by controlling
protein abundance. This task is accomplished by the coordinated
ubiquitination of proteins, followed by their destruction by the
proteasome. Proteasomal deubiquitinating enzymes alter protein
degradation rates by removing the ubiquitin side chains attached
to substrates, thus preventing their degradation [14,15,45]. Usp14
is one of three deubiquitinating enzymes found on mammalian
proteasomes, and blocking Usp14’s deubiquitinating activity has
been shown to accelerate degradation of aggregate-prone proteins
in vitro [11,14,15]. However, our investigations on the loss of
Usp14 in vivo did not reveal any significant changes in the steady-
state levels of tau or ataxin-3 in the brains of the Usp14-deficient
axJ mice compared to controls, indicating that endogenous tau and
ataxin-3 may not be substrates for Usp14, or these proteins may be
degraded through an alternative pathway in vivo [46]. In addition,
the lack of any detectable change in these proteins may be due to
decreased levels of free ubiquitin in the axJ mice that could reduce
targeting of tau and ataxin-3 to the proteasome. Further
experimentation will be required to determine if Usp14 plays a
direct role in the degradation of specific classes of proteasomal
substrates, such as proteins produced with mutated amino acid
sequences. Stress-activated kinases, which are activated by phosphoryla-
tion, have also been hypothesized to contribute to elevated
phospho-tau levels in the brains of AD patients [23,34,37–40]. We
therefore examined if loss of Usp14 alters the levels of stress-
activated kinases in the axJ hippocampus (Fig. 5b). While no
significant changes were observed in the levels of total MEK1,
JNK or ERK, we did find increased levels of phosho-MEK1, JKN
and ERK in the axJ mice (Fig. 5b). These results indicate that
altered proteasome function caused by the loss of Usp14 results in
widespread changes in the levels of activated stress kinases that
have been implicated in tau phosphorylation. Effect of Usp14 Deficiency on Tau Levels in Mice No changes in the levels of total
GSK3b, Cdk5, or the Cdk5 activator p35 were seen in the extracts
from the axJ mice as compared to controls. In contrast, a
significant increase in the levels of the inactive form of GSK3b
(pSer9) were observed in the axJ mice as compared to controls. Since the phosphorylation of GSK3b serine 9 is thought to be the
result of activated Akt, we also investigated the levels of
phosphorylated Akt (pSer473 and pThr308) in axJ and control
mice. Consistent with the increased levels of pSer9 GSK3b, we
observed an increase in both pThr308 and pSer473 Akt in the axJ
mice. These results indicate that Cdk5 and GSK3b are not likely
to be contributing to the increased levels of tau phosphorylation
observed in the axJ mice. Effect of Usp14 Deficiency on Tau Levels in Mice Figure 4. Analysis of phosphorylated tau-epitopes in wt and axJ
mice. Hippocampal extracts from two different 3 and 6-week-old mice
were probed with either the 12E8 antibody (pSer262), AT180 antibody
(pThr231), AT100 antibody (pSer212/pThr214), tau1 antibody (unpho-
sphorylated tau), or tau5 antibody (total tau). Gapdh was used as a
loading control. n = 3 to 4 mice per genotype. doi:10.1371/journal.pone.0047884.g004 axJ mice, we found that depletion of tau did not improve body
weights or increase viability of the axJtauKO mice as compared to
axJ controls (Fig. 6a and b). Our finding of PHF-1-reactive tau in the Purkinje cell processes
of the axJ mice prompted us to investigate if tau depletion affected
the number of Purkinje cell axonal swellings in the axJ mice. As
previously reported [12], the axJ mice had a large increase in the
number of Purkinje cell axonal swellings compared to wt controls
(Fig. 6c), and there were no significant changes in the number of
Purkinje cell axonal swellings in the tauKO mice as compared to
controls (Fig. 6c). When we compared the number of axonal
swellings in the axJ and axJtauKO mice, we did not observe any
significant changes in the abundance of these swellings, indicating
that the increased levels of phospho-tau did not contribute to the
formation of Purkinje cell axonal swellings in the axJ mice. In addition to the structural deficits observed in the CNS of the
axJ mice, loss of Usp14 also results in impairment of hippocampal
short-term plasticity [13,44] that is thought to be required for
normal learning and memory. When we compared short-term
plasticity in the hippocampus of wt, tauKO, and axJ mice, we only
observed a reduction in paired-pulse facilitation in the axJ mice
(Fig. 6d). No differences in paired-pulse facilitation were observed
between the axJ and axJtauKO mice (Fig. 6d), indicating that the
synaptic changes caused by Usp14 deficiency are not due to
alterations in tau phosphorylation. Figure 4. Analysis of phosphorylated tau-epitopes in wt and axJ
mice. Hippocampal extracts from two different 3 and 6-week-old mice
were probed with either the 12E8 antibody (pSer262), AT180 antibody
(pThr231), AT100 antibody (pSer212/pThr214), tau1 antibody (unpho-
sphorylated tau), or tau5 antibody (total tau). Gapdh was used as a
loading control. n = 3 to 4 mice per genotype. doi:10.1371/journal.pone.0047884.g004 by immunoblotanalysis Fig. 5a). Examination of phospho-tau epitopes in axJ mice Several studies have investigated the phosphorylation state of
tau as it relates to tau pathology, and a panel of antibodies has
been generated to detect these different phosphorylation events on October 2012 | Volume 7 | Issue 10 | e47884 PLOS ONE | www.plosone.org 5 Effect of Usp14 Deficiency on Tau Levels in Mice Tau reduction does not alter disease progression in the
axJ mice (A) Steady-state levels of total and inactive (pSer9) GSK3b, total
and activated (pSer473 and pThr308) Akt, Cdk5, p35 and b-actin in the hippocampus of two different wt and axJ mice. (B) Total and phosphorylated
MAPKs MEK1, JNK1/2 and ERK1/2 were examined in hippocampal extracts from 4-week-old wt and axJ mice. (C) Graph depicts the quantitation of
proteins from blots in A and B. n = 3 to 4 mice per genotype. doi:10.1371/journal.pone.0047884.g005 Figure 5. Immunoblot analysis of putative tau kinases in axJ and wt mice. (A) Steady-state levels of total and inactive (pSer9) GSK3b, total
and activated (pSer473 and pThr308) Akt, Cdk5, p35 and b-actin in the hippocampus of two different wt and axJ mice. (B) Total and phosphorylated
MAPKs MEK1, JNK1/2 and ERK1/2 were examined in hippocampal extracts from 4-week-old wt and axJ mice. (C) Graph depicts the quantitation of
proteins from blots in A and B. n = 3 to 4 mice per genotype. doi:10.1371/journal.pone.0047884.g005 multiple cytoskeletal proteins including tau, MAP1b and APC
[53]. Since GSK3b phosphorylation of tau and Map1b reduces
their ability to bind microtubules, we reasoned that increased
GSK3b kinase activity may contribute to the altered microtubule
structure observed in the axons of axJ mice. Surprisingly, instead of
observing increased GSK3b activity, we detected a 5-fold increase
in the levels of inactive GSK3b, indicating that GSK3b is not
likely to be responsible for the elevated levels of phosphorylated
tau in the axJ mice. In addition, our studies demonstrated that the
steady-state levels of Cdk5 and its activator p35 do not change in
the axJ mice compared to wt controls. Like Cdk5 and p35, the
steady-state levels of total GSK3b are also unchanged in the axJ
mice. Acting upstream of GSK3b, Akt and PTEN have been
shown to regulate the phosphorylation state of GSK3b. Consistent
with an increase in inactive GSK3b, we observed a 3-fold increase
in the levels of phospho-Akt. These alterations could result in domains, and these regions of tau are sites for protein phosphor-
ylation [29]. Since phosphorylated tau is less effective than non-
phosphorylated tau in binding polymerized microtubules, it is
believed that the phophorylation state of tau can directly affect
microtubule assembly and/or stability [50]. While tau is normally
restricted to axons, PHF-1-reactive tau was also detected in the
soma and dendrites of the Purkinje cells in the axJ mice. Tau reduction does not alter disease progression in the
axJ mice Proteasome dysfunction induced by the loss of Usp14 results in
a significant increase in the levels of phosphorylated tau in the
brains of the axJ mice. We found increased levels of PHF-1-
reactive tau that coincided with the presence of disorganized
microtubules and Purkinje cell axonal swellings. These changes in
tau were found in several regions of the axJ brain, suggesting that
the loss of Usp14 results in a global change in kinase and/or
phosphatase activity. The large increase in different reactive tau
phospho-epitopes seen in the axJ mice is also consistent with this
idea. The phospho-tau epitopes that were increased in the axJ mice
include Thr212, Ser214, Thr231, Ser235, Ser262, and Ser356,
and these phospho-epitopes are also detected in neurodegenerative
diseases such as AD [26,47–49]. Tau has been shown to associate
with microtubules through internal repeat microtubule-binding Hyperphosphorylated tau has been associated with several
neurological disorders, and this aberrant phosphorylation of tau is
believed to contribute to neuronal dysfunction [27,41,42]. Gene
disruption studies indicate that removal of tau increases viability
and improves synaptic function in some mouse models of AD
[42,43]. Since the axJ mice suffer from impaired neuromuscular
development and synaptic dysfunction, that reduces their lifespan
to approximately 2 months, we investigated if tau reduction would
alter the expression of these phenotypes in the axJ mice. Examination of survival curves generated for wt and tauKO mice
(Fig. 6a) demonstrated that genetic ablation of tau did not affect
viability. When we examined the effect of removing tau from the October 2012 | Volume 7 | Issue 10 | e47884 PLOS ONE | www.plosone.org October 2012 | Volume 7 | Issue 10 | e47884 6 Effect of Usp14 Deficiency on Tau Levels in Mice Figure 5. Immunoblot analysis of putative tau kinases in axJ and wt mice. (A) Steady-state levels of total and inactive (pSer9) GSK3b, total
and activated (pSer473 and pThr308) Akt, Cdk5, p35 and b-actin in the hippocampus of two different wt and axJ mice. (B) Total and phosphorylated
MAPKs MEK1, JNK1/2 and ERK1/2 were examined in hippocampal extracts from 4-week-old wt and axJ mice. (C) Graph depicts the quantitation of
proteins from blots in A and B. n = 3 to 4 mice per genotype. doi:10.1371/journal.pone.0047884.g005 Figure 5. Immunoblot analysis of putative tau kinases in axJ and wt mice. Tau reduction does not alter disease progression in the
axJ mice In
addition, we also detected PHF-1-reactive tau in discrete focal
regions in the axJ Purkinje cell axons. These changes in the
phosphorylation and distribution of tau are hallmarks of AD,
indicating that altered proteasomal function could be a contrib-
uting factor in the development of tau pathologies. Aberrant phosphorylation of tau
has been suggested to
contribute to the accumulation of tau in neurons [51], and
numerous protein kinases have been shown to interact with and
phosphorylate tau [52]. For example, GSK3b can phosphorylate October 2012 | Volume 7 | Issue 10 | e47884 7 PLOS ONE | www.plosone.org PLOS ONE | www.plosone.org Effect of Usp14 Deficiency on Tau Levels in Mice Figure 6. Effect of tau reduction on development and survival of axJ mice. (A) Four-month survival curves of wt, axJ, tauKO and axJtauKO
mice. n = 4 mice per genotype. (B) Body weights of 4 and 8-week-old wt, axJ, tauKO and axJtauKO mice. n = 4 mice per genotype. (C) Purkinje cell axonal
swellings were quantitated in wt, axJ, tauKO and axJtauKO mice. n = 4 mice per genotype. ***indicates p#1027 (D) Paired-pulse facilitation was
measured at hippocampal CA3-CA1 synapses. Facilitation was measured at 5 interpulse intervals ranging from 50 to 500 ms. n = 4 mice per genotype. doi:10.1371/journal.pone.0047884.g006
p
y Figure 6. Effect of tau reduction on development and survival of axJ mice. (A) Four-month survival curves of wt, axJ, tauKO and axJtauKO
mice. n = 4 mice per genotype. (B) Body weights of 4 and 8-week-old wt, axJ, tauKO and axJtauKO mice. n = 4 mice per genotype. (C) Purkinje cell axonal
swellings were quantitated in wt, axJ, tauKO and axJtauKO mice. n = 4 mice per genotype. ***indicates p#1027 (D) Paired-pulse facilitation was
measured at hippocampal CA3-CA1 synapses. Facilitation was measured at 5 interpulse intervals ranging from 50 to 500 ms. n = 4 mice per genotype. doi:10.1371/journal.pone.0047884.g006 widespread cellular changes in the axJ mice, including effects on
protein and glycogen synthesis, apoptosis, and intracellular-
signaling pathways such as WNT. axJ mice. Although there is no evidence that activated ERK or
JNK directly phosphorylate tau in vivo, several studies have
demonstrated that altered proteasome activity leads to the
activation of MAPKs [20]. References 25. Speed HE, Dobrunz LE (2009) Developmental Changes in Short-Term
Facilitation Are Opposite at Temporoammonic Synapses Compared to Schaffer
Collateral Synapses onto CA1 Pyramidal Cells. Hippocampus 19: 187–204. 1. Glickman MH, Ciechanover A (2002) The ubiquitin-proteasome proteolytic
pathway: destruction for the sake of construction. Physiol Rev 82: 373–428. 2. Bingol B, Sheng M (2011) Deconstruction for reconstruction: the role of
proteolysis in neural plasticity and disease. Neuron 69: 22–32. 26. Otvos L, Jr., Feiner L, Lang E, Szendrei GI, Goedert M, et al. (1994)
Monoclonal antibody PHF-1 recognizes tau protein phosphorylated at serine
residues 396 and 404. J Neurosci Res 39: 669–673. 3. Burke RE (2004) Recent advances in research on Parkinson disease: synuclein
and parkin. Neurologist 10: 75–81. 27. Grundke-Iqbal I, Iqbal K, Quinlan M, Tung YC, Zaidi MS, et al. (1986)
Microtubule-associated protein tau. A component of Alzheimer paired helical
filaments. J Biol Chem 261: 6084–6089. 4. Greer PL, Hanayama R, Bloodgood BL, Mardinly AR, Lipton DM, et al. (2010)
The Angelman Syndrome protein Ube3A regulates synapse development by
ubiquitinating arc. Cell 140: 704–716. 28. Gotz J, Ittner LM, Kins S (2006) Do axonal defects in tau and amyloid precursor
protein transgenic animals model axonopathy in Alzheimer’s disease? Journal of
Neurochemistry 98: 993–1006. q
g
5. Wilkinson KD, Urban MK, Haas AL (1980) Ubiquitin is the ATP-dependent
proteolysis factor I of rabbit reticulocytes. J Biol Chem 255: 7529–7532. 5. Wilkinson KD, Urban MK, Haas AL (1980) Ubiquitin is the proteolysis factor I of rabbit reticulocytes. J Biol Chem 255: 7 6. Dawson TM, Dawson VL (2003) Molecular pathways of neurodegeneration in
Parkinson’s disease. Science 302: 819–822. 29. Biernat J, Gustke N, Drewes G, Mandelkow EM, Mandelkow E (1993)
Phosphorylation of Ser262 strongly reduces binding of tau to microtubules:
distinction between PHF-like immunoreactivity and microtubule binding. Neuron 11: 153–163. 7. McNaught KS, Olanow CW, Halliwell B, Isacson O, Jenner P (2001) Failure of
the ubiquitin-proteasome system in Parkinson’s disease. Nat Rev Neurosci 2:
589–594. 30. Zubenko GS, Moossy J, Martinez AJ, Rao GR, Kopp U, et al. (1989) A brain
regional analysis of morphologic and cholinergic abnormalities in Alzheimer’s
disease. Arch Neurol 46: 634–638. 8. Hanna J, Finley D (2007) A proteasome for all occasions. FEBS Lett 581: 2854–
2861. 9. Finley D (2009) Recognition and processing of ubiquitin-protein conjugates by
the proteasome. Annu Rev Biochem 78: 477–513. 31. References Ahlijanian MK, Barrezueta NX, Williams RD, Jakowski A, Kowsz KP, et al. (2000) Hyperphosphorylated tau and neurofilament and cytoskeletal disruptions
in mice overexpressing human p25, an activator of cdk5. Proc Natl Acad
Sci U S A 97: 2910–2915. 10. Lee MJ, Lee BH, Hanna J, King RW, Finley D (2011) Trimming of ubiquitin
chains by proteasome-associated deubiquitinating enzymes. Mol Cell Proteomics
10: R110 003871. 32. Wagner U, Utton M, Gallo JM, Miller CC (1996) Cellular phosphorylation of
tau by GSK-3 beta influences tau binding to microtubules and microtubule
organisation. J Cell Sci 109 (Pt 6): 1537–1543. 11. Guterman A, Glickman MH (2004) Deubiquitinating enzymes are IN/(trinsic to
proteasome function). Curr Protein Pept Sci 5: 201–211. 12. Crimmins S, Jin Y, Wheeler C, Huffman AK, Chapman C, et al. (2006)
Transgenic rescue of ataxia mice with neuronal-specific expression of ubiquitin-
specific protease 14. J Neurosci 26: 11423–11431. 33. Lucas JJ, Hernandez F, Gomez-Ramos P, Moran MA, Hen R, et al. (2001)
Decreased nuclear beta-catenin, tau hyperphosphorylation and neurodegener-
ation in GSK-3beta conditional transgenic mice. Embo J 20: 27–39. 13. Wilson SM, Bhattacharyya B, Rachel RA, Coppola V, Tessarollo L, et al. (2002)
Synaptic defects in ataxia mice result from a mutation in Usp14, encoding a
ubiquitin-specific protease. Nat Genet 32: 420–425. 34. Tatebayashi Y, Planel E, Chui DH, Sato S, Miyasaka T, et al. (2006) c-jun N-
terminal kinase hyperphosphorylates R406W tau at the PHF-1 site during
mitosis. FASEB J 20: 762–764. 14. Lee BH, Lee MJ, Park S, Oh DC, Elsasser S, et al. (2010) Enhancement of
proteasome activity by a small-molecule inhibitor of USP14. Nature 467: 179–
184. 35. Jackson GR, Wiedau-Pazos M, Sang TK, Wagle N, Brown CA, et al. (2002)
Human wild-type tau interacts with wingless pathway components and produces
neurofibrillary pathology in Drosophila. Neuron 34: 509–519. 15. Hanna J, Hathaway NA, Tone Y, Crosas B, Elsasser S, et al. (2006)
Deubiquitinating enzyme Ubp6 functions noncatalytically to delay proteasomal
degradation. Cell 127: 99–111. 36. Noble W, Olm V, Takata K, Casey E, O M, et al. (2003) Cdk5 is a key factor in
tau aggregation and tangle formation in vivo. Neuron 38: 555–565. 37. Ploia C, Antoniou X, Sclip A, Grande V, Cardinetti D, et al. (2011) JNK plays a
key role in tau hyperphosphorylation in Alzheimer’s disease models. J Alzheimers
Dis 26: 315–329. g
16. References Mayer RJ, Lowe J, Lennox G, Doherty F, Landon M (1989) Intermediate
filaments and ubiquitin: a new thread in the understanding of chronic
neurodegenerative diseases. Prog Clin Biol Res 317: 809–818. 38. Yoon SY, Park JS, Choi JE, Choi JM, Lee WJ, et al. (2010) Rosiglitazone
reduces tau phosphorylation via JNK inhibition in the hippocampus of rats with
type 2 diabetes and tau transfected SH-SY5Y cells. Neurobiol Dis 40: 449–455. 17. Kosik KS, Joachim CL, Selkoe DJ (1986) Microtubule-associated protein tau
(tau) is a major antigenic component of paired helical filaments in Alzheimer
disease. Proc Natl Acad Sci U S A 83: 4044–4048. 39. Harris FM, Brecht WJ, Xu Q, Mahley RW, Huang Y (2004) Increased tau
phosphorylation in apolipoprotein E4 transgenic mice is associated with
activation of extracellular signal-regulated kinase: modulation by zinc. J Biol
Chem 279: 44795–44801. 18. Paul S (2008) Dysfunction of the ubiquitin-proteasome system in multiple disease
conditions: therapeutic approaches. Bioessays 30: 1172–1184. 19. Shi YY, Small GW, Orlowski RZ (2006) Proteasome inhibitors induce a p38
mitogen-activated protein kinase (MAPK)-dependent anti-apoptotic program
involving MAPK phosphatase-1 and Akt in models of breast cancer. Breast
Cancer Res Treat 100: 33–47. 40. Pei JJ, Braak H, An WL, Winblad B, Cowburn RF, et al. (2002) Up-regulation of
mitogen-activated protein kinases ERK1/2 and MEK1/2 is associated with the
progression of neurofibrillary degeneration in Alzheimer’s disease. Brain Res
Mol Brain Res 109: 45–55. 20. Zhang L, Ebenezer PJ, Dasuri K, Bruce-Keller AJ, Liu Y, et al. (2009)
Proteasome inhibition modulates kinase activation in neural cells: relevance to
ubiquitination, ribosomes, and survival. J Neurosci Res 87: 3231–3238. 41. Morris M, Maeda S, Vossel K, Mucke L (2011) The Many Faces of Tau. Neuron 70: 410–426. 21. Um JW, Im E, Park J, Oh Y, Min B, et al. (2010) ASK1 negatively regulates the
26 S proteasome. J Biol Chem 285: 36434–36446. 42. Roberson ED, Scearce-Levie K, Palop JJ, Yan F, Cheng IH, et al. (2007)
Reducing endogenous tau ameliorates amyloid beta-induced deficits in an
Alzheimer’s disease mouse model. Science 316: 750–754. p
J
22. Jantas D, Lorenc-Koci E, Kubera M, Lason W (2011) Neuroprotective effects of
MAPK/ERK1/2 and calpain inhibitors on lactacystin-induced cell damage in
primary cortical neurons. Neurotoxicology 32: 845–856. 43. Roberson ED, Halabisky B, Yoo JW, Yao J, Chin J, et al. Tau reduction does not alter disease progression in the
axJ mice Therefore, while increased tau
phosphorylation caused by the activation of these signaling
pathways does not contribute to pathogenesis, these altered The ERK, JNK and p38 MAPK pathways are activated in
neurons of patients with AD, which suggests that the MAPK
pathways are involved in the pathogenesis of AD [54]. Our data
indicate that the ERK and JNK pathways are also activated in the October 2012 | Volume 7 | Issue 10 | e47884 PLOS ONE | www.plosone.org 8 Effect of Usp14 Deficiency on Tau Levels in Mice signaling pathways may affect synaptic function in the axJ mice
through a tau-independent mechanism. synaptic plasticity were not due to changes in tau. Even in the
context of neurological disease, the presence of phosphorylated tau
may only represent a marker of neuronal stress, as opposed being
to a major factor contributing to neurological disease. Hyperphosphorylation of tau has been reported in many animal
models of neurological disease [31,33,55,56], but the relationship
between increased levels of hyperphosphorylated tau and disease is
not clear. Tau reduction provides a protective effect in several
models of neurodegeneration, supporting the notion that tau can
be deleterious to neurons. For example, genetic ablation of tau in
an animal model of AD prevented the behavioral deficits and
neuronal excitotoxicity induced by beta amyloid [42,43]. Analysis
of axJ mice with a genetic ablation of tau demonstrated that the
reduced life span, Purkinje cell axonal swellings, and altered Author Contributions Conceived and designed the experiments: SMW EDR LED RS YNJ GVJ. Performed the experiments: YNJ PC J.A. Watson BJW SEP KG J.A. Wilson. Analyzed the data: SEP SMW J.A. Watson. Contributed reagents/
materials/analysis tools: EDR GVJ. Wrote the paper: J.A. Wilson SMW. 55. Asuni AA, Perry VH, O’Connor V (2010) Change in tau phosphorylation
associated with neurodegeneration in the ME7 model of prion disease. Biochem
Soc Trans 38: 545–551. References (2011) Amyloid-beta/
Fyn-induced synaptic, network, and cognitive impairments depend on tau levels
in multiple mouse models of Alzheimer’s disease. J Neurosci 31: 700–711. 23. Zhu X, Lee HG, Raina AK, Perry G, Smith MA (2002) The role of mitogen-
activated protein kinase pathways in Alzheimer’s disease. Neurosignals 11: 270–
281. 44. Walters BJ, Campbell SL, Chen PC, Taylor AP, Schroeder DG, et al. (2008)
Differential effects of Usp14 and Uch-L1 on the ubiquitin proteasome system
and synaptic activity. Mol Cell Neurosci 39: 539–548. 24. Dawson HN, Ferreira A, Eyster MV, Ghoshal N, Binder LI, et al. (2001)
Inhibition of neuronal maturation in primary hippocampal neurons from tau
deficient mice. J Cell Sci 114: 1179–1187. 45. Lam YA, Xu W, DeMartino GN, Cohen RE (1997) Editing of ubiquitin
conjugates by an isopeptidase in the 26S proteasome. Nature 385: 737–740. PLOS ONE | www.plosone.org October 2012 | Volume 7 | Issue 10 | e47884 October 2012 | Volume 7 | Issue 10 | e47884 9 Effect of Usp14 Deficiency on Tau Levels in Mice Effect of Usp14 Deficiency on Tau Levels in Mice Effect of Usp14 Deficiency on Tau Levels in Mice 46. Berger Z, Ravikumar B, Menzies FM, Oroz LG, Underwood BR, et al. (2006)
Rapamycin alleviates toxicity of different aggregate-prone proteins. Hum Mol
Genet 15: 433–442. 51. Grundke-Iqbal I, Iqbal K, Tung YC, Quinlan M, Wisniewski HM, et al. (1986)
Abnormal phosphorylation of the microtubule-associated protein tau (tau) in
Alzheimer cytoskeletal pathology. Proc Natl Acad Sci U S A 83: 4913–4917. y
p
gy
52. Dolan PJ, Johnson GV (2010) The role of tau kinases in Alzheimer’s disease. Curr Opin Drug Discov Devel 13: 595–603. 47. Biernat J, Mandelkow EM, Schroter C, Lichtenberg-Kraag B, Steiner B, et al. (1992) The switch of tau protein to an Alzheimer-like state includes the
phosphorylation of two serine-proline motifs upstream of the microtubule
binding region. Embo J 11: 1593–1597. 53. Scales TME, Lin S, Kraus M, Goold RG, Gordon-Weeks PR (2009) Nonprimed
and DYRK1A-primed GSK3 beta-phosphorylation sites on MAP1B regulate
microtubule dynamics in growing axons. J Cell Sci 122: 2424–2435. 48. Porzig R, Singer D, Hoffmann R (2007) Epitope mapping of mAbs AT8 and
Tau5 directed against hyperphosphorylated regions of the human tau protein. Biochemical and Biophysical Research Communications 358: 644–649. 54. Zhu XW, Castellani RJ, Takeda A, Nunomura A, Atwood CS, et al. (2001)
Differential activation of neuronal ERK, JNK/SAPK and p38 in Alzheimer
disease: the ‘two hit’ hypothesis. Mechanisms of Ageing and Development 123:
39–46. 49. Goedert M, Jakes R, Crowther RA, Cohen P, Vanmechelen E, et al. (1994)
Epitope mapping of monoclonal antibodies to the paired helical filaments of
Alzheimer’s disease: identification of phosphorylation sites in tau protein. Biochem J 301 (Pt 3): 871–877. 55. Asuni AA, Perry VH, O’Connor V (2010) Change in tau phosphorylation
associated with neurodegeneration in the ME7 model of prion disease. Biochem
Soc Trans 38: 545–551. 50. Trinczek B, Biernat J, Baumann K, Mandelkow EM, Mandelkow E (1995)
Domains of tau protein, differential phosphorylation, and dynamic instability of
microtubules. Mol Biol Cell 6: 1887–1902. 56. Treiber-Held S, Distl R, Meske V, Albert F, Ohm TG (2003) Spatial and
temporal distribution of intracellular free cholesterol in brains of a Niemann-
Pick type C mouse model showing hyperphosphorylated tau protein. Implications for Alzheimer’s disease. J Pathol 200: 95–103. PLOS ONE | www.plosone.org October 2012 | Volume 7 | Issue 10 | e47884 PLOS ONE | www.plosone.org 10
|
https://openalex.org/W2774619557
|
https://link.springer.com/content/pdf/10.1007%2Fs40037-017-0390-9.pdf
|
English
| null |
Hospital-wide education committees and high-quality residency training: A&nbsp;qualitative study
|
Perspectives on medical education
| 2,017
|
cc-by
| 6,512
|
1
Professional Performance Research group, Department for
Educational Support, Academic Medical Center/University of
Amsterdam, Amsterdam, The Netherlands 2
Department of Educational Development and Research,
Maastricht University, Maastricht, The Netherlands Milou E. W. M. Silkens
m.e.silkens@amc.uva.nl Abstract Keywords Postgraduate medical education · Residency training · Educational governance · Hospital-wide education
committees Keywords Postgraduate medical education · Residency training · Educational governance · Hospital-wide education
committees A qualitative study Milou E. W. M. Silkens1
· Irene A. Slootweg1 · Albert J. J. A. Scherpbier2 · Maas Jan Heineman1 ·
Kiki M. J. M. H. Lombarts1 Milou E. W. M. Silkens1
· Irene A. Slootweg1 · Albert J. J. A. Scherpbier2 · Maas Jan Heineman1 ·
Kiki M. J. M. H. Lombarts1 Published online: 11 December 2017
© The Author(s) 2017. This article is an open access publication. Published online: 11 December 2017
© The Author(s) 2017. This article is an open access publication. Abstract Introduction
High-quality residency training is of utmost importance for residents to become competent medical spe-
cialists. Hospital-wide education committees have been adopted by several healthcare systems to govern postgraduate
medical education and to support continuous quality improvement of residency training. To understand the functioning
and potential of such committees, this study examined the mechanisms through which hospital-wide education committees
strive to enable continuous quality improvement in residency training. Methods
Focus group studies with a constructivist grounded theory approach were performed between April 2015 and
August 2016. A purposeful sample of hospital-wide education committees led to seven focus groups. August 2016. A purposeful sample of hospital-wide education committees led to seven focus groups. Results
Hospital-wide education committees strived to enable continuous quality improvement of residency training by
the following mechanisms: creating an organization-wide quality culture, an organization-wide quality structure and by
collaborating with external stakeholders. However, the committees were first and foremost eager to claim a strategic
position within the organization they represent. All identified mechanisms were interdependent and ongoing. Discussion
From a governance perspective, the position of hospital-wide education committees in the Netherlands is
uniquely contributing to the call for institutional accountability for the quality of residency training. When implementing
hospital-wide education committees, shared responsibility of the committees and the departments that actually provide
residency training should be addressed. Although committees vary in the strategies they use to impact continuous quality
improvement of residency training they increasingly have the ability to undertake supporting actions and are working step Results
Hospital-wide education committees strived to enable continuous quality improvement of residency training by
the following mechanisms: creating an organization-wide quality culture, an organization-wide quality structure and by
collaborating with external stakeholders. However, the committees were first and foremost eager to claim a strategic
position within the organization they represent. All identified mechanisms were interdependent and ongoing. Discussion
From a governance perspective, the position of hospital-wide education committees in the Netherlands is
uniquely contributing to the call for institutional accountability for the quality of residency training. When implementing
hospital-wide education committees, shared responsibility of the committees and the departments that actually provide
residency training should be addressed. Although committees vary in the strategies they use to impact continuous quality
improvement of residency training, they increasingly have the ability to undertake supporting actions and are working step
by step to contribute to high-quality postgraduate medical education. https://doi.org/10.1007/s40037-017-0390-9
Perspect Med Educ (2017) 6:396–404 https://doi.org/10.1007/s40037-017-0390-9
Perspect Med Educ (2017) 6:396–404 ORIGINAL ARTICLE Setting In the Netherlands residency training is organized in eight
geographical regions, all consisting of a coordinating aca-
demic hospital and regional affiliated teaching hospitals. Regional affiliated hospitals are either top clinical hospitals
(providing PGME, scientific research and specialized care;
a top clinical qualification can be achieved by hospitals ful-
filling requirements set by the Organization of Top Clinical
Hospitals) or general hospitals (providing PGME and pa-
tient care). Residents work in a clinical department and are
trained and supervised by a team of clinical teachers jointly
responsible for training. Each training program is headed by
a program director appointed by the Royal Dutch Medical
Association. The Dutch College for Medical Specialties (the legisla-
tive body for PGME accreditation) officially launched the
modernization program of PGME in 2003 [13]. Since 2011,
the Directive of the Central College of Medical Specialists,
which defines rules and regulations for (i) the curriculum
of residency training, (ii) the registration of program di-
rectors and (iii) the training institute, mandates teaching
hospitals to have an operational HEC responsible for mon-
itoring and promoting the quality of residency training [6]. Without such an HEC, hospitals cannot be assigned the sta-
tus of teaching hospital. The core tasks of HECs include the
protection of individual residents, facilitating high-quality
residency training for the collective group of residents in
the teaching hospital, and guaranteeing that training pro-
grams have supportive learning climates [6]. The directive
states that HEC members should represent all residency
program directors, residents and the hospital board. The
internal quality management performed by HECs is addi-
tional and supportive to the accreditation process performed
every 5 years by external auditors. When it comes to patient care, the concept of institutional
accountability is not new to the medical world. Clinical gov-
ernance is a well-established vehicle through which health
organizations as a whole are accountable for the quality
of patient care [10]. Clinical governance aims to integrate
fragmented CQI efforts for patient care into a centralized in-
ternal quality management system [11]. Similarly, PGME
should entail a solid internal quality management system
that facilitates educational CQI to benefit the quality of res-
idency training. Healthcare systems such as in the United
States and the Netherlands have contributed to centralized
internal systems for PGME by requiring teaching hospitals
to install hospital-wide education committees (HECs) for
overseeing the quality of residency programs. What this paper adds healthcare systems have implemented hospital-wide educa-
tion committees (HECs) to govern quality improvement of
residency training at the organizational level. This paper
explored the mechanisms through which HECs strive to
enable continuous quality improvement in residency train-
ing and thereby added to current insights on the potential of
institutional accountability. The main message for health-
care systems that implemented or wish to implement HECs
is: HECs can contribute to the quality of residency training,
but a strategic position from which they have the authority
to intervene is crucial to their success. The traditional specialty-specific organization of residency
training lacks institutional accountability. Therefore, several Milou E. W. M. Silkens
m.e.silkens@amc.uva.nl 397 Education committees in medical education Setting These com-
mittees monitor residency training and support CQI efforts
for PGME within the hospital. A case study by Curry et al. (2008) reports positive ef-
fects for educational quality as a result of installing an HEC,
showing HECs can guide changes in residency programs,
closing of unsuccessful programs and development of new
programs [12]. Despite these promising observations, com-
prehensive insight into mechanisms through which HECs
may contribute to PGME is missing. This study addresses
this knowledge gap by answering the following research
question: what are the mechanisms through which hospi-
tal-wide education committees strive to enable continuous Introduction quality improvement in postgraduate medical education? In
answering this question, we aim to gain a deeper under-
standing of the value of HECs for the quality of residency
training. quality improvement in postgraduate medical education? In
answering this question, we aim to gain a deeper under-
standing of the value of HECs for the quality of residency
training. High-quality postgraduate medical education (PGME) is
of utmost importance for residents to become competent
medical specialists [1]. During the worldwide moderniza-
tion of medical education, methods for continuous quality
improvement of residency training (CQI) (e. g. the Plan-
Do-Check-Act cycle (PDCA cycle)) [2] have been intro-
duced to assure and improve the quality of PGME. Al-
though such CQI efforts have shown promising improve-
ments in teaching performance and learning climate [3, 4],
the traditional specialty-specific organization of residency
training means CQI of PGME does not receive attention
at the organizational level. Therefore, exchange of poli-
cies, facilities and best practices for CQI of PGME are not
shared within a teaching hospital, leaving unfulfilled po-
tential to improve PGME quality. For this reason, bodies
such as the American Accreditation Council on Graduate
Medical Education (ACGME), the British General Medical
Council (GMC) and the Royal Dutch Medical Association
are calling for a shift from solely specialty-specific respon-
sibility towards centralized institutional accountability in
which residency training is a shared responsibility of ed-
ucators and the teaching hospital [5–9]. Insight into how
institutional accountability for PGME takes form is absent,
hindering clarity about its potential for the quality of resi-
dency training. Study design 398 Type of hospitala and
geographical regionb
Formal positions of participants (P)
HEC1: top clinical
teaching hospital,
region A
P 1: Coordinating staff for PGME
P 2: Advisor to the HEC
P 3: Chair of the HEC
P 4: Educational supporting staff
P 5: Formal educator at the department
HEC2: general teaching
hospital, region B
P 6: Formal educator at the department
P 7: Formal educator at the department
P 8: Vice formal educator at the departments
P 9: Advisor to the HEC
P 10: Successive chair of the HEC
HEC3: top clinical
teaching hospital,
region C
P 11: Vice chair of the HEC
P 12: Resident representative
P 13: Educational supporting staff
P 14: Hospital board member and formal educator at the department
HEC4: top clinical
teaching hospital,
region D
P 15: Chair of the HEC
P 16: Formal educator at the department and successive chair of the HEC
P 17: Coordinating staff for PGME and formal educator at the department
P 18: Secretary staff
HEC5: academic
teaching hospital,
region B
P 19: Coordinating staff for PGME
P 20: Resident representative
P 21: Vice chair of the HEC
P 22: Formal educator at the department
HEC6: general teaching
hospital, region C
P 23: Educational supporting staff
P 24: Coordinating staff for PGME
P 25: Secretary staff
HEC7: top clinical
teaching hospital,
region E
P 26: Vice formal educator at the department
P 27: Resident representative
P 28: Formal educator at the department
P 29: Chair of the HEC and formal educator at the department
aAcademic hospital (coordinating PGME for affiliated hospitals); top clinical hospital (providing specialized
clinical care, scientific research and PGME); general hospital (providing patient care and PGME)
bCovered a total of 5 out of 8 geographical regions aAcademic hospital (coordinating PGME for affiliated hospitals); top clinical hospital (providing specialized
clinical care, scientific research and PGME); general hospital (providing patient care and PGME)
bCovered a total of 5 out of 8 geographical regions cussions between participants to capture a wide array of
feelings, attitudes and opinions [16, 17]. research to the chairs by email. Chairs then disseminated
this information to the participating HEC members. Based on extensive discussions within the research
group, a discussion guide with five open-ended questions
was designed to structure the focus groups. Two ques-
tions were intended to explore and define internal quality
management systems and the role of the HEC within this
process. Study design We performed a focus group study with a constructivist
grounded theory approach, because of the explorative na-
ture of the study. All researchers participated in the iter-
ative development of the initial research question and the
co-construction of meaning and knowledge during data col-
lection and analysis in order to create an understanding of
HEC functioning [14, 15]. We allowed for interactive dis- M. E. W. M. Silkens et al. Study design Two questions investigated processes and factors
underlying and evolving from internal quality manage-
ment. One wrap-up question asked participants for sug-
gestions for other relevant topics that could be discussed
and whether they could formulate take-home messages
(Table 2). A skilled moderator (IS) facilitated the discus-
sion and an observer (either the first author (MS) or fellow
researchers from the research group) captured nonverbal
communication. All focus groups were audio-taped and
transcribed verbatim. Table 1
Type of hospital from
which hospital-wide educational
committees (HECs) originated
and formal positions of partici-
pants Study population and data collection We purposefully selected a variety of teaching hospitals,
based on type of hospital (academic, top clinical and gen-
eral teaching hospitals), size and geographical location, to
capture a diversity and richness of data ([15]; Table 1). Selection was performed iteratively in which every new
inclusion was driven by previous inclusions. For every se-
lected teaching hospital, chairs of the corresponding HECs
were approached from February 2014 until January 2016 by
a member of the research team (KL) via email to partici-
pate with daily board members of the HEC in a focus group
study. Upon interest, the main researcher (MS) provided ex-
tensive information about the goals and procedures of the Education committees in medical education 399 2
Focus group guide
during the study
Number:
End:
Date:
Moderator
Start:
Observer
Five main questions
Topic lista
What is internal quality
management and what is the role of
the HEC herein? Definition:
– Quality a
performan
– Quality im
– Internal a
Internal qua
Parties invo
– Hospital b
– HEC
– Education
– Departme
– Residents
– External c
Role of HEC
– Promote/
– Birds-eye
– Policy ma
– Take resp
How do you feel about internal
quality management? Positive:
– Good/imp
– Added va
– Supportiv
Negative:
– Bad/takes
– Too much
– Hierarchi
What are achievements of internal
quality management and the HEC’s
effort? Levels of im
– Level of t
– Level of d
ers)
– Level of r
– Level of t
Achievemen
– More awa
– Impact of
power, fin
– Education
– Patient ca
What is needed to make internal
quality management work? What
impairs? Needed:
– Culture (c
exchange
– Systems (
– Commun
– HEC pow
Impairing:
– External p
– Excess w
– Compulsi
– Pressure t
– Lack of fi
– Content o Table 2
Focus group guide
used during the study – Birds-eye view/monitoring – Policy making How do you feel about internal
quality management? – Good/important/enthusiastic/activating p
g
– Added value (improves quality of residency training) – Added value (improves quality of residency training)
– Supportive to external quality management (
p
q
y
y
g)
– Supportive to external quality management Negative: – Bad/takes time/necessity – Too much work/not useful – Too much work/not useful
– Hierarchical – Hierarchical – Hierarchical What are achievements of internal
quality management and the HEC’s
effort? What are achievements of internal
quality management and the HEC’s
effort? What is needed to make internal
quality management work? What
impairs? Study population and data collection Levels of impact: – Level of teaching hospital (hospital board, finances, HEC) – Level of departments (leadership, teaching teams, clinical teach-
ers) – Level of residents (more residents, better residents, satisfaction) – Level of residents (more residents, better residents, satisfaction)
– Level of the patient (change in care) – Level of the patient (change in care) – Level of the patient (change in care) Achievements: – More awareness/attention/interest – Impact of HEC (meetings, content, collaborations, facilities,
power, finance) – Education (better programs, development, innovations)
P i – Education (better programs, development, innovations)
– Patient care What is needed to make internal
quality management work? What
impairs? – Culture (collective vision, everybody on board, representatives,
exchange of best practices) – Systems (routines) – Communication/collaboration/support – HEC power (right to intervene, freedom, trust)
Impairing: – External pressure – Compulsivity/coercion/pressure – Lack of finance/support pp
– Content of some tools used in internal quality management M. E. W. M. Silkens et al. 400 Table 2
Focus group guide used
during the study (Continued) Table 2
Focus group guide used
during the study (Continued) Are there remaining topics that
were not discussed? What is your
take home message? – Innovations
– Hospital merges
– Large scale projects (building, expanding hospital)
– Accreditation
aThe topic list was used only after initial discussion of the main question. In line with the iterative approach
adopted in the study, the topic list was adjusted (mainly extended) for each new focus group, based on topics
addressed in previous focus groups. The topic list was used to introduce points for discussion that were not
mentioned during initial discussion (to broaden the focus group discussion) aThe topic list was used only after initial discussion of the main question. In line with the iterative approach
adopted in the study, the topic list was adjusted (mainly extended) for each new focus group, based on topics
addressed in previous focus groups. The topic list was used to introduce points for discussion that were not
mentioned during initial discussion (to broaden the focus group discussion) A focus group consisted of members of an HEC within
one hospital only. Focus groups were held between April
2015 and August 2016 and lasted a maximum of 75 min. Overall, seven focus group discussions with a total of
29 participants were conducted. No new codes emerged
after the sixth focus group, confirming data saturation at
the seventh focus groups. Study population and data collection The number of participants per
focus group and their formal positions varied (Table 1). tion to build the authority to intervene. This claim for power
was a prerequisite for HECs to be successful in enabling
CQI. All four mechanisms were ongoing and interdepen-
dent and HECs needed to balance these to maximize impact
on the quality of residency training. Mechanism 1: HECs claim a strategic position in the
organization Despite legislation dictating the role of the HEC, we iden-
tified that HEC members had to claim a position within
teaching hospitals. Respondents mentioned the struggle of
prioritizing education in hospitals and of getting a hold as
an education committee in the hospital context. An exam-
ple was provided by HEC 2, stating ‘Until now, it (the value
of PGME) was not mentioned in the strategy of the hospi-
tal’, which illustrates PGME is often under prioritized at
the hospital level. Therefore, HEC members stressed for-
malizing their role in hospitals’ statutory documents and
negotiating their financial independency as crucial to gain-
ing the right to intervene in residency programs in diffi-
culty. Since hospital boards were felt not to prioritize ed-
ucation and sometimes lacked the will to invest in PGME,
respondents identified these boards as the party to negoti-
ate their formal position with. Participants also mentioned
that a solid strategic position enhanced the credibility of
the HEC, which was especially required to guarantee the
success of interventions aimed at residency programs of
colleague medical specialists. Data analysis Data collection, analysis (coding) and interpretation was it-
erative to allow adaptations and refinement of the research
question, focus group guides, sampling strategy and codes
[15]. After each focus group, a debriefing was performed
between the moderator (IS) and the observer to reflect on
the session. The first author (MS) read and open coded
the transcripts. Independently, two researchers from the re-
search group also read and open coded the first two tran-
scripts. Initial codes were then clustered into overarching
codes by using axial and selective coding. All stages of
coding were discussed and codes were adapted until agree-
ment was reached. Furthermore, memos and a logbook were
kept and along with frequent discussions with the research
team guided the process of creating understanding of and
assigning meaning to the data. The research team consisted
of experts with various backgrounds (e. g. doctors, poli-
cymakers, professors), with the first author being a health
scientist. Eventually, 217 codes resulted in 52 organizing
codes grouped into four mechanisms that describe how the
HECs strive to enable educational CQI in residency train-
ing. Software for qualitative research (MAXQDA) was used
to support coding. The achievements of HECs seemed to vary: some HECs
formulated clear policies or obtained their own budgets,
others were still working on the contents of such policy or
were awaiting approval from the hospital board for their
financial independency. Mechanism 2: HECs create an organization-wide
educational CQI culture We found that HECs are striving to enable CQI for PGME
by (i) creating an organization-wide educational CQI cul-
ture, (ii) an organization-wide educational CQI structure
and by (iii) collaborating with external stakeholders. Al-
though we did not see any chronological order in these
mechanisms, we did identify that HECs were foremost ea-
ger to (iv) strategically position themselves in the organiza- HEC members mentioned an organization-wide educational
CQI culture as core to achieving excellent PGME. HECs
were said to work on culture amongst, for example, pro-
gram directors, educators and residents, with the aim to
engage them in CQI. HECs enabled awareness about the
relevance of qualitatively sound education by continuously 401 Education committees in medical education addressing strengths and weaknesses of residency programs
and their implications, but also by facilitating exchange of
best practices between residency programs. The importance
of this exchange for PGME was stressed by HEC 1, stating:
‘Things are shared, even when they are not going well. So
you can learn from each other. But it also means you can be
called upon aspects so you can improve them’. This quote
underlines how sharing practices contributes to improve-
ment of PGME through learning about strengths and weak-
nesses. To facilitate this exchange, HECs supported a psy-
chologically safe environment to achieve openness, trans-
parency, approachability and ‘speaking-up’. HEC members
further expressed the importance of goodwill of the staff
in the teaching hospital, to assure willingness to cooperate
with the HEC. It was explicitly mentioned that HECs will
not start a quality movement on their own and need support
from people throughout the organization. for such a backbone was illustrated by HEC 2, stating: ‘But
that whole PDCA cycle, it requires an investment in human
resources.. You need to have a structure for that’. Therefore,
respondents indicated that human resources are a neces-
sary condition to enable CQI in PGME. To create these re-
sources, respondents collaborated with the hospital board to
acquire funds and to discuss national and international de-
velopments in PGME and patient care. Collaboration with
both educators and residents was necessary to set up appro-
priate data collection concerning the quality of residency
programs as well as to receive performance feedback from
educators and residents themselves. Furthermore, respon-
dents underlined the value of empowering residents in CQI
for PGME, since they are the ‘consumers’ of the education. Mechanism 2: HECs create an organization-wide
educational CQI culture Finally, it was stated that educationalists, along with secre-
tarial staff, often took part in HEC meetings to support and
inform the HEC on educational policy. We noted that some HECs have already succeeded in
creating an open environment in which speaking-up and
transparency had become common practice. These HECs
often had a proactive attitude towards CQI and were con-
tinuously working towards improvement of PGME. Others
had just started to work towards transparency about their
CQI policies. These HECs mainly displayed reactive be-
haviour and intervened only when residency programs had
quality issues. Some HECs deemed themselves successful in imple-
menting approaches such as PDCA cycles at various levels
of the organization, others were mainly working at the level
of the residency programs. Some HECs had access to sup-
porting staff and as a result lighter administrative and exec-
utive loads, whereas other HECs were practically operating
on their own. Mechanism 3: HECs create an organization-wide
educational CQI structure W. M. Silkens et al. 402 Strategy HECs are striving to enable CQI in PGME by (i) creating
an educational CQI culture, (ii) as well as an educational
CQI structure and (iii) by collaborating with external stake-
holders. These mechanisms are continuously supported by
(iv) HECs building a strategic position in the organization. We found these identified mechanisms to strengthen each
other in a positive way, potentially creating a virtuous cycle
towards high-quality PGME. Program directors are members of the HECs. These pro-
gram directors fulfil simultaneously the role of committee
member, ergo monitoring peers providing PGME, and of
clinical educator, thus being colleagues of their peers pro-
viding PGME. This might imply that HEC members are
placed in a hierarchical position towards colleague medical
specialists. As a result, equivalent relationships between
medical specialists may become pressured when members
of the HECs execute their right to speak-up and even inter-
vene in residency programs in which fellow specialists are
involved. A solid strategic position as HEC then becomes
even more important to assure credibility and decisional
power. Implications for research and practice With a call for strong educational governance and increased
local management of PGME [18], we think that research
should focus on the extent to which HEC-like bodies can
meet such requirements. To support the potential of HECs,
committees in the Netherlands are becoming more impor-
tant and are assigned a more prominent leadership role in
PGME. Moreover, the Dutch Registration Committee for
Medical Specialties recently announced its plan for exter-
nal audits of the functioning of the HECs [25]. Research
could investigate whether committees that perform well in
these audits are providing better PGME. For healthcare systems that wish to implement HEC-
like bodies to govern PGME, this study informs on mech-
anisms through which such committees can positively im-
pact PGME. For these healthcare systems, guaranteeing that
HEC-like bodies have a strong strategic position through
which they can exert power and influence on CQI in PGME
is a first important step towards successful HECs. Governance Many countries are initiating stronger and more holistic
governance structures for PGME to improve the quality
of education and to keep up with changes in the clinical
field [18]. In many leading healthcare systems governance
of PGME is set and assessed by external bodies [7, 18,
19]. In Canada, for example, residency programs operate
in a university-based organizational structure from which
quality is governed per specialty [18, 19]. In many of these
systems there is a lack of accountability for teaching hospi-
tals and leaders of these hospitals should be more engaged
in PGME and resulting issues [18]. From this perspective,
HECs in the Netherlands are uniquely contributing to this
issue by creating a central position from which teaching
hospitals can be held accountable for the quality of PGME. This centrality might facilitate exchange of resources, en-
hance interdisciplinary programs and support CQI efforts
to guarantee high-quality PGME. Mechanism 3: HECs create an organization-wide
educational CQI structure HEC members collaborated with external stakeholders
to improve their policy within their own hospital and to
strengthen the position of the teaching hospital itself. Re-
spondents mentioned collaborations with other HECs and
hospitals from which HECs could organize educational
events for residents or educators (e. g. joint trainings). Ex-
ternal coaches were involved in training or departmental
group discussions about the results of quality evaluations. Such coaches were considered to provide impartiality and
anonymity and therefore stimulate speaking-up. Besides creating a culture, respondents mentioned the need
for an educational CQI structure to evaluate and improve
PGME organization-wide and to define when and what ac-
tions concerning PGME are taken. Within this structure,
respondents described the HECs as an overarching body
from which CQI efforts for PGME are monitored and co-
ordinated. As a result, HECs were said to initiate the use of
systematic approaches, such as PDCA cycles, at the level of
individual departments as well as at the organizational level. HMC 6 illustrated this as follows: ‘We are working on (...)
a PDCA cycle at the level of the training programs and (...)
at the level of the teaching hospital as a whole. We have in-
stalled cycles throughout the organization (...) to reach im-
provement in education’. Moreover, respondents mentioned
the importance of encouraging people to commit to CQI for
PGME in order to consolidate quality initiatives. HMC 1
posed: ‘You can make plans, but without the support of clin-
ical teachers and residents those plans are not going to hap-
pen. (...) Making sure people actually commit to long-term
improvement is difficult’. Respondents mentioned that hospitals have the respon-
sibility to guarantee that their residency programs live up
to the standards that are set by external bodies and soci-
ety. HEC 1 illustrated this by stating: ‘If you want a vision
on what should be happening in those postgraduate medi-
cal training programs you will need a voice from the outside
world’. This quote shows that respondents were of the opin-
ion that education should fit societal demands. However,
HECs indicated that priorities set by society are not always
acknowledged by educators or residents and, therefore, the
HEC has to take the lead in incorporating this voice into
their policy. Next, HEC members were found to collaborate with
PGME stakeholders throughout the hospital, which sup-
ported the translation of HEC policy into action. The need M. E. Implementation Implementation of CQI for PGME might be challenging,
since it might be perceived as burdensome, bureaucratic or
as interfering by clinical departments [20–22]. The clinical
environment has proven to be rather unreceptive to educa-
tional policy when education is considered to be secondary
to patient care and medical research [23]. Examining imple-
mentation of similar hospital-wide bodies for clinical gov-
ernance, literature shows the importance of aligning organi-
zational and departmental levels to impact patient outcomes
[24]. Such a cooperation was also stressed by our respon-
dents, stating CQI efforts for PGME will not succeed with-
out involvement of departments. Implementing committees
such as HECs, shared responsibility of the HEC and depart-
ments to achieve high-quality PGME should be addressed. HECs keep the bird’s eye view and intervene when nec-
essary, but medical specialists managing residency training
at departments should take on the final responsibility to
deliver high-quality education. Furthermore, governance of PGME should commit to
maintaining the connection between education and patient
care [18]. Especially since residency training takes place
in the same context as patient care, we imply that those
interested in working towards institutional accountability
for PGME can learn from the experience and knowledge
assimilated by research into clinical governance. We think
collaboration between the two fields might lead to a more
integrated system that can benefit education as well as pa-
tient care. This research refrains from investigating effects of
HECs’ actions, but respondents addressed the positive ef-
fects of CQI on the quality of education (e. g. an increase in
resident applications and significant changes in residency
programs when compared with previous years). Effects on
patient care were mentioned by only one HEC. To decide
whether the implementation of HECs actually contributes 403 Education committees in medical education Ethical approval
The institutional ethics review board of the Aca-
demic Medical Centre of the University of Amsterdam confirmed that
the Medical Research Involving Human Subjects Act (WMO) did
not apply to the current study on the 3 December 2014 (W14_322
# 14.17.0386) and as such provided a waiver for the current study. However, to safeguard our participants, we took several necessary pre-
cautions. We fully informed our participants on the goal, methodology,
advantages and disadvantages and possible consequences of our study
by means of a detailed participant information letter. Conclusions 5. ACGME Institutional Requirements Chicago, IL: Accreditation
Council for Graduate Medical Education. 2014. http://www.acgme. org. Accessed 1 Nov 2014. HECs strive to enable CQI for PGME by creating an organi-
zation-wide educational CQI culture, structure and by col-
laborating with external stakeholders. However, HEC mem-
bers need to claim a strategic position before they are able
to bring CQI for PGME forward. HECs seem to increas-
ingly have the ability to undertake supporting actions and
are working step by step to contribute to high-quality resi-
dency training. 6. Royal Dutch Medical Association. Directive of the Central College
of Medical Specialists. Utrecht: Royal Dutch Medical Association;
2009. In Dutch. 7. General Medical Council. Quality improvement framework for un-
dergraduate and postgraduate medical education and training in the
UK. London: General Medical Council; 2010. 8. Global Standards KH, Medical Education A. A view from the
WFME. Acad Med. 2006;81(12):43–S8. https://doi.org/10.1097/
01.ACM.0000243383.71047.c4. Acknowledgements The authors would like to thank L. Hoogenboom
for her help with transcribing the audio and G. Bressers, MSc, for her
extensive support with coding and interpreting the data as well as her
availability to observe during focus groups. Similarly, the authors thank
M. van den Goor, MD, for making time to observe one of the focus
groups. Additionally, the authors are grateful for the respondents that
freed up their valuable time to participate in this study. 9. Afrin LB, Arana GW, Medio FJ, Ybarra AF, Clarke HS Jr.. Im-
proving oversight of the graduate medical education enterprise: one
institution’s strategies and tools. Acad Med. 2006;81:419–25. 10. Scally G, Donaldson LJ. The NHS’s 50 anniversary. Clinical gov-
ernance and the drive for quality improvement in the new NHS in
England. BMJ. 1998;317:61–5. 11. Halligan A, Donaldson L. Implementing clinical governance: turn-
ing vision into reality. BMJ. 2001;322:1413–7. Conflict of interest
K.M.J.M.H. Lombarts received a grant provided
by the Dutch Ministry of Health, Welfare and Sports to finance,
amongst others, this study. The ministry had no role in the study
design, data collection, analysis, interpretation and reporting of data. M.E.W.M. Silkens, I.A. Slootweg, A.J.J.A. Scherpbier and M.J. Heine-
man declare that they have no competing interests. 12. Curry RH, Burgener AJ, Dooley SL, Christopher RP. Collabora-
tive Governance of Multiinstitutional Graduate Medical Education:
lessons from The McGaw Medical Center of Northwestern Univer-
sity. Acad Med. 2008;83:568–73. y
13. Scheele F, Van Luijk S, Mulder H, et al. Implementation In this letter, we
stressed that participation was voluntary at all times, data were treated
confidentially and published anonymously, and withdrawal from the
study was possible at any time throughout the whole research process
without any consequences for the participants. We repeated this infor-
mation verbally preceding the focus groups. Furthermore, participants
were asked to sign informed consent prior to the focus group or in-
terview, stating they had been informed about and were aware of the
abovementioned details of the study. to the quality of PGME, research into the effects of HECs’
actions is necessary. to the quality of PGME, research into the effects of HECs’
actions is necessary. Strengths and limitations This study was performed in PGME in the Netherlands,
so results are bound to the context in which the data were
gathered. However, HEC-like bodies implemented in other
healthcare systems (e. g. the Graduate Medical Education
Committees in the United States) received comparable roles
and purposes and have to function in a comparable environ-
ment (teaching hospitals). Therefore, we consider the four
identified mechanisms to be relevant to HEC-like bodies
outside the Netherlands. Open Access
This article is distributed under the terms of the
Creative Commons Attribution 4.0 International License (http://
creativecommons.org/licenses/by/4.0/),
which permits unrestricted
use, distribution, and reproduction in any medium, provided you give
appropriate credit to the original author(s) and the source, provide a
link to the Creative Commons license, and indicate if changes were
made. Since this research provides a first insight into how HECs
are striving to enable CQI in PGME, we only focused on the
perspectives of the members of the HECs who seem to be
engaged and enthusiastically involved in CQI for PGME. Therefore, we stress that investigating other perspectives,
such as from hospital boards, medical specialists and resi-
dents, may create additional knowledge about the value of
HECs for PGME. References A strength of this study is the iterative and systematic
approach we used to analyze the data, thereby refraining
from enforcing existing assumptions and models on the
data. This approach contributed to the aim of developing
an understanding of the value of HECs for the quality of
residency training. 1. van der Leeuw RM, Lombarts KM, Arah OA, Heineman MJ. A
systematic review of the effects of residency training on patient out-
comes. Bmc Med. 2012;10:65. 2. Deming WE. Out of the crisis. Cambridge: Massachusetts Institute
of Technology Center for Advanced Engineering Study; 1986. 3. Boerebach BC, Arah OA, Heineman MJ, Busch OR, Lombarts KM. The impact of resident- and self-evaluations on surgeon’s subse-
quent teaching performance. World J Surg. 2014;38:2761–9. 4. Silkens ME, Arah OA, Scherpbier AJ, Heineman MJ, Lombarts
KM. Focus on quality: investigating residents’ learning climate per-
ceptions. PLoS ONE. 2016;11:e147108. Conclusions Is the modernisation of
postgraduate medical training in the Netherlands successful? Views 404 M. E. W. M. Silkens et al. hospital and pathway levels. Int J Qual Health Care. 2014;26(Suppl
1):66–73. of the NVMO Special Interest Group on Postgraduate Medical Ed-
ucation. Med Teach. 2014;36:116–20. )
25. Dutch Registration Committee for Medical Specialties (RGS). Modernisering
visitatieproces
door
instellingsvisitaties
RGS. 2017. https://www.knmg.nl/opleiding-herregistratie-carriere/rgs/
wat-doet-de-rgs/nieuwsbericht-rgs/modernisering-visitatieproces-
door-instellingsvisitaties-rgs.htm. Accessed 30 Apr 2017. 14. Mills J, Bonner A, Francis K. Adopting a constructivist approach to
grounded theory: Implications for research design. Int J Nurs Pract. 2006;12:8–13. 15. Watling CJ, Lingard L. Grounded theory in medical education re-
search: AMEE Guide No. 70. Med Teach. 2012;34:850–61. organ DL. Focus groups. Annu Rev Sociol. 1996;22:129–5 17. Stalmeijer RE, McNaughton N, Van Mook WN. Using focus groups
in medical education research: AMEE Guide No. 91. Med Teach. 2014;36:923–39. Milou E.W.M. Silkens is a PhD student at the Department for Edu-
cational Support in the Academic Medical Center/University of Am-
sterdam. Her research interests mainly focus on learning climates in
postgraduate medical education. 18. Dowton SB, Stokes ML, Rawstron EJ, Pogson PR, Brown MA. Postgraduate medical education: rethinking and integrating a com-
plex landscape. Med J Aust. 2005;182:177–80. Irene A. Slootweg is an advisor at the Department for Educational
Support in the Academic Medical Center and senior researcher for the
Professional Performance Research Group in the Academic Medical
Center/University of Amsterdam. Irene A. Slootweg is an advisor at the Department for Educational
Support in the Academic Medical Center and senior researcher for the
Professional Performance Research Group in the Academic Medical
Center/University of Amsterdam. 19. Canada RCoPaSo. General information concerning accredita-
tion of residency programs [updated June 2006]. 2006. http://
www.royalcollege.ca/portal/page/portal/rc/common/documents/
accreditation/genaccred_e.pdf. Accessed 15 Jun 2017. Albert J.J.A. Scherpbier is professor for the Department of Educa-
tional Development and Research at Maastricht University and dean of
the Faculty of Health, Medicine and Life Sciences at Maastricht Uni-
versity. 20. Skochelak SE. A decade of reports calling for change in medical ed-
ucation: what do they say? Acad Med. 2010;85(9 Suppl):S26–S33. 21. Centers CFTFo A. Training Tomorrow’s Doctors: the Medical Ed-
ucation Mission of Academic Health Centers: a Report of the Com-
monwealth Fund Task Force on Academic Health Centers: Com-
monwealth Fund. 2002. Conclusions Maas Jan Heineman is professor for the Department of Obstetrics
and Gynecology in the Academic Medical Center and vice-dean of
the Faculty of Medicine at the Academic Medical Center/University
of Amsterdam. 22. Cooke M, Irby DM, Sullivan W, Ludmerer KM. American med-
ical education 100 years after the Flexner report. N Engl J Med. 2006;355:1339–44. 23. Engbers R, Fluit CR, Bolhuis S, de Visser M, Laan RF. Implement-
ing medical teaching policy in university hospitals. Adv Health Sci
Educ Theory Pract. 2017;22:985–1009. Kiki M.J.M.H. Lombarts is professor of Professional Performance
for the Department of Educational Support in the Academic Medical
Center/University of Amsterdam. 24. Wagner C, Groene O, Thompson CA, et al. DUQuE quality man-
agement measures: associations between quality management at
|
https://openalex.org/W4247151094
|
https://bmcmusculoskeletdisord.biomedcentral.com/track/pdf/10.1186/s12891-020-03448-2
|
English
| null |
The risk of fracture and prevalence of osteoporosis is elevated in patients with idiopathic inflammatory myopathies: cross-sectional study from a single Hungarian center
|
Research Square (Research Square)
| 2,020
|
cc-by
| 6,381
|
The risk of fracture and prevalence of
osteoporosis is elevated in patients with
idiopathic inflammatory myopathies: cross-
sectional study from a single Hungarian
center Anett Vincze1,2, Levente Bodoki3,2, Katalin Szabó1,2, Melinda Nagy-Vincze1,2, Orsolya Szalmás4, József Varga5,
Katalin Dankó1,2, János Gaál1,6,2 and Zoltán Griger1,2* (2020) 21:426 (2020) 21:426 Vincze et al. BMC Musculoskeletal Disorders
https://doi.org/10.1186/s12891-020-03448-2 Open Access Abstract Background: The prevalence of osteoporosis and risk of fractures is elevated in rheumatoid arthritis (RA), but we
have limited information about the bone mineral density (BMD) and fracture risk in patients with inflammatory
myopathies. We intended to ascertain and compare fracture risk, bone mineral density and the prevalence of
vertebral fractures in patients with inflammatory myositis and rheumatoid arthritis and to assess the effect of
prevalent fractures on the quality of life and functional capacity. Methods: Fifty-two patients with myositis and 43 patients with rheumatoid arthritis were included in the study. Fracture Risk was determined using FRAX® Calculation Tool developed by the University of Sheffield. Dual energy X-
ray absorptiometry and bidirectional thoracolumbar radiographs were performed to assess BMD and vertebral
fractures. Quality of life was measured with Short Form-36 (SF-36) and physical function assessment was performed
using Health Assessment Questionnaire (HAQ). Results: We found a significantly elevated fracture risk in RA as compared to myositis patients if the risk assessment
was performed without the inclusion of the BMD results. If BMD results and glucocorticoid dose adjustment were
taken into account, the differences in fracture risk were no longer significant. The prevalence of osteoporosis was
found to be significantly higher in the myositis group (7% vs. 13.5%, p: 0.045), but the fracture prevalence was similar in
the two groups (75% vs. 68%). The fracture rates were independently associated with age in the myositis group, and
with lower BMD results in the RA patients. The number of prevalent fractures was significantly correlated to poorer
physical function in both groups, and poorer health status in the myositis group, but not in the RA group. (C
i
d
) (Continued on next page) * Correspondence: grigerz@gmail.com
1Division of Clinical Immunology, Faculty of Medicine, University of
Debrecen, Móricz Zsigmond út 22, Debrecen H-4032, Hungary
2Gyula Petrányi Doctoral School of Clinical Immunology and Allergology,
University of Debrecen, Debrecen, Hungary
Full list of author information is available at the end of the article Background Osteoporosis is a common metabolic skeletal disorder
characterized by decreased bone mass and deteriorated
bone structure, leading to increased fracture rate [1]. Since the average age and the proportion of elderly per-
sons in the population is increasing continuously, osteo-
porosis and consecutive fractures have become a global
public health problem with enormous socioeconomic
consequences [2, 3]. It is widely known that rheumatoid
arthritis (RA) is one of the most important causes of sec-
ondary osteoporosis. Osteoporotic bone fractures are of
crucial importance in the functioning and quality of life
of patients. The pathogenesis of bone loss in auto-
immune disorders is multifactorial. The increased serum
and tissue levels of pro-inflammatory mediators lead to
the increased expressions of receptor activator of nuclear
factor
kappa-Β
ligand (RANKL)
by osteoblasts,
T-
lymphocytes and synovial fibroblasts. The RANKL-
RANK binding is the main pathogenetic event of osteo-
clastogenesis, osteoclast maturation and functioning. Another, not at all negligible reason, is the synergism of
several factors that negatively affect the bone mass: diet-
ary factors (decreased calcium /Ca/ and vitamin D3 in-
take), decreased muscle mass/strength and functional
capacity, immobilization, deficient intestinal Ca absorp-
tion, reduced levels of sexual steroids, avoidance of sun-
light and use of sunscreens and, last but not least,
glucocorticoid (GC) use [4–10]. The chronic GC expos-
ure leads to decreased calcium absorption, increased
renal Ca loss, secondary hyperparathyroidism, decreased
sexual hormone levels, decreased number and function
of osteoblasts and eventually increased bone resorption
and reduced bone formation [11, 12]. In patients with
RA, systemic osteoporosis coincides with local bone re-
sorption, as a typical consequence of inflammatory syno-
vitis. Vertebral fractures are important but yet under
recognized manifestations of osteoporosis. Most of them
are asymptomatic, which makes their recognition more
difficult, consequently, they might remain unnoticed for
years. Clinical observations show that 30% of patients
taking steroids for more than 3 years suffer from an
osteoporotic fracture. Moreover, literature data also
demonstrated that patients with polymyositis (PM) or
dermatomyositis (DM) had significantly lower BMD in
both the hip and lumbar (L) spine compared to the Our present work is a cross-sectional observational
study, in which we intended to answer the following
questions: 1. What is the prevalence of low BMD, verte-
bral fracture and high fracture risk in our patients with
inflammatory myositis and RA? 2. Which factors are as-
sociated with higher fracture rates in myositis and RA
patients? Background 3. How do the vertebral fractures influence the
physical function and quality of life of patients? (Continued from previous page) (Continued from previous page)
Conclusions: Our findings suggest that inflammatory myopathies carry significantly elevated risks for osteoporosis and
fractures. These higher risks are comparable to ones detected with RA in studies and strongly affect the physical
function and quality of life of patients. Therefore further efforts are required to make the fracture risk assessment
reliable and to facilitate the use of early preventive treatments. Keywords: Fracture risk, Vertebral fractures, Myositis, Rheumatoid arthritis healthy, age- and gender-matched population [10, 13]. The Fracture Risk Assessment Tool (FRAX®) developed
and validated by Kanis et al., more than 10 years ago, is
still the most widely accepted and used method in clin-
ical practice to estimate the 10-year probability of osteo-
porotic fractures [14]. FRAX score takes into account
the relevant risk factors for a bone fracture, e.g., the
presence of RA, but not myositis. © The Author(s). 2020 Open Access This article is licensed under a Creative Commons Attribution 4.0 International License,
which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give
appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if
changes were made. The images or other third party material in this article are included in the article's Creative Commons
licence, unless indicated otherwise in a credit line to the material. If material is not included in the article's Creative Commons
licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain
permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by/4.0/. The Creative Commons Public Domain Dedication waiver (http://creativecommons.org/publicdomain/zero/1.0/) applies to the
data made available in this article, unless otherwise stated in a credit line to the data. Vincze et al. BMC Musculoskeletal Disorders (2020) 21:426 Page 2 of 8 Methods This scientific cross-sectional study was conducted on
our own initiative, in 52 consecutive patients with myo-
sitis and 43 patients with RA under the care of the Na-
tional Myositis Center, in the Division of Clinical
Immunology, Faculty of Medicine, at the University of
Debrecen between January 2017–June 2018. This study
meets, and is in compliance with all ethical standards of
medicine. Informed consent was obtained from all of the
subjects. This study is ethically compliant and was car-
ried out in compliance with the Declaration of Helsinki. The eligibility criteria were the diagnosis of probable, or
definitive idiopathic inflammatory myopathy (IIM) based
on the Bohan and Peter criteria [15], and rheumatoid
arthritis according to the 2010 American College of
Rheumatology-European League Against Rheumatism
(ACR-EULAR) classification criteria [16]. The patients
with confounders of bone health were excluded from the
study: if the patient took any drug affecting bone mineral
density (including bisphosphonates, thiazide diuretics,
anticoagulants, anticonvulsants, glitazones, etc.) except
for vitamin D3 and Ca, but including secondary osteo-
porosis and those patients suffering from malignancies. In total, 121 patients were included at the start of the
study, and finally 26 individuals were excluded based on
the presence of exclusion criteria, or missing BMD and/
or FRAX data. Page 3 of 8 Page 3 of 8 Vincze et al. BMC Musculoskeletal Disorders (2020) 21:426 Vincze et al. BMC Musculoskeletal Disorders (2020) 21:426 Laboratory tests included the measurements of cal-
cium, alkaline phosphatase, C-reactive protein (CRP),
thyroid-stimulating hormone, serum 25-OH Vitamin D3
levels and bone turnover markers (BTM): (parathyroid
hormone, osteocalcin /OC/, beta-crosslaps, C-terminal
telopeptides of type-I collagen /CTX-I/). Blood sampling
was done after overnight fasting to measure levels of
PTH, OC and CTX-I. Plasma 25-OH-D3 level was ana-
lyzed by high pressure liquid chromatography (HPLC)
using a Jasco HPLC system (Jasco, Tokyo, Japan) and
Bio-Rad reagent kit (Bio-Rad Laboratories, Hercules,
CA, USA). Serum PTH, OC and CTX-I were measured
using electrochemiluminescence immunoassay (Roche
Diagnostics GmbH, Mannheim, Germany). The inter-
assay CV was < 7% for PTH (lower detection limit:
0.127 pmol/L, upper detection limit: 530 pmol/L), < 4%
for OC (lower detection limit: 0.5 μg/L, upper detection
limit: 300 μg/L) and < 7% for CTX-I (lower detection
limit: 0.010 μg/L, upper detection limit: 6 μg/L). individually instructed to undergo X-ray examinations,
on personally scheduled dates, and finally 40 myositis
and 35 RA patients were able to complete the study. Methods On
standard radiographs the Genant’s semi-quantitative as-
sessment was used to evaluate vertebral fractures [21]. Vertebral shape (wedge, concave, or crush) and de-
creases in anterior, posterior, and/or middle vertebral
height (grade 0: no reduction; grade 1: minimal fracture,
20–25% height decrease; grade 2: moderate fracture, 25–
40% height decrease; and grade 3: severe fracture, greater
than 40% height decrease) were determined by two inde-
pendent assessors (A.V. and O.S.). Their differences
were resolved by repeated analysis and by consensus. Statistical analysis was performed with version 26 of
the SPSS software package (IBM Corp., Armonk, NY,
USA). The normality of the distributions in case of con-
tinuous variables was tested using the Shapiro-Wilk test. Normally distributed continuous variables were de-
scribed by mean and standard deviation values (SD). Categorical variables were described using frequencies
(case number) and percentages. For comparing the
groups we used independent samples t-test, or Mann-
Whitney test depending on the distribution. The con-
nection of two scalar variables was characterized by
Spearman’s correlation, while in case of binary variables
we used Fisher’s exact test. To identify the risk factors of
vertebral fracture, we applied stepwise discriminant ana-
lysis (Wilks), also. The multivariate general linear model
was applied to determine the factors that influence HAQ
and SF36 results. P values of less than 0.05 were
regarded as statistically significant. We measured the BMD of the lumbar spine (L1–4 ver-
tebrae) and the left femoral neck by AP-DXA. The scan
was performed with a DPX Pro bone densitometer (GE-
Lunar Radiation Corporation, Madison, WI, USA), ac-
cording to the manufacturer’s protocol. In patients with
a history of a previous hip fracture, hip replacement sur-
gery, or severe joint destruction, we measured bone min-
eral density in the right femoral neck. Osteoporosis was
diagnosed according to the criteria proposed by the
World Health Organization Study Group, when the
BMD was 2.5 or more standard deviations below the
young-adult mean, and osteopenia was diagnosed when
the BMD was between −1 and −2.5 [17]. Results FRAX scores were calculated 1: without the DEXA levels and without steroid
dose adjustment (steroid yes/no only); 2: With DEXA levels without steroid dose adjustment (steroid yes/no only); 3: With DEXA levels and steroid
dose adjustment Table 1 Summary of the most relevant clinical data and FRAX scores of the patients gy
y
p
y
Significances were calculated with independent samples t-test or Mann-Whitney test, according to the distribution. Normality of the distributions was checked
using Shapiro-Wilk test. Data are presented as mean and standard deviation (SD) with normal distribution (a on the upper corner of the variable) and median,
minimum, maximum with non-gaussian distribution (b on the upper corner of the variable); 95% CIoD: 95% Confidence Interval of the Difference (lower-upper). Categorical variables were described using frequency (case number) and percentage. FRAX scores were calculated 1: without the DEXA levels and without steroid
dose adjustment (steroid yes/no only); 2: With DEXA levels without steroid dose adjustment (steroid yes/no only); 3: With DEXA levels and steroid
dose adjustment after adjustment to the dose of glucocorticoids according
to Kanis et al. [22]. With this correction the magnitude
of the difference further decreased: the risk of major
osteoporotic and hip fractures were found to be 9.96%
vs 9.54% (p = 0.884) and 2.46% vs. 2.87% (p = 0.128)
(Table 1). of the patients, respectively, and osteoporosis found in
13.5 and 7% of the patients, respectively, and the differ-
ence in frequency of osteoporosis found to be statisti-
cally significant between the two groups (Fisher’s exact
test, p = 0.045). The fracture risk assessment was calculated first with-
out applying the BMD values. Regarding the other major
and femoral neck fractures the fracture risk in RA pa-
tients was significantly higher than in myositis patients
(15.58% vs. 9.68 and 6.23% vs 3.06%; p = 0.008 and p =
0.022). As a second step, the fracture risk calculation
was repeated, and this time with the BMD values taken
into account, with the earlier significant difference in the
fracture probability disappearing (13.25% vs. 9.44 and
3.57% vs. 2.77%; p = 0.053 and p = 0.811). During the
third step, the fracture risk assessment was performed As previously mentioned 75 patients underwent bidir-
ectional vertebral X-ray examinations, 40 myositis pa-
tients (8 males and 32 females, mean age 60.97 years)
and 35 RA patients (all female, mean age 59.71 years). Results The FRAX, Health Assessment Questionnaire (HAQ)
and Short Form 36 (SF-36) questionnaires were com-
pleted by a personal interviewer. The web-based algo-
rithm at http://www.shef.ac.uk/FRAX® was applied as the
FRAX® algorithm (version 3.6) adapted for Hungary [18]. Special risk factors (age, sex, weight, height, previous
fracture, parental hip fracture, current smoking, GCs,
RA, secondary osteoporosis, alcohol 3 or more units/
day, femoral neck BMD) were recorded into this calcula-
tor, for every single patient. The output was a 10-year
probability of hip fracture and a 10-year probability of a
major osteoporotic fracture (clinical spine, forearm, hip
or shoulder fracture) [18]. We measured quality of life
including mental health with the SF-36 questionnaire
validated for use in Hungary [19]. The assessment of the
patients’ physical function was performed using the
HAQ questionnaire [20]. One hundred and twenty-one Caucasian patients partici-
pated in the study and fullfilled the inclusion criteria,
while 26 patients were excluded based on the presence
of exclusion criteria, or missing BMD, or FRAX results. The final myositis group consisted of 52 patients (9
males and 43 females, with a mean age of 57.46 years),
while the RA group consisted of 43 patients (2 males
and 41 females, with a mean age of 58.58 years). There
were no significant differences between the two groups
in the basic clinical data (including the mean BMD and
25 OH Vitamin D3 level) as indicated in Table 1. The
proportion of patients receiving oral Ca and vitamin D
substitution did not differ significantly between the two
groups (34 vs. 29 patients). We could not find any sig-
nificant differences between the two groups in terms of
other factors included in the FRAX® tool (previous frac-
ture, parental hip fracture, smoking, glucocorticoids, al-
cohol consumption) except the presence of rheumatoid
arthritis (data not shown). In the myositis and RA
groups normal BMD was found in 27 and 53.5%, re-
spectively, whilst osteopenia was found in 60 and 39.5% To assess the prevalence of vertebral fractures patients
underwent a bidirectional (anteroposterior and lateral)
X-ray imaging of the thoracic (Th) and lumbar (L) spine
on separate cassettes for each picture. To decrease prob-
ability of potential selection bias, all of the patients were Vincze et al. Results BMC Musculoskeletal Disorders (2020) 21:426 Page 4 of 8 Table 1 Summary of the most relevant clinical data and FRAX scores of the patients
Myositis
N = 52
Rheumatoid arthritis
N = 43
P-value
(95% CIoD)
Meana/Medianb
SDa/Min-Maxb
Meana/Medianb
SDa/Min-Maxb
Age (years)a
57.46
11.168
58.58
10.486
0.618 (−5.6–3.3)
Men (N)
Women (N)
9 (17.3%)
43 (82.7%)
2 (4.7%)
41 (95.3%)
0.104
Menopause at examination (N)
33/43 (76.67%)
35/41% (85.3%)
0.314
Weight (kg)a
70.88
14.38
73.74
13.77
0.328 (−2.9–2.9)
Height (cm) a
164.12
7.56
163.7
7.1
0.308
(0.4–1.5)
BMI (kg/m2) a
26.39
5.58
27.5
4.6
0.318
(−1.1–1.1)
Vitamin D and calcium supplement (N)
34 (65.4%)
29 (67.4%)
0,833
25 OH Vitamin D3 level (nmol/L) b
59.5
15.2–125.2
62.5
27.5–129.2
0.196
BMD L1–4 (g/cm2)a
1.04
0,238
1.07
0.159
0.557 (−0.08–0.07)
BMD femur (g/cm2) a
0.83
0,108
0.85
0.125
0.294 (−0.09–0.01)
Normal (N)
14 (27%)
23 (53.5%)
0.045
Osteopenia (N)
31 (60%)
17 (39.5%)
Osteoporosis (N)
7 (13.5%)
3 (7%)
FRAX score: MOF
Without DEXA1a
9.68%
7.42
15.58%
10.91
0.008
With DEXA2a
9.44%
6.723
13.25%
9.43
0.053
Steroid dose adjusted with DEXA3a
9.54%
7.475
9.96%
7.968
0.884
FRAX score: HF
Without DEXA1a
3.06%
3.97
6.23%
7.20
0.022
With DEXA2a
2.77%
3.01
3.57%
5.08
0.811
Steroid dose adjusted with DEXA3a
2.87%
3.393
2.46%
3.382
0.128
BMI body mass index, BMD L1–4 Bone Mineral Density from lumbal 1–4 vertebrae, BMD femur Bone Mineral Density in the left femoral neck, Normal-Osteopenia-
Osteoporosis: the condition of the bone according to the T-score result, FRAX Fracture Risk Assessment Tool, MOF Major Osteoporotic Fracture, HF Hip Fracture,
DEXA dual-energy x-ray absorptiometry
Significances were calculated with independent samples t-test or Mann-Whitney test, according to the distribution. Normality of the distributions was checked
using Shapiro-Wilk test. Data are presented as mean and standard deviation (SD) with normal distribution (a on the upper corner of the variable) and median,
minimum, maximum with non-gaussian distribution (b on the upper corner of the variable); 95% CIoD: 95% Confidence Interval of the Difference (lower-upper). Categorical variables were described using frequency (case number) and percentage. Results 52,0; p = 0.022 in the RA group; Table 3), which was
similarly found with stepwise discriminant analysis in
myositis (p = 0.042), but age was not an independent
predictor of vertebral fractures in the RA group. In
addition significantly lower lumbar and femur neck
BMD were seen in fractured patients in the RA group
(1.0 g/cm2 vs 1.19 g/cm and 0.83 g/cm2 vs 0.94 g/cm2; p = 0.008 and p = 0.01. The correlation of lower lumbar
BMD and fractures in RA was further confirmed by
stepwise discriminant analysis (p = 0.001), but with this
method femur-neck BMD and age of the RA patients
were not profound independent significant factors, ac-
cording to their interdependency. The mean 25-OH
Vitamin D3 levels showed no correlation with the pres-
ence of vertebral fractures (Table 3). Finally we investigated the influence of vertebral frac-
tures on these patients’ physical function and quality of
life using HAQ and SF-36 questionnaires (Fig. 1a-b). It
was found that the decrease in physical function and
quality of life was proportional to the number of verte-
bral fractures if we analyzed the two groups together. In Table 3 Clinical and laboratory data of myositis and RA patients with and without fractures
Characteristics of myositis patients
With fracture N = 30
Without fracture N = 10
P-value
Meana/Medianb
SDa/Min-Maxb
Meana/Medianb
SDa/Min-Maxb
Age (years)a
62.83
9.858
55.4
9.057
0.034
Duration (years)b
13.5
1–28
9.5
1–21
0.16
Cumulative steroid (g)b
20.2
0–135.4
14.6
0.2–71.3
0.79
BMD L1–4 (g/cm2) b
1.08
0.79–1.54
1.04
0.84–1.15
0.20
BMD femur (g/cm2) b
0.82
0.62–1.05
0.84
0.67–1.02
0.76
25OH-Vitamin D3 level (nmol/L)b
57.4
27.8–125.2
66.2
24.2–90
0.61
ß-CTx (ug/L) b
0.28
0.05–0.81
0.21
0.1–0.46
0.43
Characteristics of RA patients
With fracture N = 24
Without fracture N = 11
P-value
Age (years) a
63.25
9.175
52
11.610
0.022
Duration (years) b
6.5
1–29
7
1–26
0.97
Cumulative steroid (g)b
4
0–55
4.3
0–28
0.430
BMD L1–4 (g/cm2) b
1.00
0.85–1.42
1.19
1–1.41
0.008
BMD femur (g/cm2) b
0.83
0.7–1.02
0.94
0.84–1.11
0.010
25OH-Vitamin D3 level (nmol/L) b
69.0
29.2–129.2
53.2
27.5–85.5
0.11
ß-CTx (ug/L) b
0.26
0.07–0.57
0.22
0.07–0.45
0.3
Duration of the disease in years, BMD L1–4 Bone Mineral Density from the lumbar 1–4 vertebrae, BMD femur Bone Mineral Density in the left femoral neck,
ß-CTx beta-crosslaps
Significances were calculated with independent samples t-test or Mann-Whitney test, according to the distribution. Results Patients with myositis had significantly longer disease
duration (13 vs. 7 years, p = 0.021) and higher cumulative
steroid dose (17.6 g vs. 4.1 g, p = 0.009) (Table 2). Overall
194 vertebral fractures were discovered in 54 patients
(115 fractures in 30 myositis and 79 fractures in 24 RA
patients), with these patients representing 75% of the Vincze et al. BMC Musculoskeletal Disorders (2020) 21:426 Page 5 of 8 Table 2 Basic clinical data of patients available for vertebral X-ray assessments
Myositis
N = 40
Rheumatoid arthritis
N = 35
P-value
Age (years) a
60.97
10.09
59.71
11.16
0.795
Female/male (N)
32/8
35/0
–
Duration of disease (years) b
13
1–28
7
1–29
0.021
Cumulative steroid dose (g) b
17.6
0–13.5
4.1
0–55
0.009
Patients with vertebral fractures (N)
30
24
0.375
Number of all fractures (N)
115
79
0.206
Significances were calculated with independent samples t-test or Mann-Whitney test, according to the distribution. Normality of the distributions was checked
using Shapiro-Wilk test. Data are presented as mean and standard deviation (SD) with normal distribution (a on the upper corner of the variable) and median,
minimum, maximum with non-gaussian distribution (b on the upper corner of the variable); 95% CIoD: 95% Confidence Interval of the Difference (lower-upper). Categorical variables were described using frequency (case number) and percentage Table 2 Basic clinical data of patients available for vertebral X-ray assessments Significances were calculated with independent samples t-test or Mann-Whitney test, according to the distribution. Normality of the distributions was checked
using Shapiro-Wilk test. Data are presented as mean and standard deviation (SD) with normal distribution (a on the upper corner of the variable) and median,
minimum, maximum with non-gaussian distribution (b on the upper corner of the variable); 95% CIoD: 95% Confidence Interval of the Difference (lower-upper). Categorical variables were described using frequency (case number) and percentage myositis group and 68% the RA group, and the differ-
ence was not statistically significant (Table 2). As a next
step the myositis and RA patients were divided into two
groups according to the presence of vertebral fractures. The mean age of the fractured patients was significantly
higher in both groups using Mann-Whitney test (62.83
vs. 55.4; p = 0.034 in the myositis group, and 63.25 vs. ß CTx beta crosslaps
Significances were calculated with independent samples t-test or Mann-Whitney test, according to the distribution. Normality of the distributions was checked
using Shapiro-Wilk test. Data are presented as mean and standard deviation (SD) with normal distribution (a on the upper corner of the variable) and median,
minimum, maximum with non-gaussian distribution (b on the upper corner of the variable). Stepwise discriminant analysis (Wilks) was also performed, data are
found in the text years, BMD L1–4 Bone Mineral Density from the lumbar 1–4 vertebrae, BMD femur Bone Mineral Density in the left femoral neck, Results Normality of the distributions was checked
using Shapiro-Wilk test. Data are presented as mean and standard deviation (SD) with normal distribution (a on the upper corner of the variable) and median,
minimum, maximum with non-gaussian distribution (b on the upper corner of the variable). Stepwise discriminant analysis (Wilks) was also performed, data are
found in the text Table 3 Clinical and laboratory data of myositis and RA patients with and without fractures
Characteristics of myositis patients
With fracture N
30
Without fracture N Duration of the disease in years, BMD L1–4 Bone Mineral Density from the lumbar 1–4 vertebrae, BMD femur Bone Mineral Density in
ß-CTx beta-crosslaps Age (years)a Page 6 of 8 Vincze et al. BMC Musculoskeletal Disorders (2020) 21:426 Vincze et al. BMC Musculoskeletal Disorders addition female gender was significantly associated with
poor SF-36 results (p = 0.015). The worsening of physical
function was more pronounced in the myositis group
compared to the RA group (R = 0.457; p = 0.008 vs. R =
0.376; p = 0.041). (Fig. 1a - Correlation between number
of fractures and the results of the functional tests
/HAQ/). Surprisingly, we could not detect any signifi-
cant correlations with regard to the SF-36 data of pa-
tients with RA, but in myositis patients and in the total
patient group the number of bone fractures was strongly
associated with poor SF-36 results (Fig. 1b - Correlation
between number of fractures and the results of the pa-
tient’s
health
/SF-36/). Furthermore,
general
linear
model analysis showed that in the RA group the history of
previous
fractures
(p = 0.024),
and
also
its
co-
occurrence with steroid treatment (p = 0.032) were sig-
nificant factors for poor SF-36 results (data not shown). The same results were found, if we examined separately
the mental and the physical components of the ques-
tionnaire (data not shown). Discussion To our knowledge this is the first study, which investi-
gates and compares bone fracture risks in IIM and RA,
and also the first work that correlates BMD, FRAX and
vertebral fracture data of myositis patients with rheuma-
toid arthritis patients. Data from a recent population
based study in Taiwan also showed a higher osteoporosis Fig. 1 a The effect of the number of bone fractures on the physical function (HAQ) and b quality of life (SF-36) in myositis, rheumatoid arthritis,
and in all patients. HAQ: Health Assessment Questionnaire, SF-36: Short Form-36.Regression analysis (Data of multivariate general linear model
analysis are presented in the text) Fig. 1 a The effect of the number of bone fractures on the physical function (HAQ) and b quality of life (SF-36) in myositis, rheumatoid arthritis,
and in all patients. HAQ: Health Assessment Questionnaire, SF-36: Short Form-36.Regression analysis (Data of multivariate general linear model
analysis are presented in the text) Vincze et al. BMC Musculoskeletal Disorders (2020) 21:426 Page 7 of 8 Vincze et al. BMC Musculoskeletal Disorders material, and a patient advisory card has been con-
structed, with an aim to increasing the patients’ aware-
ness and adherence to preventive pharmacological and
non-pharmacological antiporotic treatments. prevalence rate among patients with DM/PM. The in-
creased osteoporosis risk was independent of the cor-
ticosteroid
and
immunosuppressant
treatments
[9]. Gupta et al. recently published a study about prevalence
of vertebral deformities in patients with inflammatory
myositis and found a high prevalence of asymptomatic
vertebral fractures, but fracture risk and the conse-
quences of fracture on physical function and quality of
life were not examined [23]. Basically, fracture risk
assessed without taking BMD into consideration showed
a greater risk of fracture in patients with rheumatoid
arthritis than in myositis patients. If the BMD data were
applied as well, it showed there was no longer any sig-
nificant differences between the values of the two
groups. This might support the argument for the lower
BMD – which is more frequent in patients with myositis
- counterbalances the “confounding” effect of RA as a
risk factor in the FRAX tool. With an adjustment in
FRAX according to the dose of glucocorticoids, the
remaining non-significant differences were further de-
creased. Acknowledgments
W
h
k K
li
H d g
We thank Katalin Hodosi for her help with the statistical analysis. The authors
would like to thank Dr. Thomas Ben for the proof reading of the manuscript. Abbreviations
ACR A
i
C ACR: American College of Rheumatology; ß-CTx: Beta-crosslaps; BMD: Bone
mineral density; BMI: Body mass index; BTM: Bone turnover markers;
Ca: Calcium; CRP: C-reactive protein; CTX-I: C-terminal telopeptides of type-I
collagen; DEXA: Dual-energy x-ray absorptiometry; DM: Dermatomyositis;
EULAR: European League Against Rheumatism; FRAX: Fracture Risk
Assessment Tool; GC: Glucocorticoid; HAQ: Health Assessment Questionnaire;
HPLC: High pressure liquid chromatography; IIM: Idiopathic inflammatory
myopathy; L: Lumbar; OC: Osteocalcin; PM: Polymyositis; PTH: Parathyroid
hormone; RA: Rheumatoid arthritis; RANK: Receptor activator of nuclear factor
kappa-Β; RANKL: Receptor activator of nuclear factor kappa-Β ligand;
SD: Standard deviation; SF-36: Short Form-36; TNF-α: Tumor necrosis factor
alpha; ug: Microgram Discussion Taking into account the high prevalence of
osteoporosis/osteopenia in the myositis group, it seems
logical to consider incorporating a factor that modifies
the FRAX tool and allows for a more reliable risk calcu-
lation in patients with myositis. Of course, this requires
studies with a larger patient population and with bone
fracture endpoints. In addition, it would generate a ne-
cessity for multiple, disease dependent modifying factor
development according to other systemic musculoskel-
etal diseases (lupus, Sjögren’s syndrome, vasculitis, etc). We showed that the fractured patients were significantly
older in myositis, but had lower lumbar BMD levels in
RA. The occurrence of vertebral fractures in both myo-
sitis and rheumatoid arthritis were very common and
seriously affected the patients’ physical function and
quality of life, especially in those with multiple fractures. It is interesting to observe that this effect was more pro-
nounced in females and in myositis patients with regard
to the HAQ results, and, surprisingly, the fractures did
not significantly modify the health status of the RA pa-
tients. This latter phenomenon could be explained by
the frequent joint damage and secondary fibromyalgia
seen in RA, which might bias the results of the question-
naire. We found a similarly high prevalence rate of ver-
tebral fracture as Gupta et al. [22], but in their myositis
population the median age and disease duration were
shorter than in our population, and only patients with
myositis were investigated. Despite the longer duration
of the disease in our population the prevalence of frac-
tures was not more frequent, therefore it is logical to
speculate that the majority of fractures occur in the early
phase of the disease, when the administration of higher
corticosteroid doses is more frequent. Based on the re-
sults
of
our
study,
a
national
patient
educational The possible limitations of this study should be ac-
knowledged. This work was a single center study from a
national myositis unit in Hungary, and the number of
participants in the study was relatively low. The lower
number of patients with vertebral X-ray examinations
could be a cause for selection bias, and due to the cross
sectional nature of the investigation the calculated and
the real fracture risks were not comparable. Conclusions It can be concluded that osteoporosis and consequential
fractures in myositis are common and probably underes-
timated, and that their examination is often neglected. Therefore, it would be important to pay greater atten-
tion to the recognition of low BMD and high fracture
risk and adherence to preventive measures. Our results
showed a good concordance with data of groups from
other regions of the world, suggesting that the high frac-
ture prevalence is a global myositis dependent feature. Beyond that, in our opinion, the use of validated, inter-
nationally accepted patient warning cards could increase
patients’ compliance and will lead to decreased fracture
rates. We believe, that through the collaboration of myo-
sitis centers we can integrate myositis into the FRAX
calculator as a unique risk factor. This way we can make
the risk assessment more reliable worldwide in patients
with inflammatory myositis. Author details
1 1Division of Clinical Immunology, Faculty of Medicine, University of
Debrecen, Móricz Zsigmond út 22, Debrecen H-4032, Hungary. 2Gyula
Petrányi Doctoral School of Clinical Immunology and Allergology, University
of Debrecen, Debrecen, Hungary. 3Division of Rheumatology, Faculty of
Medicine, University of Debrecen, Debrecen, Hungary. 4Department of
Medical Imaging, Faculty of Medicine, University of Debrecen, Debrecen,
Hungary. 5Department of Medical Imaging, Division of Nuclear Medicine,
University of Debrecen, Debrecen, Hungary. 6Department of Medicine, 22. Kanis JA, Johansson H, Oden A, McCloskey EV. Guidance for the adjustment
of FRAX according to the dose of glucocorticoids. Osteoporos Int. 2011;
22(3):809–16. 23. Gupta L, Lawrence A, Edavalath S, Misra R. Prevalence and predictors of
asymptomatic vertebral fractures in inflammatory myositis. Int J Rheum Dis. 2018;21(3):725–31. Kenézy Gyula University Hospital, University of Debrecen, Debrecen, Hungary. Consent for publication
Not applicable. 18. John A. Kanis: Fracture Risk Assessment Tool. 2011. https://www.sheffield.ac. uk/FRAX/. Accessed 1 June 2011. 18. John A. Kanis: Fracture Risk Assessment Tool. 2011. https://www.sheffield.ac. uk/FRAX/. Accessed 1 June 2011. 19. Vereckei E, Susanszky E, Kopp M, Ratko I, Czimbalmos A, Nagy Z, et al. Psychosocial, educational, and somatic factors in chronic nonspecific low
back pain. Rheumatol Int. 2013;33(3):587–92. Funding 13. de Andrade DC, de Magalhães Souza SC, de Carvalho JF, et al. High
frequency of osteoporosis and fractures in women with dermatomyositis/
polymyositis. Rheumatol Int. 2012;32(6):1549–53. g
This work was supported by the ÚNKP-17-2 New National Excellence
Program of the Ministry of Human Capacities”. 14. Kanis JA, Johnell O, Oden A, Johansson H, McCloskey E. FRAX and the
assessment of fracture probability in men and women from the UK. Osteoporos Int. 2008;19(4):385–97. Authors’ contributions AV: performed the fracture risk assessment, data capture, statistical analysis,
radiographs assessment; LB participated in data capture and manuscript
preparation, KS participated in the patients selection and in the organization
of the study; MN-V participated in data acquisition and processing, OS: as-
sessment of radiographs; KD participated in the design of the study and
helped in drafting the manuscript JG and ZG conceived of the study and
participated in its design and coordination. All authors read and approved
the final manuscript. JG and ZG equally contributed to this work. Page 8 of 8 Page 8 of 8 Page 8 of 8 Vincze et al. BMC Musculoskeletal Disorders Vincze et al. BMC Musculoskeletal Disorders (2020) 21:426 Competing interests Anett Vincze, Levente Bodoki, Katalin Szabó, Melinda Nagy-Vincze, Orsolya
Szalmás, Katalin Dankó, János Gaál and Zoltán Griger declare that they have
no conflict of interest. 20. Fries JF. The hierarchy of quality-of-life assessment, the health assessment
questionnaire (HAQ), and issues mandating development of a toxicity index. Control Clin Trials. 1991;12(4 Suppl):106S–17S. 21. Grados F, Fechtenbaum J, Flipon E, Kolta S, Roux C, Fardellone P. Radiographic methods for evaluating osteoporotic vertebral fractures. Joint
Bone Spine. 2009;76(3):241–7. Publisher’s Note Received: 1 November 2019 Accepted: 23 June 2020 Received: 1 November 2019 Accepted: 23 June 2020 Received: 1 November 2019 Accepted: 23 June 2020 Springer Nature remains neutral with regard to jurisdictional claims in
published maps and institutional affiliations. Springer Nature remains neutral with regard to jurisdictional claims in
published maps and institutional affiliations. Springer Nature remains neutral with regard to jurisdictional claims in
published maps and institutional affiliations. Availability of data and materials The datasets used and/or analysed during the current study are available
from the corresponding author on reasonable request. 15. Bohan A, Peter JB. Polymyositis and dermatomyositis (first of two parts). N
Engl J Med. 1975;292:344–7. 16. Aletaha D, Neogi T, Silman AJ, , Funovits J, Felson DT, Bingham CO 3rd,
et al. 2010 rheumatoid arthritis classification criteria: an American College of
Rheumatology/European league against rheumatism collaborative initiative. Ann Rheum Dis 2010; 69(9):1580–1588. Ethics approval and consent to participate The study protocol was approved by the Scientific and Research Ethics
Committee of the University of Debrecen under number DE RKEB/IKEB 5101. Prior to participation, written consent was obtained from all patients. 17. Consensus development conference: Diagnosis, prophylaxis, and treatment
of osteoporosis. Am J Med. 1993;94:646–50.. 17. Consensus development conference: Diagnosis, prophylaxis, and treatment
of osteoporosis. Am J Med. 1993;94:646–50.. References 1. Kanis JA, Melton LJ, Christiansen C, Johnston CC, Khaltaev N. The diagnosis
of osteoporosis. J Bone Miner Res. 1994;9(8):1137–41. 1. Kanis JA, Melton LJ, Christiansen C, Johnston CC, Khaltaev N. The diagnosis
of osteoporosis. J Bone Miner Res. 1994;9(8):1137–41. 1. Kanis JA, Melton LJ, Christiansen C, Johnston CC, Khaltaev N. The diagnosis
of osteoporosis. J Bone Miner Res. 1994;9(8):1137–41. 2. Bliuc D, Nguyen ND, Nguyen TV, Eisman JA, Center JR. Compound risk of
high mortality following osteoporotic fracture and refracture in elderly
women and men. J Bone Miner Res. 2013;28(11):2317–24. 2. Bliuc D, Nguyen ND, Nguyen TV, Eisman JA, Center JR. Compound risk of
high mortality following osteoporotic fracture and refracture in elderly
women and men. J Bone Miner Res. 2013;28(11):2317–24. 3. Frost SA, Nguyen ND, Center JR, Eisman JA, Nguyen TV. Excess mortality
attributable to hip-fracture: a relative survival analysis. Bone. 2013;56(1):23–9
4. Lacativa PG, Farias ML. Osteoporosis and inflammation. Arq Bras Endocrinol
Metabol 2010;54(2):123 32 3. Frost SA, Nguyen ND, Center JR, Eisman JA, Nguyen TV. Excess mortality
attributable to hip-fracture: a relative survival analysis. Bone. 2013;56(1):23–9
4. Lacativa PG, Farias ML. Osteoporosis and inflammation. Arq Bras Endocrinol
Metabol. 2010;54(2):123–32. 3. Frost SA, Nguyen ND, Center JR, Eisman JA, Nguyen TV. Excess mortality
attributable to hip-fracture: a relative survival analysis. Bone. 2013;56(1):23–9 4. Lacativa PG, Farias ML. Osteoporosis and inflammation. Arq Bras Endocrinol
Metabol. 2010;54(2):123–32. 5. Rehman Q, Lane NE. Bone loss. Therapeutic approaches for preventing
bone loss in inflammatory arthritis. Arthritis Res. 2001;3(4):221–7. 5. Rehman Q, Lane NE. Bone loss. Therapeutic approaches for preventing
bone loss in inflammatory arthritis. Arthritis Res. 2001;3(4):221–7. 6. Santiago RA, Silva CA, Caparbo VF, Sallum AM, Pereira RM. Bone mineral
apparent density in juvenile dermatomyositis: the role of lean body mass
and glucocorticoid use. Scand J Rheumatol. 2008;37(1):40–7. 7. van Staa TP, Leufkens HG, Cooper C. The epidemiology of corticosteroid-
induced osteoporosis: a meta-analysis. Osteoporos Int. 2002;13(10):777–87. 7. van Staa TP, Leufkens HG, Cooper C. The epidemiology of corticosteroid-
induced osteoporosis: a meta-analysis. Osteoporos Int. 2002;13(10):777–87. 8. Mazess RB, Whedon GD. Immobilization and bone. Calcif Tissue Int. 1983;
35(3):265–7. p
y
p
;
(
)
8. Mazess RB, Whedon GD. Immobilization and bone. Calcif Tissue Int. 1983;
35(3):265–7. 9. Lee CW, Muo CH, Liang JA, Sung FC, Hsu CY, Kao CH. Increased
osteoporosis risk in dermatomyositis or polymyositis independent of the
treatments: a population-based cohort study with propensity score. Endocrine. 2016;52(1):86–92. References 9. Lee CW, Muo CH, Liang JA, Sung FC, Hsu CY, Kao CH. Increased
osteoporosis risk in dermatomyositis or polymyositis independent of the
treatments: a population-based cohort study with propensity score. Endocrine. 2016;52(1):86–92. 10. LeBlanc CM, Ma J, Taljaard M, Roth J, Scuccimarri R, Miettunen P, et al. Incident vertebral fractures and risk factors in the first three years following
glucocorticoid initiation among pediatric patients with rheumatic disorders. J Bone Miner Res. 2015;30(9):1667–75. 11. Book C, Karlsson M, Akesson K, Jacobsson L. Disease activity and disability
but probably not glucocorticoid treatment predicts loss in bone mineral
density in women with early rheumatoid arthritis. Scand J Rheumatol. 2008;
37(4):248–54. 12. Rouster-Stevens KA, Langman CB, Price HE, Seshadri R, Shore RM, Abbott K,
Pachman LM, et al. RANKL:osteoprotegerin ratio and bone mineral density
in children with untreated juvenile dermatomyositis. Arthritis Rheum. 2007;
56(3):977–83.
|
https://openalex.org/W2782182674
|
http://www.journalijar.com/uploads/229_IJAR-20849.pdf
|
English
| null |
IMPACT OF RISK MANAGEMENT ON THE PERFORMANCE OF COMMERCIAL BANKS , IN SRI LANKA.
|
International journal of advanced research
| 2,017
|
cc-by
| 6,892
|
IMPACT OF RISK MANAGEMENT ON THE PERFORMANCE OF COMMERCIAL BANKS , IN SRI
LANKA. The focus sample
is covered by thirteen commercial banks for the period of six years
from 2011 to 2016 and data analysed through SPSS package. The
results of the study revealed that, operating risk has significant
relationship with financial performance and capital management risk
has significant relationship with only return on equity of commercial
banks in Sri Lanka. Based on the research findings study recommended
commercial banks have to control their operating risk, through
operating expense level. Copy Right, IJAR, 2017,. All rights reserved. Copy Right, IJAR, 2017,. All rights reserved. Copy Right, IJAR, 2017,. All rights reserved. ………………………………………………………………………………
Introduction:-
All banks in the present-day context survive in the unstable environment and facing a l
credit risk, liquidity risk, foreign exchange risk, market risk and interest rate risk etc. all banks to survive and success in the market therefore efficient risk management
banks. According to Richard and Safra (1996) there should maintain a balance betwe
consider about the banking sector they have to face those problems higher than other s
factor when to consider about the banking sector. Therefore banks must protect their t
risk management practices. The main aim of financial institutions is to maximize reven
to shareholders by offering a variety of financial services, and especially by administe
Lankan banking sector has growth continuously. For this situation mainly cause to e
Lanka after the civil war. And also competition has been very high due to all of the bank
individually and new banks comes to the banking industry on day by day. Currently
commercial banks. The banking sector accounted for 83% of financial system assets
licensed commercial banks (LCBs) and licensed specialized banks (LSBs).Due to th
organizations many commercial banks have appointed senior staff members to overse
function. According to Richard and Safra (1996) there should maintain a balance betw
Corresponding Author:- W.Percy Wijewardana. Address:- Department of Accountancy and Finance, Faculty of Management Studies,
University of Sri Lanka, Mihintale , Sri Lanka. ISSN: 2320-5407 ISSN: 2320-5407 Int. J. Adv. Res. 5(11), 1441-1449 Journal Homepage: - www.journalijar.com
Article DOI: 10.21474/IJAR01/5919
DOI URL: http://dx.doi.org/10.21474/IJAR01/5919 IMPACT OF RISK MANAGEMENT ON THE PERFORMANCE OF COMMERCIAL BANKS , IN SRI
LANKA. W. Percy Wijewardana1 and P. D. Wimalasiri2. W. Percy Wijewardana1 and P. D. Wimalasiri2. 1. Department of Accountancy and Finance, Faculty of Management Studies, Rajarata University of Sri Lanka,
Mihintale , Sri Lanka. 2. Department of Business Management, Faculty of Management Studies, Rajarata University of Sri Lanka,
Mihintale , Sri Lanka. …………………………………………………………………………………………………….... Manuscript Info
Abstract
……………………. ………………………………………………………………
Manuscript History
Received: 19 September 2017
Final Accepted: 21 October 2017
Published: November 2017
Key words:-
Risk Management, Commercial Banks,
Credit Risk. Risk management is a cornerstone of practical banking practice and
mismatch of return is raised in economic issues in any country. Therefore the study examined how risk affects to the performance of
commercial banks in Sri Lanka. For this purpose the current study
selects credit risk, liquidity risk, operational risk and capital
management risk as independent variables while return on assets and
return on equity are utilized as dependent variables. The focus sample
is covered by thirteen commercial banks for the period of six years
from 2011 to 2016 and data analysed through SPSS package. The
results of the study revealed that, operating risk has significant
relationship with financial performance and capital management risk
has significant relationship with only return on equity of commercial
banks in Sri Lanka. Based on the research findings study recommended
commercial banks have to control their operating risk, through
operating expense level. Copy Right, IJAR, 2017,. All rights reserved. …………………………………………………………………………………………………….... I t
d
ti W. Percy Wijewardana and P. D. Wimalasiri . 1. Department of Accountancy and Finance, Faculty of Management Studies, Rajarata University of Sri Lanka,
Mihintale , Sri Lanka. 2. Department of Business Management, Faculty of Management Studies, Rajarata University of Sri Lanka,
Mihintale , Sri Lanka. …………………………………………………………………………………………………….... Manuscript Info
Abstract
……………………. ………………………………………………………………
Manuscript History
Received: 19 September 2017
Final Accepted: 21 October 2017
Published: November 2017
Key words:-
Risk Management, Commercial Banks,
Credit Risk. Risk management is a cornerstone of practical banking practice and
mismatch of return is raised in economic issues in any country. Therefore the study examined how risk affects to the performance of
commercial banks in Sri Lanka. For this purpose the current study
selects credit risk, liquidity risk, operational risk and capital
management risk as independent variables while return on assets and
return on equity are utilized as dependent variables. Literature Review:- There have been a large number of studies published about risk management in general. However, the number of the
empirical studies on risk management practices in financial institutions was found to be relatively small. The study
theoretical review and empirical review as follows. Stakeholder theory is most relevant to this study. Because a major objective of bank management is to increase
shareholders’ return signifying performance. Stakeholder theory helps to identify different risks and how to increase
shareholder’s return. Therefore stakeholder theory helps to identify the relationship between risk management and
financial performance of banks. Freeman (1984) introduced the concept as a managerial instrument, has since
evolved into a theory of the firm with high explanatory potential. Stakeholder theory focuses explicitly on
equilibrium of stakeholder interest as the main determinant of corporate policy. According to Miller and Modigliani
(1958), the theory of financial economic theory states that corporate risk management is appropriate to increase firm
value in the presence of capital market imperfections such as bankruptcy costs, a convex tax schedule, or
underinvestment problems. According to Williamson (1998), New Institutional Economics theory predicts that risk
management practices may be determined by institutions or accepted practice within a market or industry. Further,
the theory links security with specific assets purchase, which implies that risk management can be important in
contracts which bind two sides without allowing diversification, such as large financing contract or close
cooperation within a supply chain. According to Yousfi (2012) study aims to determine the impact of risk management practices on Jordanian Islamic
bank’s performance. For this purpose, the current study selects credit risk, operational risk, liquidity risk and market
risks as explanatory variables while return on assets (ROA) and return on equity (ROE) are utilized as dependent
variables for the period of fifteen years from 1998 to 2012. Liquidity, credit and operational risk management
practices have a negative and significant statistical impact on Islamic bank’s performance, and these banks failed at
the same time in managing these risks. Market risk management practices have a positive and significant statistical
impact on bank’s performance which, means that these banks don’t suffer either from the operational risk during the
study period or from managing this type of risk. According to Santomero (1997), Liquidity risk can best be described as the risk of a funding crisis. Introduction:- All banks in the present-day context survive in the unstable environment and facing a large number of risks such as
credit risk, liquidity risk, foreign exchange risk, market risk and interest rate risk etc. Those risks may threaten for
all banks to survive and success in the market therefore efficient risk management is absolutely required for all
banks. According to Richard and Safra (1996) there should maintain a balance between risk and return. When to
consider about the banking sector they have to face those problems higher than other sectors. Trust is the important
factor when to consider about the banking sector. Therefore banks must protect their trust by using the appropriate
risk management practices. The main aim of financial institutions is to maximize revenues and offer the most value
to shareholders by offering a variety of financial services, and especially by administering risks. In past decade Sri
Lankan banking sector has growth continuously. For this situation mainly cause to economic development in Sri
Lanka after the civil war. And also competition has been very high due to all of the banks try to maximize the wealth
individually and new banks comes to the banking industry on day by day. Currently, Sri Lanka has 25 licensed
commercial banks. The banking sector accounted for 83% of financial system assets at end 2016 and comprises
licensed commercial banks (LCBs) and licensed specialized banks (LSBs).Due to the high competition most of
organizations many commercial banks have appointed senior staff members to oversee a formal risk management
function. According to Richard and Safra (1996) there should maintain a balance between risk and return. When to 1441
Corresponding Author:- W.Percy Wijewardana. Address:- Department of Accountancy and Finance, Faculty of Management Studies, Rajarata
University of Sri Lanka, Mihintale , Sri Lanka. 14
Corresponding Author:- W.Percy Wijewardana. Address:- Department of Accountancy and Finance, Faculty of Management Studies, Rajarata
University of Sri Lanka, Mihintale , Sri Lanka. Corresponding Author:- W.Percy Wijewardana. Address:- Department of Accountancy and Finance, Faculty of Management Studies, Rajarata
University of Sri Lanka, Mihintale , Sri Lanka. 1441 Int. J. Adv. Res. 5(11), 1441-1449 Int. J. Adv. Res. 5(11), 1441-1449 Int. J. Adv. Res. 5(11), 1441-1449 Int. J. Adv. Res. 5(11), 1441-1449 ISSN: 2320-5407 consider about the banking sector they have to face those problems higher than other sectors. Trust is the important
factor when to consider about the banking sector. Introduction:- Therefore banks must protect their trust by using the appropriate
risk management practices. Risk can be classified into systematic and unsystematic risk. Systematic risk is
associated with the overall market or the economy, whereas unsystematic risk is related to a specific asset or firm. Some of the systematic risk can be reduced through the use of risk mitigation and transmission techniques. The
objective of this study is to examine the impact of risk management on the Performance of Commercial Banks in
Sri Lanka. consider about the banking sector they have to face those problems higher than other sectors. Trust is the important
factor when to consider about the banking sector. Therefore banks must protect their trust by using the appropriate
risk management practices. Risk can be classified into systematic and unsystematic risk. Systematic risk is
associated with the overall market or the economy, whereas unsystematic risk is related to a specific asset or firm. Some of the systematic risk can be reduced through the use of risk mitigation and transmission techniques. The
objective of this study is to examine the impact of risk management on the Performance of Commercial Banks in
Sri Lanka. Literature Review:- While some
would include the need to plan for growth and unexpected expansion of credit, the risk here is seen more correctly as
the potential for a funding crisis. Such a situation would inevitably be associated with an unexpected event, such as a
large charge off, loss of confidence, or a crisis of national proportion such as a currency crisis. Kithinji (2010) examined the relationship between the credit risk management and profitability of commercial banks
in Kenya. The findings reveal that the level of credit was high in the early years of the implemented of Basel II but
decreased significantly in 2007 and 2008, probably when the Basel II was implemented by commercial banks. Notably, the level of nonperforming loans given by nonperforming loans to total loans decreased during the period
2004 to 2008. Thus on average the profits of the banking industry increased during the period 2004 to 2008. However, profitability of the commercial banks fluctuated during the period but on average increased marginally
during the period 2004 to 2008. Kargi (2011) conducted a study on credit risk and the performance of Nigerian
banks. Kargi used non-performing loan portfolios and these significantly contributed to financial distress in the
banking sector. Financial ratios as measures of bank performance and credit risk were the data collected from
secondary sources mainly the annual reports and accounts of sampled banks from 2004-2008. The author concluded
that credit risk management has a significant impact on the profitability of Nigeria banks Fan and Yijun (2014) investigated if there is a relationship between credit risk management and profitability of
commercial banks in Europe. The research collects data from the largest 47 commercial banks in Europe from 2007 1442 Int. J. Adv. Res. 5(11), 1441-1449 Int. J. Adv. Res. 5(11), 1441-1449 ISSN: 2320-5407 to 2012. The findings reveal that credit risk management does have positive effects on profitability of commercial
banks. to 2012. The findings reveal that credit risk management does have positive effects on profitability of commercial
banks. Sulieman (2014) conducted a study aims at examining the effect of credit risk management on the financial
performance of the Jordanian commercial banks during the period (2005-2013), thirteen commercial banks have
been chosen to express on the whole Jordanian commercial banks. Methodology:- In this section we describe our data and analytical aspect on the impact of risk management on the performance of
risk management and basically this research data type is qualitative. Although we are not able to identify and
analysis of all the type of risk which are likely used by the commercial bank, we believe that our approach provides
a moving the current paradigm forward. By incorporating risk and performance into our analysis, we are able to
evaluate the comparative advantages of commercial bank operation towards the effective and efficient. The study
developed a conceptual framework presented on table 01 based on extensive literature survey and it presented at
figure 1. According to the framework it is expected to use financial ratio on non-performing loan, liquidity, cost to
income, capital ratio and return on asset and return on equity ratios. The dependent variable in this study is financial
performance. It measure to return on assets (ROA) and return on equity (ROE). Independent variable in this study is
risk management and the four aspects of risk management process. There are Credit risk, Liquidity risk, Operational
risk and Capital management risk. Figure 1:- Conceptual Framework
When considering previous researchers, Adeusi (2013) identify in his study dependent variable which represents
bank profitability measured by the return on equity and return on asset. Independent variables which are represent
liquidity, credit, and capital risks. Muteti (2014) state to dependent variable in his study is financial performance and
independent variables are the credit risk, interest rate risk, foreign exchange risk, liquidity risk, bank size, bank
deposits. Credit Risk (CR)
Financial Performance
Liquidity Risk (LR)
Operational Risk (OR)
Capital Management Risk (MR) Figure 1:- Conceptual Framework Liquidity Risk (LR) Financial Performance Capital Management Risk (MR) When considering previous researchers, Adeusi (2013) identify in his study dependent variable which represen
bank profitability measured by the return on equity and return on asset. Independent variables which are represen
liquidity, credit, and capital risks. Muteti (2014) state to dependent variable in his study is financial performance an
independent variables are the credit risk, interest rate risk, foreign exchange risk, liquidity risk, bank size, ban
deposits. When considering previous researchers, Adeusi (2013) identify in his study dependent variable which represents
bank profitability measured by the return on equity and return on asset. Independent variables which are represent
liquidity, credit, and capital risks. Literature Review:- Two mathematical models have been designed to
measure this relationship, the research revealed that the credit risk management effects on financial performance of
Jordanian commercial banks as measured by ROA and ROE. The research further concludes that the credit risk
management indicators considered in this research have a significant effect on the financial performance of the
Jordanian Commercial banks. According to the above empirical analysis studies reflected different kind of
relationship between risk and commercial bank performance based on their result findings. Methodology:- Muteti (2014) state to dependent variable in his study is financial performance and
independent variables are the credit risk, interest rate risk, foreign exchange risk, liquidity risk, bank size, bank
deposits. This study will test the following hypotheses to identify the relationship among variables. If credit risk is not completely impact the performance, however, credit risk may impact higher performance for
higher risk. In this case, higher bank credit risk increases and the bank performance moderate by other risk, leading
bank shareholders to adhere to a lower credit risk ratio in order to reduce the cost of funding and predict the credit 1443 Int. J. Adv. Res. 5(11), 1441-1449 Int. J. Adv. Res. 5(11), 1441-1449 ISSN: 2320-5407 ISSN: 2320-5407 risk will have a positive influence on Return on equity. Credit risk issues may, in the case of convey positive
information to the Return on the bank equity holders. risk will have a positive influence on Return on equity. Credit risk issues may, in the case of convey positive
information to the Return on the bank equity holders. Moreover, increasing credit risk via increases in credit may require a bank to expect positive net return. Commercial
Banks may thus prefer to hold a ‘‘buffer” of excess credit availability to make it higher likely they will higher below
the legal credit requirements. This reason would be enough to encourage banks to establish liquidity buffers even if
funds were totally invest the banking operations, and there were no cost of financial distress associated with the loss
of charter value. Brown and Davis (2009) show the relevance of adjustments when a limited access to external
equity capital in mutual financial institutions leads them to have higher capital ratios. It directly affects the credit
risk. On the basis of above our first hypothesis is: H1: Commercial Bank credit risk is negatively related to return on equity and return on assets if bank credit risk
changed its affect to change negatively on Return on equity and assets. We postulate that the expected positive influence of capital management risk and return on equity and assets,
depending on bank regulation, bank supervision, and a country’s institutions. We analyze the impact of the quality
of capital management, the generosity of return on assets and equity, restrictions on management activities, official
supervisory power, and the quality of a banking system and institutions. Therefore, if the quality of capital
management favors greater risk related discipline, we expect the capital management risk to have a greater positive
influence on bank return on assets and equity. We therefore expect to note higher capital management risk in
commercial bank return. This study will test the following hypotheses to identify the relationship among variables. Accounting capital management requirements may also have an effect on bank return; the
quality of bank services is negatively related to capital management risk. Reductions in spreads may increase bank
return and thus mitigate bank risk-taking capital management related risk. In this context, the second hypothesis is: H2. Severer capital management requirement standards encourage reduce capital management risk to increase both
the return on equity and return on assets with the increase of overall banking performance. It has long been suggested that more substantial liquidity weakens the market discipline enforced by depositors and
encourages banks to take greater risks of liquidity and ultimate influence to the return on assets and equity. It is
showing that deposit insurance increases the likelihood of banking liquidity crises and that liquidity risk-shifting
incentives are positively related to the overall banking performance. If more generous deposit reduces liquidity risk,
it will also make the cost of deposits less sensitive to bank risk and reduce the optimum liquidity capital ratio for
commercial banks. For this reason, we forecast that the positive relation between the cost of deposits and the
liquidity risk will be weaker depending on the bounty of the deposit strengthening. There is little attention on the
influence of the generosity of deposit strengthening on bank risk. To our knowledge, offers evidence on this issue,
finding a positive relation between the presence of explicit liquidity risk and the return. A positive relation would
provide banks that have explicit deposit increase with incentives to maintain liquidity capital adequacy to preserve
their higher liquidity risk value both return on assets and return on equity. If liquidity diminishes liquidity risk
diminishes market discipline but increases bank return on equity, its predominant effect on liquidity capital
management becomes an empirical question. The third hypothesis is: H3: More generous deposits make liquidity capital defenses less important in increasing liquidity risk but more
important in reducing the cost of liquidity related distress positively related to increase the return on assets and
equity. One further regulatory variable is whether commercial banks are allowed to undertake operational activities that
generate non-interest income this study identify as operational risk. Formulated the restrictions on bank operational activities may affect returns, as constraints on a commercial bank’s
range of operational activities limit the opportunities for commercial bank managers to operate risky investments. This study will test the following hypotheses to identify the relationship among variables. Relaxation of market discipline will make it less beneficial for a commercial bank to hold capital adequacy. It shows
that when long-standing operational risk on permissible bank activities was controlled, banks raised their income as
a consequence of enhanced market discipline. We thus expect higher the operational risk limited on bank income to
make the return of assets and equity less sensitive to bank operational risk. Based on those restrictions on bank
operation activities have a negative influence on return. We would accordingly expect that the stricter the 1444 Int. J. Adv. Res. 5(11), 1441-1449 ISSN: 2320-5407 restrictions on bank activities, the greater the positive influence of operational risk on commercial bank return on
both assets and equity. restrictions on bank activities, the greater the positive influence of operational risk on commercial bank return on
both assets and equity. The theoretical embarking effect that restriction on bank operational risk activities may have on returns both equity
and assets are considered in our fourth hypothesis, making their analysis an empirical question. H4: Higher restrictions on bank operational activities make operational risk higher important in reducing the return
on equity and assets but more important in reducing the operational risk of bank activities. Study primary data from the 2011 to 2016 collected from annual statements of listed commercial banks traded in
Colombo Stock exchange, and entire population in this sector contained 25 commercial banks. Study sample
comprised on thirteen commercial banks has selected as a study sample and based on the random sample technique. Although in this sector more firm have other financial, insurance and finance companies. First section discuss the
descriptive statistics of the study to understand the primal behavior of the data using mean, maximum, minimum
and standard deviation parameters of the study variables. Second section is devoted to analysis the correlation
metrics to identify the relationship among the study variables. We employed the generalized method of multiple
regression estimator developed for dynamic models of study data. This methodology is specifically designed to
address three relevant econometric issues: (1) the presence of unobserved bank-specific effects; (2) the
autoregressive process in the data regarding the behavior of commercial bank behavior; and (3) the likely
endogeneity of the explanatory variables. This study will test the following hypotheses to identify the relationship among variables. Study empirical methodological analytical model is designed to test the
empirical prediction of the current paradigm whether risk categorical impact on the equity and operational
performance of the banks, how those risk measures are influence the equity and operational performance. Our first
model specification have designed to analysis the risk associate with the four overall risk aspects of the bank
influence on bank performance as identified as equity measurement. The second model specification has designed to
analysis same risk variables on bank performance identify as asset measurement. Study Model 1:- Study Model 1:-
2
..... 4
4
3
3
2
2
1
1
Eq
MR
OR
LR
CR
ROA
Study Model 1:-
2
..... 4
4
3
3
2
2
1
1
Eq
MR
OR
LR
CR
ROA
Study Model 2:- 3
..... 4
4
3
3
2
2
1
1
Eq
MR
OR
LR
CR
ROA
Descriptive Analysis:- Descriptive analysis provides simple summaries about the sample and the observations that have been made to
describe the data distribution of the study. Graph 1:-
Descriptive Statistic 4
3
2
1
Man
Capital
Risk
Operating
Risk
Liquidity
Risk
Credit
e
Performanc
Bank
Where; Y is the commercial bank financial performance is labels returns on assets (ROA) and equity (ROE) as the
dependent variables for two models. We measures the credit risk (CR), liquidity risk (LR), operating risk (OR) and
capital management risk (MR) as the independent variables. Result and Discussion:- Descriptive Analysis:- Graph 1:- 1445 Int. J. Adv. Res. 5(11), 1441-1449 ISSN: 2320-5407 According to the descriptive statistical analysis graph 1 highest maximum value (2.823) shows the operating risk
and lowest minimum value shows (0.004) by the credit risk. All risk were between 0.004 – 2.823. The mean value of
credit risk indicates that when the increase of total loan in hundred, nonperforming loan is increased 2.9%. The
mean value of capital management indicates that when increase of total assets hundred, capital and reserves are
increased 8.1%. The mean value of liquidity risk indicates that when the increase in total deposits hundred, the total
loan is increased 92.1%. When considering mean value of operational risk gross earnings increased hundred,
operational expense increased in 60.8%. -0.5
0
0.5
1
1.5
2
2.5
3
CR
LR
OR
MR
ROA
ROE
Minimum
Maximum
Mean
Std. Deviation -0.5
0
0.5
1
1.5
2
2.5
3
CR
LR
OR
MR
ROA
ROE
Minimum
Maximum
Mean
Std. Deviation Minimum
Maximum
Mean
Std. Deviation According to the descriptive statistical analysis graph 1 highest maximum value (2.823) shows the operating risk
and lowest minimum value shows (0.004) by the credit risk. All risk were between 0.004 – 2.823. The mean value of
credit risk indicates that when the increase of total loan in hundred, nonperforming loan is increased 2.9%. The
mean value of capital management indicates that when increase of total assets hundred, capital and reserves are
increased 8.1%. The mean value of liquidity risk indicates that when the increase in total deposits hundred, the total
loan is increased 92.1%. When considering mean value of operational risk gross earnings increased hundred,
operational expense increased in 60.8%. According to the descriptive statistical analysis graph 1 highest maximum value (2.823) shows the operating risk
and lowest minimum value shows (0.004) by the credit risk. All risk were between 0.004 – 2.823. The mean value of
credit risk indicates that when the increase of total loan in hundred, nonperforming loan is increased 2.9%. The
mean value of capital management indicates that when increase of total assets hundred, capital and reserves are
increased 8.1%. The mean value of liquidity risk indicates that when the increase in total deposits hundred, the total
loan is increased 92.1%. When considering mean value of operational risk gross earnings increased hundred,
operational expense increased in 60.8%. According to the correlation analysis, credit risk, capital management risk and operational risk have the negative
relationship with ROA and credit risk has significant relationship with ROA. Operational risk and capital
management risk have the higher negative correlation with ROA. It indicates that operational risk and capital
management risk highly affect the decrease ROA. Credit risk, operational risk and capital management risk have the
negative relationship with ROE. Considering liquidity risk has low positive and significant relationship with ROE.
Capital management risk and operating risk highly affect the decrease ROE. Considering both of ROA and ROE,
operating risk and capital management risk negatively affect decrease the bank’s performance. Multiple Regression Results:- p
g
We present the results of estimating the multiple regressions for bank financial performance sample in table 3 and 4. The model 1 table 3 appears to fit the data well (F – value = 0.00), and ROA is significantly and negatively related
operating risk. Since the other risk factors such as credit, liquidity, and capital management risk negatively related
but not significant. The problem of heteroscedasticity also does not exist in the data set as Durbin Watson value
(1.758) gets closer to the critical value 2. The expletory power (R2) of the model was 46.8. Accordingly, variation of
46.8% on the financial performance in term of ROA of commercial bank due to changes in credit risk, liquidity risk,
operational risk and capital management risk. Table 3:- Coefficients of Model ROA
Model
Unstandardized Coefficients
Standardized
Coefficients
T
Sig. B
Std. Error
Beta
(Constant)
0.011
0.007
1.583
0.119
CR
-0.034
0.056
-0.061
-0.619
0.538
LR
0.010
0.006
0.184
1.664
0.101
OR
-0.009
0.002
-0.470
-3.833
0.000
MR
-0.022
0.012
-0.208
-1.875
0.066
a. Dependent Variable: ROA
Model Summary Statistics:-
Model
R
R Square
Adjusted R Square
Std. Error of the Estimate
ROA
0.6840a
0.4680
0.4330
0.0064
a. Predictors: (Constant), Capital management risk, Liquidity risk, Credit risk, Operational risk Table 3:- Coefficients of Model ROA Table 3:- Coefficients of Model ROA The above regression equation it was revealed that holding credit risk, liquidity risk, operational risk, capital
management risk, return on assets of commercial banks would be at 0.011, a unit increase in credit risk would lead
to decrease in return on assets of commercial banks by factors 0.034, a unit increase in liquidity risk would lead to
increase in return on assets by factors of 0.010, a unit increase operational risk would lead to decrease in return on
assets of commercial banks by a factors 0.009, a unit increase in capital management risk would lead to decrease in
return on assets of commercial banks by a factors 0.022. In table 4, we present the results from a second regression
for return on equity sample. The model appears to fit the data well (p value-0.00), and, as expected the dependent
variable in terms of return on assets is significantly and positively related to operating and capital management
ratios. Correlation Analysis:- Correlation Analysis:-
A correlation coefficient is a statistical measure of the degree to which changes to the value of one variable predict
change to the value of another. And also it represents the relationship between two variables. y
A correlation coefficient is a statistical measure of the degree to which changes to the value of
change to the value of another. And also it represents the relationship between two variables. y
A correlation coefficient is a statistical measure of the degree to which changes to the value of one variable predict
change to the value of another. And also it represents the relationship between two variables. Table 2:- Descriptive statistics and Correlations Analysis
CR
LR
OR
CMR
ROA
ROE
Minimum
Maximum
Mean
0.004
0.078
0.029
0.438
1.358
0.921
0.099
2.823
0.608
0.007
0.335
0.080
-0.019
0.038
0.012
-0.160
0.452
0.150
Standard Deviation
0.015
0.150
0.464
0.082
0.008
0.108
Credit Risk (CR)
1
Liquidity Risk (LR)
-0.225
(0.72)
1
Operating Risk(OR)
-0.050
(0.695)
-0.417**
(0.001)
1
Capital Management Risk(CMRC)
-0.106
(0.401)
-0.009
(0.940)
0.483**
(0.000)
1
ROA
-0.057
(0.654)
0.395**
(0.001)
0.644**
(0.000)
0.430**
(0.000)
1
ROE
-0.051
(0.688)
0.081
(0.521)
-0.590**
(0.000)
-0.661**
(0.000)
0.679**
(0.000)
1
Source: Study analysis 2017 ,**. Correlation is significant at the 0.01 level (2-tailed). According to the correlation analysis, credit risk, capital management risk and operational risk have the negative
relationship with ROA and credit risk has significant relationship with ROA. Operational risk and capital
management risk have the higher negative correlation with ROA. It indicates that operational risk and capital
management risk highly affect the decrease ROA. Credit risk, operational risk and capital management risk have the
negative relationship with ROE. Considering liquidity risk has low positive and significant relationship with ROE. Capital management risk and operating risk highly affect the decrease ROE. Considering both of ROA and ROE,
operating risk and capital management risk negatively affect decrease the bank’s performance. 1446 Int. J. Adv. Res. 5(11), 1441-1449 ISSN: 2320-5407 Model
R
R
Square
Adjusted
R Square
Std. Error of
the Estimate
ROE
0.735a
0.541
0.510
0.0456
a. Predictors: (Constant), Capital management risk, Liquidity risk, Credit risk, Operational risk The above regression equation it was revealed that holding credit risk, liquidity risk, operational risk, capital
management risk, return on assets of commercial banks would be at 0.336, a unit increase in credit risk would lead
to decrease in return on equity by factors of 0.305, a unit increase in liquidity risk would lead to decrease in return
on equity by factors of 0.075, a unit increase operational risk would lead to decrease in return on equity of
commercial banks by a factors 0.095, a unit increase in capital management risk would lead to decrease in return on
equity of commercial banks by a factors 0.616. According to above regression results, the effect of credit risk and
liquidity risk were significant relationship with return on equity. Operational risk, capital management risks
significant relationship with return on equity of commercial banks. Further, therefor we find that previously
unrecognized capital management risk factor impact on bank performance significantly. Under the hypothesis
testing analysis study hypothesis one is; Commercial Bank credit risk is negatively related to return on equity and
return on assets if bank credit risk changed its affect to change negatively on return on equity and assets accepted. Second hypothesis is; severer capital management requirement standards encourage reduce capital management risk
to increase both the return on equity and return on assets with the increase of overall banking performance is
accepted. Third hypothesis is; more generous deposits makes liquidity capital defenses less important in increasing
liquidity risk but more important in reducing the cost of liquidity related distress positively related to increase the
return on assets and equity, hypothesis is accepted. This implication is more important when increase the liquidity
basically bank can maintain the liquidity levels in low cost at the same time liquidity risk minimizes to help to
increase the bank performance. Study hypothesis four is higher restrictions on bank operational activities may
operational risk higher important in reducing the return on equity and assets but more important in reducing the
operational risk of bank activities. Study is accepted the hypothesis. Continues management alert policy in the
banking sector is compulsory when loose the managing alert simultaneously risk level of the capital management
does not contribute to optimize the bank performance. Close monitoring and managing drivers are essential and
compulsory part of the banking industry. Multiple Regression Results:- We also do not found that credit and liquidity ratios significantly associate with return on equity, but
insignificantly and negatively associated with return on equity. The expletory power (R2) of the model was 54.1%. Accordingly, variation of 54.1% on the financial performance in term of ROE of commercial bank due to changes in
credit risk, liquidity risk, operational risk and capital management risk. Table 4:- Coefficients and model summary of Model ROE
Model
Unstandardized Coefficients
Standardized Coefficients
T
Sig. B
Std. Error
Beta
(Constant)
0.336
0.082
4.085
0.000
CR
-0.305
0.653
-0.042
-
0.466
0.643
LR
-0.075
0.074
-0.103
-
1.009
0.317
OR
-0.095
0.027
-0.409
-
3.588
0.001
MR
-0.616
0.135
-0.469
-
4.554
0.000
a. Dependent Variable: ROE
Model Summary 1447 Int. J. Adv. Res. 5(11), 1441-1449 ISSN: 2320-5407 Reference:- 1. Adeusi, S., Akeke, A., & Adebisi, O. (2013). Risk managent and financial performance of Banks in Nigeria. Journal of Business and Management . 1. Adeusi, S., Akeke, A., & Adebisi, O. (2013). Risk managent and financial performance of Banks in Nigeria. Journal of Business and Management . 2. Brown, C., Davis, K., 2009. Capital management in mutual financial institutions. Journal of Banking and
Finance 33, 443–455. 2. Brown, C., Davis, K., 2009. Capital management in mutual financial institutions. Journal of Banking and
Finance 33, 443–455. 3. Fan, L., & Yijun, Z. (2014). The impact of credit risk management on profitability of commercial banks. . A
study of Europe,Umea School of Business and Economics . 3. Fan, L., & Yijun, Z. (2014). The impact of credit risk management on profitability of commercial banks. . A
study of Europe,Umea School of Business and Economics . 5. Kargi, , H. (2011). (pp. Credit risk and the performance of Nigerian banks. an unpublished MBA project,). Ahmadu Bello University,Zaria,Nigeria . 5. Kargi, , H. (2011). (pp. Credit risk and the performance of Nigerian banks. an unpublished MBA project,). Ahmadu Bello University,Zaria,Nigeria . 6. Kithinji, A. (2010). Credit risk and the performance of Nigerian banks. an unpublished MBA project, . School
of business,University of Nairobi,Kenya. 6. Kithinji, A. (2010). Credit risk and the performance of Nigerian banks. an unpublished MBA project, . School
of business,University of Nairobi,Kenya. 7. Miller, M., & Modigliani, F. (1958). A correction,American Economic Review ., (p. Corporate Income Taxes
and the Cost of Capital.). 7. Miller, M., & Modigliani, F. (1958). A correction,American Economic Review ., (p. Corporate Income Taxes
and the Cost of Capital.). p
8. Muteti, , S. (2014). Relationship between financial risk management and financial performance of commercial
banks in Kenya. An unpublished MBA project,School of Business,University of Nairobi,Kenya. 8. Muteti, , S. (2014). Relationship between financial risk management and financial performance of commercial
banks in Kenya. An unpublished MBA project,School of Business,University of Nairobi,Kenya. y
p
p
j
y
y
9. Richard, J., & Safra, J. (1996). . : The Implications for Risk Management and Regulation, . The Collapse of
Continental Illinois National Bank and Trust Company Professor of International Banking. The Wharton
School, University of Pennsylvania. 9. Richard, J., & Safra, J. (1996). . : The Implications for Risk Management and Regulation, . Conclusion and Recommendation:- Res. 5(11), 1441-1449 Int. J. Adv. Res. 5(11), 1441-1449 ISSN: 2320-5407 ISSN: 2320-5407 financial stability and profitability. Bank capital management requirement and supervisors should recognize that the
effectiveness of regulatory capital requirements, depending on current bank regulation, official supervision, and the
quality of accounting information and institutions. It is concluded that bank capital management requirements is
essential management requirement to increase the performance of the banks. There is strongly understood
conclusion validated in this study too, who control a country’s banking matters. Controlling for capital market
development and initial economic issues, we found that national banking systems entrusted more predominant to
tycoons and business performance, hence banking performance subsequently seeks to enhance the country
development. Conclusion and Recommendation:- This study analyzes the impact risk Management on the Performance of Commercial Banks in using a data of 13
banks between 2011 and 2011. We employ the multiple regression estimators to control for credit risk, operating
risk, liquidity and capital management risk of the explanatory variables. The results suggest that banks hold capital
management risk, the lower the return on equity and the negative affect their return. The negative influence of the
operating risk mirrors the return on assets in our sample. The negative influence of operating risk is consistent with
evidence exploring that banks with higher risk value have fewer low return on assets and need higher supervision
and control to reduce the non-performing loans. Moreover, our results highlight that bank regulation, supervision,
and institutions alter the influence of the credit risk management as well as other risks and market power on credit
management power across competitors. Restrictions on bank activities and official supervision increase the loan
collecting power to increase repayment ability by weakening non-performing loan balance accumulating, but at the
same time promote higher repayment mechanism by reducing the non-payment loan balance market power. The net
effect is negative for risk variables. Commercial bank operational quality with regards to overall risk aspect has two
opposite effects to restrictions on bank operational activities and bank managers and other official supervision. The
net influence on capital buffers is negative. Inflexible information disclosure requirements and less substantial non-
performance loan recovery, however, have a clear positive effect on ROA by both strengthening market discipline
and making competitive value better able to reduce overall risk-taking bank management. These results relate to the
literature that finds a negative relation between operating risk and financial performance in term of return on equity
because they suggest that better bank disclosure or less performing in non-performing loan may be ways to make
greater banking competition compatible with greater financial stability and strength. Risk related aspect of the
liquidity, credit and capital management on return on assets are negatively associated with return on assets. Our
analysis has three basic implications for risk policies. First, one should not always assume that credit risk capital
standards inevitably constrain a bank performance. Bank credit risk by themselves may induce banks to hold return
above the minimum stipulated, thereby reducing the risk optimizing of bank return requirement as instruments of 1448 Int. J. Adv. Res. 5(11), 1441-1449 Int. J. Adv. Reference:- The Collapse of
Continental Illinois National Bank and Trust Company Professor of International Banking. The Wharton
School, University of Pennsylvania. y
y
10. Santomero. (1997). Commercial Bank Risk Management. An Analysis of the process, Journal of
FinancialCommercial Bank Risk Management. An Analysis of the process, . Journal of Financial Services
Research. 10. Santomero. (1997). Commercial Bank Risk Management. An Analysis of the process, Journal of
FinancialCommercial Bank Risk Management. An Analysis of the process, . Journal of Financial Services
Research. 11. Sulieman, A. (2014). The effect of credit risk management on financial performance. Investment Management
and Financial Innovations. 11. Sulieman, A. (2014). The effect of credit risk management on financial performance. Investment Management
and Financial Innovations. 12. Williamson, O. (1998). The Institution of Governance. The AEA Papers and Proceedings . 1449
|
https://openalex.org/W4282000116
|
https://www.pure.ed.ac.uk/ws/files/280623359/PDF.pdf
|
English
| null |
The delivery of an emergency audit response to a communicable disease outbreak can inform future orthopaedic investigations and clinical practice
|
Bone & joint research
| 2,022
|
cc-by
| 2,496
|
General rights
C
i h f
h General rights
Copyright for the publications made accessible via the Edinburgh Research Explorer is retained by the author(s)
and / or other copyright owners and it is a condition of accessing these publications that users recognise and
abide by the legal requirements associated with these rights. Digital Object Identifier (DOI): Digital Object Identifier (DOI):
10.1302/2046-3758.116.BJR-2022-0061 Link:
Link to publication record in Edinburgh Research Explorer Document Version:
Publisher's PDF, also known as Version of record Published In:
Bone & Joint Research Edinburgh Research Explorer
The delivery of an emergency audit response to a communicable
disease outbreak can inform future orthopaedic investigations
and clinical practice Edinburgh Research Explorer lessons from IMPACT Hip Fracture Global Audits lessons from IMPACT Hip Fracture Global Audits Citation for published version:
Hall, AC, Clement, ND, MacLullich, AMJ, Simpson, AHWR, White, TO & Duckworth, A 2022, 'The delivery
of an emergency audit response to a communicable disease outbreak can inform future orthopaedic
investigations and clinical practice: lessons from IMPACT Hip Fracture Global Audits', Bone & Joint
Research, vol. 11, no. 6, pp. 346-348. https://doi.org/10.1302/2046-3758.116.BJR-2022-0061 Twitter Follow A. J. Hall @andrewhallortho, @IM-
PACTaudits Follow A. M. J. MacLullich @A_MacLullich
Follow A. H. R. W. Simpson @EdinOrthopae-
dic Follow T. O. White @EdinburghTrauma
Follow A. D. Duckworth @Duckwort-
hOrthEd The delivery of an emergency audit
response to a communicable disease
outbreak can inform future orthopaedic
investigations and clinical practice lessons from IMPACT Hip Fracture Global Audits A. J. Hall,
N. D. Clement,
A. M. J. MacLullich,
A. H. R. W. Simpson,
A. Johansen,
T. O. White,
A. D. Duckworth Keywords: COVID-19, Collaborative audit, Trauma & Orthopaedics, Meta-audit, Hip fracture, Frailty, Research methodology, Public health,
Communicable disease, Infection The
International
Multicentre
Project
Auditing COVID-19 in Trauma & Orthopae-
dics (IMPACT) was established in March 2020
to provide information about COVID-19
in orthopaedic patients.1 The IMPACT Hip
Fracture Audits included two national and
one international multicentre retrospective
cohort studies, and a worldwide survey
of clinical service disruption during the
summer of 2020.2-4 A systematic review and
meta-analysis examined the existing litera-
ture,5 and a number of observational studies
examined COVID-19 in general and planned
orthopaedics.6,7 the quality and robustness of future studies,
facilitate
audit-based
investigation
of
topical issues using existing and common
datasets, and permit a modern approach
to health research using the meta-audit
principles.9,10 Correspondence should be sent to
Andrew J Hall; email:
andrew.hall@ed.ac.uk
doi: 10.1302/2046-3758.116.BJR-
2022-0061
Bone Joint Res 2022;11(6):346–348. These lessons are relevant to improving
the evidence-based management of commu-
nicable disease outbreaks, including further
COVID-19
waves,
novel
diseases,
and
seasonal viral illnesses such as norovirus and
influenza. They will also inform the examina-
tion of wider topical issues, such as peripros-
thetic femur fractures, orthopaedic infection,
and major geriatric trauma. IMPACT reported important findings
that included: mortality risk associated with
COVID-19 at the time of injury; risk factors
for acquiring and dying of the disease; and
the role of nosocomial transmission.8 This
global collaboration highlighted practical
lessons that might guide the research and
audit response to future specialty-wide
issues. Reporting standards for future
studies are suggested and include adequate
reporting of pathogen, healthcare, and
patient factors.f
Infographic BJR The delivery of an emergency audit
response to a communicable disease
outbreak can inform future orthopaedic
investigations and clinical practice Take down policy Take down policy
The University of Edinburgh has made every reasonable effort to ensure that Edinburgh Research Explorer
content complies with UK legislation. If you believe that the public display of this file breaches copyright please
contact openaccess@ed.ac.uk providing details, and we will remove access to the work immediately and
investigate your claim. Download date: 24. Oct. 2024 Freely available online
Follow us @BoneJointRes
BJR
A. J. Hall,
N. D. Clement,
Infographic
The delivery of an emergency audit
response to a communicable disease
outbreak can inform future orthopaedic
investigations and clinical practice
lessons from IMPACT Hip Fracture Global Audits
Keywords: COVID-19, Collaborative audit, Trauma & Orthopaedics, Meta-audit, Hip fracture, Frailty, Research methodology, Public health,
Communicable disease, Infection 2. Hall AJ, Clement ND, Farrow L, et al. IMPACT-scot report
on COVID-19 and hip fractures: A multicentre study assessing Author information:
A. J. Hall, BMedSci (Hons), MRCS(Ed), Doctoral Research Fellow, Specialty Regis-
trar, Deputy Chair, Department of Orthopaedics, Golden Jubilee National Hospital,
Clydebank, UK; Edinburgh Orthopaedics, Royal Infirmary of Edinburgh, Edinburgh,
UK; Scottish Orthopaedic Research Trust into Trauma (SORT-IT), Edinburgh, UK;
Scottish Hip Fracture Audit (SHFA), NHS National Services Scotland, Edinburgh, UK;
Department of Orthopaedics & Trauma, University of Edinburgh, Edinburgh, UK. DELIVERING A CLINICAL AUDIT RESPONSE TO A COMMUNICABLE DISEASE OUTBREAK: LESSONS FROM IMPACT HIP FRACTURE GLO 347 mortality, predictors of early SARS-cov-2 infection, and the effects of social lockdown
on epidemiology. Bone Joint J. 2020;102-B(9):1219–1228. mortality, predictors of early SARS-cov-2 infection, and the effects of social lockdown
on epidemiology. Bone Joint J. 2020;102-B(9):1219–1228.
N. D. Clement, MD, PhD, FRCS (Tr&Ortho), Consultant Orthopaedic Surgeon,
Edinburgh Orthopaedics, Royal Infirmary of Edinburgh, Edinburgh, UK; Scottish
Orthopaedic Research Trust into Trauma (SORT-IT), Edinburgh, UK; Scottish Hip
Fracture Audit (SHFA), NHS National Services Scotland, Edinburgh, UK.
N. D. Clement, MD, PhD, FRCS (Tr&Ortho), Consultant Orthopaedic Surgeon,
Edinburgh Orthopaedics, Royal Infirmary of Edinburgh, Edinburgh, UK; Scottish
Orthopaedic Research Trust into Trauma (SORT-IT), Edinburgh, UK; Scottish Hip
Fracture Audit (SHFA), NHS National Services Scotland, Edinburgh, UK. 3. Hall AJ, Clement ND, MacLullich AMJ, White TO, Duckworth AD. IMPACT-Scot 2 report on COVID-19 in hip fracture patients. Bone Joint J. 2021;103-B(5):888–897.
A. M. J. MacLullich, PhD, MRCP(UK), Professor of Geriatric Medicine, Chair, Scottish
Hip Fracture Audit (SHFA), NHS National Services Scotland, Edinburgh, UK;
Department of Ageing & Health, Usher Institute, University of Edinburgh, Edinburgh,
UK. 4. Hall AJ, Clement ND, MacLullich AMJ, et al. IMPACT of COVID-19 on hip fracture
services: A global survey by the International Multicentre Project Auditing COVID-19 in
Trauma & Orthopaedics. Surgeon. 2021;S1479-666X(21)00092-5.
A. H. R. W. Simpson, DM, FRCS (Tr&Orth), Professor of Orthopaedics, Department of
Orthopaedics & Trauma, University of Edinburgh, Edinburgh, UK.f 5. Clement ND, Ng N, Simpson CJ, et al. The prevalence, mortality, and associated
risk factors for developing COVID-19 in hip fracture patients: A systematic review and
meta-analysis. Bone Joint Res. 2020;9(12):873–883. p
,
y
g ,
g ,
A. Johansen, MB BChir, Consultant Geriatrician, University Hospital of Wales, Cardiff
University, Cardiff, UK; National Hip Fracture Database, Royal College of Physicians,
London, UK.
T. O. White, MD, FRCS (Tr&Orth), Consultant Orthopaedic Trauma Surgeon,
Edinburgh Orthopaedics, Royal Infirmary of Edinburgh, Edinburgh, UK; Scottish
Orthopaedic Research Trust into Trauma (SORT-IT), Edinburgh, UK; Department of
Orthopaedics & Trauma, University of Edinburgh, Edinburgh, UK. 6. Clement ND, Hall AJ, Makaram NS, et al. IMPACT-Restart: the influence
of COVID-19 on postoperative mortality and risk factors associated with
SARS-CoV-2 infection after orthopaedic and trauma surgery. Bone Joint J. 2020;102-B(12):1774–1781. p
y
g
g
A. D. Author contributions: Author contributions:
A. J. Hall: Conceptualization, Project administration, Visualization, Writing – original
draft, Writing – review & editing. Author contributions:
A. J. Hall: Conceptualization, Project administration, Visualization, Writing – original
draft, Writing – review & editing. 8. Hall AJ, Clement ND, Ojeda-Thies C, et al. IMPACT-Global Hip Fracture Audit:
Nosocomial infection, risk prediction and prognostication, minimum reporting
standards and global collaborative audit: Lessons from an international multicentre
study of 7,090 patients conducted in 14 nations during the COVID-19 pandemic. Surgeon. 2022;S1479-666X(22)00049-X. ,
g
g
N. D. Clement: Conceptualization, Writing – review & editing, Supervision. g
g
N. D. Clement: Conceptualization, Writing – review &
A. M. J. MacLullich: Writing – review & editing.
A. H. R. W. Simpson: Writing – review & editing. p
g
A. Johansen: Writing – review & editing.
T. O. White: Writing – review & editing. g
g
A. D. Duckworth: Conceptualization, Writing – review & editing, Supervision. g
g
A. D. Duckworth: Conceptualization, Writing – review & editing, Supervision. 9. Hall AJ, Duckworth AD, MacLullich AMJ, Farrow L, White TO, Clement ND. Meta-audit: a novel approach to healthcare improvement clinical audit and health
data. Bone Jt 360. 2021;10(6):3–5. DELIVERING A CLINICAL AUDIT RESPONSE TO A COMMUNICABLE DISEASE OUTBREAK: LESSONS FROM IMPACT HIP FRACTURE GLO Duckworth, PhD, FRCS (Tr&Orth), Consultant Orthopaedic Trauma Surgeon,
Senior Clinical Lecturer, Edinburgh Orthopaedics, Royal Infirmary of Edinburgh,
Edinburgh, UK; Scottish Orthopaedic Research Trust into Trauma (SORT-IT),
Edinburgh, UK; Department of Orthopaedics & Trauma, University of Edinburgh,
Edinburgh, UK; Usher Institute, University of Edinburgh, Edinburgh, UK. 7. Clement ND, Scott CEH, Murray JRD, Howie CR, Deehan DJ, IMPACT-Restart
Collaboration. The number of patients “worse than death” while waiting for a hip
or knee arthroplasty has nearly doubled during the COVID-19 pandemic. Bone Joint J. 2021;103-B(4):672–680. References Rapid collaboration and effective data-
sharing were central to producing a dataset
of sufficient size and diversity to answer
urgent questions relating to COVID-19. A
strategy for establishing collaborative clin-
ical audits is outlined in order to improve 1. Hall AJ. Scotland Makes an IMPACT on Global Clinical Audit. 2021. https://www.shfa.scot.nhs.uk/Reports/_docs/2021/shfa-
annual-report-2021-scotland-makes-an-impact.pdf
(date
last
accessed 12 May 2022). 2. Hall AJ, Clement ND, Farrow L, et al. IMPACT-scot report
on COVID-19 and hip fractures: A multicentre study assessing 2. Hall AJ, Clement ND, Farrow L, et al. IMPACT-scot report
on COVID-19 and hip fractures: A multicentre study assessing VOL. 11, NO. 6, JUNE 2022 346 ICMJE COI statement: ICMJE COI statement:
A. H. R. W. Simpson is the Editor-in-Chief of Bone & Joint Research, and reports in-
stitutional grants from RCUK, Charities, and Stryker, unrelated to this article. A. D. Duckworth reports research grants from NIHR, OTA, and SORT-IT, institutional grants
from Stryker, Smith & Nephew, and Acumed, royalties from Elsevier and Taylor &
Francis, a speaker payment from AgNovos Healthcare, all unrelated to this article. A. D. Duckworth is also an editorial board member of The Bone & Joint Journal, Bone &
Joint Research, and Bone & Joint 360, an Associate Editor for Trials, OTAI, and JBJS Case
Connector, and sits on the research committees for OTA Research and OTS Research. T. White reports royalties from Elsevier, unrelated to this article, and is President Elect
of the Orthopaedic Trauma Society. A. J. Hall is an Honorary Research Fellow at the
Scottish Hip Fracture Audit. 10. Johansen A, Ojeda-Thies C, Poacher A, et al. Developing a minimum common
dataset for hip fracture audit to help counties set up national audits that can support
international comparisons. Bone Joint J. 2022. © 2022 Author(s) et al. This is an open-access article distributed under the terms of
the Creative Commons Attribution Non-Commercial No Derivatives (CC BY-NC-ND 4.0)
licence, which permits the copying and redistribution of the work only, and provided
the original author and source are credited. See https://creativecommons.org/licenses/
by-nc-nd/4.0/ VOL. 11, NO. 6, JUNE 2022 Delivery of an Emergency Audit Response to a Communicable Disease Outbreak
Can Inform Future Orthopaedic Investigations and Clinical Practice:
Lessons from IMPACT Hip Fracture Global Audits Can Inform Future Orthopaedic Investigations and Clinical Practice:
Lessons from IMPACT Hip Fracture Global Audits
Hall AJ, Clement ND, MacLullich AMJ, Simpson AHRW, Johansen A, White TO, Duckworth AD. Bone Joint Res. 2022 Lessons from IMPACT Hip Fracture Global Audits
Hall AJ, Clement ND, MacLullich AMJ, Simpson AHRW, Johansen A, White TO, Duckworth AD. Bone Joint Res. 2022 p
Hall AJ, Clement ND, MacLullich AMJ, Simpson AHRW, Johansen A, White TO, Duckworth AD. Bone Joint Res. 20 Reporting Standards for studies of communicable disease Key Principles of successful emergency clinical audit
Collection
Utilise common datasets
Ensure sufficient power
Generalisability
Coordination
Rapid collaboration
Effective data-sharing
Agile & flexible approach
Communication
Timely reporting
Accessibility
Stakeholder involvement Reporting Standards for studies of communicable disease uccessful emergency c
Collection
Utilise common datasets
Ensure sufficient power
Generalisability Current Issues amenable to collaborative audit
Communicable Disease
COVID-19 outbreaks
Novel pathogens or variants
Seasonal viral illnesses
General Orthopaedics
Periprosthetic femur fractures
Orthopaedic infection
Major geriatric trauma Current Issues amenable to collaborative audit General Orthopaedics Collaborative Audit Model for multicentre observational studies in Orthopaedics Infographic design by:
Andrew J Hall
@IMPACTaudits
@andrewhallortho
Identify knowledge gaps
•
Literature search
•
Clinical/management liaison
•
Identify priorities & existing standards
Develop methodology
•
Multicentre retrospective cohort
•
Determine method of collection
•
Anticipate analytic methods
Local pilot
•
Demonstrate feasibility
•
Make iterative improvements
Develop audit materials
•
Information for participants
•
Authorship policy
•
Data collection tool
•
Research ethics materials
Determine scope of participation
•
Population of interest
•
Geopolitical context
•
Distribution of centres
•
Generalisability of findings
Communication
•
Open channel communication for collaborators
•
‘Query clinics’ to address FAQs
•
Website/social media updates for collaborators
•
Waypoint reports to governing bodies
Data submission
•
Dedicated secure mailbox
•
Cleaning & compilation
•
Two-stage quality assurance
Data analysis
•
According to a priori methods
•
Early peer-review discussion
Information dissemination
Early sharing of key findings:
•
Audit collaborators
•
Senior clinical groups
•
Representative bodies
•
Government
Definitive dissemination
•
Scientific meetings
•
Peer-review publications
Recruit collaborators
•
Affiliated hospitals
•
Research & audit networks
•
Special interest groups
•
Representative bodies
•
Social media & word of mouth
Defining data variables
•
Establish core variables
•
Utilise minimum common datasets
•
Ensure feasibility & acceptability
•
Adhere to reporting standards
@BoneJointRes
bjr.boneandjoint.org.uk/ Identify knowledge gaps
•
Literature search
•
Clinical/management liaison
•
Identify priorities & existing standards
Develop methodology
•
Multicentre retrospective cohort
•
Determine method of collection
•
Anticipate analytic methods
Local pilot
•
Demonstrate feasibility
•
Make iterative improvements
Defining data variables
•
Establish core variables
•
Utilise minimum common datasets
•
Ensure feasibility & acceptability
•
Adhere to reporting standards Identify knowledge gaps
•
Literature search
•
Clinical/management liaison
•
Identify priorities & existing standards Data submission
•
Dedicated secure mailbox
•
Cleaning & compilation
•
Two-stage quality assurance Determine scope of participation
•
Population of interest
•
Geopolitical context
•
Distribution of centres
•
Generalisability of findings Develop audit materials
•
Information for participants
•
Authorship policy
•
Data collection tool
•
Research ethics materials
Determine scope of participation
•
Population of interest
•
Geopolitical context
•
Distribution of centres
•
Generalisability of findings
Recruit collaborators
•
Affiliated hospitals
•
Research & audit networks
•
Special interest groups
•
Representative bodies
•
Social media & word of mouth Develop methodology
•
Multicentre retrospective cohort
•
Determine method of collection
•
Anticipate analytic methods Develop audit materials
•
Information for participants
•
Authorship policy
•
Data collection tool
•
Research ethics materials Information dissemination
Early sharing of key findings:
•
Audit collaborators
•
Senior clinical groups
•
Representative bodies
•
Government
Definitive dissemination
•
Scientific meetings
•
Peer-review publications Communication
•
Open channel communication for collaborators
•
‘Query clinics’ to address FAQs
•
Website/social media updates for collaborators
•
Waypoint reports to governing bodies @IMPACTaudits
@andrewhallortho
@BoneJointRes
bjr.boneandjoint.org.uk/ Infographic design
Andrew J Hall
@IMPACTaudits
@andrewhallortho
@BoneJointRes
bjr.boneandjoint.org.uk/ Infographic design by:
Andrew J Hall
|
https://openalex.org/W2395581958
|
https://www.bjbms.org/ojs/index.php/bjbms/article/download/3233/906
|
English
| null |
Topical treatment of standardised burns with herbal remedies in model rats
|
Bosnian journal of basic medical sciences
| 2,005
|
cc-by
| 4,559
|
Fahir Bečić1*, Nedžad Mulabegović1,
Zakira Mornjaković2, Elvedina Kapić1,
Senad Prašović3, Ervina Bečić4, Jasna Kusturica1 Fahir Bečić1*, Nedžad Mulabegović1, . Institute of Pharmacology, clinical pharmacology and toxicology, Faculty of . Institute of Histology and embriology, Faculty of Medicine, Čekaluša , Sarajevo, Bosnia and Herzegovina
. Institute of Pathology, Faculty of Veterinary Medicine, Zmaja od Bosne , Sarajevo, Bosnia and Herzegovina
. Department of Analytical of Drugs, Faculty of Pharmacy, Čekaluša , Sarajevo, Bosnia and Herzegovina * Corresponding author Abstract Experimental studies of burns require the use of different animal models. The aim of this work was
to establish experimental model of thermal injuries and to evaluate the effects of topical agents
on healing of the burn wounds. Forty female Wistar rats were randomly classified in groups and
isolated for weeks before the onset of experiment. Animals were primarily anaesthetized with
pentobarbital-sodium and then shaved (skin area of their back with diameters cm x cm). A
round metal stamp with contact area of cm and total weight of g was heated up to oC
and then applied without additional pressure on the depilated skin of the back for seconds. This procedure produced a standardized burn wound. Induced burn wounds were immediately
drowned in the oC- water for s in order to maintain microcirculation. After the inducement
of thermal injures, all rats were treated with silver sulfadiazine cream, herbal topical prepara-
tions or were not treated at all. Burn wounds were treated twice a day until the healing completion. The result of treatment application was a significant reduction of burn wound diameters. Herbal top-
ical preparations expressed positive therapeutic effects on the parameters of burn wounds. The effi-
ciency of silver sulfadiazine cream in burn wound healing was significantly more expressed in com-
parison to healing process in control group of animals (p≤,). We conclude that herbal topical
preparations efficiently caused the completion of burn wound healing process without scar formation. Experimental studies of burns require the use of different animal models. The aim of this work was
to establish experimental model of thermal injuries and to evaluate the effects of topical agents
on healing of the burn wounds. Forty female Wistar rats were randomly classified in groups and
isolated for weeks before the onset of experiment. Animals were primarily anaesthetized with
pentobarbital-sodium and then shaved (skin area of their back with diameters cm x cm). A
round metal stamp with contact area of cm and total weight of g was heated up to oC
and then applied without additional pressure on the depilated skin of the back for seconds. This procedure produced a standardized burn wound. Induced burn wounds were immediately
drowned in the oC- water for s in order to maintain microcirculation. &
Topical Treatment of
Standardised Burns
with Herbal Remedies
in Model Rats Fahir Bečić1*, Nedžad Mulabegović1,
Zakira Mornjaković2, Elvedina Kapić1,
Senad Prašović3, Ervina Bečić4, Jasna Kusturica1 Fahir Bečić1*, Nedžad Mulabegović1,
Zakira Mornjaković2, Elvedina Kapić1,
Senad Prašović3, Ervina Bečić4, Jasna Kusturica1 Abstract After the inducement
of thermal injures, all rats were treated with silver sulfadiazine cream, herbal topical prepara-
tions or were not treated at all. Burn wounds were treated twice a day until the healing completion. The result of treatment application was a significant reduction of burn wound diameters. Herbal top-
ical preparations expressed positive therapeutic effects on the parameters of burn wounds. The effi-
ciency of silver sulfadiazine cream in burn wound healing was significantly more expressed in com-
parison to healing process in control group of animals (p≤,). We conclude that herbal topical
preparations efficiently caused the completion of burn wound healing process without scar formation. KEY WORDS: burns, animal model, rats, herbal remedies BOSNIAN JOURNAL OF BASIC MEDICAL SCIENCES 2005; 5 (4): 50-57 STUDY PROTOCOL AND
EXPERIMENTAL MODEL Forty Wistar rat females were randomly sorted in four
groups and put in quarantine for two weeks before
the initiation of experiment. Each group consisted of
animals treated according to following schedule: •
Group I - topical treatment with
sulphadiazine silver cream; sulphadiazine silver cream; •
Group I - topical treatment with
Jomelop (extract of herbs); •
Group I - topical treatment with Jomelop (extract of herbs); •
Group III - topical treatment with HD ointment
(contains an aqueous extract of following plants
(Sambucus nigra L. Caprifoliaceae, Sanguisorba
minor Bertol. Rosaceae, Teucrium chamaedrys L. Labiateae, Polypodium vulgarae L. Polypodiaceae); •
Group IV - control group. •
Group IV - control group. All animals were anaesthetised by intraperitoneally
applied pentobarbital sodium, in a dose of mg/kg. Subsequent to that, animal hair was removed from the
scapula zone of right side of the beck ( cm x cm). Burns were caused by adherence of the round metal
seal heated on oC in water bath and through contact
thermometer control for seconds. Metal seal was
just reclined on the rat skin without additional pres-
sure. The dimensions of the utilized metal seal were:
radius . cm, weight . g. In order to maintain the
microcirculation of the burn wounds and to stimulate
the conversion from burns of partial thickness to burns
of full thickness, burns were immediately sunk in oC-
water for s. Following the inducement of burns, each
animal was put individually in proper cage and left to
rest for four hours. Though that period, careful monitor-
ing of the burn expansion and animal general state were
performed. Test preparations were applied twice a day
until the complete re-epithelization of burn wound was
achieved. Burn wound radius was measured daily by
ruler throughout determination of its widest diameter. For the period of monitoring of the parameters of burn
wound healing, two animals from each group were ex-
cluded in order to monitor changes of the burn wound
surface. Animals were photographed each day using
Olimpus digital camera, while burn wound surface Introduction mals). These were all female animals with body weight
of approximately g. Animals originated from their
own brood. During experiment performed during
summertime, all animals were situated in appropriate
cages, in rooms with natural light regime and humid-
ity level, on controlled temperature ( ± C). Experi-
mental animals were given food and water ad libitum. Pharmacological-toxicological studies present an im-
portant contribution in the evaluation of protective
effects of the new drugs in treatment of burns. These
researches consist of preliminary (in vitro), preclinical
(in vivo), and clinical researches (,). Animal models
are successfully used in investigation of the different
diseases and diverse pathophysiologic processes. The
utilization of animal models provides careful control of
alterations, because all animals are modified in similar
manner. At the same time, it should be kept in mind
that such alterations might affect the course of investi-
gated diseases. Beside the limited utilization of labora-
tory animals, animal models have some advantages in
comparison to similar studies in humans. Little mam-
mals have accelerated life cycle. That fact has been suc-
cessfully used in shortening of the duration and costs of
complex researches. The advantage of animal models
is possibility to observe a great number of subjects al-
lowing the generation of adequate data and confirma-
tion of the complete importance of statistical analysis. In addition, using animal models make possible the
realisation of more precise control of experiment con-
duction. The best option is to use the most investigated
and most available animals. However, the most avail-
able animals are not necessarily the most appropriate
ones for required animal model (). An increased in-
terest in development and utilization of in vitro and in
vivo experimental models appeared in the beginning of
nineties of the past century, particularly interest in in-
vestigation of the pathophysiologic skin disorders. For
the majority of dermatological problems there are no
models that completely reflect all symptoms and mech-
anisms normally present in clinical praxis (). A unique
problem in dermatological pharmacology is skin burn. Experimental variables might be better controlled in
animal models reflecting, in significant extent, the situa-
tion in humans (). Burns induce complex skin damages. STUDY PLAN AND EXPERIMENTAL ANIMALS Prospective, cross-sectional, intervention animal
study was designed to evaluate the influence of topi-
cal herbal preparations in burn wound healing pro-
cess. Animals were raised and cared according to
European Convention for the protection of vertebrate
animals used for experimental and other scientific pur-
poses (). Healthy laboratory Rattus rats of Wistar type
and albino sort were used in the study (rodents, mam- BOSNIAN JOURNAL OF BASIC MEDICAL SCIENCES 2005; 5 (4): 50-57 FAHIR BEČIĆ ET AL.: TOPICAL TREATMENT OF STANDARDISED BURNS WITH HERBAL REMEDIES IN MODEL RATS CHANGES IN BODY WEIGHT The percentage of body weight changes was anal-
ysed throughout t-test in order to determine the sig-
nificance of differences. Obtained values are presented
on Graph , with additional note, that every point
is representing the percentage of average change in
body weight of all animals from each animal group. STATISTICAL ANALYSIS The following parameters were monitored in each
group of animals: change in body weight, burn heal-
ing time, burn healing speed, morbidity parameters
and mortality rate , level of achieved skin reparation
that is histological status of burns. The initial animal
body weights were measured just prior the induce-
ment of burns. Subsequent to that, body weight deter-
mination was performed daily throughout first days,
and then in day-intervals until the complete burn
wound healing was achieved. All obtained results were
presented and documented through digital Figures
taken by a high-resolution Olympus camera. All differ-
ences pointed as p<, were considered as significant. Results was detected throughout software program AutoCAD. Macroscopic and microscopic analyses were performed
by classic morphofunctional method of examination. BURN HEALING TIME h The study results of cumulative sizes of burn radiuses
indicate the presence of statistically significant differ-
ence between treated and control animals. Bimodal
grouping might be observed, as well. Animal groups
treated with tested preparations are on one side
while control group of animals is on the other side. BOSNIAN JOURNAL OF BASIC MEDICAL SCIENCES 2005; 5 (4): 50-57
BOSNIAN JOURNAL OF BASIC MEDICAL SCIENCES 2005 5 (4) 50 57 BOSNIAN JOURNAL OF BASIC MEDICAL SCIENCES 2005; 5 (4): 50-57 BURN SURFACE Day : An improvement in wound healing process is ob-
served (Figure A,D,G). Control burn is dry and shriv-
elled together with the presence of necrotic tissue. Its size
approximately corresponds to size of three day-old treat-
ed burn (Figure J). Also, a difference in inflammatory
reaction between treated and control animals is evident. Day : A significant improvement in healing process is
evident. Formed incrustation is already dropped down
or is in process of dropping down. Additionally, a dif-
ference in healing process between burns treated with
herbal preparations and those treated with sulphadi- Exponential reduction in sizes of surfaces of burn
wounds alongside with time distances from the mo-
ment of burn inducement is evident on Graph . MACROSCOPIC MANIFESTATION Macroscopic expression of burn wounds was
observed daily while the improvement in heal-
ing process was recorded. Burn wounds were de-
fined as completely healed when eshar was en-
tirely separated and wound surface re-epithelized. BOSNIAN JOURNAL OF BASIC MEDICAL SCIENCES 2005; 5 (4): 50-57 BOSNIAN JOURNAL OF BASIC MEDICAL SCIENCES 2005; 5 (4): 50-57 FAHIR BEČIĆ ET AL.: TOPICAL TREATMENT OF STANDARDISED BURNS WITH HERBAL REMEDIES IN MODEL RATS in control animals (Figure C,F,I). Rough healing pro-
cess, accompanied with the formation of uneven and
bumpy scars, is evident in control animals (Figure L). azine silver is evident (Figure B,E,H). Healing of the
control burns is characterised by irregular contraction
of the newly-formed tissue (Figure K). Burn wounds
treated with tested preparations are visibly less ex-
pressed in comparison to control ones, while being char-
acterised by noticeable epithelization of the margins
and uniform contraction of the newly-formed tissue. Day : Treated burn wounds are completely healed
with minor and cosmetically acceptable scars. Cheloids,
formed in rats treated with herbal preparations, are cos-
metically much more acceptable if compared to those
observed in rats treated with sulphadiazine silver or FAHIR BEČIĆ ET AL.: TOPICAL TREATMENT OF STANDARDISED BURNS WITH HERBAL REMEDIES IN MODEL RATS observed. In maintained epidermis, especially along-
side the margins of burn, there is a hypertrophic stra-
tum spinosum. A granulation tissue with numerous
blood vessels is below the burn zone. In healthy skin,
stratum spinosum cells are multiplied and large while
containing circular and huge nuclei. Stratum granulo-
sum is a bit wider while stratum lucidum is practically
invisible. Stratum granulosum is directly followed by
hyperkeratotic stratum corneum, particularly in the
zone of margin with superficial desquamation part. Day : In histological tissue slices made of burned
skin treated with HD ointment, masses of destructed
cellular debris are separated from multiplied granu-
lation tissue (Figure H). Connective tissue compo-
nents of corium are destructed in the form of areoles. Day : A crust formation, in the form of unstruc-
tured line beneath the skin tissue made of fibroblasts
and inflammation cellular infiltration, is evident in
control burn tissue slices (Figure K). Neutrophilic
granulocytes, cellular detritus and mass of coccoid
bacteria are recognisable in necrotic mass. Signs of oe-
dema are visible in deeper layers of dermis. Connec-
tive fibres are swelled and separated while rich cellular
infiltrations, mostly made of leukocytes, are evident. Day : Burns treated with sulphadiazine silver are
covered with thin epidermis including keratin lay-
er. In granulation tissue, somewhere below surface
and somewhere a bit deeper, there is a disposal of
calcium salts in the form of long continuous or ir-
regular stripes parallel with epithelium (Figure C). Day : There is cellular infiltration forming a wide zone in
superficial parts of dermis beneath the preserved epider-
mis of burns treated with Jomelop. Deep dermis is oede-
matous with moderately dilated blood vessels. In central
wound zone, masses of necrotising tissue and significant
amounts of cellular detritus are present (Figure F). Day : Thin epidermis with unevenly wide layer of cor-
neal cells is present in burns treated with HD ointment. Margins are formed of closely serried fibres allocated in
unequal directions and containing numerous fibroblasts. Thin layer of inflammatory cellular infiltration reaches
the formed epidermis, but only between the zones
newly-formed and previous epithelium (Figure I). Day : Tightly serried fibres in asymmetrical direc-
tions with many fibroblasts and mobile cells are evi-
dent in tissue slices made of control burns (Figure L). visible keratinous superficial layer is evident in ani-
mals treated with sulphadiazine silver. MORBIDITY PARAMETERS AND
MORTALITY RATE There were no signs of morbidity in our study of ex-
perimental burns. Additionally, no mortality was
recorded during the study period. Achieved results
are congruent with literature reports (mortality rate
ranging from to in burns of partial thickness). Day : A wide zone of coagulated necrotic mass with BOSNIAN JOURNAL OF BASIC MEDICAL SCIENCES 2005; 5 (4): 50-57 FAHIR BEČIĆ ET AL.: TOPICAL TREATMENT OF STANDARDISED BURNS WITH HERBAL REMEDIES IN MODEL RATS It is separated
from healthy skin by demarcation line formed of in-
flammatory-cell infiltration with prevailing presence
of the leukocytes. Stripes of immature granulation
tissue full of cells spread from the burn margins bel-
low the burn zone. Connective fibres in skin corium
are swelled while vessels were mostly hyperaemic. Normal skin with numerous hair follicles and sweat
glands are visible near the burn zone (Figure A). Day : Evidently distracted epidermis partially homog-
enous and with acidophilic properties is evident in ani-
mals treated with Jomelop. It is surrounded with massive
cell detritus. Occasionally, in the deeper zone of cellu-
lar-inflammatory infiltration, numerous erythrocytes are
present (Figure D). Some parts of demarcation zone, in
the form of a slightly wider area, are characterised by
coagulation exudate with accumulation of cellular-in-
flammatory infiltration and residues of necrotic tissue. In deep wound zone no structure alteration is evident. Day : A coagulation necrotic mass with abundant cel-
lular detritus is present in animals treated with HD oint-
ment (Figure C). Demarcation zone is formed of numer-
ous inflammatory cells, mostly leukocytes. Connective
tissue components are not present below the centre of
burn wound. Scab material is directly reclined on granu-
lation tissue with some cellular-inflammatory infiltra-
tion and blood vessels. Relatively rich, multiplied granu-
lation tissue is present beside and below the burn zone. Day : There are numerous hyperkeratotic cells in fat-
tened stratum corneum of control burns. Massive
granulation inflammatory infiltration in the form of
demarcation line is bond to corneal layer, beneath
which, necrotic coagulated masses containing pro-
fuse cellular detritus are evident. Bellow the burn zone,
some corium blood vessels are thrombosed while
granulation tissue fills almost entire corium (Figure J). Day : A crust material, in the form of cellular detri-
tus islets accompanied with deep-below demarcation
zone, is observed in histological slices of burn wound
treated with sulphadiazine silver (Figure B). In pro-
found layers of connective tissue there are numerous
of blood vessels surrounded by moderate number of
fixed and mobile cells. Epidermis is thin and made
of just few cell rows. Hair follicles are activated in the
process of support in reconstruction of epidermis. Day : In histological tissue slices of burned skin
treated with Jomelop, intensive fibroblast activation
i
id
i
l l i
h
i i
b BOSNIAN JOURNAL OF BASIC MEDICAL SCIENCES 2005; 5 (4): 50-57 FAHIR BEČIĆ ET AL.: TOPICAL TREATMENT OF STANDARDISED BURNS WITH HERBAL REMEDIES IN MODEL RATS tion. Burn wound healing can be divided to primary,
secondary and tertiary healing. The significance of sec-
ondary burn wound healing has the most important
role in investigation of the new topical preparations (). The results of our study show that tested preparations
reduce the duration of exudative and inflammatory
phases of the wound healing in rats with standardised
burns. Obtained results are consistent with those found
in literature, representing the favourable effects of ap-
plication of different herbal extracts on the processes
of wound healing (,,). The improvement in
burn healing is monitored throughout the wound ra-
dius measuring method. Our study results demonstrate
that burns treated with sulphadiazine silver heal more
rapidly than control burns. Additionally, burns treated
with herbal preparations heal more rapidly than control
burns. The intensification of general protective response
observed in rats after the application of tested prepa-
rations (day vs. day ) may explain the reduction of
the burn radius. Applied herbal preparations probably
improve anabolic phase of the burn wound healing by
enabling a simpler resorption of the disintegration prod-
ucts and better tissue nourishment. The results of our
study show that inception of the epithelisation of wound
margins is more rapid in the group of animals treated
with herbal preparations than in the control group of
animals. In addition, wound healing is accompanied
with more uniformed contraction and slighter cicatricial
tissue than in control group. Scars formed in animals
treated with herbal preparations are less visible and have,
from the cosmetic point of view, better appearance. Neuromuscular deficit, hypermetabolism and immu-
nosuppression are processes causing the weakening
of organism while possessing a crucial role in morbid-
ity and mortality of the patients suffering from burns. The prominent metabolic response in thermal wounds
is manifested by the increased oxygen consumption
(). Almost identical results are achieved in human
and animal studies (). Neither mortality nor mor-
tality in treated and control animals are noted in our
study of experimental burns. Our study results are
consistent with already published ones obtained from
studies elaborating the same parameters (). Histo-
logical findings confirm results achieved throughout
monitoring of the healing parameters of treated and
control burns. Approximately one third of traditional
“medicines” are purposed to treat wounds in relation to
only - of all contemporary medicines (). FAHIR BEČIĆ ET AL.: TOPICAL TREATMENT OF STANDARDISED BURNS WITH HERBAL REMEDIES IN MODEL RATS Our re-
sults clearly point the efficacy of topical herbal prepa- model depends on numerous factors including animal
availability, costs, handling, as well as their anatomi-
cal and functional similarity with human body. Small
animals are more frequently used in wound heal-
ing studies. Nevertheless, there are a lot of anatomi-
cal and physiological differences between frequently
utilized small mammalians and humans. Pig skin is
the most resembled to human skin from the anatomi-
cal and physiological point of view. But, the size of
pigs is limiting factor for their larger utilization as ani-
mal models (). In our model, a very available animal
sort accomplishing all requirements is used (,,). Body weight loss is a crucial characteristic of post-burn-
ing period. Hypermetabolic response of burned ani-
mals is frequently up to two times increased comparing
to the state of normal metabolism. Body weight loss
in post-burning period is in a direct proportion with
depth and size of burn, physical state of burns, as well
as the choice of treatment. In initial post-burning period,
body weight loss is a result of metabolism and excre-
tion of separated fluid inside the burns. Comparison
between body masses of animals with identical burns
is in direct relation with energy differences or meta-
bolic balance (). In our study, a statistically significant
difference in the change of body mass between differ-
ently treated experimental animals is not established. Two probable curing phases in thermal wound heal-
ing processes are noticed in our histologically analysed
material. The first one would be the removal of ne-
crotic material, while the second one would be wound
cicatrisation, that is, the appearance of newly-formed
tissue. Control animals, not treated with protective
medical preparations, have prolonged both first and
second phase. From the reparation point of view, the
most efficient healing is observed in wounds treated
with sulphadiazine silver. Our finding is entirely in ac-
cordance with results noted by Isler et al., Snelling and
Roberts, as well as by Fox et al. (,,). In addition,
subsequent healing phases are in accordance with the
intensity of initial catabolism, wound purification and
resorption of disintegration products. In the same time,
these are basic determinants of anabolism extent and
reparation status, as well (). It is necessary to empha-
sise that minor skin injuries provide a greater possibil-
ity for inner immune system to accomplish more suc-
cessful infection inhibition. Discussion There are many in vitro and in vivo models of wound
healing described in literature. The choice of animal BOSNIAN JOURNAL OF BASIC MEDICAL SCIENCES 2005; 5 (4): 50-57 BOSNIAN JOURNAL OF BASIC MEDICAL SCIENCES 2005; 5 (4): 50-57 Conclusion A simple, reproductive, low-cost and recommendable animal model for the investigation of thermal wound healing
is adopted allowing insignificantly low morbidity and mortality in tested animals. The efficacy if sulphadiazine silver
versus tested topical herbal preparations in the treatment of burn wounds is confirmed, as well. Additionally, tested
herbal preparations appear to be efficacious in the healing of experimental burns comparing to control burn wounds. FAHIR BEČIĆ ET AL.: TOPICAL TREATMENT OF STANDARDISED BURNS WITH HERBAL REMEDIES IN MODEL RATS Denatured eschar products,
in II degree burns, primarily make a solid cover repre-
senting temporary blockage of the wound. Burn wound
healing represent a dynamic relation between necrosis BOSNIAN JOURNAL OF BASIC MEDICAL SCIENCES 2005; 5 (4): 50-57 References () Webb M.L. Research. In: Alfonso R. Gennaro, et al. eds. Remington:
The Science and Practice of Pharmacy, th ed. Philadelphia, Col-
lege of Pharmacy and Science, ; -. () Thorton J.W., Hess C.A., Cassingham V., Bartlett R. Epidermal
growth factor in the healing of second degree burns: a controlled
animal study. Burns ; : -. () Luellmann H., Mohr K. Pharmakologie und Toxikologie, . kom-
plett ueberarbeitete und neu gestaltete Auflage, Georg Thieme Ver-
lag Stuttgart-New York, ; -. () Fox C.L., Rao T.N., Azmeth R., Gandhi S.S., Modak S. Comparative
evaluation of zinc sulfadiazine and silver sulfadiazine in burn wound
infection. J. Burn. Care. Rehabil. ; (): -. infection. J. Burn. Care. Rehabil. ; (): -. () Brown R., Kinsty D. Strategies for cell engineering in tissue repair. Wound. Rep. Reg. ; : -. () Rouse M.S., Wilson, W.R. General Methodologies for Animal
Models. In: Oto Zak (ed) Handbook of animal models of infection
Experimental Models in Antimicrobial Chemotherapy. San Diego,
London, Boston, New York, Sydney, Tokyo, Toronto, Academic
Press, ; -. () Mensah A.Y., Sampson J., Houghton P.J. et al. Effects of Buddleja
globosa leaf and its constituents relevant to wound healing. J. Ethno-
pharmacol. ; (-): -. () Bouclier M., Cavey D., Kail N., Hensby C. Experimental models in
skin pharmacology: Pharmacological reviews. ; (): -. () Parck E.H., Chun M.J. Wound healing activity of Opuntia ficus-in-
dica. Fitoterapia ; (): -. () Bouclier M., Cavey D., Kail N., Hensby C. Experimental models in
skin pharmacology: Pharmacological reviews. ; (): -. () Choi S.W., Son B.W., Son Y.S., Park Y.I., Lee S.K., Chung M.H. The
wound-healing effect of glycoprotein fraction isolated from aloe
vera. Br. J. Dermatol. ; (): -. () Isler H., Bauen A., Baschong W. Topical treatment of standardized
burns with a protein-free haemodialysate. Burns ; (): -. vera. Br. J. Dermatol. ; (): -. () European Convention for the protection of vertebrate animals used
for experimental and other scientific purposes. Official Journal L
. (Document A-) , - () European Convention for the protection of vertebrate animals used
for experimental and other scientific purposes. Official Journal L
. (Document A-) , - () Herndon D.N., Wilmore D.W., Mson A.D. Development an analy-
sis of a small animal model simulating the human post burn hypo
metabolic response. J. Surg. Res. ; : () Sullivan T.P., Eaglstein W.H., Davis S.C., Mertz P. References The pig as a model
for human wound healing. Wound. Repair. Regen. ; (): -
. () Sullivan T.P., Eaglstein W.H., Davis S.C., Mertz P. The pig as a model
for human wound healing. Wound. Repair. Regen. ; (): -
. () Barrow R.E., Meyer N.A., Jeschke M.G. Effect of varying burn size
and ambient temperature on the hyper metabolic rate in thermally
injured rats. J. Surg. Res. ; (): -. () Walker H.L., Mason A.D. A standard animal burn. J. Trauma. ;
: -. () Mantle D., Gok M.A., Lennard T.W. Adverse and beneficial effects
of plant extracts an skin and skin disorders. Adverse Drug React. Toxicol. Rev. ; (): -. () Mantle D., Gok M.A., Lennard T.W. Adverse and beneficial effects
of plant extracts an skin and skin disorders. Adverse Drug React. Toxicol. Rev. ; (): -. () Mikuš D., Sikirić P., Seiwerth S., et al. Pentadecapeptide BPC
cream improves burn-wound healing and attenuates burn-gastric
lesions in mice. Burns ; (): -. () Mikuš D., Sikirić P., Seiwerth S., et al. Pentadecapeptide BPC
cream improves burn-wound healing and attenuates burn-gastric
lesions in mice. Burns ; (): -. () Snelling C.F., Roberts F.J. Comparison of sulfadiazine with and
without chlorhexidine digluconate for topical antibacterial effect
in the burnt infected rat. J. Burn. Care. Rehabil. ; (): -. () Snelling C.F., Roberts F.J. Comparison of sulfadiazine with and
without chlorhexidine digluconate for topical antibacterial effect
in the burnt infected rat. J. Burn. Care. Rehabil. ; (): -. BOSNIAN JOURNAL OF BASIC MEDICAL SCIENCES 2005; 5 (4): 50-57 BOSNIAN JOURNAL OF BASIC MEDICAL SCIENCES 2005; 5 (4): 50-57
|
https://openalex.org/W4396662646
|
https://ojs.unida.ac.id/karimahtauhid/article/download/12373/4799
|
Indonesian
| null |
Upaya Peningkatan Hasil Belajar Siswa Melalui Pembelajaran Kooperatif Tipe Stad ( Student Team Achievement Division) Pada Materi Mahluk Hidup
| null | 2,024
|
cc-by-sa
| 9,570
|
Karimah Tauhid, Volume 3 Nomor 3 (2024), e-ISSN 2963-590X | Azizah dan Aliyyah Karimah Tauhid, Volume 3 Nomor 3 (2024), e-ISSN 2963-590X | Azizah dan Aliyyah Neng Siti Farhatul Azizah1, Rusi Rusmiati Aliyyah2
Universitas Djuanda 1Neng siti farhatul azizah, farha170501@gmail.com
2Rusi Ruamiati Aliyyah, rusi.rusmiati@unida.ac.id Kata Kunci: Hasil Belajar, Kooperatif Tipe Stad, Makhluk hidup. Karimah Tauhid, Volume 3 Nomor 3 (2024), e-ISSN 2963-590X | Azizah dan Aliyyah Karimah Tauhid, Volume 3 Nomor 3 (2024), e-ISSN 2963-590X | Azizah dan Aliyyah ABSTRAK Ilmu Pengetahuan Alam (IPA) didefinisikan sebagai kumpulan pengetahuan yang
tersusun Salah satu model pembelajaran yang dapat mengatasi yaitu model
pembelajaran kooperatif tipe STAD yang dilaksanakan di SD Negeri Cimande 01. Penelitian ini berjumlah dua siklus di desain dalam setiap siklusnya terdapat
pelaksanaan, tindakan, observasi dan refleksi, dengan nilai KKM yang ditentukan adalah
66. Hasil penelitian ini dibagi menjadi dua siklus dengan menggunakan tingkat
persentase keberhasilan dalam kemampuan siswa dalam mengerjakan tes disetiap akhir
pembelajaran serta aktivitas siswa. Hasil penelitian menunjukkan bahwa hasil belajar
siswa meningkat pada setiap siklusnya. Hal ini dapat dilihat dari hasil belajar siswa pada
setiap siklusnya yakni: siklus pertama siswa hanya memperoleh nilai rata-rata siswa
sebesar 67,85, pada siklus kedua mengalami peningkatan menjadi 84.82 maka penelitian
dikatakan berhasil mencapai nilai KKM yang ditentukan yaitu 66. Berdasarkan hasil dari
penelitian yang telah dilakukan diperoleh data bahwa model pembelajaran STAD dapat
dikatakan berhasil dalam meningkatkan hasil belajar siswa, hal ini terlihat dari nilai yang
diperoleh siswa dari siklus 1 sampai dengan siklus2 peningkatan secara signifikan. 3458 PENDAHULUAN IPA adalah mata pelajaran pokok yang diberikan pada pendidikan dasar,
khususnya dalam konteks SD. IPA, atau Alfabet Fonetik Internasional, adalah
kerangka kerja untuk mempelajari dan merepresentasikan bunyi bahasa manusia. Ini
memiliki hubungan yang luas dengan beberapa aspek keberadaan manusia. Pendidikan sains mempunyai peranan penting baik dalam proses pendidikan
maupun kemajuan teknologi. Tujuan pendidikan sains adalah untuk memungkinkan
siswa mengeksplorasi dan memahami diri mereka sendiri dan alam, sekaligus
meningkatkan kemampuan mereka untuk menggunakan pengetahuan ilmiah dalam
keseharian. Siswa yang menjadi penerima utama pendidikan diharapkan terlibat aktif
dalam proses perolehan pengetahuan dengan cara mencari informasi secara mandiri
dan melakukan eksplorasi baik secara individu maupun kolaboratif. Peran guru hanya sebatas memfasilitasi dan membimbing optimalisasi
perolehan pengetahuan. Selama proses pembelajaran diharapkan siswa akan
menunjukkan
kemauan
dan
kemampuan
mengartikulasikan
pemikirannya
berdasarkan pemahamannya. Selain itu, mereka diharapkan untuk terlibat dalam
interaksi yang konstruktif dengan teman sejawat dan guru jika ada tantangan. Namun, sebenarnya tingkat keterlibatan siswa dalam kegiatan pembelajaran masih
kurang. Misalnya, kurangnya antusiasme siswa dalam mengikuti pembelajaran
kelompok, dan masih jarangnya pelaksanaan pembelajaran lapangan melalui
kegiatan kelompok. Akibatnya, hasil yang dicapai juga di bawah standar. Biasanya,
siswa menunjukkan kepasifan dengan hanya menyetujui pernyataan guru tanpa
secara aktif terlibat dalam mengungkapkan pemikiran mereka sendiri, mengajukan
pertanyaan, atau menanggapi pertanyaan. Ketika guru mengajukan pertanyaan,
siswa menahan diri untuk menjawab jika ukuran kelas dibatasi hanya 4-5 siswa. Selain
itu, dengan adanya hambatan, siswa kurang berani bertanya. Hasil belajarnya masih
di bawah ambang batas ketuntasan belajarnya dengan tolak ukur 66. Namun
demikian, sebagian besar siswa, khususnya 60%, pada bidang pendidikan sains,
memperoleh nilai yang berada di bawah kisaran tolak ukur yang ditetapkan yaitu 25-
60.Permasalahan tersebut perlu menjawab tantangan meningkatkan hasil belajar
siswa pada mata pelajaran pendidikan sains, khususnya pada domain makhluk
hidup, di SD Negeri 1 Penpen. Model pembelajaran kooperatif dipandang sebagai
salah satu strategi pembelajaran yang potensial untuk memecahkan tantangan. Gaya
belajar kooperatif ini memfasilitasi ekspresi sudut pandang siswa secara aktif dan 3459 Karimah Tauhid, Volume 3 Nomor 3 (2024), e-ISSN 2963-590X | Azizah dan Aliyyah menumbuhkan perkembangan informasi, sikap, keaktifan, dan keterampilan sosial
seperti kerja sama, yang bermanfaat bagi kehidupan bermasyarakat. PENDAHULUAN Pernyataan
Slavin (dalam Nurasma, 2008 : 1) “Cooperative learning approaches involve the
collaborative effort of students to acquire knowledge, with the added responsibility
of ensuring their colleagues' learning alongside their own” yang berarti Pendekatan
pembelajaran kooperatif melibatkan upaya kolaboratif siswa untuk memperoleh
pengetahuan, dengan tanggung jawab tambahan untuk memastikan pembelajaran
rekan-rekan mereka sama dengan pembelajaran mereka sendiri”. Berbagai model
pembelajaran kolaboratif dapat digunakan selama proses perolehan pengetahuan. Ruang lingkup pembelajaran kooperatif dibatasi pada penerapan paradigma STAD
(Student Teams-Achievement Divisions). Model STAD diterapkan untuk memfasilitasi pencapaian tujuan pendidikan,
merangkul keberagaman antar individu, dan mendorong pertumbuhan sosial. Sebagaimana dikemukakan oleh Slavin (dalam Nurasma, 2008:50), model STAD
melibatkan pengorganisasian siswa ke dalam kelompok kecil dengan anggotanya 4-5
anak. Kelompok-kelompok ini sengaja dibuat beragam, terdiri dari individu- individu
dengan kemampuan akademis, jenis kelamin, ras, etnis, atau latar belakang sosial
lainnya yang berbeda-beda. Berdasarkan pandangan di atas, kesimpulannya model
pembelajaran tipe STAD ialah model kolaboratif yang menghubungkan beragam
peserta untuk mencapai tujuan pendidikan. Memperoleh pengetahuan atau keterampilan melalui studi, praktik, atau
pengalaman.Paradigma STAD meningkatkan dan merangsang kegembiraan siswa
untuk secara kolaboratif dan efektif mengatasi suatu masalah. Model STAD ialah
model yang paling dasar sehingga cocok bagi guru pemula yang baru mengenal
metode pembelajaran kooperatif. Pembelajaran kooperatif tipe STAD memerlukan
kerja kolaboratif antar siswa. Upaya kolaboratif antar siswa, yang difasilitasi oleh
guru, meningkatkan pemahaman dan mempercepat perolehan pengetahuan dengan
memanfaatkan pembelajaran teman sebaya. Nur Asma (2008: 3) berpendapat bahwa
“siswa dapat lebih mudah memahami dan menangkap suatu gagasan ketika mereka
terlibat dalam diskusi kolaboratif”. Selain itu, Ari dkk. (2007: 96) menyatakan “anak-
anak memiliki pemahaman yang lebih baik terhadap bahasa teman sebayanya
dibandingkan dengan yang dilontarkan orang dewasa.”dari pnyampaian tersebut,
kesimpulannya pembelajaran kolaboratif mempermudah pemahaman siswa terhadap 3460 Karimah Tauhid, Volume 3 Nomor 3 (2024), e-ISSN 2963-590X | Azizah dan Aliyyah Karimah Tauhid, Volume 3 Nomor 3 (2024), e-ISSN 2963-590X | Azizah dan Aliyyah Karimah Tauhid, Volume 3 Nomor 3 (2024), e-ISSN 2963-590X | Azizah dan Aliyyah suatu topik tertentu, sehingga menghasilkan pemahaman yang lebih signifikan baik
bagi dirinya sendiri maupun teman-temannya. Lebih lanjut menurut penelitian Slavin
(sebagaimana dikutip dalam Nur Asma, 2008:44), pembelajaran kelompok
memberikan dampak yang menguntungkan bagi siswa. Kehadiran tujuan kelompok
dan tanggung jawab individu terbukti secara signifikan meningkatkan prestasi
akademik siswa. a. Perencanaan Sebelum melaksanakan penelitian tindakan kelas, tahap perencanaan meliputi
pembuatan persiapan-persiapan yang diperlukan. Pada tahap ini dilakukan penyusunan
RPP dan pembuatan media pembelajaran yang akan dimanfaatkan. b. Pelaksanaam Tindakan Tahap ini merupakan penggambaran langkah-langkah yang harus dilaksanakan,
situasi kerja di mana tindakan perbaikan akan dilaksanakan, dan protokol
yang harus diikuti dalam melaksanakan kegiatan. Saat ini, tindakan perbaikan yang
dimaksudkan sedang dipraktikkan di kelas. Observasi Tahap observasi dilalui agar memantau secara dekat pelaksanaan rencana
yang telah disusun dengan cermat. Observasi dapat dilakukan dengan cara
membagikan lembar observasi kepada peserta didik atau dengan menggunakan teknik
lain yang sesuai untuk memperoleh data yang diperlukan. Tahap refleksi ialah tahap akhir dalam PTK, dimana dilakukan penilaian terhadap
perubahan atau hasil yang dicapai sebagai konsekuensi dari tindakan yang
dilaksanakan selama PTK. Pada tahap ini, peneliti melakukan introspeksi, menganalisis
pengalaman, mencermati pencapaian atau tantangan, dan memikirkan tindakan
selanjutnya yang harus diambil. Penelitian tindakan kelas dapat dilakukan secara 3461 Karimah Tauhid, Volume 3 Nomor 3 (2024), e-ISSN 2963-590X | Azizah dan Aliyyah Karimah Tauhid, Volume 3 Nomor 3 (2024), e-ISSN 2963-590X | Azizah dan Aliyyah terkendali dan metodis untuk meningkatkan proses pembelajaran dengan mengikuti
tahapan-tahapan tersebut. terkendali dan metodis untuk meningkatkan proses pembelajaran dengan mengikuti
tahapan-tahapan tersebut. Gambar 1. Alur PTK Model Kemmis & McTaggart(Afandi, 2013) Gambar 1. Alur PTK Model Kemmis & McTaggart(Afandi, 2013) Pada peneliti ini peneliti akan menggunakan model peneliti tindaka kelas dengan model
Kemmisdan Mc Taggart. Alasan menggunakan model ini dikarenakan melakukan 4
tahap proses penelitian.hal ini dikarenakan saat melakukan Tindakan dan pengamatan
langsung dilapangan tidak dipisahkan artinya penelitian bisa dilaksanakan dalam satu
waktu . Karimah Tauhid, Volume 3 Nomor 3 (2024), e-ISSN 2963-590X | Azizah dan Aliyyah Perubahan kegiatan yang dteramati dan terukur. (Aliyyah et al. 2019
Nasution 1995: 25) menyatakan outcome mengarah pada transformasi pada individu. Maksud dari perubahan tidak cuma mencakup transformasi pengetahuan, tetapi juga
mencakup modifikasi kemampuan, sikap, dan peng peng penguan. Mencakup
perubahan dalam kemampuan, pola pikir, pemahaman, dan harga diri dalam diri
individu. Pencapaian hasil belajar dari proses belajar mengajar yang efektif biasanya
menunjukkan hasil yang ditandai dengan kepuasan dan harga diri, yang bisa
memotivasi siswa dan meningkatkan rasa percaya diri mereka. Hasil pembelajaran
yang dicapai memiliki arti penting bagi individu, karena hasil tersebut tersimpan
dalam ingatannya, mempengaruhi perilakunya, terbukti bermanfaat untuk
memperoleh pengetahuan di bidang lain, dan berfungsi sebagai sarana untuk
mengakses informasi dan pemahaman tambahan. Pengaturan diri siswa mengacu
pada kapasitas siswa untuk mengevaluasi dan mengelola perilaku mereka sendiri,
khususnya dalam hal menilai prestasi akademik mereka, proses pembelajaran, dan
tingkat usaha mereka. Hasil belajar mengacu pada transformasi yang terjadi pada diri
seseorang yang terlibat dalam kegiatan belajar, yang mencakup tidak hanya
perolehan pengetahuan tetapi juga pengembangan keterampilan, kebiasaan,
pemahaman, keahlian, dan apresiasi. a. Pengertian Hasil Belajar Istilah “hasil belajar” merupakan kombinasi dari dua kata berbeda, “hasil” dan “belajar”,
yang masing-masing mempuyai definisi tersendiri. Dengan begitu dalam pemahaman
makna hasil belajar, terlebih dahulu akan membahas pengertian hasil dan pembelajaran. Sebagaimana dikemukakan Djamarah (2000: 45), hasil merujuk pada pencapaian suatu
tindakan yang sudah dilakukan atau dihasilkan, baik oleh individu maupun kelompok. Hasil tidak akan terlihat sampai individu mengambil tindakan. Untuk mencapai prestasi
yang signifikan memerlukan usaha yang besar dan penyangkalan diri. Kesuksesan
hanya bisa diraih dengan ketekunan, keikhlasan, tekad kuat, dan pandangan positif. Jumlahnya 18. Menurut Arikunto (1990:133), hasil belajar merujuk pada hasil akhir yang
terjadi setelah melalui proses belajar. Hasil-hasil ini diwujudkan dalam bentuk 3462 Karimah Tauhid, Volume 3 Nomor 3 (2024), e-ISSN 2963-590X | Azizah dan Aliyyah b. Pengertian Pembelajaran Koperatif Pembelajaran
kooperatif
merupakan
pendekatan
pendidikan
yang
menekankan kolaborasi siswa dalam mencapai tujuan pembelajaran, karena
dianggap penting oleh para strukturalis. Teknik pembelajaran harus disusun sebagai
pembelajaran kooperatif untuk memberikan siswa kesempatan memperoleh
pengalaman kolaboratif, perspektif yang beragam, dan belajar satu sama lain. Mengembangkan pemahaman tentang kesenjangan pengalaman, pengetahuan, dan
minat antara siswa dan orang lain akan meningkatkan kapasitas mereka untuk
menyelidiki dan mengadopsi cara dan perspektif baru (Zainuddin, 2008:35). Pembelajaran kolaboratif mengharuskan siswa berkolaborasi dalam kelompok dalam
penyelesaian tugas, pemecahan permasalahan, atau mencapai tujuan bersama. Namun demikian, agar suatu pengalaman belajar dapat digolongkan sebagai
pembelajaran kooperatif, diperlukan adanya komponen-komponen tambahan yang
menjadi landasan mendasar agar dapat disebut sebagai pembelajaran kooperatif. Aspek ini mendorong kerja kolaboratif di antara siswa ketika mereka ditugaskan 3463 Karimah Tauhid, Volume 3 Nomor 3 (2024), e-ISSN 2963-590X | Azizah dan Aliyyah proyek kelompok. Hal ini menyiratkan bahwa seluruh partisipan diberikan peluang
yang sama dalam kontribusi terhadap tujuan kolektif kelompok, daripada hanya
mengandalkan individu tertentu yang memiliki kemampuan untuk berkontribusi dan
terlibat aktif dalam proses pembelajaran. Ini adalah balasan bagi individu yang
melakukan perbuatan baik. Referensi informasi tersebut adalah Isjoni, 2007 halaman
66-67. Pembelajaran Kooperatif ialah suatu pendekatan yang mana siswanya
dikelompokan menjadi 4-6 individu dengan komposisi kelompok yang beragam,
untuk belajar dan bekerja sama. Jumlahnya adalah 2. Pendekatan pembelajaran
kooperatif melibatkan tim pembelajar yang kohesif yang berkolaborasi untuk secara
kolektif mengatasi dan menyelesaikan suatu masalah, menyelesaikan tugas, atau
terlibat dalam tugas bersam (Menurut Solihatin Aliyyah et al., 2019) 3464 Karimah Tauhid, Volume 3 Nomor 3 (2024), e-ISSN 2963-590X | Azizah dan Aliyyah Karimah Tauhid, Volume 3 Nomor 3 (2024), e-ISSN 2963-590X | Azizah dan Aliyyah Karimah Tauhid, Volume 3 Nomor 3 (2024), e-ISSN 2963-590X | Azizah dan Aliyyah c. Pengertian Materi Makhluk Hidup Makhluk hidup adalah entitas yang menjalankan banyak fungsi
penunjang kehidupan. Organisme hidup terbagi menjadi manusia, hewan, dan
tumbuhan. Setiap organisme hidup memiliki karakteristi tersendiri yang
berbeda dengan benda mati
Makhluk hidup berkemampuan melindungi dirinya dari berbagai
fluktuasi
lingkungan
dan
melakukan
reproduksi
untuk
menjamin
kelangsungan spesiesnya. Dalam bidang biologi yang mencakup semua
spesies makhluk hidup, termasuk mikroorganisme seperti manusia,
tumbuhan, hewan, dan bakteri. Makhluk hidup adalah entitas yang menjalankan banyak fungsi
penunjang kehidupan. Organisme hidup terbagi menjadi manusia, hewan, dan
tumbuhan. Setiap organisme hidup memiliki karakteristi tersendiri yang
berbeda dengan benda mati Makhluk hidup berkemampuan melindungi dirinya dari berbagai
fluktuasi
lingkungan
dan
melakukan
reproduksi
untuk
menjamin
kelangsungan spesiesnya. Dalam bidang biologi yang mencakup semua
spesies makhluk hidup, termasuk mikroorganisme seperti manusia,
tumbuhan, hewan, dan bakteri. d. Langkah-Langkah Pembelajaran Kooperatif (Aliyyah et al., 2019 Ibrahim 2000: 10)
mengusulkan model pembelajaran kooperatif yang memiliki 5
langkah berbeda: 1. Tercapainya tujuan dan menginspirasi siswa. 2. Menyampaikan informasi. 2. Menyampaikan informasi. 3. Mengelompokkan siswa ke dalam kelompok pelajaran. 4. Memfasilitasi kerja kelompok dan mendorong pembelajaran. 5. Pemberian penghargaan melalui evaluasi. Tahapan di atas menggambarkan bahwa pembelajaran diawali dari
guru menguraikan tujuan pembelajaran dan membujuk siswa belajar. Selanjutnya, terjadilah penyajian pengetahuan, biasanya menggunakan
materi
tertulis
dan
bukan
komunikasi
lisan. Selanjutnya,
siswa
diorganisasikan ke dalam tim belajar. Selanjutnya, siswa menerima instruksi
dari guru dan berkolaborasi untuk mencapai tujuan bersama dengan sukses. d. Langkah-Langkah Pembelajaran Kooperatif Penerapan pembelajaran kooperatif memerlukan kepatuhan terhadap proses
atau prosedur tertentu. Karli dan Yuliariatiningsih (2002: 72) mengusulkan
serangkaian prosedur pelaksanaan pembelajaran kooperatif, yang meliputi:
Guru merumuskan pembelajaran, mempertimbangkannya, dan menentukan
pencapaian tujuan belajarnya. Guru merumuskan lembar observasi untuk
memantau aktivitas belajar kolaboratif siswa dalam kelompok kecil. Guru
memberikan instruksi individu dan kelompok, menawarkan bimbingan dan
pengawasan kepada siswa. Memfasilitasi jalan bagi siswa untuk
memamerkan karyanya. Teks berikut memberikan penjelasan rinci tentang
empat proses yang terlibat dalam pembelajaran kooperatif: 1)
Guru merumuskan pelajaran, dengan mengamati dan menetapkan
tujuan pembelajaran yang ingin diraih siswa sesuai dengan kebutuhan
bahan ajar. Guru bertanggung jawab untuk menentukan sikap dan
keterampilan sosial yang diinginkan yang seharusnya diperoleh siswa
selama proses pembelajaran. Selanjutnya, guru memfasilitasi tugas-
tugas kolaboratif yang memerlukan partisipasi aktif dari seluruh
anggota kelompok, sehingga mendorong pembelajaran kolektif. 2) Guru membuat formulir terstruktur untuk mencatat dan memantau
aktivitas siswa saat mereka terlibat dalam pembelajaran kolaboratif
dalam kelompok kecil. Siswa akan meningkatkan pemahaman dan
memperdalam pengetahuannya dengan terlibat dalam pembelajaran
kelompok
kolaboratif
selama
penyampaian
materi
pelajaran. Guru membuat formulir terstruktur untuk mencatat dan memantau
aktivitas siswa saat mereka terlibat dalam pembelajaran kolaboratif
dalam kelompok kecil. Siswa akan meningkatkan pemahaman dan
memperdalam pengetahuannya dengan terlibat dalam pembelajaran
kelompok
kolaboratif
selama
penyampaian
materi
pelajaran. 3465 Karimah Tauhid, Volume 3 Nomor 3 (2024), e-ISSN 2963-590X | Azizah dan Aliyyah Pemahaman dan persepsi guru terhadap setiap siswa mempengaruhi
kekompakan
kelompok
yang
dibentuk
guru
selama
proses
pembelajaran. Pemahaman dan persepsi guru terhadap setiap siswa mempengaruhi
kekompakan
kelompok
yang
dibentuk
guru
selama
proses
pembelajaran. Saat mengamati siswa, guru memberikan instruksi dan dukungan
untuk
membantu
siswa
memahami
materi. Mereka
juga
memperhatikan sikap dan perilaku siswa ketika sedang belajar, baik
ketika bekerja sendiri maupun dalam kelompok. 3) Saat mengamati siswa, guru memberikan instruksi dan dukungan
untuk
membantu
siswa
memahami
materi. Mereka
juga
memperhatikan sikap dan perilaku siswa ketika sedang belajar, baik
ketika bekerja sendiri maupun dalam kelompok. 4) Selanjutnya, guru wajib memberi peluang bagi siswa dalam
melakukan presentasi kerjaannya. Guru juga menggarisbawahi cita-
cita sosial, sikap, dan perilaku yang telah ditumbuhkan dan
diterapkan siswa di kelas. (Aliyyah et al., 2019 Ibrahim 2000: 10)
mengusulkan model pembelajaran kooperatif yang memiliki 5
langkah berbeda: Selanjutnya, guru wajib memberi peluang bagi siswa dalam
melakukan presentasi kerjaannya. Guru juga menggarisbawahi cita-
cita sosial, sikap, dan perilaku yang telah ditumbuhkan dan
diterapkan siswa di kelas. e. Kelebihan Pembelajaran Kooperatif Manfaat Pembelajaran Kooperatif sebagai pendekatan pembelajaran
adalah sebagai berikut: Pembelajaran
kooperatif
memungkinkan
siswa
mengembangkan 3466 Karimah Tauhid, Volume 3 Nomor 3 (2024), e-ISSN 2963-590X | Azizah dan Aliyyah kemandirian dalam kemampuan kognitifnya, mengakses beragam
sumber informasi, dan memperoleh pengetahuan dari teman sebayanya,
sehingga mengurangi ketergantungannya pada guru. b) Pembelajaran kooperatif menumbuhkan perolehan keterampilan ekspresi
verbal dan kapasitas untuk menyandingkan dan mengevaluasi sudut
pandang alternatif. b) Pembelajaran kooperatif menumbuhkan perolehan keterampilan ekspresi
verbal dan kapasitas untuk menyandingkan dan mengevaluasi sudut
pandang alternatif. c)
Pembelajaran kooperatif memfasilitasi pengembangan rasa hormat
terhadap orang lain, kesadaran diri akan keterbatasan, dan penerimaan
segala bentuk keberagaman pada diri anak. c)
Pembelajaran kooperatif memfasilitasi pengembangan rasa hormat
terhadap orang lain, kesadaran diri akan keterbatasan, dan penerimaan
segala bentuk keberagaman pada diri anak. d) Pembelajaran kolaboratif menumbuhkan keterlibatan dan meningkatkan
proses kognitif, sehingga memfasilitasi hasil pendidikan jangka panjang. e) Pembelajaran kooperatif memfasilitasi pemberdayaan setiap peserta didik,
menumbuhkan rasa tanggung jawab yang lebih besar bagi pembelajarannya. f)
Pembelajaran kooperatif adalah teknik yang efektif untuk meningkatkan
kinerja akademik dan mengembangkan keterampilan sosial, seperti
memupuk hubungan interpersonal yang positif dan mengasah kemampuan
manajemen waktu. g)
Pembelajaran
kooperatif
meningkatkan
kapasitas
siswa
untuk
mengevaluasi konsep dan memperoleh kritik yang membangun. Siswa
dapat terlibat dalam latihan pemecahan masalah tanpa takut melakukan
kesalahan, karena kelompok secara kolektif mengasumsikan akuntabilitas
atas penilaian yang dibuat. h) Pembelajaran
kooperatif
meningkatkan
kapasitas
siswa
untuk
memanfaatkan informasi dan memperoleh kemahiran dalam memahami
konsep-konsep abstrak, sehingga memungkinkan penerapan praktisnya. Karimah Tauhid, Volume 3 Nomor 3 (2024), e-ISSN 2963-590X | Azizah dan Aliyyah Memahami dan memahami filosofi pembelajaran kooperatif memerlukan waktu dan
tidak dapat diharapkan dapat dipahami secara otomatis oleh siswa. Siswa yang mempunyai
kelebihan mungkin akan merasakan terkekang oleh siswa berkemampuan lebih rendah. Tentunya mengakibatkan situasi seperti ini berpotensi merusak suasana kolaboratif dalam
organisasi. Pembelajaran kooperatif ditandai dengan siswa saling memberi instruksi. Oleh
karena itu, tanpa adanya pengajaran sejawat yang efisien, siswa mungkin gagal memperoleh
tingkat pemahaman dan penguasaan yang diinginkan yang dapat dicapai dengan instruksi
langsung dari guru. Penilaian pembelajaran kooperatif diperoleh dari hasil kegiatan
kelompok kolaboratif. Meskipun demikian, penting bagi para pendidik untuk menyadarkan
jika hasil atau pencapaian yang ditujukan berkaitan secara khusus dengan keberhasilan setiap
siswa. Kemanjuran pembelajaran kooperatif dalam menumbuhkan rasa sadar kolektif
membutuhkan jangka waktu yang lama; hasil ini tidak memungkinkan dicapai melalui
penggunaan kerangka pembelajaran kooperatif secara soliter atau sporadis. Kemanjuran pembelajaran kooperatif dalam menumbuhkan rasa sadar kolektif
membutuhkan jangka waktu yang lama; hasil ini tidak memungkinkan dicapai melalui
penggunaan kerangka pembelajaran kooperatif secara soliter atau sporadis. Meskipun kolaborasi adalah keterampilan yang penting bagi siswa, banyak aktivitas dalam
kehidupan hanya bergantung pada bakat individu. Jadi, dalam skenario yang ideal,
pembelajaran kooperatif seharusnya tidak hanya menumbuhkan kolaborasi antar siswa,
namun juga memfasilitasi pengembangan rasa percaya diri. Namun, mencapai tujuan-tujuan
ini dalam konteks pembelajaran kooperatif adalah tugas yang menantang. Meskipun kolaborasi adalah keterampilan yang penting bagi siswa, banyak aktivitas dalam
kehidupan hanya bergantung pada bakat individu. Jadi, dalam skenario yang ideal,
pembelajaran kooperatif seharusnya tidak hanya menumbuhkan kolaborasi antar siswa,
namun juga memfasilitasi pengembangan rasa percaya diri. Namun, mencapai tujuan-tujuan
ini dalam konteks pembelajaran kooperatif adalah tugas yang menantang. f. Kelemahan Pembelajaran Kooperatif Selain kelebihannya, pembelajaran kooperatif juga mempunyai
keterbatasan atau kekurangan, yaitu: Selain kelebihannya, pembelajaran kooperatif juga mempunyai
keterbatasan atau kekurangan, yaitu: Selain kelebihannya, pembelajaran kooperatif juga mempunyai
keterbatasan atau kekurangan, yaitu: 3467 3467 Karimah Tauhid, Volume 3 Nomor 3 (2024), e-ISSN 2963-590X | Azizah dan Aliyyah dengan topik yang dipelajari, dengan pembatasan kerja kolaboratif antar siswa (Wardana,
Ika: 2017). dengan topik yang dipelajari, dengan pembatasan kerja kolaboratif antar siswa (Wardana,
Ika: 2017). B. Hasil Penelitian Temuan penelitian Penelitian ini menggunakan metodologi PTK. PTK dipandang
tepat dan efisien karena fokusnya menjawab masalah pembelajaran yang berkembang di
dalam kelas. Penelitian ini dilakukan dengan tujuan untuk menyempurnakan proses
belajar mengajar, serta menyelesaikan segala kesulitan belajar yang mungkin timbul. PTK
mudah dimanfaatkan dikarenakan tidak harus memperbandingkan model
pembelajarannya. Selanjutnya guru dapat sekaligus melakukan penelitian terhadap
permasalahan kelas sambil sekaligus melakukan kegiatan belajar mengajar. Tujuan
dilakukannya penelitian yakni untuk menentukan keefektifan diterapkannya STAD
untuk mambuat hasil belajar IPA siswa kelas IV meningkat di SD Negeri Cimande 01
yang terletak di Kecamatan Caringin Kabupaten Bogor. Berikut ini adalah garis besar
jadwal penelitian; A. Deksripsi Hasil Penelitian A. Deksripsi Hasil Penelitian Peneitian ini dilakukan di SDN Cimande 01 Kecamatan caringin, Kabupaten
Bogor. Penelitian ini di fokuskan dalam mata Pelajaran IPA memanfaatkan STAD di kelas
IV siswa yang ada di SDN Cimande 01 ini khususnya di kelas IV berjumlah 30 orang yang
terdiri dari 18 laki-laki dan 12 perempuan hal ini menjadikan pertimbangan seberapa jauh
Tingkat berhasil siswa dalam pembelajaran IPA yang di berikan oleh guru. Metode STAD
mengedepankan pembelajaran kooperatif dengan mengorganisasikan siswa ke dalam
kelompok. Pendekatan ini bertujuan untuk menumbuhkan kolaborasi, pemikiran kritis,
motivasi, dan rasa tanggung jawab dalam kelompok. Siswa memiliki kapasitas untuk
membantu teman-temannya dan diri mereka sendiri dengan berkolaborasi dalam kuis di
lain waktu, dengan tujuan akhir untuk mendapatkan penghargaan tim yang bergengsi. Melalui penilaian, siswa dapat secara ringkas mensintesis pengetahuan yang diperoleh
dari pengajaran guru dan hasil kegiatan kolaboratif. Guru menilai hasil belajar berkaitan 3468 Karimah Tauhid, Volume 3 Nomor 3 (2024), e-ISSN 2963-590X | Azizah dan Aliyyah PRA SIKLUS Analisis awal data sebelum dimulainya suatu siklus. Data hasil ujian pada saat akhir
pembelajaran mata pelajaran IPA prasiklus meliputi siswa yang tidak menyelesaikan
studinya dan tidak memenuhi KKM sebesar 65 yang sudah ditentukan. Dari 28 siswa,
yang belum lulus sejumlah 28,57% atau 8 siswa tidak tuntas, sedangkan 71,42% atau 20
siswa berhasil tuntas. Nilai maksimum yang dicapai adalah 90, sedangkan nilai minimum
adalah 60. Nilai rata-ratanya ialah 75,71. Hasil belajar prasiklus didistribusikan menurut
frekuensinya sebagai berikut: Gambar 2 Tingkat Ketuntasan Belajar Pra Siklus
Rentang
Kategori
Pra Siklus
Keterangan
Frekuensi
Persentase
86 - 100
Tinggi
0
0 %
Tuntas
65 - 85
Sedang
20
71,42 %
45 - 64
Rendah
8
28,57 %
Tidak Tuntas
25 - 44
Sangat Rendah
0
0 % Gambar 2 Tingkat Ketuntasan Belajar Pra Siklus Gambar 2 Tingkat Ketuntasan Belajar Pra Siklus 3469 Karimah Tauhid, Volume 3 Nomor 3 (2024), e-ISSN 2963-590X | Azizah dan Aliyyah Karimah Tauhid, Volume 3 Nomor 3 (2024), e-ISSN 2963-590X | Azizah dan Aliyyah 100 %
85
60
67,85
65
oleh ketuntasan belajar KKM 65,
mperoleh ketuntasan sama sekali. ediakan sebelumnya. Nilai rata-rata
ategi penelitian yang tercantumnya
ketidaktuntasan pembelajaran yang
meningkatkan pengetahuan ilmiah
TAD melalui dua siklus: satu siklus
da pemeliharaan kesehatan tulang. a yang telah menyelesaikan studinya
bel di bawah:
ajar Ipa Siswa Kelas Iv Cimande
an Belajar Pra
28.57
Tuntas
Tidak
k Total
28
100 %
Nilai Max
85
Nilai Min
60
Rata-rata
67,85
KKM
65
Dua puluh siswa (71,42%) memperoleh ketuntasan belajar KKM 65,
sedangkan delapan siswa (28,57%) tidak memperoleh ketuntasan sama sekali. Informasi ini berasal dari tabel yang sudah disediakan sebelumnya. Nilai rata-rata
kelas hanya 67,85. Peneliti akan mengikuti strategi penelitian yang tercantumnya
sebelumnya melalui PTK menyikapi tingkat ketidaktuntasan pembelajaran yang
mencapai 28,57%. Penelitian bertujuan dalam meningkatkan pengetahuan ilmiah
siswa kelas IV semester I melalui pemodelan STAD melalui dua siklus: satu siklus
pada makhluk hidup dan siklus lainnya pada pemeliharaan kesehatan tulang. Diagram batang menampilkan persentase siswa yang telah menyelesaikan studinya
sebelum siklus dimulai, seperti terlihat pada tabel di bawah: Total
28
100 %
Nilai Max
85
Nilai Min
60
Rata-rata
67,85
KKM
65 Dua puluh siswa (71,42%) memperoleh ketuntasan belajar KKM 65,
sedangkan delapan siswa (28,57%) tidak memperoleh ketuntasan sama sekali. Informasi ini berasal dari tabel yang sudah disediakan sebelumnya. Nilai rata-rata
kelas hanya 67,85. Peneliti akan mengikuti strategi penelitian yang tercantumnya
sebelumnya melalui PTK menyikapi tingkat ketidaktuntasan pembelajaran yang
mencapai 28,57%. Penelitian bertujuan dalam meningkatkan pengetahuan ilmiah
siswa kelas IV semester I melalui pemodelan STAD melalui dua siklus: satu siklus
pada makhluk hidup dan siklus lainnya pada pemeliharaan kesehatan tulang. Diagram batang menampilkan persentase siswa yang telah menyelesaikan studinya
sebelum siklus dimulai, seperti terlihat pada tabel di bawah: ar 3. Grafik Tingkat Ketuntasan Belajar Ipa Siswa Kelas Iv Cima
Tingkat Ketuntasan Belajar Pra
80 Siklus
60
40
28.57
20
Tuntas
Tidak
0
Tuntas
Tidak bar 3. Grafik Tingkat Ketuntasan Belajar Ipa Siswa Kelas Iv Cima Gambar 3. Grafik Tingkat Ketuntasan Belajar Ipa Siswa Kelas Iv Cimande
01 01 Tingkat Ketuntasan Belajar Pra Tuntas Tidak 3470 Karimah Tauhid, Volume 3 Nomor 3 (2024), e-ISSN 2963-590X | Azizah dan Aliyyah Karimah Tauhid, Volume 3 Nomor 3 (2024), e-ISSN 2963-590X | Azizah dan Aliyyah
Dari grafik tingkat ketuntasan belajar IPA siswa kelas IV SDN CIMANDE 01, ada 20
siswa yang melampaui KKM yaitu 65 atau 71,42% dari total. Sementara yang belum
tuntas KKM sebanyak 8 siswa yaitu sebesar 28,57%. Karimah Tauhid, Volume 3 Nomor 3 (2024), e-ISSN 2963-590X | Azizah dan Aliyy
Dari grafik tingkat ketuntasan belajar IPA siswa kelas IV SDN CIMANDE 01, ada 20
siswa yang melampaui KKM yaitu 65 atau 71,42% dari total. Sementara yang belum
tuntas KKM sebanyak 8 siswa yaitu sebesar 28,57%. Karimah Tauhid, Volume 3 Nomor 3 (2024), e-ISSN 2963-590X | Azizah dan Aliyyah arimah Tauhid, Volume 3 Nomor 3 (2024), e-ISSN 2963-590X | Azizah dan Aliyyah Dari grafik tingkat ketuntasan belajar IPA siswa kelas IV SDN CIMANDE 01, ada 20
siswa yang melampaui KKM yaitu 65 atau 71,42% dari total. Sementara yang belum
tuntas KKM sebanyak 8 siswa yaitu sebesar 28,57%. Dari grafik tingkat ketuntasan belajar IPA siswa kelas IV SDN CIMANDE 01, ada 20
siswa yang melampaui KKM yaitu 65 atau 71,42% dari total. Sementara yang belum
tuntas KKM sebanyak 8 siswa yaitu sebesar 28,57%. 3471 Karimah Tauhid, Volume 3 Nomor 3 (2024), e-ISSN 2963-590X | Azizah dan Aliyyah 1) Pelaksanaan Siklus I Penelitian ini dilakukan peneliti bekerjasama dengan SDN Cimande 01 yang
terletak di Kecamatan Caringin Kabupaten Bogor. Penelitian tersebut secara
khusus mata pelajaran IPA siswa kelas IV. 2) Perencanaan SIKLUS I 1) Pelaksanaan Siklus I Karimah Tauhid, Volume 3 Nomor 3 (2024), e-ISSN 2963-590X | Azizah dan Aliyyah Karimah Tauhid, Volume 3 Nomor 3 (2024), e-ISSN 2963-590X | Azizah dan Aliyyah Siklus awal dilakukan di hari Senin tanggal 16 Oktober 2023 yang
meliputi rangkaian tugas: a. Kegiatan Awal
Kegiatan diawali dari menata ruang kelas, memberi salam, selanjutnya doa
sebelum mengikuti pembelajaran, dan diakhiri dengan verifikasi kehadiran untuk
memastikan
kehadiran
siswa. Selanjutnya,
instruksikan
siswa
untuk
mengumpulkan dan mengatur bahan tulis dan buku teks yang diperlukan untuk
tugas pendidikan mereka. Apersepsi dan motivasi berupaya merangsang
keterlibatan kognitif siswa dengan aktivitas keseharian berdasarkan tematik
dengan materi pelajaran yang dipelajari. Kemudian guru mengkkordinasikan
tujuan pembelajaran yang perlu dicapai. a. a. Kegiatan Awal
Kegiatan diawali dari menata ruang kelas, memberi salam, selanjutnya doa
sebelum mengikuti pembelajaran, dan diakhiri dengan verifikasi kehadiran untuk
memastikan
kehadiran
siswa. Selanjutnya,
instruksikan
siswa
untuk
mengumpulkan dan mengatur bahan tulis dan buku teks yang diperlukan untuk
tugas pendidikan mereka. Apersepsi dan motivasi berupaya merangsang
keterlibatan kognitif siswa dengan aktivitas keseharian berdasarkan tematik
dengan materi pelajaran yang dipelajari. Kemudian guru mengkkordinasikan
tujuan pembelajaran yang perlu dicapai. b. Dalam tahapan ini peneliti melakukan beberapa langkah atau tugas
persiapan yang akan dilaksanakan selama penelitian, yakni: Dalam tahapan ini peneliti melakukan beberapa langkah atau tugas
persiapan yang akan dilaksanakan selama penelitian, yakni: a) Penelitian akan dilakukan di SDN Cimande 01 yang terletak di
Kecamatan Caringin Kabupaten Bogor. a) Penelitian akan dilakukan di SDN Cimande 01 yang terletak di
Kecamatan Caringin Kabupaten Bogor. b) Peneliti mengumpulkan data melalui observasi dan wawancara yang
dilakukan kepada guru kelas IV dan kepala sekolah. Tujuannya adalah
untuk
mengidentifikasi
titik
fokus
pembelajaran
dengan
menggunakan paradigma pembelajaran STAD pada mapel IPA kelas
IV. c) Peneliti menentukan Standar Kompetensi (SK) yang akan diteliti,
khusus berfokus pada 1. Organisme dan atributnya. Identifikasi
Kompetensi Dasar (KD) khusus yang akan diperiksa, khususnya 1.1. Makhluk hidup mengacu pada organisme yang memiliki kemampuan
untuk
tumbuh,
berkembang
biak,
merespons
rangsangan,
mempertahankan homeostatis, dan memperoleh serta memanfaatkan
energi. Ciri-ciri makhluk hidup antara lain organisasi seluler, materi
genetik, metabolisme, adaptasi, dan kemampuan berevolusi. d) Menyusun RPP mata pelajaran IPA dengan model pembelajaran
STAD dalam rentang waktu dua kali pertemuan. d) Menyusun RPP mata pelajaran IPA dengan model pembelajaran
STAD dalam rentang waktu dua kali pertemuan. d) Menyusun RPP mata pelajaran IPA dengan model pembelajaran
STAD dalam rentang waktu dua kali pertemuan. e) Membuat kuesioner yang menanyakan mengenai topik Makhluk
Hidup. e) Membuat kuesioner yang menanyakan mengenai topik Makhluk
Hidup. Membentuk
tabel untuk mendokumentasikan
observasi
yang
dilakukan guru dan siswa pada saat pembelajaran IPA Pertemuan 1 3472 a. Kegiatan Awal Mengawali pembelajaran dengan mengatur kelas, memberi salam, dan
berdoa sebelum memulai kegiatan belajar mengajar, dilanjutkan dengan verifikasi
kehadiran siswa melalui proses absensi. Selanjutnya, instruksikan siswa untuk
mengumpulkan alat tulis dan buku teks yang akan digunakan untuk tugas-tugas
pendidikan. Apersepsi dan motivasi berupaya merangsang keterlibatan kognitif
siswa dengan aktivitas keseharian berdasarkan tematik yang berkenaan dengan
materi pelajaran yang dipelajari. Selanjutnya guru mengkomunikasikan tujuan
pembelajaran yang ingin dicapai sebagai pengingat. Pertemuan 2 Pertemuan kedua siklus I dilakukan di hari Selasa tanggal 17 Oktober 2023
yang mencakup; Kegiatan Inti Kegiatan utama dimulai dengan guru menjelaskan materi pelajaran yang akan
diperiksa, menggunakan alat bantu pengajaran yang bertujuan dalam
memudahkan pemahaman siswa bagi materi yang ada. Sesuai instruksi guru,
siswa disusun menjadi kelompok yang berjumlah 4-6 siswa per meja. Setiap
kelompoknya dibekali bahan diskusi dan diawasi oleh guru. Setelah siswa
menyelesaikan diskusinya dalam kelompok masing-masing, guru menunjuk
satu kelompok agar berpresentasi mengenai hasil diskusinya, sedang kelompok
lainnya diinstruksikan untuk memberikan umpan balik kepada kelompok yang
menyajikan. Guru memerintahkan siswa untuk kembali ke tempat yang telah
ditentukan. 3473 Karimah Tauhid, Volume 3 Nomor 3 (2024), e-ISSN 2963-590X | Azizah dan Aliyyah c.
Kegiatan Akhir Selama kegiatan penutup, guru dan siswa terlibat dalam sesi
tanya jawab untuk merangkum materi pelajaran, menyoroti poin-poin
utama. Menugaskan siswa tugas melakukan observasi tambahan di
rumah dan menyimpulkan pembelajaran b. Kegiatan Inti Kegiatan inti pada pertemuan kedua berkisar pada diskusi mengenai materi
sebelumnya. Guru memberikan pedoman dalam melakukan evaluasi kepada siswa. Kemudian, guru memberi kuesioner dan lembar respon kepada siswa,
menginstruksikan 3474 Karimah Tauhid, Volume 3 Nomor 3 (2024), e-ISSN 2963-590X | Azizah dan Aliyyah mereka untuk memberikan informasi pribadi mereka di lembar jawaban. Ada total 20 soal evaluasi yang harus diselesaikan dalam jangka waktu 40
menit. Sesudah soal evaluasi terjawab, siswa mengumpulkan hasil
pekerjaannya secara berurutan dari kursi belakang hingga kedepan
(estapet). c. Kegiatan Akhir Pada latihan penutup, guru dan siswa terlibat dalam sesi tanya jawab
sesuai dengan isi pelajaran. Sebelum mengakhiri pelajaran, guru
menginstruksikan siswa untuk mengatur alat tulis mereka, kemudian
mnyudahi pelajaran dengan salam akhir. d. Hasil Observasi Siklus I Pada kajian ini, peneliti juga mengkaji proses pengajaran dan
pendidikan yang berlangsung antara guru dan siswa. Penelitian
observasional ini sejalan dengan temuan peneliti yang tertuang dalam RPP. Meskipun demikian, peneliti masih menghadapi berbagai tantangan, seperti
ditemuinya siswa yang pasif, siswa yang menolak kerja kelompok, siswa
yang terlalu sibuk dengan tugas individu dan kurang perhatian, serta belum
lengkapnya pelaksanaan langkah-langkah tertentu dalam RPP. Menganalisis
data yang diperoleh dari observasi kegiatan pengajaran dilakukan oleh
pengajar kelas IV dalam dua sesi pada siklus I, dengan memanfaatkan model
STAD untuk kelas sains. Standar Kompetensi 1. Memahami hubungan antara
susunan organ tubuh manusia dengan tujuannya, serta pemeliharaannya. Kemahiran mendasar: Makhluk hidup adalah organisme yang memiliki sifat
dasar kehidupan. Ciri-ciri tersebut meliputi kemampuan untuk tumbuh,
bereproduksi, merespon rangsangan, mempertahankan homeostatis, serta
memperoleh dan menggunakan energi. Informasinya ditampilkan pada tabel
di bawah ini: 3475 Karimah Tauhid, Volume 3 Nomor 3 (2024), e-ISSN 2963-590X | Azizah dan Aliyyah 3476
Gambar 4. Hasil Observasi Kegiatan Guru Siklus I
No
Aspek pengamatan
Pertemuan I
Pertemuan II
1
2
3
4
1
2
3
4
1
Guru menilai kesiapan kelas dan
peralatan, serta materi pendidikan. √
√
2
Guru menilai kesiapan siswa. √
√
3
Guru menanamkan pemahaman dan
semangat kepada siswa. √
√
4
Guru menyampaikan tujuan
pembelajaran kepada siswa. √
√
5
Guru memberikan penjelasan singkat
yang berfokus pada aspek penting
dari bidang materi. √
√
6
Guru mengkomunikasikan informasi
secara efektif menggunakan sumber
daya pendidikan yang sesuai. √
√
7
Guru membentuk siswa menjadi
individu-individu dalam kelompok
kecil yang terdiri dari 4-6 siswa. √
√
8
Guru melibatkan siswa dalam proses
pembelajaran melalui penggunaan
media ini. √
√
9
Guru memberikan bimbingan
individual dan adil kepada kelompok
dalam melaksanakan strategi. √
√ Gambar 4. Hasil Observasi Kegiatan Guru Siklus I 3476 10
Guru menawarkan siswa kesempatan
untuk berkolaborasi dan
menyampaikan presentasi kepada
seluruh kelas. √
√
11
Guru mendorong kelompok untuk
menumbuhkan kreativitas dan
keterlibatan siswa dalam lingkungan
kelompok. √
√
12
Guru menginstruksikan siswa dalam
proses mensintesis dan mengatur
ringkasan isi materi. √
√
13
Guru terlibat dalam diskusi reflektif
dengan siswa untuk mengatasi
setiap anomali dalam mata
pelajaran. Selain itu, guru
memberikan umpan balik kepada
siswa sehingga membina hubungan
timbal balik. √
√
14 Guru memberikan umpan balik kepada
siswa sehingga membina hubungan
timbal balik. Karimah Tauhid, Volume 3 Nomor 3 (2024), e-ISSN 2963-590X | Azizah dan Aliyyah Karimah Tauhid, Volume 3 Nomor 3 (2024), e-ISSN 2963-590X | Azizah dan Aliyyah Karimah Tauhid, Volume 3 Nomor 3 (2024), e-ISSN 2963-590X | Azizah dan Aliyyah e. Refleksi Siklus I Berdasarkan tabel 4.3 hasil observasi aktivitas belajar siswa siklus I
selama dua sesi menghasilkan total skor 30 pada pertemuan pertama, dengan
skor rata-ratanya 3 dengan kategori baik. Untuk pertemuan kedua, total
skornya adalah 38 dengan rata-ratanya 3,8 dalam kategori baik. Catatan
peneliti pada lembar observasi guru dan siswa menunjukkan bahwa pada
pertemuan awal siswa kurang aktif mengemukakan pendapatnya dan masih
terdapat beberapa siswa yang enggan berkolaborasi dalam kelompoknya. Akibatnya penerapan model pembelajaran STAD kurang lancar. Sebagai
ilustrasi, Sebelum memulai kerja kelompok, guru mengatur siswa menjadi
satu kelompok yang terdiri dari empat orang. Selanjutnya guru memberikan
arahan kepada siswa tentang tata cara bekerja dalam kelompok, khususnya
menggunakan langkah-langkah model STAD. Pada tahap awal, setiap siswa
dalam kelompok secara mandiri menyelesaikan soal-soal pada lembar
kerjanya (Think) dengan durasi waktu lima menit. Setelah selesai, guru
memintakan siswanya agar melakuakn diskusi dengan berpasangan (pair),
di mana mereka membandingkan jawaban masing-masing. Jika ada
perbedaan yang muncul, siswa menjelaskan alasannya satu sama lain untuk
mencapai kesepakatan. Setelah diskusi berpasangan, guru menginstruksikan
siswa untuk membentuk kelompok yang lebih besar (Share) yang terbagi
menjadi 4 siswa. Pada kelompok ini, siswa terlibat dalam diskusi mengenai
hasil pekerjaan mereka. Apabila terdapat perbedaan jawaban dalam suatu
kelompok, siswa bekerjasama dalam kelompoknya untuk mengidentifikasi
dan menentukan jawaban yang tepat. Metode optimal untuk menyelesaikan
pertanyaan. Selama penerapan model pembelajaran STAD, beberapa siswa
terus menerus mengalami kebingungan mengenai proses bekerja dalam
kelompok. Dalam konferensi tersebut, peneliti juga mengungkapkan
ketidakpuasannya terhadap terbatasnya waktu yang dialokasikan untuk
berdiskusi, karena sebagian siswa belum memanfaatkan kesempatan
tersebut secara maksimal untuk mencapai hasil diskusi yang optimal. Peneliti berpendapat bahwa penerapan model pembelajaran STAD belum
berhasil. Meskipun demikian, pendidik dan peneliti yang memainkan peran
menjadi guru dan fasilitator pada saat berlangsungnya pembelajaran melalui Berdasarkan tabel 4.3 hasil observasi aktivitas belajar siswa siklus I
selama dua sesi menghasilkan total skor 30 pada pertemuan pertama, dengan
skor rata-ratanya 3 dengan kategori baik. Untuk pertemuan kedua, total
skornya adalah 38 dengan rata-ratanya 3,8 dalam kategori baik. Catatan
peneliti pada lembar observasi guru dan siswa menunjukkan bahwa pada
pertemuan awal siswa kurang aktif mengemukakan pendapatnya dan masih
terdapat beberapa siswa yang enggan berkolaborasi dalam kelompoknya. Akibatnya penerapan model pembelajaran STAD kurang lancar. Sebagai
ilustrasi, Sebelum memulai kerja kelompok, guru mengatur siswa menjadi
satu kelompok yang terdiri dari empat orang. Selanjutnya guru memberikan
arahan kepada siswa tentang tata cara bekerja dalam kelompok, khususnya
menggunakan langkah-langkah model STAD. d. Hasil Observasi Siklus I √
√
15
Guru menilai hasil belajar pada akhir
periode pembelajaran
√
√ 3477 16
Guru mengakhiri pelajaran dengan
meminta siswa merapihkan alat
tulisannya
√
√
Jumlah
18
40
9
52
Total skor
58
61
Rata-rata
3,63
3,81
Kategori
Baik
Baik Berdasarkan data pada tabel 4.2 hasil observasi aktivitas mengajar
guru pada siklus I diperoleh hasil sebagai berikut: total skor pertemuan
pertamanya ialah 58 dengan skor rata-ratanya 3,63 termasuk dalam kategori
baik. Pada pertemuan kedua, total skornya adalah 61 dengan rata-rata skor
3,81 juga termasuk dalam kategori baik. Gambar 5. Hasil Observasi Kegiatan Siswa Siklus I
No
Aspek pengamatan
Pertemuan I
Pertemuan II
1
2
3
4
1
2
3
4
1
Kesiapan siswa untuk mengikuti
pelajaran
√
√
2
Perhatikan baik-baik saat kompetensi
dan tujuan Pencapaian pengetahuan
dan keterampilan
√
√
√
3
Penting untuk fokus secara hati-hati
ketika informasi pembelajaran sedang
dibahas. √
√ Gambar 5. Hasil Observasi Kegiatan Siswa Siklus I 3478 3479
4
Siswa hendaknya berpartisipasi aktif
dan menunjukkan semangat selama
proses pembelajaran. √
√
5
Terdapat korelasi yang baik antara
siswa dengan model pembelajaran
yang diterapkan. √
√
6
Siswa dapat berkolaborasi secara
efektif untuk menyelesaikan lembar
kerja sebagai kelompok. √
√
7
Siswa menunjukkan tingkat tanggung
jawab yang tinggi ketika melakukan
kegiatan presentasi di depan kelas. √
√
8
Siswa mempunyai kemampuan untuk
menanggapi pertanyaan yang
diajukan oleh guru. √
√
9
Siswa menunjukkan keterlibatan dan
partisipasi ketika memadatkan dan
menyajikan poin-poin utama materi
pelajaran
√
√
10
Siswa menunjukkan reaksi yang baik
ketika evaluasi dilakukan. √
√
Jumlah
10
2
8
Total skor
30
38
Rata-rata
3,0
3,8 3479 Karimah Tauhid, Volume 3 Nomor 3 (2024), e-ISSN 2963-590X | Azizah dan Aliyyah Karimah Tauhid, Volume 3 Nomor 3 (2024), e-ISSN 2963-590X | Azizah dan Aliyyah Karimah Tauhid, Volume 3 Nomor 3 (2024), e-ISSN 2963-590X | Azizah dan Aliyyah
Kategori
Baik
Baik Dari tabel 4.3 hasil observasi aktivitas belajar siswa siklus I selama dua sesi
menghasilkan total skor 30 pada pertemuan pertama, dengan skor rata-rata 3
dengan kategori baik. Untuk pertemuan kedua, total skornya adalah 38 dengan
rata-rata 3,8 dalam kategori baik. 3480 Karimah Tauhid, Volume 3 Nomor 3 (2024), e-ISSN 2963-590X | Azizah dan Aliyyah e. Refleksi Siklus I Pada tahap awal, setiap siswa
dalam kelompok secara mandiri menyelesaikan soal-soal pada lembar
kerjanya (Think) dengan durasi waktu lima menit. Setelah selesai, guru
memintakan siswanya agar melakuakn diskusi dengan berpasangan (pair),
di mana mereka membandingkan jawaban masing-masing. Jika ada
perbedaan yang muncul, siswa menjelaskan alasannya satu sama lain untuk
mencapai kesepakatan. Setelah diskusi berpasangan, guru menginstruksikan
siswa untuk membentuk kelompok yang lebih besar (Share) yang terbagi
menjadi 4 siswa. Pada kelompok ini, siswa terlibat dalam diskusi mengenai
hasil pekerjaan mereka. Apabila terdapat perbedaan jawaban dalam suatu
kelompok, siswa bekerjasama dalam kelompoknya untuk mengidentifikasi
dan menentukan jawaban yang tepat. Metode optimal untuk menyelesaikan
pertanyaan. Selama penerapan model pembelajaran STAD, beberapa siswa
terus menerus mengalami kebingungan mengenai proses bekerja dalam
kelompok. Dalam konferensi tersebut, peneliti juga mengungkapkan
ketidakpuasannya terhadap terbatasnya waktu yang dialokasikan untuk
berdiskusi, karena sebagian siswa belum memanfaatkan kesempatan
tersebut secara maksimal untuk mencapai hasil diskusi yang optimal. Peneliti berpendapat bahwa penerapan model pembelajaran STAD belum
berhasil. Meskipun demikian, pendidik dan peneliti yang memainkan peran
menjadi guru dan fasilitator pada saat berlangsungnya pembelajaran melalui 3481 Karimah Tauhid, Volume 3 Nomor 3 (2024), e-ISSN 2963-590X | Azizah dan Aliyyah Karimah Tauhid, Volume 3 Nomor 3 (2024), e-ISSN 2963-590X | Azizah dan Aliyyah STAD
berupaya
mengoptimalkan
suasana
kelas
untuk
menjamin
pengendalian yang efektif, sehingga memungkinkan siswa untuk aktif
terlibat dan memahami materi pembelajaran. Untuk memperbaiki
kekurangan yang diamati di awal. konferensi, peneliti merancang
penyempurnaan dengan tujuan mencapai hasil yang lebih baik pada
pertemuan berikutnya. Peneliti memodifikasi alokasi waktu bagi siswa
untuk terlibat dalam diskusi kelompok dan membuat konten media yang
menarik, seperti film yang menyoroti struktur dan fungsi kerangka manusia. Selain itu, peneliti memfasilitasi koordinasi instruksi kepada guru guna
meningkatkan
kemampuannya
dalam
mengkomunikasikan
proses
pembelajaran secara efektif dengan menggunakan model STAD. Dalam
pertemuan 2 siklus I peneliti masih memanfaatkan model STAD sebagai
metodenya. Selama pertemuan ini, hasilnya menyatakan lebih baik. Model
pembelajaran STAD. Secara efektif meningkatkan keterlibatan siswa dalam
mengungkapkan sudut pandangnya selama proses pembelajaran. Hal ini
ditunjukkan melalui latihan kelompok, di mana peneliti mencatat bahwa
sebagian besar siswa mahir dalam berbagi tanggapan dan mengartikulasikan
alasan yang mendasarinya kepada pasangan dan kelompok mereka. Untuk
menjamin kelancaran pelaksanaannya, disarankan untuk menyajikan hasil
observasi guru dan siswa dalam sebuah tabel pada pertemuan kedua pada
lembar observasi. Hasilnya melebihi hasil pertemuan awal. Guru secara
efektif mencapai kemajuan dalam pertemuan ini dengan berhasil
mengkomunikasikan instruksi penerapan paradigma pembelajaran STAD
yang dikembangkan dari pertemuan sebelumnya. Pertemuan 1 Siklus II dilakukan di hari Selasa tanggal 2 Agustus 2016 dengan kegiatan antara
lain kegiatan Awal Kegiatan diawali dengan persiapan ruang kelas, saling bertukar sapa,
dilanjutkan dengan doa sebelum perkuliahan dimulai, dan selanjutnya
melakukan absensi untuk memverifikasi kehadiran siswa. Selanjutnya,
instruksikan siswa untuk mempersiapkan peralatan yang diperlukan. SIKLUS II Penelitian tersebut dilakukan oleh peneliti yang bekerjasama dengan SD
Negeri Cimande 01 yang terletak di Kecamatan Caringin Kabupaten Bogor. Dipusatkan pada mata pelajaran IPA untuk kelas IV. Dalam fase ini peneliti
melakukan beberapa langkah atau tugas persiapan yang akan dilaksanakan
selama penelitian berlangsung, yaitu: a) Identifikasi lokasi spesifik yang akan dimanfaatkan untuk
penelitian, khususnya SDN CIMANDE 01. a) Identifikasi lokasi spesifik yang akan dimanfaatkan untuk
penelitian, khususnya SDN CIMANDE 01. 3482 Karimah Tauhid, Volume 3 Nomor 3 (2024), e-ISSN 2963-590X | Azizah dan Aliyyah b) Peneliti mengambil data dari temuan penelitian pada siklus I. c) Identifikasi bidang minat penelitian tertentu dengan memanfaatkan
paradigma pembelajaran STAD di kelas IPA kelas empat. d) Peneliti menentukan Standar Kompetensi (SK) yang akan diperiksa,
khususnya Memahami organisme dan sifat-sifatnya. Identifikasi
Kompetensi Dasar (KD) khusus yang akan dipelajari, khususnya
fokus pada konsep Kewarganegaraan pada makhluk hidup. e) Membuat RPP mata pelajaran IPA melalui pemodelan STAD dalam
rentang waktu dua kali pertemuan. e) Membuat RPP mata pelajaran IPA melalui pemodelan STAD dalam
rentang waktu dua kali pertemuan. f) Membuat kuesioner yang yang menanyakan berkenaan topik
Makhluk Hidup. g) Buat format tabulasi untuk mencatat pengamatan yang dilakukan
oleh guru dan siswa selama pelajaran IPA Pertemuan 1 Pertemuan 1 A. Kegiatan Awal Mengawali pembelajaran dengan mengorganisasi kelas, memberi
salam, dan berdoa sebelum memulai kegiatan belajar mengajar, dilanjutkan
dengan mencatat kehadiran untuk memverifikasi kehadiran siswa. Selanjutnya, instruksikan siswa untuk mengumpulkan alat tulis dan buku
pelajaran yang dipakai selama tugas pendidikan. Apersepsi dan motivasi
berupaya merangsang keterlibatan kognitif siswa dengan aktivitas
keseharian berdasarkan tematik yang berkenaan dengan materi pelajaran
yang dipelajari. kemudian, guru mengkomunikasikan tujuan pendidikan
yang ingin dicapai kembali. b). Kegiatan Inti Tugas utama dimulai dengan guru mengkomunikasikan tujuan
pendidikan. Perwakilan siswa dipanggil ke depan dimana tersedia dua kursi. Mereka diinstruksikan untuk berlatih duduk dan berdiri, sementara siswa
lainnya mengamati dan bereaksi. Memberikan gambaran singkat tentang
pokok bahasan yang akan dibahas secara kolektif. Siswa dikelompokan
dengan beranggotakan empat orang. Setelah siswa duduk berpasangan
dalam kelompoknya masing-masing, guru memberikan petunjuk mengenai 3483 Karimah Tauhid, Volume 3 Nomor 3 (2024), e-ISSN 2963-590X | Azizah dan Aliyyah Karimah Tauhid, Volume 3 Nomor 3 (2024), e-ISSN 2963-590X | Azizah dan Aliyyah kegiatan yang hendak dikerjakan siswa secara bersamaan.Memberi arahan
pada setiap kelompoknya, kemudian perintahkan siswa untuk mencari
informasi dari materi yang diberikan (Think). Guru menyebarkan lembar
diskusi kepada masing-masing kelompok untuk kerja sama antar
anggotanya
masing-masing
(pair). Setelah
itu,
perwakilan
setiap
kelompoknya wajib menyampaikan presentasi hasil kerjanya, sedangkan
kelompok lain diharapkan memberikan masukan (Share). Selanjutnya guru
dan siswa melakukan proses introspeksi untuk meningkatkan pemahaman
siswa terhadap mata pelajaran yang dipelajari. a. Pertemuan 2 Pertemuan kedua siklus II dilakukan di hari Rabu tanggal 3 Agustus
2016 dengan tugas seperti: c). Kegiatan Akhir Pada tahap kegiatan menyimpulkan, guru dan siswa menilai hasil
pembelajaran. Selanjutnya, instruksikan siswa untuk mengumpulkan alat
tulis dan buku mereka. Guru mengakhiri sesi pembelajaran dengan doa. B. Kegiatan Inti Kegiatan utama pertemuan kedua berkisar pada diskusi mengenai
materi sebelumnya. Guru memperkenalkan siswa dengan protokol untuk
melakukan evaluasi. Selanjutnya, guru membagikan kuesioner dan lembar 3484 Karimah Tauhid, Volume 3 Nomor 3 (2024), e-ISSN 2963-590X | Azizah dan Aliyyah Karimah Tauhid, Volume 3 Nomor 3 (2024), e-ISSN 2963-590X | Azizah dan Aliyyah jawaban kepada siswa, menginstruksikan mereka untuk memberikan
informasi pribadi mereka di halaman jawaban. Ada total 20 soal evaluasi
yang harus diselesaikan dalam jangka waktu 40 menit. Sesudah menjawab
soal evaluasi, siswa mengumpulkan hasil pekerjaannya secara berurutan,
dimulai dari kursi belakang dan berlanjut ke kursi depan (estapet). C. Kegiatan Akhir Selama fase penutup latihan, guru dan siswa terlibat dalam sesi tanya
jawab untuk merangkum materi pelajaran, menyoroti topik-topik utama. Menugaskan siswa tugas untuk melakukan lebih banyak observasi di rumah,
dan menyimpulkan pembelajaran D. Hasil Pengamatan Siklus II Selama penelitian ini, peneliti juga mengkaji pertukaran instruksional
dan pendidikan antara pendidik dan peserta didik. Data dianalisis dari
observasi aktivitas mengajar guru sepanjang siklus II dalam dua pertemuan
yang dilaksanakan oleh pengamat yakni guru kelas IV. Observasi terfokus
pada kelas IPA yang memanfaatkan pembelajaran kooperatif bentuk STAD
Standar Kompetensi 1. Memahami makhluk hidup beserta Ciri-cirinya. Kompetensi dasar menerapkan cara pemeliharan kesehatan kerangka tubuh. Tercantum pada table dibawah: Gambar 6. Hasil Observasi Kegiatan Guru Siklus Ii
No
Aspek pengamatan
Pertemuan I
Pertemuan II
1 2 3
4
1
2 3
4
1
Guru menilai kesiapan kelas dan
peralatan, serta materi pendidikan. √
√
2
Guru menilai kesiapan siswa. √
√
3
Guru menanamkan pemahaman dan
semangat kepada siswa. √
√ Gambar 6. Hasil Observasi Kegiatan Guru Siklus Ii
No
Aspek pengamatan
Pertemuan I
Pertemuan II
1 2 3
4
1
2 3
4
1
Guru menilai kesiapan kelas dan
peralatan, serta materi pendidikan. √
√
2
Guru menilai kesiapan siswa. √
√
3
Guru menanamkan pemahaman dan
semangat kepada siswa. √
√ Gambar 6. Hasil Observasi Kegiatan Guru Siklus Ii 3485 Karimah Tauhid, Volume 3 Nomor 3 (2024), e-ISSN 2963-590X | Azizah dan Aliyyah 4
Guru menyampaikan tujuan
pembelajaran kepada siswa. √
√
5
Guru memberikan penjelasan singkat
yang berfokus pada aspek penting
dari bidang materi. √
√
6
Guru mengkomunikasikan informasi
secara efektif menggunakan sumber
daya pendidikan yang sesuai. √
√
7
Guru membentuk siswa menjadi
individu-individu dalam kelompok
kecil yang terdiri dari 4-6 siswa. √
√
8
Guru melibatkan siswa dalam proses
pembelajaran melalui penggunaan
media ini. √
√
9
Guru memberikan bimbingan
individual dan adil kepada kelompok
dalam melaksanakan strategi. √
√
10
Guru menawarkan siswa kesempatan
untuk berkolaborasi dan
menyampaikan presentasi kepada
seluruh kelas. √
√ 3486 Karimah Tauhid, Volume 3 Nomor 3 (2024), e-ISSN 2963-590X | Azizah dan Aliyyah 11
Guru mendorong kelompok untuk
menumbuhkan kreativitas dan
keterlibatan siswa dalam lingkungan
kelompok. √
√
12
Guru menginstruksikan siswa dalam
proses mensintesis dan mengatur
ringkasan isi materi. √
√
13
Guru terlibat dalam diskusi reflektif
dengan siswa untuk mengatasi
setiap anomali dalam mata
pelajaran. Selain itu, guru
memberikan umpan balik kepada
siswa sehingga membina hubungan
timbal balik. D. Hasil Pengamatan Siklus II √
√ 3487 Karimah Tauhid, Volume 3 Nomor 3 (2024), e-ISSN 2963-590X | Azizah dan Aliyyah 14
Guru memberikan umpan balik
kepada siswa sehingga membina
hubungan timbal balik
√
√
15
Guru menilai hasil belajar pada akhir
periode pembelajaran
√
√
16
Guru mengakhiri pelajaran dengan
meminta siswa merapihkan alat
tulisannya
√
√
Jumlah
16
16
Total skor
64
64
Rata-rata
4
4
Kategori
Sangat Baik
Sangat Baik 3488 Karimah Tauhid, Volume 3 Nomor 3 (2024), e-ISSN 2963-590X | Azizah dan Aliyyah Dari hasil pengobservasian guru pada siklus I diperoleh hasil yakni:
total skor pertemuan pertama adalah 64 yang skor rata-ratanya 4 termasuk
dalam kategori “sangat baik”. Begitu pula pada pertemuan kedua yang juga
memperoleh skor total 64 dengan skor rata-rata 4 yang terkategori “baik”. Gambar 7. Hasil Observasi Kegiatan Siswa Siklus II Gambar 7. Hasil Observasi Kegiatan Siswa Siklus II Gambar 7. Hasil Observasi Kegiatan Siswa Siklus II
No
Aspek pengamatan
Pertemuan I
Pertemuan II
1
2 3
4
1
2 3
4
1
Kesiapan siswa untuk mengikuti
pelajaran
√
√
2
Perhatikan baik-baik saat kompetensi
dan tujuan Pencapaian pengetahuan
dan keterampilan
√
√
3
Penting untuk fokus secara hati-hati
ketika informasi pembelajaran sedang
dibahas. √
√
4
Siswa hendaknya berpartisipasi aktif
dan menunjukkan semangat selama
proses pembelajaran. √
√
5
Terdapat korelasi yang baik antara
siswa dengan model pembelajaran
yang diterapkan. √
√
6
Siswa dapat berkolaborasi secara
efektif untuk menyelesaikan lembar
kerja sebagai kelompok. √
√ 3489 Karimah Tauhid, Volume 3 Nomor 3 (2024), e-ISSN 2963-590X | Azizah dan Aliyyah 7
Siswa menunjukkan tingkat tanggung
jawab yang tinggi ketika melakukan
kegiatan presentasi di depan kelas. √
√
8
Siswa mempunyai kemampuan untuk
menanggapi pertanyaan yang
diajukan oleh guru. √
√
9
Siswa menunjukkan keterlibatan dan
partisipasi ketika memadatkan dan
menyajikan poin-poin utama materi
pelajaran
√
√
10
Siswa menunjukkan reaksi yang baik
ketika evaluasi dilakukan. √
√
Jumlah
10
10
Total skor
40
40
Rata-rata
4
4
Kategori
Sangat Baik
Sangat Baik Berdasarkan tabel yang tersedia, hasil pemantauan aktivitas belajar
siswa Siklus I selama dua sesi adalah sebagai berikut: Total nilai yang dicapai
pada pertemuan pertama adalah 40, dengan rata-rata nilai 4 yang terkategori
“sangat baik”. Pada pertemuan kedua, total skornya juga 40 dengan rata-rata
skor 4 dengan kategori “baik”. E. Refleksi Siklus II 3490 Karimah Tauhid, Volume 3 Nomor 3 (2024), e-ISSN 2963-590X | Azizah dan Aliyyah Karimah Tauhid, Volume 3 Nomor 3 (2024), e-ISSN 2963-590X | Azizah dan Aliyyah Penelitian
yang
dilakukan
peneliti
melalui
pemodelan
STAD
menunjukkan adanya kesenjangan yang mencolok pada hasil belajar siswa kelas
IV SDN Cimande 01 pada siklus II. Peneliti mencapai keberhasilan dalam
meningkatkan kinerja akademis dalam disiplin sains, khususnya dalam domain
sistem kerangka manusia dan proses fisiologisnya, sekaligus meningkatkan
kesehatan kerangka. Hal ini tercantum dari indikator-indikator yang diperoleh dari observasi
yang dilakukan selama siklus II yang berlangsung dalam dua sesi. Peneliti
melakukan rancangan siklus II menjadi dua sesi. Pada sesi awal siklus I, terlihat
bahwa siswa berprestasi dalam tes evaluasi. Berdasarkan tabel observasi, sebagian
besar siswa menunjukkan peningkatan keterlibatan dalam proses pembelajaran. Secara khusus, siswa menunjukkan minat yang besar terhadap bahan ajar yang
digunakan guru, serta keinginan yang kuat untuk berpartisipasi aktif dalam
pembelajaran melalui penerapan model STAD. 3491 Karimah Tauhid, Volume 3 Nomor 3 (2024), e-ISSN 2963-590X | Azizah dan Aliyyah a. Analisis Data Ketuntasan Siklus I Tabel di bawah ini menggambarkan hasil belajar siswa yang meningkat
yang diamati pada tes siklus II pada materi menjaga kesehatan kerangka tubuh: Gambar 8. Analisis Ketuntasan Hasil Ipa Siswa Kelas Iv S Gambar 8. Analisis Ketuntasan Hasil Ipa Siswa Kelas Iv S
Negeri Cimande 01semester I Siklus I
No
Ketuntasan
Frekwensi
Persentase
1
Tuntas
22
78,57
2
Tidak Tuntas
6
21,42
Rerata
68,57
Maksimum
90
Minimum
45 Informasi di atas dapat direpresentasikan secara visual dengan
menggunakan diagram batang untuk menampilkan persentase ketuntasan
siswa pada siklus I: Informasi di atas dapat direpresentasikan secara visual dengan
menggunakan diagram batang untuk menampilkan persentase ketuntasan
siswa pada siklus I: Tingkat Ketuntasan
78,57
21,42
100
%
80%
Tidak Tuntas
Tuntas
40
%
20%
Tuntas
Tidak Tuntas Tingkat Ketuntasan 3492 Karimah Tauhid, Volume 3 Nomor 3 (2024), e-ISSN 2963-590X | Azizah dan Aliyyah Karimah Tauhid, Volume 3 Nomor 3 (2024), e-ISSN 2963-590X | Azizah dan Aliyyah Gambar 9. Grafik Ketuntasan Nilai Hasil Belajar Ipa Siswa Kelas Iv
Cimande 01 Berdasarkan grafik ketuntasan yang tersedia, ada 22 siswa yang
sudah melampaui ambang batas KKM 65 atau setara dengan 78,57%. Sedangkan siswa yang belum memenuhi syarat ketuntasan belajar masih
berjumlah 6 orang atau mewakili 21,42% dari total siswa. b. b. Analisis Data Ketuntasan Siklus II Peningkatan hasil belajar siswa pada materi menjaga kesehatan
krangka tubuh terlihat pada hasil ujian siklus II seperti yang tergambar
dalam tabel dibawah: Gambar 10. Analisis Ketuntasan Hasil Ipa Siswa Kelas Gambar 10. Analisis Ketuntasan Hasil Ipa Siswa Kelas
Iv Sd Negeri Cimande 01 Semester 1 Siklus II
No
Ketuntasan
Frekwensi
Persentase
1
Tuntas
27
96,42%
2
Tidak Tuntas
1
2,57%
Rerata
84.82
Maksimum
95
Minimum
60 Informasi di atas dapat ditampilkan dalam diagram batang untuk
menggambarkan persentase ketuntasan siswa pada siklus II: 3493 Karimah Tauhid, Volume 3 Nomor 3 (2024), e-ISSN 2963-590X | Azizah dan Aliyyah Gambar 11. Grafik Ketuntasan Nilai Hasil Belajar Ipa Siswa Kelas
Iv Sdn Cimande 01
Tingkat Ketuntasan Siklus II
96,42%
3,57%
100
%
80%
Tidak Tuntas
Tuntas
60%
40%
Tuntas
Tidak Tuntas Tingkat Ketuntasan Siklus II Gambar 11. Grafik Ketuntasan Nilai Hasil Belajar Ipa Siswa Kelas
Iv Sdn Cimande 01 Dari grafik ketuntasan yang tersedia, ada 27 siswa yang memperoleh nilai
ketuntasan belajar lebih tinggi dari nilai ketuntasan minimal 65 atau setara dengan
96,42%. Sebaliknya, hanya ada 1 siswa atau 3,57% yang belum memperoleh nilai
ketuntasan belajar di bawah 65. a. Analisis Komparatif Analisis Komparatif dilakukan sesudah penerapan pendekatan
pembelajaran kooperatif TPS (Think Pair and Share). Pendekatan ini berguna
dalam menganalisis hasil belajar siswa melalui tahapan yang berbeda, yaitu
Pra Siklus, Siklus I, dan Siklus II. Hal ini berfokus pada penilaian pencapaian
indikator-indikator yang sudah ditetapkan sebelumnya yang ditetapkan oleh
peneliti sebelumnya. 3494
Gambar 12. Analisis Komparatif Ketuntasan Hasil IPA Siswa Kelas IV SD
Negeri Cimande 01 Semester I
Pra Siklus
Siklus I
Siklus II 3494
Gambar 12. Analisis Komparatif Ketuntasan Hasil IPA Siswa Kelas IV SD
Negeri Cimande 01 Semester I
Pra Siklus
Siklus I
Siklus II Karimah Tauhid, Volume 3 Nomor 3 (2024), e-ISSN 2963-590X | Azizah dan Aliyy
No
Ketuntasan
F
%
F
%
f
%
1
Tuntas
20
71,42
22
78,57
27
96,42%
2
Tidak Tuntas
8
28,57
6
21,42
1
3,57%
Rerata
67,85
68,57
84.82
Maksimum
85
90
95
Minimum
60
45
60 Tabel tersebut menunjukkan peningkatan hasil belajar yang menyatakan
dengan persentase ketuntasan dari pra siklus ke siklus I dan siklus II. Dari 28 siswa, 20
di antaranya mencapai skor 65 atau lebih tinggi, yang setara dengan 71,42% dari total. Sebaliknya, 8 siswa tidak memperoleh nilai kelulusan minimal 65, yaitu 28,57% dari
total. Setelah selesainya tindakan awal atau siklus I, 22 dari 28 siswa mencapai nilai 65
atau lebih tinggi, setara dengan persentase 78,57%. Sementara itu, 6 siswa masih
berada di bawah nilai ketuntasan minimal 65 dengan persentase 21,42%. Selanjutnya
peneliti melakukan pengukuran sekunder terhadap sampel sebanyak 28 siswa. Dari
jumlah tersebut, 27 siswa memperoleh nilai minimal 65, yang setara dengan tingkat
keberhasilan 96,42%. Namun, ada satu siswa yang tidak memenuhi nilai kelulusan
minimum atau mendapat nilai di bawah 65, yang presentasenya 3,57%. KESIMPULAN PTK tersebut dilakukan di SDN Cimande 01 yang terletak di Kecamatan
Caringin Kabupaten Bogor. Penelitian ini memanfaatkan pemodel STAD yang terbagi
kedalam 4 pertemuan yang dilaksanakan dalam dua siklus. Peneliti dalam studi ini
mencapai kemajuan penting dalam pendidikan IPA. Siswa mencapai prestasi melebihi
nilai kelulusan minimal 65. Dalam setiap pertemuan, peneliti memperkenalkan
beberapa tugas, termasuk terlibat dalam diskusi berpasangan, berpartisipasi dalam
diskusi kelompok besar, dan menyampaikan presentasi kelompok. Penelitian ini
menyoroti
keunggulan
pendekatan
kooperatif
STAD,
yang
meliputi:
(1)
menumbuhkan kemandirian siswa; (2) mendorong keterlibatan siswa secara aktif
dengan mendorong mereka untuk bebas mengungkapkan pemikirannya; dan (3)
meningkatkan kelincahan kognitif siswa. Pada siklus I, sebelum melaksanakan
kegiatan belajar mengajar melalui pemodelan tipe STAD, terlebih dahulu guru
memberikan petunjuk secara tegas kepada siswa tentang cara efektif memanfaatkan 3495 Karimah Tauhid, Volume 3 Nomor 3 (2024), e-ISSN 2963-590X | Azizah dan Aliyyah Karimah Tauhid, Volume 3 Nomor 3 (2024), e-ISSN 2963-590X | Azizah dan Aliyyah pemodelan tipe STAD. Ini memfasilitasi pemahaman siswa tentang pelaksanaan
tugas. Pada
penerapannya,
siswa
bisa
mengikuti
pembelajaran
dengan
menggunakan model pedagogi yang selaras dengan petunjuk yang diberikan oleh
guru dan peneliti. Efektivitas paradigma model STAD dalam membuat hasil belajar
IPA meningkat, dapat ditunjukkan dengan mengamati peningkatan nilai tes evaluasi
pada setiap siklus. Hasil analisis menunjukkan bahwa peningkatan kinerja guru dan keterlibatan
siswa yang lebih tinggi selama proses belajar mengajar bisa membuat hasil belajar
siswa meningkat. Sebanyak 27 siswa (96,42%) berhasil menyelesaikan siklus II,
melampaui KKM. Hanya 1 siswa (3,57%) yang tidak memenuhi syarat KKM. Dari data
tersebut kesimpulannya ada peningkatan dalam pembelajaran. Sebanyak 27 siswa
atau 96,42% angkatan berhasil menyelesaikan siklus II di atas ambang batas
ketuntasan minimal (KKM). Hanya satu siswa yang mewakili 3,57% kelompok yang
tidak memenuhi syarat KKM. Dari data tersebut, keimpulannya sudah terjadi
peningkatan pembelajaran, dengan tercapainya penanda keberhasilan yang
ditentukan sebesar 80%, dibuktikan dengan tingkat ketuntasan hasil pembelajaran
sebesar 96,42%. Hasil pemeriksaan lembar observasi guru menunjukkan adanya
peningkatan. Siswa menunjukkan tingkat aktivitas yang lebih tinggi dibandingkan
dengan guru. Siswa menunjukkan tingkat keingintahuan yang tinggi dalam proses
memperoleh pengetahuan. Ketidak lengkapan siswa muncul karena kurangnya fokus
belajar dan kecenderungan mengganggu teman sekelas saat sesi belajar. Pada proses
pembelajaran tahap kedua, tingkat ketuntasan pembelajaran sudah mencapai 96,42%,
melampaui indikator berhasil yang sudah ditetapkan yakni 80%. Dengan begitu, PTK
ini terbukti efektif. Meningkatnya hasil belajar IPA tidak lepas dari penerapan model kooperatif
STAD yang efektif melibatkan siswa dalam proses pembelajaran dan lebih
menekankan pembelajaran yang dipusatkan kesiswa. KESIMPULAN Hasil penelitian menyatakan penggunaan model STAD pada kelas IV SDN
Cimande 01 Kecamatan Caringin Kabupaten Bogor pada semester I tahun ajaran
2023/2024 bisa membuat hasil belajar IPA meningkat. Selaras pada penelitian
terdahulu, penelitian ini menyajikan variasi dan manfaat yang berbeda, khususnya:
siswa menerima bimbingan tidak hanya dalam lingkungan kolektif tetapi juga secara
individu. Dengan demikian,proses penelitian dengan memanfaaatkan media
pembelajaran kooperatip tipe stad pada pembelajaran makhluk hidup di kelas
IV
di SDN CIMANDE 01 dianggap efektif dan berjalan sesuai dengan rencana, artinya
dapat disimpulkan dan tidak memerlukan kelanjutan pada siklus selanjutnya karena
telah memenihi keberhasilan indicator. 3496 Karimah Tauhid, Volume 3 Nomor 3 (2024), e-ISSN 2963-590X | Azizah dan Aliyyah REFERENSI Aliyyah, R. R., Amini, A., Subasman, I., Herawati, E. S. B., & Febiantina, S. (2021). Upaya Meningkatkan Hasil Belajar Ipa Melalui Penggunaan Media Video
Pembelajaran. Jurnal Sosial Humaniora, 12(1), 54-72.Afifah, D. S. N. (2012). Interaksi belajar matematika siswa dalam pembelajaran kooperatif tipe
STAD. Pedagogia: Jurnal Pendidikan, 1(2), 145-152. Aliyyah, R. R., Puteri, F. A., & Kurniawati, A. (2017). Pengaruh kemandirian belajar
terhadap hasil belajar IPA. Jurnal Sosial Humaniora, 8(2), 126-143. Gita, S. D., Annisa, M., & Nanna, W. I. (2018). Pengembangan modul IPA materi
hubungan makhluk hidup dan lingkungannya berbasis pendekatan
kontekstual. LENSA (Lentera Sains): Jurnal Pendidikan IPA, 8(1). Gita, S. D., Annisa, M., & Nanna, W. I. (2018). Pengembangan modul IPA materi
hubungan makhluk hidup dan lingkungannya berbasis pendekatan
kontekstual. LENSA (Lentera Sains): Jurnal Pendidikan IPA, 8(1). Hazmiwati, H. (2018). Penerapan Model Pembelajaran Kooperatif Tipe Stad Untuk
Meningkatkan Hasil Belajar Ipa Siswa Kelas Ii Sekolah Dasar. Primary:
Jurnal Pendidikan Guru Sekolah Dasar, 7(1), 178-184. Maslina, A. S. A. A., Saputro, B., & Kusumawati, E. R. (2020). Pengembangan Bahan
Ajar Ipa Dengan Media Audio Visual Pada Tema Selamatkan Makhluk
Hidup Di Pendidikan Dasar. Jurnal JPSD (Jurnal Pendidikan Sekolah Dasar),
7(1), 70-80. Nugroho, U., & Edi, S. S. (2009). Penerapan pembelajaran kooperatif tipe STAD
berorientasiketerampilan proses. Jurnal Pendidikan Fisika Indonesia, 5(2). Nur, F. M. (2012). Pemanfaatan sumber belajar dalam pembelajaran sains kelas V SD
pada pokok bahasan makhluk hidup dan proses kehidupan. Jurnal
penelitian pendidikan, 13(1), 67-78. Permatasari, D. N., & Desstya, A. (2022). Analisis Kebutuhan Modul Pembelajaran
Tematik Peduli Terhadap Makhluk Hidup Berbasis Penguat Karakter IPA
Siswa Sekolah Dasar. Jurnal Basicedu, 6(4), 5638-5645. Prananda, G. (2019). Pengaruh Model Pembelajaran Kooperatif Tipe STAD Dalam
Pembelajaran IPA Siswa Kelas V SD. Pedagogik: Jurnal Ilmiah Pendidikan dan
Pembelajaran Fakultas Tarbiyah UniversitasMuhammadiyah Aceh, 6(2, 3497 Karimah Tauhid, Volume 3 Nomor 3 (2024), e-ISSN 2963-590X | Azizah dan Aliyyah Karimah Tauhid, Volume 3 Nomor 3 (2024), e-ISSN 2963-590X | Azizah dan Aliyyah Oktober), 122-130. Sudana, I. P. A., & Wesnawa, I. G. A. (2017). Penerapan model pembelajaran kooperatif
tipe STAD untukmeningkatkan hasil belajar IPA. Jurnal Ilmiah Sekolah Dasar,
1(1), 1-8. Utami, S. (2015). Peningkatan hasil belajar melalui pembelajaran kooperatif tipe stad
pada pembelajaran dasar sinyal video. Jurnal Pendidikan Teknologi dan
Kejuruan, 22(4), 424-431. Wijaya, H., & Arismunandar, A. (2018). Pengembangan model pembelajaran kooperatif
tipe stad berbasismedia sosial. Jurnal Jaffray, 16(2), 175-196. Suryantika, I., & Aliyyah, R. R. (2023). Implementasi Kurikulum Merdeka: Strategi
Pembelajaran di Luar Kelas pada Sekolah Dasar. KARIMAH TAUHID, 2(6),
3103- 3134. 3498
|
https://openalex.org/W4280515856
|
https://www.frontiersin.org/articles/10.3389/fimmu.2022.841254/pdf
|
English
| null |
Gpr174 Knockout Alleviates DSS-Induced Colitis via Regulating the Immune Function of Dendritic Cells
|
Frontiers in immunology
| 2,022
|
cc-by
| 8,516
|
Background: Dysfunction of the immune system would disturb the intestinal
homeostasis and lead to inflammatory bowel disease (IBD). Dendritic cells (DCs) help
maintain intestinal homeostasis and immediately respond to pathogens or injuries once
the mucosa barriers are destroyed during IBD. G protein-coupled receptors(GPR)174 is
an essential regulator of immunity that is widely expressed in most immune cells, including
DCs. However, the role of GPR174 in regulating the immune function of DC in colitis has
not been investigated. Edited by:
Chih-Hao Chang,
Jackson Laboratory, United States Edited by:
Chih-Hao Chang,
Jackson Laboratory, United States
Reviewed by:
Wenru Su,
Sun Yat-Sen University, China
Shubha Priyamvada,
University of Illinois at Chicago,
United States
*Correspondence:
Zhenju Song
song.zhenju@zs-hospital.sh.cn
Yilin Yang
yang.yilin@zs-hospital.sh.cn
Yiqun Zhang
zhang.yiqun@zs-hospital.sh.cn
†These authors share first authorship Reviewed by:
Wenru Su,
Sun Yat-Sen University, China
Shubha Priyamvada,
University of Illinois at Chicago,
United States Reviewed by:
Wenru Su,
Sun Yat-Sen University, China
Shubha Priyamvada,
University of Illinois at Chicago,
United States Methods: Dextran sodium sulfate (DSS) was administered to establish the mice colitis
model. Data of weight, length of colon, disease activity index (DAI), and macroscopic
scores were collected. The flow cytometry was used to detect the infiltrations of T cells
and DCs, the mean fluorescence intensity (MFI) of CD80, CD86, CD40, and major
histocompatibility complex-II (MHC-II). And T cells proliferataion was measured by
carboxyfluorescein diacetate succinimidyl ester (CFSE). The expression of cytokines
(tumor necrosis factor-a (TNF-a), interleukin-6 (IL-6), interleukin-10 (IL-10), interferon-g
(IFN-g), interleukin -4 (IL-4)) and GPR174 mRNA were measured by Elisa, quantitative
polymerase chain reaction (qPCR), and immunofluorescence. RNA of bone-marrow-
derived dendritic cells (BMDCs) was extracted for sequencing. Adoptive transfer of
BMDCs was administrated intravenously. *Correspondence:
Zhenju Song
song.zhenju@zs-hospital.sh.cn
Yilin Yang
yang.yilin@zs-hospital.sh.cn
Yiqun Zhang
zhang.yiqun@zs-hospital.sh.cn
†These authors share first authorship Specialty section:
This article was submitted to
T Cell Biology,
a section of the journal
Frontiers in Immunology Results: Gpr174-/- mice exposed to 3% DSS showed significant alleviation characterized
by reduced loss of weight, more minor colon damage, and better DAI and macroscopic
scores. The expression of pro-inflammatory cytokines (TNF-a, IL-6) decreased, while anti-
inflammatory cytokine (IL-10) increased compared with WT mice. In vitro, Gpr174-/-
BMDCs showed less maturity, with a declined expression of MHC-II, CD80, CD86 and
reduced TNF-a, higher IL-10 after LPS stimulation. Gpr174 Knockout Alleviates DSS-
Induced Colitis via Regulating the
Immune Function of Dendritic Cells
Wei Wei 1†, Sucheng Mu 1†, Yi Han 1†, Yao Chen 1, Zhongshu Kuang 1, Xingyue Wu 1,
Yue Luo 1, Chaoyang Tong 1, Yiqun Zhang 2*, Yilin Yang 1* and Zhenju Song 1,3,4* 1 Department of Emergency Medicine, Zhongshan Hospital, Fudan University, Shanghai, China, 2 Endoscopy Center and
Endoscopy Research Institute, Zhongshan Hospital, Fudan University, Shanghai, China, 3 Shanghai Key Laboratory of Lung
Inflammation and Injury, Shanghai, China, 4 Shanghai Institute of Infectious Disease and Biosecurity, School of Public Health,
Fudan University, Shanghai, China ORIGINAL RESEARCH
published: 20 May 2022
doi: 10.3389/fimmu.2022.841254 Gpr174-/- BMDCs were less capable
of activating OT-II naïve CD4+ T cells than WT BMDCs and induced more Th0 cells to
differentiate into Treg while less into Th1. Furthermore, the transcriptome sequencing
analysis exhibited that Gpr174 participated in TNF-a (NF-kB) signaling, leukocyte Received: 22 December 2021
Accepted: 25 April 2022
Published: 20 May 2022 Animals Inflammatory bowel disease (IBD), consisting of Crohn’s disease
(CD) and ulcerative colitis (UC), is a multifaceted disease
resulting from genetic, autoimmune, and environmental
factors, which in fact, is an inflammatory condition of the
gastrointestinal tract (1). Although precise pathogenesis is not
clear, the immune response caused by the invasion of intestinal
pathogens has been regarded as a pathogenic factor. The
sustaining and aberrant inflammation responses disturb the
mucosa and mucosal immunity homeostasis, leading to
inappropriate and magnified intestinal inflammation (2). Male wild-type C57BL/6 and male Gpr174 knockout mice
(Gpr174-/-) (8-12 weeks, 20-25g) were obtained from the
Southern Model Biological Technology Development Co. (Shanghai, China) and bred under pathogen-free conditions. Animals were housed separately in a temperature-controlled
room with a 12-hour light/12-hour dark cycle and free food
and water access. All experimental procedures and animal
operations followed the international guidelines for the Care
and Use of Laboratory Animals (NO. 201804001Z). Dendritic cells (DCs) maintain intestinal homeostasis and
respond at the first time to pathogens or injuries once the mucosa
barriers are destroyed during IBD (3). Clinical evidence indicates
that the numbers and subsets of CD103+ dendritic cells change in
inflamed mucosa, display loss of tolerogenic function, and disorder
of cytokine profiles in IBD patients (4, 5). Similarly, dextran sulfate
sodium salt (DSS)-induced colitis is developed in severe combined
immunodeficiency (SCID) mice, suggesting that acute DSS colitis
did not require the presence of either T cells or B cells, while
dendritic cells participate in the pathogenesis of IBD independently
(6, 7). Abbreviations: AAD, autoimmune Addison’s disease; BMDCs, bone-marrow-
derived dendritic cells; DAI, disease activity index; DC, dendritic cells; GPCRs, G
protein-coupled receptors; IBD, inflammatory bowel disease; SCFAs, short-chain
fatty acids; TJ, tight junction. Citation: Wei W, Mu S, Han Y, Chen Y, Kuang Z,
Wu X, Luo Y, Tong C, Zhang Y, Yang Y
and Song Z (2022) Gpr174
Knockout Alleviates DSS-Induced
Colitis via Regulating the Immune
Function of Dendritic Cells. Front. Immunol. 13:841254. doi: 10.3389/fimmu.2022.841254 May 2022 | Volume 13 | Article 841254 Frontiers in Immunology | www.frontiersin.org Wei et al. Gpr174 Influence Colitis via DCs transendothelial migration, and Th1/Th2 cell differentiation pathways. Adoptive transfer of
Gpr174-/- BMDCs to WT mice ameliorated DSS-induced colitis. transendothelial migration, and Th1/Th2 cell differentiation pathways. Adoptive transfer of
Gpr174-/- BMDCs to WT mice ameliorated DSS-induced colitis. Conclusion: Our study indicated that GPR174 was involved in the pathogenesis of IBD
by regulating the maturation of the dendritic cells to maintain immune homeostasis. TNF-a
(NF-kB) signaling pathway, leukocyte transendothelial migration, and Th1/Th2 cell
differentiation pathways may be the target pathway. Keywords: GPR174, inflammatory bowel disease, dendritic cells, intestinal barrier function, T cell activation Induction and Assessment of
DSS-Induced Colitis The acute colitis mice model (n=25) was induced by oral
administration of 3% DSS (35-50 kDa)(Sigma-Aldrich, St. Louis, MO, United States) for seven days and two days on
regular drinking water (13). The control group of mice (n=25)
was only fed with regular drinking water during the study. The
health condition was monitored daily, and the disease activity
index (DAI), including weight loss, stool consistency, and stool
bleeding, was scored as described (14). All mice were sacrificed
on day 9 after the intraperitoneal injection of avertin, and the
colon was collected for length and gross macroscopic
appearance (15). G protein-coupled receptors (GPCRs) are critical signaling
molecules in immune response, cell proliferation, inflammation
regulation, and intestinal barrier maintenance (8). One of the
GPCRs, GPR174, is widely expressed in the intestine, spleen,
thymus, and lymph nodes (9), is widely expressed in most
immune cells, including T/B lymphocytes, and DCs, and is
involved in many infectious diseases, including sepsis (10, 11). Gpr174 has also been reported to play an important role in
gender dimorphism of humoral immunity and be more
susceptible to experimental autoimmune encephalomyelitis. It
suppresses germinal center formation in male mice and is a
chemokine receptor for the CCL21 ligand (12). Intestinal Permeability Assay Intestinal permeability was measured by fluorescein
isothiocyanate (FITC)-dextran (3,000-5,000 kDa) (Sigma-
Aldrich, St. Louis, MO, United States) (16). Briefly, 9 days after
oral administration of DSS, mice were fasted for 4 h and then
gavaged with FITC-dextran (0.5 mg/g body weight at 125 mg/
mL). Four hours later, blood taken from the abdominal aorta was
centrifuged at 12,000 g for 5 min. The serum was collected to
detect the FITC-dextran by a microplate reader with an
excitation wavelength of 490 nm and an emission wavelength
of 520 nm. However, whether GPR174 could regulate the immune
function of DCs in IBD has not been investigated. The present
study aimed to identify GPR174 involved in colitis by both in
vivo and vitro studies. In addition, we focused on the relationship
between GPR174 and DCs function, which may play a critical
role in regulating the intestinal injury of colitis. Patient Enrollment The Research and Ethics committee of Zhongshan Hospital,
Fudan University, approved this study (NO: B2020-016R). All
procedures were directed under guidelines. Samples from ten
patients with ulcerative colitis (UC) collected under colonoscopy
were used for qPCR and immunofluorescence. Myeloperoxidase (MPO) Assay y
p
(
)
y
The MPO activity assay was applied to assess the neutrophil
infiltration into the colon. The distal colon was homogenized in 4
volumes of MPO Assay buffer (Jiancheng Co. Ltd., Nanjing,
China) and centrifuged at 13,000 g for 10 min at 4 °C. Prepare a
Master Mix of the reagent according to the instructions, then add Abbreviations: AAD, autoimmune Addison’s disease; BMDCs, bone-marrow-
derived dendritic cells; DAI, disease activity index; DC, dendritic cells; GPCRs, G
protein-coupled receptors; IBD, inflammatory bowel disease; SCFAs, short-chain
fatty acids; TJ, tight junction. May 2022 | Volume 13 | Article 841254 Frontiers in Immunology | www.frontiersin.org 2 Wei et al. Gpr174 Influence Colitis via DCs RNA Sequencing and Functional Analysis
Treated with LPS for 12 h or left untreated, BMDCs were harvested
forRNAextractionbyQIAGENRNeasyMiniKit.TheAgilent4200
TapeStation System was used to check RNA integrity to ensure
RNA integrity was ≥8.9. The TruSeqStranded mRNA Library Prep
Kit (Illumina) made libraries for RNASeq using 1-1.5 mg of RNA as
input. The HiSeq 4000 Sequencing Systems (Illumina) was used to
run across two lanes to yield 2 × 151 bp paired-end reads with an
average yield of ~55 million reads/sample. a the reaction mix to each well containing the controls and samples. Mix and incubate at 25°C for hours. Add stop mix to all wells and
incubate at room temperature and TNB for 5-10 min. At last,
measure output (OD412 nm) on a microplate reader. Flow Cytometry An anti-FcR antibody was used to block the non-specific
staining. Cells were subsequently immunostained with APC-
conjugated anti-CD11c, APC-cy7-conjugated anti-CD86, and
FITC-conjugated anti-MHC-II antibodies (Biolegend, USA). Quantitative RT-PCR RNA was isolated from the whole colonic tissues using the
Qiagen RNeasy Mini Kit following the manufacturer’s
instructions. Quantitative polymerase chain reaction (qPCR)
was performed for expressions of mRNAs using the primers
for hGPR174, mIl-6, mIl-10, mTnf-a, mOccludin, mZo-1, and
b-actin. HE Staining and Histopathological
Evaluation About2cminlength,thecolonictissuewasexcisedontheninthday
after drinking DSS-water, washed in phosphate-buffered saline
(PBS), embedded in 4% formaldehyde in paraffin sectioned about
2 mm, and stained with hematoxylin and eosin (H&E) (17). Fisher’s exact test was used to calculate differential gene
expression. If it met the cut-off criteria, it was considered
significantly differentially. An R Bioconductor package
heatmap3 was used to visualize the differential gene expression
(https://cran.r-project.org/package=heatmap3). DAVID (version
6.8) was used for gene term enrichment. BMDCs and Naïve CD4+ Cell In Vitro
Co-Culture Assays BMDCs were prepared as previously described and purified by
CD11c+ magnetic beads. BMDCs were co-cultured with isolated
OT-II naïve CD4+ T cells at a 1:10 ratio in complete RPMI in 96-
well V-bottom plates. Cells were harvested 3 days later for
further analysis. OVA peptide 323-339 (GenScript, USA) was
added to wells at 1 mg/mL as a positive control. RESULTS Gpr174-/- Mice Were Resistant to
DSS-Induced Colitis Statistical Analysis Data were expressed as mean ± standard error (SEM). One-way
ANOVA, t-test followed by Bonferroni tests was performed
using GraphPad Prism 7.0. Transcriptome analysis was
visualized using the R prcomp function and ggplot package. A
P value < 0.05 was considered statistically significant. Bone Marrow-Derived Dendritic
Cells (BMDCs) Bone marrow cells were extracted from femurs and tibiae of
C57BL/6(n=5) and Gpr174-/- mice(n=5). After incubated in Red
blood cell (RBC) lysis buffer at 37°C for 5 min, cells were seeded
in six-well plates at 106 cells/well concentration with RPMI-1640
medium supplemented with 20 ng/mL of granulocyte-
macrophage colony-stimulating factor, 10% fetal bovine serum,
100 U/mL penicillin, and 100 mg/mL streptomycin. On day 3,
the culture medium was changed to remove the nonadherent
granulocyes without dislodging the cluster of dendritic cells and
macrophages. On day 5, the suspension and loosely adherent
dendritic cells were harvested by gently swirling the dishes to get
rid of the bone-marrow derived macrophages which is tightly
adherent to the dishes (18). Gpr174-/- Mice Were Resistant to
DSS-Induced Colitis To explore whether the GPR174 receptor is involved in
the pathophysiology of IBD, we established a DSS-induced
colitis animal model as previously described. After oral
administration of DSS for 7 days and 2 days on regular
drinking water, Gpr174 -/- mice showed improved acute colitis
symptoms compared with WT mice. Gpr174-/- mice significantly
decreased DAI scores (Figure 1B, P < 0.01), with less weight loss
(Figure 1A, P < 0.05) and later appearance of diarrhea loose
feces. Decreased loose bloody stools and a lower reduction in
colon length were shown in Gpr174-/- mice compared with the
WT mice (Figures 1C, D, P < 0.01). Furthermore, compared
with the WT mice, Gpr174-/- mice exhibited minor architecture
damage or epithelial barrier disruption, leading to the
improved macroscopic scores with DSS treatment (Figures 1E, F,
P < 0.01). CFSE Proliferation Assay Knockout of Gpr174 Reduced Gut
Inflammatory Response and Maintained
Intestinal Barrier in DSS-Induced Colitis
The intestinal barrier damage is caused by inflammatory
responses and is characterized by epithelial barrier dysfunction Knockout of Gpr174 Reduced Gut
Inflammatory Response and Maintained
Intestinal Barrier in DSS-Induced Colitis The number of
DCs in LP showed no difference between WT and Gpr174-/-
mice, while the percentage of CD11c+ DC, one of the most potent
antigen-presenting cells, increased in Gpr174-/- mice than in WT
mice (Figures 3A, B, P < 0.05). However, the expression of the
MHC-II molecule decreased (Figure 3C). and increased mucosal permeability. Myeloperoxidase (MPO),
reflecting the degree of neutrophil infiltration, was decreased in
the distal colonic tissue of Gpr174-/- mice compared with WT
mice (Figure 2A, P < 0.01). To analyze why the Gpr174-/- mice
reduced susceptibility to DSS, we evaluated the production of
pro-inflammatory and anti-inflammatory cytokines in the
colonic tissue. The results showed that Gpr174-/- mice treated
with DSS demonstrated decreased levels of TNF-a (P < 0.01), IL-
6 (P < 0.01), and increased levels of IL-10 (P < 0.001) compared
with WT mice (Figure 2B). In addition, our study indicated that
the expression of Zo-1 mRNA (Figure 2C, P < 0.05) and
Occludin (Figure 2D, P < 0.01) in Gpr174-/- mice was
significantly higher than that in WT mice. Besides, we
observed that serum FITC-dextran was markedly decreased in
Gpr174-/- mice compared with WT ones 6 h after gavaging with
FITC-dextran (Figure 2E, P < 0.01), suggesting that epithelial
permeability was reduced in the Gpr174-/- mice. and increased mucosal permeability. Myeloperoxidase (MPO),
reflecting the degree of neutrophil infiltration, was decreased in
the distal colonic tissue of Gpr174-/- mice compared with WT
mice (Figure 2A, P < 0.01). To analyze why the Gpr174-/- mice
reduced susceptibility to DSS, we evaluated the production of
pro-inflammatory and anti-inflammatory cytokines in the
colonic tissue. The results showed that Gpr174-/- mice treated
with DSS demonstrated decreased levels of TNF-a (P < 0.01), IL-
6 (P < 0.01), and increased levels of IL-10 (P < 0.001) compared
with WT mice (Figure 2B). In addition, our study indicated that
the expression of Zo-1 mRNA (Figure 2C, P < 0.05) and
Occludin (Figure 2D, P < 0.01) in Gpr174-/- mice was
significantly higher than that in WT mice. Besides, we
observed that serum FITC-dextran was markedly decreased in
Gpr174-/- mice compared with WT ones 6 h after gavaging with
FITC-dextran (Figure 2E, P < 0.01), suggesting that epithelial
permeability was reduced in the Gpr174-/- mice. Knockout of Gpr174 Reduced Gut
Inflammatory Response and Maintained
Intestinal Barrier in DSS-Induced Colitis Knockout of Gpr174 Reduced Gut
Inflammatory Response and Maintained
Intestinal Barrier in DSS-Induced Colitis Isolate naïve CD4+ T cells from OT-II mice(n=5), label with
carboxyfluorescein N-succinimidyl ester (CFSE, C34554; Life
Technologies), and culture in the presence of WT BMDCs and
Gpr174-/- DCs. Unstimulated T cells serve as a negative control. The percentage of divided cells was analyzed by FlowJo V10. The intestinal barrier damage is caused by inflammatory
responses and is characterized by epithelial barrier dysfunction May 2022 | Volume 13 | Article 841254 Frontiers in Immunology | www.frontiersin.org 3 Wei et al. Gpr174 Influence Colitis via DCs A
B
D
E
F
C
FIGURE 1 | Gpr174-/- mice were resistant to DSS-induced colitis. (A) Body weight loss in each group, (B) DAI scores in each group, (C, D) Colon length of each
group, (E) Macroscopic score of each group, (F) Morphology and HE of the intestine of each group. Data were exhibited as mean ± SEM of 5 mice per group. (Black arrow: epithelial barrier disruption. *P < 0.05, **P < 0.01). B A B D
E D
E
C C C D F
FIGURE 1 | Gpr174-/- mice were resistant to DSS-induced colitis. (A) Body weight loss in each group, (B) DAI scores in each group, (C, D) Colon length of each
group, (E) Macroscopic score of each group, (F) Morphology and HE of the intestine of each group. Data were exhibited as mean ± SEM of 5 mice per group. (Black arrow: epithelial barrier disruption. *P < 0.05, **P < 0.01). F FIGURE 1 | Gpr174-/- mice were resistant to DSS-induced colitis. (A) Body weight loss in each group, (B) DAI scores in each group, (C, D) Colon length of each
group, (E) Macroscopic score of each group, (F) Morphology and HE of the intestine of each group. Data were exhibited as mean ± SEM of 5 mice per group. (Black arrow: epithelial barrier disruption. *P < 0.05, **P < 0.01). Gpr174 Knockout Influenced CD11c+
Dendritic Cells Infiltration in Intestinal
Lamina Propria(LP) in DSS-Induced Colitis
Due to intestinal inflammation, both WT and Gpr174-/- mice
exhibited increased CD11c+ DCs accumulation in LP compared
with the WT ones after model establishment. Gpr174 Knockout Led to Difficulty of
BMDC Maturation To further analyze the role of Gpr174 on the immune function of
CD11c+ DC, we examined the expression of co-stimulatory
molecules of BMDCs. BMDCs extracted from the bone May 2022 | Volume 13 | Article 841254 Frontiers in Immunology | www.frontiersin.org 4 Wei et al. Gpr174 Influence Colitis via DCs A
B
D
E
C
FIGURE 2 | Gpr174 knockout reduced intestinal inflammation and protected the intestinal barrier. (A) Intestinal MPO expression in each group, (B) Intestinal
cytokine expression of TNF-a, IL-6, IL-10 in each group, (C, D) Intestinal tight junctions of Zo-1 and Occludin mRNA expression in each group, (E) Intestinal
permeability in each group. Data were exhibited as mean ± SEM of 5 mice per group. (*P < 0.05, **P < 0.01, ***P < 0.001). A
C E C FIGURE 2 | Gpr174 knockout reduced intestinal inflammation and protected the intestinal barrier. (A) Intestinal MPO expression in each group, (B) Intestinal
cytokine expression of TNF-a, IL-6, IL-10 in each group, (C, D) Intestinal tight junctions of Zo-1 and Occludin mRNA expression in each group, (E) Intestinal
permeability in each group. Data were exhibited as mean ± SEM of 5 mice per group. (*P < 0.05, **P < 0.01, ***P < 0.001). Gpr174 Knockout Suppressed
BMDCs to Naïve T Cell Proliferation
and Differentiation marrow of Gpr174-/- mice were less prone to differentiate into
CD11c+ cells, especially on the seventh day of culture
(Figure 4A). Both Gpr174+/+ and Gpr174-/- BMDCs
upregulated the expression of MHC II, CD80, CD86, and
CD40 with LPS stimulation. However, the expression of MHC
II, CD80, and CD86 increased much less in Gpr174-/- BMDCs
than that in Gpr174+/+ (Figure 4B). The naïve CD4+ T cells purified from the spleens of OT-II OVA-
specific T cell receptor transgenic mice were used to examine the
effects of Gpr174 on BMDCs to stimulate naïve T cell cytokine
production. Gpr174-/- BMDCs induced less naïve T cell proliferation A
B
C
FIGURE 3 | Gpr174 knockout increased CD11c+ dendritic cells infiltration while suppressed maturation in the lamina propria. (A, B) Infiltration of CD11c+ dendritic
cells in the lamina propria in each group, (C) MHC-II expression of CD11c+ dendritic cells in each group. Data were exhibited as mean ± SEM of 5 mice per group. (*P < 0.05). B A B C C FIGURE 3 | Gpr174 knockout increased CD11c+ dendritic cells infiltration while suppressed maturation in the lamina propria. (A, B) Infiltration of CD11c+ dendritic
cells in the lamina propria in each group, (C) MHC-II expression of CD11c+ dendritic cells in each group. Data were exhibited as mean ± SEM of 5 mice per group. (*P < 0.05). May 2022 | Volume 13 | Article 841254 5 Frontiers in Immunology | www.frontiersin.org Wei et al. Gpr174 Influence Colitis via DCs than Gpr174+/+ BMDCs (Figure 5A, P < 0.01). Co-culturing with
Gpr174-/- BMDCs led to a significant reduction of IFN-g
Adoptively Transfer of Gpr174-/- BMDCs
Alleviated DSS-Induced Colitis
A
B
FIGURE 4 | Gpr174 knockout inhibited maturation of BMDCs. (A) CD11c+ expression of BMDCs in each group, (B) Co-stimulatory molecular MHC-II, CD80, CD86,
CD40 BMDCs. Data were exhibited as mean ± SEM of 5 mice per group. A B B FIGURE 4 | Gpr174 knockout inhibited maturation of BMDCs. (A) CD11c+ expression of BMDCs in each group, (B) Co-stimulatory molecular MHC-II, CD80, CD86,
CD40 BMDCs. Data were exhibited as mean ± SEM of 5 mice per group. Gpr174 Knockout Suppressed
BMDCs to Naïve T Cell Proliferation
and Differentiation Adoptively Transfer of Gpr174-/- BMDCs
Alleviated DSS-Induced Colitis
To verify the function of Gpr174-/- BMDCs in alleviating DSS-
induced colitis, we adoptively transfer Gpr174-/- BMDCs and
Gpr174+/+ BMDCs into C57/BL6 mice intravenously at 5 days and
3 days before and 1 day after the induction of colitis, flow chart of
procedurewereshowninFigure6C.MicetransferredwithGpr174-/- Adoptively Transfer of Gpr174-/- BMDCs
Alleviated DSS-Induced Colitis
To verify the function of Gpr174-/- BMDCs in alleviating DSS-
induced colitis, we adoptively transfer Gpr174-/- BMDCs and
Gpr174+/+ BMDCs into C57/BL6 mice intravenously at 5 days and
3 days before and 1 day after the induction of colitis, flow chart of
procedurewereshowninFigure6C.MicetransferredwithGpr174-/- Adoptively Transfer of Gpr174-/- BMDCs
Alleviated DSS-Induced Colitis than Gpr174+/+ BMDCs (Figure 5A, P < 0.01). Co-culturing with
Gpr174-/- BMDCs led to a significant reduction of IFN-g
(Figure 5B, P < 0.01) and an increase in the production of IL-10
in CD4+ T cells (Figure 5D, P < 0.05), while without any change in
IL-4 (Figure 5C), indicating that Gpr174-/- BMDCs induced more
naïve CD4+ T cells to differentiate into Treg cells, while less into
Th1 cells. than Gpr174+/+ BMDCs (Figure 5A, P < 0.01). Co-culturing with
Gpr174-/- BMDCs led to a significant reduction of IFN-g
(Figure 5B, P < 0.01) and an increase in the production of IL-10
in CD4+ T cells (Figure 5D, P < 0.05), while without any change in
IL-4 (Figure 5C), indicating that Gpr174-/- BMDCs induced more
naïve CD4+ T cells to differentiate into Treg cells, while less into
Th1 cells. than Gpr174+/+ BMDCs (Figure 5A, P < 0.01). Co-culturing with
Gpr174-/- BMDCs led to a significant reduction of IFN-g
(Figure 5B, P < 0.01) and an increase in the production of IL-10
in CD4+ T cells (Figure 5D, P < 0.05), while without any change in
IL-4 (Figure 5C), indicating that Gpr174-/- BMDCs induced more
naïve CD4+ T cells to differentiate into Treg cells, while less into
Th1 cells. To verify the function of Gpr174-/- BMDCs in alleviating DSS-
induced colitis, we adoptively transfer Gpr174-/- BMDCs and
Gpr174+/+ BMDCs into C57/BL6 mice intravenously at 5 days and
3 days before and 1 day after the induction of colitis, flow chart of
procedurewereshowninFigure6C.MicetransferredwithGpr174-/- May 2022 | Volume 13 | Article 841254 Frontiers in Immunology | www.frontiersin.org 6 Wei et al. Gpr174 Influence Colitis via DCs A
B
D
C
FIGURE 5 | Gpr174 knockout inhibited T cell proliferation and influenced T cell differentiation. Gpr174 Knockout Suppressed
BMDCs to Naïve T Cell Proliferation
and Differentiation (A) Degree of naive T cells proliferation, which was labeled as
CFSE after stimulation in each group, (B) IFN-g expression of naive CD4+ T cells after stimulation in each group, (C) IL-4 expression of naive CD4+ T cells after
stimulated in each group, (D) IL-10 expression of naive CD4+ T cells after stimulated in each group. Data were exhibited as mean ± SEM of 5 mice per group. (*P < 0.05, **P < 0.01). A B C D D C FIGURE 5 | Gpr174 knockout inhibited T cell proliferation and influenced T cell differentiation. (A) Degree of naive T cells proliferation, which was labeled as
CFSE after stimulation in each group, (B) IFN-g expression of naive CD4+ T cells after stimulation in each group, (C) IL-4 expression of naive CD4+ T cells after
stimulated in each group, (D) IL-10 expression of naive CD4+ T cells after stimulated in each group. Data were exhibited as mean ± SEM of 5 mice per group. (*P < 0.05, **P < 0.01). BMDCs were in a better condition after administration of DSS,
characterized by the decreased weight loss (Figure 6A, P<0.05),
ameliorative DAI scores (Figure 6B, P<0.01), better macroscopic
scores(Figure 6D, P<0.01),andreducedshorteningofcolon length
(Figures 6E, F P<0.05). Less disruption of the epithelium in mice
treated with Gpr174-/- BMDCs than the Gpr174+/+ BMDCs was
exhibited by H&E staining (Figure 6G). Besides, the inflammatory
response was mitigated in the ones transferred with Gpr174-/-
BMDCs, with MPO (Figure 7A, P < 0.05), TNF-a (Figure 7B,
P < 0.05), and IL-6 (Figure 7C) declined, IL-10 (Figure 7D, P <
0.001)increasedinthecolon.Inaddition,Tcellswerereducedinthe
LP transferred with Gpr174-/- BMDCs (Figure 7E, P < 0.01) differentiation and maturation of BMDCs. LPS led to significant
modulation of gene expression between Gpr174-/- BMDCs and
Gpr174+/+ BMDCs. However, no significant differences were
observed between Gpr174-/- BMDCs and Gpr174+/+ BMDCs
without LPS stimulation (Figures 8A, B). TNF-a was extracted
from protein to protein interaction (PPI) network analysis of DEGs
to be a hub gene (Figure 8C). The enrichment of KEGG analysis
suggested that the TNF-a (NF-kB) signaling pathway and leukocyte
transendothelial migration pathway may be the potential target
pathway of Gpr174 (Figure 8D). Gpr174 Knockout Altered Dendritic Cells
Transcriptome After Maturation Gpr174 Knockout Altered Dendritic Cells
Transcriptome After Maturation To further confirm the role of GPR174 in the pathogenesis of UC,
the inflamed mucosas of the UC patients(n=10) were collected. The expression of GPR174 mRNA was relatively lower in
the pathological site than in the healthy control (Figure 9A, p
Gpr174+/+ BMDCs and Gpr174-/- BMDCs were subjected to
transcriptome analysis to reveal the role of Gpr174 in the May 2022 | Volume 13 | Article 841254 Frontiers in Immunology | www.frontiersin.org 7 Wei et al. Gpr174 Influence Colitis via DCs A
B
D
E
F
G
C
FIGURE 6 | Adoptive transfer of Gpr174-/- BMDCs alleviated DSS-induced colitis. (A) Body weight loss in each group after adoptive transfer of Gpr174-/- DC,
(B) DAI scores in each group, (C) Flow chart of the procedure of adoptive transfer of BMDCs and establishing DSS-induced colitis animal models, (D) Macroscopic
score of each group, (E, F) Colon length of each group, (G) Morphology and HE of the intestine of each group. Data were exhibited as mean ± SEM of 5 mice per
group. (*P < 0.05, **P < 0.01). B A D E C E D C G F F G FIGURE 6 | Adoptive transfer of Gpr174-/- BMDCs alleviated DSS-induced colitis. (A) Body weight loss in each group after adoptive transfer of Gpr174-/- DC,
(B) DAI scores in each group, (C) Flow chart of the procedure of adoptive transfer of BMDCs and establishing DSS-induced colitis animal models, (D) Macroscopic
score of each group, (E, F) Colon length of each group, (G) Morphology and HE of the intestine of each group. Data were exhibited as mean ± SEM of 5 mice per
group. (*P < 0.05, **P < 0.01). P < 0.05). Likewise, the immunofluorescence showed less
GPR174 expression in the inflamed tissues (Figure 9B). (20), and intestinal barrier maintenance (8, 21). Researchers have
revealed that GPR43 could bind with short-chain fatty acids
(SCFAs) to regulate the inflammatory responses in intestinal
diseases, which indicated that GPCRs took a critical part in the
pathogenesis of IBD. Our previous study demonstrated that
Gpr174 knockout elevated the number of marginal zone B cells
in the spleen of mice and promoted the function of regulatory T
cells, cytotoxic T lymphocytes, and M2 macrophage polarization
(11, 22). Gpr174 Knockout Altered Dendritic Cells
Transcriptome After Maturation In this study, we found that deletion of Gpr174 could
generally alleviate intestinal injury in the DSS-induced colitis,
which considered Gpr174 participated in the pathogenesis
of IBD. Frontiers in Immunology | www.frontiersin.org DISCUSSION Our study showed that Gpr174 knockout reduced inflammatory
response and altered the phenotype of DCs in colitis intestines. Gpr174 knockout lowered the expression of co-stimulatory
molecules and influenced the ability of BDMCs to mature,
which further altered the proliferation and differentiation of
naïve T cells. T cell activation is a critical step in the intestinal
immune system, and only the activated T cells exert an efficient
mucosa immune response. Moreover, we adoptively transferred
Gpr174-/- BMDCs to alleviate DSS-induced colitis and gut injury. The transcriptomic analysis revealed NF-kB signaling pathway
participated in the process related to BMDCs maturation. Furthermore, the expression of GPR174 mRNA was decreased
in the inflamed mucosa in IBD patients. Moreover, GPCRs also act as a navigator in the migration of
dendritic cells from peripheral to draining Lymph nodes (LNs)
and their maturation process (23). GPCRs such as chemokine
receptors CCR7 and CCR4 are highly upregulated on DCs upon
maturation. Moreover, GPCRs like GPR183 and CXCR5 help
direct DCs to certain areas upon inflammation (24). DCs, a
bridge between innate and acquired immune systems, maintain
the homeostasis of intestinal immunity in the LP, were found to
decrease colitis after Gpr174 knockout in this study. GPCRs are critical signaling molecules in immune response,
inflammation regulation, cell proliferation (19), cell migration May 2022 | Volume 13 | Article 841254 Frontiers in Immunology | www.frontiersin.org 8 Wei et al. Gpr174 Influence Colitis via DCs A
B
D
E
C
FIGURE 7 | Adoptive transfer of Gpr174-/- BMDCs alleviated inflammatory response of colitis. (A) Colonic MPO expression in each group, (B) Colonic cytokine
expression of TNF-a in each group, (C) Intestinal cytokine expression of IL-6 in each group, (D) Intestinal cytokine expression of IL-10 in each group, (E) CD3+ T
cells infiltration in the lamina propria in each group (*P < 0.05, ***P < 0.001). A
B
D
C E E FIGURE 7 | Adoptive transfer of Gpr174-/- BMDCs alleviated inflammatory response of colitis. (A) Colonic MPO expression in each group, (B) Colonic cytokine
expression of TNF-a in each group, (C) Intestinal cytokine expression of IL-6 in each group, (D) Intestinal cytokine expression of IL-10 in each group, (E) CD3+ T
cells infiltration in the lamina propria in each group (*P < 0.05, ***P < 0.001). A
B
D
C
FIGURE 8 | Gpr174 knockout influences BMDCs transcriptome after LPS stimulation. DISCUSSION Immature DCs possess a solid
ability to phagocytose and digest antigens and then process and
present antigens and gain a mature phenotype. Once BMDCs
acquire mature phenotypes, they are empowered to activate naïve
T/B cells (26, 27). Mature DCs migrate from LP to lymph nodes. They initiate naïve T/B cells and induce them to proliferate and
differentiate. Many studies indicated that tolerant DCs could
reduce the severity of DSS-induced colitis, probably due to their
cytokine profiles and weakened immunity. Activated DCs also
secrete pro-inflammatory cytokines to intercede the inflammatory
regulation in UC by activating TLRs, which induce infiltration of
neutrophils and activation of other innate immune cells (28). Like
DCs and macrophages, cytokines released by APCs and
macrophages trigger naïve T/B cell differentiation into various
subsets. Consistent with the alteration of maturity, Gpr174 -/-
BMDCs exhibited the impaired ability to stimulate naïve T cells to
proliferation and differentiation. adaptive immune response. In addition, effector T cell subsets
which were differentiated after binding with DCs also changed. Following activation, naïve CD4+ T cells differentiated into
several types of effector T cells, subdivided into either Th1,
Th2, or Treg cells according to cytokine profiles. The pathogenesis of IBD may be an excessive activation of
effector T cells and alteration of T cell-mediated tolerance, which
is related to Treg development or alteration (32). An animal
experiment showed that SAMP1/YitFc mice, which had
massively expanded B cells in the intestine, developed
transmural gut inflammation by blocking Tregs’ immune
function (33). Helper T cells are critical mediators of the
immune response, while regulatory T cells have a robust
immune suppression function. Recent studies demonstrated
that the immune imbalance between CD4+ T cells, Th1/Th2,
and Treg might be the most direct and vital factor in the
pathogenesis of IBD. It also played a role in mediating the
occurrence of ulcerative colitis (34, 35). The current study
suggested a reducing ratio of Th1/Th2 and Treg in LP
alleviated inflammatory response in colonic tissue. Less effector
T cells reduced pro-inflammatory factors release, thereby
alleviating intestinal damage. Pro-inflammatory cytokines released in the gut are involved
in IBD and have different functions, including cellular adhesion,
differentiation, and transmigration (36). Our results strongly
suggested that increased pro-inflammatory cytokines in the gut
of mice treated with DSS accounted for the observed
exacerbation of colitis. DISCUSSION (A) Venn diagram of differential expression genes (DEGs) of four groups of
BMDCs before and after LPS application, (B) Heatmap of DEGs of BMDCs in each group after LPS application, (C) Protein-protein interactions (PPI) networks
analysis of DEGs of Gpr174-/- BMDCs and Gpr174+/+ BMDCs after LPS application, (D) KEGG analysis of DEGs of Gpr174-/- BMDCs and Gpr174+/+ BMDCs after
LPS application (the first twenty 20 pathways with ease < 0.05, count ≥10, and FDR < 0.01). A B C D D D FIGURE 8 | Gpr174 knockout influences BMDCs transcriptome after LPS stimulation. (A) Venn diagram of differential expression genes (DEGs) of four groups of
BMDCs before and after LPS application, (B) Heatmap of DEGs of BMDCs in each group after LPS application, (C) Protein-protein interactions (PPI) networks
analysis of DEGs of Gpr174-/- BMDCs and Gpr174+/+ BMDCs after LPS application, (D) KEGG analysis of DEGs of Gpr174-/- BMDCs and Gpr174+/+ BMDCs after
LPS application (the first twenty 20 pathways with ease < 0.05, count ≥10, and FDR < 0.01). May 2022 | Volume 13 | Article 841254 Frontiers in Immunology | www.frontiersin.org 9 Wei et al. Gpr174 Influence Colitis via DCs A
B
FIGURE 9 | Expression of GPR174 between healthy and pathogenic tissues. (A) The mRNA expression of GPR174 between healthy and pathogenic tissues from
the ulcerative patients(n=10), (B) The protein expression of GPR174 receptor between healthy and pathogenic tissues from the ulcerative patients (Blue: DAPI;
Green: GPR174). (*P < 0.05). A
B B B FIGURE 9 | Expression of GPR174 between healthy and pathogenic tissues. (A) The mRNA expression of GPR174 between healthy and pathogenic tissues from
the ulcerative patients(n=10), (B) The protein expression of GPR174 receptor between healthy and pathogenic tissues from the ulcerative patients (Blue: DAPI;
Green: GPR174). (*P < 0.05). We chose BMDCs to investigate the role of GPR174 in regulating
the function of DCs because they could be abundantly generated
in vitro and were widely applied as a model myeloid DC (25). In
our study, Gpr174 modulated murine BMDCs maturation and
migration. We found that deletion of the Gpr174 in BMDCs
hardly affected the expression of CD86, CD80, or MHC-II on non-
stimulated BMDCs. However, after being stimulated by LPS,
Gpr174-/- BMDCs showed a different expression profile of co-
stimulatory molecules, such as CD80, CD86, and MHC-II, which
were significantly decreased, indicating that Gpr174-/- BMDCs
remained an immature phenotype. Frontiers in Immunology | www.frontiersin.org DISCUSSION Our study showed that Gpr174-/- mice
with DSS treatment suppressed the pro-inflammatory factors like
TNF-a and IL-6 and promoted the anti-inflammatory factor IL-
10. A delicate balance of pro-and anti-inflammatory factors is
necessary to preserve intestinal homeostasis. The increase of IL-
10 in the Gpr174-/- mice might provide compensative
mechanisms for improvements in inflammation status. Preliminary studies demonstrated that stimulation of OT-II
CD4+ T cells co-cultured with DCs and 50 nM OVA (329-337
peptides) induced a high expression level of both IFN-g and IL-4
a modest level of IL-10 (29, 30). In vitro experiments, our study
indicated that Gpr174-/- BMDCs were less capable of stimulating
naïve CD4+ T cells. Lumen antigens were endocytosed, digested
by DCs, and then presented to naïve T/B cells to initiate systemic
and mucosal immunity (31). Activated DCs in LP would acquire
maturity and migrate to mesenteric lymph node (MLN) and
secondary lymphoid tissues with antigens to translate innate to Cytokines could disrupt the intestinal epithelium, promote
barrier permeability, and modulate the tight junctions (TJs). TJs
are important for barrier integrity and play an essential role in May 2022 | Volume 13 | Article 841254 Frontiers in Immunology | www.frontiersin.org 10 Wei et al. Gpr174 Influence Colitis via DCs FUNDING This work was supported by the National Natural Science
Foundation of China (Grant No. 82072214), National Key
R e s e a r c h a n d D e v e l o p m e n t
P r o g r a m o f
C h i n a
(2021YFC2501800), Science and Technology of Shanghai
Committee (Grant No. 20Y11900100, Grant No. 21MC1930400, Grant No.20DZ2261200), and Shanghai
Municipal Health Bureau (Grant No. ZXYXZ-201906, Grant
No. GWV-10.2-XD04). Furthermore, we did some transcriptome analysis of Gpr174+/+
BMDCs and Gpr174-/- BMDCs, which enhanced the speculation
that Gpr174 mainly participated in the complete maturation
process. Studies showed that TNF-a cytokines and the NF-kB
signaling pathway were indispensable elements in the differentiation
and maturation process in BMDCs (45, 46). In this study, we found
that the expression of TLR4, MYD88, IKKa, IkB-a, and p-NF-kB
were decreased in Gpr174-/- with LPS stimulation. Taken together,
we considered that deletion of Gpr174 might inhibit the NF-kB
pathway and suppress BMDCs maturation. ACKNOWLEDGMENTS The authors wish to thank all the patients enrolled in this study. AUTHOR CONTRIBUTIONS ZS and WW designed the research plan. WW, SM and YH
conducted the experiments, analyzed all data, and wrote the
manuscript. WW, SM and YH were co-first authors. ZK helped
with the mice experiment. YC and YL assisted with some data
analysis. YZ and XW collaborated to collect endoscopy biopsies. CT and YY supervised the experimental work and data analysis. ZS, YY and YZ were co-corresponding authors. All authors
participated in revising the manuscript and agreed to the
final version. The progression of IBD is closely related to the immune
system (44). Animal studies showed that the GPR174 was
involved in IBD’s pathogenic course; some clinical evidence
must be unearthed. Therefore, we took samples from ulcerative
patients and set the non-lesion site as healthy controls to
combine with clinical practice. The mRNA expression of
GPR174 decreased in the inflamed tissues compared with
healthy controls, which was consistent with the
immunofluorescence results. This might be because the
samples we collected were the patients in the late disease
courses of colitis. We stimulated BMDCs of WT mice with
LPS to validate this idea and found Gpr174 increased in the
first 24h. Moreover, we collected the healthy controls’ PBMC to
induce the monocyte-derived DCs (MODCs) and found the
GPR174 began to increase in the 12h after LPS stimulation in
vitro. Therefore, in the early course of LPS stimulation and
colitis, the GPR174 might increase, while in the late course of the
disease, the GPR174 expression decreased (Figure S1). In-depth
studies are required to confirm these observations further. ETHICS STATEMENT The studies involving human participants were reviewed and
approved by NO:B2020-016R. The patients/participants
provided their written informed consent to participate in this
study. The animal study was reviewed and approved by
NO. 201804001Z. To further validate whether Gpr174-/- BMDCs have protective
effects on IBD, we transferred Gpr174-/- BMDCs into DSS-
induced colitis animals. Adoptive transfer of Gpr174-/- BMDCs
could significantly improve intestinal damage and suppress the
inflammatory response in both innate and acquired immunity. Some clinical research revealed that patients with active IBD
lacked immature peripheral blood plasmacytoid and myeloid
DCs (42). DCs with immature phenotype could induce immune
tolerance by inducing T regulatory cells and amplifying
suppressor activity, which may have anti-inflammatory
effects (43). DATA AVAILABILITY STATEMENT intestinal homeostasis. It is reported that TNF-a on the intestinal
epithelium could disrupt the epithelial barrier and decrease the
tight junction proteins expression, including occludin and Zo-1
(37, 38). Once immune cells entered intestinal mucosa through
broken TJs, the raised TNF-a and IFN-g triggered cytokine
storm made the intestinal barrier disruption heavier (39–41). This study showed increased occludin and Zo-1 mRNA levels
and reduced macromolecule permeability in Gpr174-/- intestinal
mucosa compared with Gpr174 +/+ ones, indicating Gpr174
knockout helped maintain the intensity of the intestinal
mucosa in DSS treated mice. The datasets presented in this study can be found in online
repositories. The names of the repository/repositories and accession
number(s) can be found below: https://www.ncbi.nlm.nih.gov/
bioproject/, PRJNA819571. Frontiers in Immunology | www.frontiersin.org REFERENCES Hammer GE, Ma A. Molecular Control of Steady-State Dendritic Cell
Maturation and Immune Homeostasis. Annu Rev Immunol (2013) 31:743–
91. doi: 10.1146/annurev-immunol-020711-074929 7. Qualls JE, Tuna H, Kaplan AM, Cohen DA. Suppression of Experimental
Colitis in Mice by CD11c+ Dendritic Cells. Inflammation Bowel Dis (2009)
15:236–47. doi: 10.1002/ibd.20733 27. Merad M, Sathe P, Helft J, Miller J, Mortha A. The Dendritic Cell Lineage:
Ontogeny and Function of Dendritic Cells and Their Subsets in the Steady
State and the Inflamed Setting. Annu Rev Immunol (2013) 31:563–604. doi:
10.1146/annurev-immunol-020711-074950 8. Lämmermann T, Kastenmüller W. Concepts of GPCR-Controlled Navigation
in the Immune System. Immunol Rev (2019) 289:205–31. doi: 10.1111/
imr.12752 28. Geremia A, Biancheri P, Allan P, Corazza GR, Di Sabatino A. Innate and
Adaptive Immunity in Inflammatory Bowel Disease. Autoimmun Rev (2014)
13:3–10. doi: 10.1016/j.autrev.2013.06.004 9. Chu X, Shen M, Xie F, Miao XJ, Shou WH, Liu L, et al. An X Chromosome-
Wide Association Analysis Identifies Variants in GPR174 as a Risk Factor for
Graves' Disease. J Med Genet (2013) 50:479–85. doi: 10.1136/jmedgenet-2013-
101595 29. Connor LM, Tang SC, Camberis M, Le Gros G, Ronchese F. Helminth-
Conditioned Dendritic Cells Prime CD4+ T Cells to IL-4 Production In Vivo. J Immunol (2014) 193:2709–17. doi: 10.4049/jimmunol.1400374 10. Napier C, Mitchell AL, Gan E, Wilson I, Pearce SH. Role of the X-Linked Gene
GPR174 in Autoimmune Addison's Disease. J Clin Endocrinol Metab (2015)
100:E187–90. doi: 10.1210/jc.2014-2694 30. Keshavarz Shahbaz S, Varasteh AR, Koushki K, Ayati SH, Mashayekhi K,
Sadeghi M, et al. Sublingual Dendritic Cells Targeting by Aptamer: Possible
Approach for Improvement of Sublingual Immunotherapy Efficacy. Int
Immunopharmacol (2020) 85:106603. doi: 10.1016/j.intimp.2020.106603 11. Zhu M, Li C, Song Z, Mu S, Wang J, Wei W, et al. The Increased Marginal
Zone B Cells Attenuates Early Inflammatory Responses During Sepsis in
Gpr174 Deficient Mice. Int Immunopharmacol (2020) 81:106034. doi:
10.1016/j.intimp.2019.106034 31. Neurath MF. Cytokines in Inflammatory Bowel Disease. Nat Rev Immunol
(2014) 14:329–42. doi: 10.1038/nri3661 32. Larmonier CB, Shehab KW, Ghishan FK, Kiela PR. T Lymphocyte Dynamics
in Inflammatory Bowel Diseases: Role of the Microbiome. BioMed Res Int
(2015) 2015:504638. doi: 10.1155/2015/504638 12. Zhao R, Chen X, Ma W, Zhang J, Guo J, Zhong X, et al. A GPR174-CCL21
Module Imparts Sexual Dimorphism to Humoral Immunity. Nature (2020)
577:416–20. doi: 10.1038/s41586-019-1873-0 33. Olson TS, Bamias G, Naganuma M, Rivera-Nieves J, Burcin TL, Ross W, et al. REFERENCES Expanded B Cell Population Blocks Regulatory T Cells and Exacerbates Ileitis
in a Murine Model of Crohn Disease. J Clin Invest (2004) 114:389–98. doi:
10.1172/JCI200420855 13. Breynaert C, de Bruyn M, Arijs I, Cremer J, Martens E, Van Lommel L, et al. Genetic Deletion of Tissue Inhibitor of Metalloproteinase-1/TIMP-1 Alters
Inflammation and Attenuates Fibrosis in Dextran Sodium Sulphate-Induced
Murine Models of Colitis. J Crohns Colitis (2016) 10:1336–50. doi: 10.1093/
ecco-jcc/jjw101 34. Bai A, Lu N, Zeng H, Li Z, Zhou X, Chen J, et al. All-Trans Retinoic Acid
Ameliorates Trinitrobenzene Sulfonic Acid-Induced Colitis by Shifting Th1 to
Th2 Profile. J Interferon Cytokine Res (2010) 30:399–406. doi: 10.1089/
jir.2009.0028 14. Ghia JE, Blennerhassett P, Collins SM. Impaired Parasympathetic Function
Increases Susceptibility to Inflammatory Bowel Disease in a Mouse Model of
Depression. J Clin Invest (2008) 118:2209–18. doi: 10.1172/JCI32849 35. Li J, Hou W, Lin S, Wang L, Pan C, Wu F, et al. Polydopamine Nanoparticle-
Mediated Dopaminergic Immunoregulation in Colitis. Adv Sci (Weinh)
(2022) 9:e2104006. doi: 10.1002/advs.202104006 15. Kimball ES, Wallace NH, Schneider CR, D'Andrea MR, Hornby PJ. Vanilloid
Receptor 1 Antagonists Attenuate Disease Severity in Dextran Sulphate
Sodium-Induced Colitis in Mice. Neurogastroenterol Motil (2004) 16:811–8. doi: 10.1111/j.1365-2982.2004.00549.x 36. Marafini I, Sedda S, Dinallo V, Monteleone G. Inflammatory Cytokines: From
Discoveries to Therapies in IBD. Expert Opin Biol Ther (2019) 19:1207–17. doi: 10.1080/14712598.2019.1652267 16. An G, Wei B, Xia B, McDaniel JM, Ju T, Cummings RD, et al. Increased
Susceptibility to Colitis and Colorectal Tumors in Mice Lacking Core 3-
Derived O-Glycans. J Exp Med (2007) 204:1417–29. doi: 10.1084/
jem.20061929 37. Ye X, Sun M. AGR2 Ameliorates Tumor Necrosis Factor-a-Induced Epithelial
Barrier Dysfunction via Suppression of NF-kb P65-Mediated MLCK/p-MLC
Pathway Activation. Int J Mol Med (2017) 39:1206–14. doi: 10.3892/
ijmm.2017.2928 17. Fischer AH, Jacobson KA, Rose J, Zeller R. Hematoxylin and Eosin Staining of
Tissue and Cell Sections. CSH Protoc (2008) 2008:pdb.prot4986. doi: 10.1101/
pdb.prot4986 38. Watari A, Sakamoto Y, Hisaie K, Iwamoto K, Fueta M, Yagi K, et al. Rebeccamycin Attenuates TNF-a-Induced Intestinal Epithelial Barrier
Dysfunction by Inhibiting Myosin Light Chain Kinase Production. Cell
Physiol Biochem (2017) 41:1924–34. doi: 10.1159/000472367 18. Janelsins BM, Mathers AR, Tkacheva OA, Erdos G, Shufesky WJ, Morelli AE,
et al. Proinflammatory Tachykinins That Signal Through the Neurokinin 1
Receptor Promote Survival of Dendritic Cells and Potent Cellular Immunity. Blood (2009) 113:3017–26. doi: 10.1182/blood-2008-06-163121 Physiol Biochem (2017) 41:1924–34. doi: 10.1159/000472367 39. REFERENCES 20. Williams IR. CCR6 and CCL20: Partners in Intestinal Immunity and
Lymphorganogenesis. Ann N Y Acad Sci (2006) 1072:52–61. doi: 10.1196/
annals.1326.036 1. Head KA, Jurenka JS. Inflammatory Bowel Disease Part 1: Ulcerative Colitis– 1. Head KA, Jurenka JS. Inflammatory Bowel Disease Part 1: Ulcerative Colitis–
Pathophysiology and Conventional and Alternative Treatment Options. Altern Med Rev (2003) 8:247–83. 21. Ferrer-Picón E, Dotti I, Corraliza AM, Mayorgas A, Esteller M, Perales JC,
et al. Intestinal Inflammation Modulates the Epithelial Response to Butyrate
in Patients With Inflammatory Bowel Disease. Inflammation Bowel Dis (2020)
26:43–55. doi: 10.1093/ibd/izz119 Pathophysiology and Conventional and Alternative Treatment Options. Altern Med Rev (2003) 8:247–83. 2. Jiang LY, Zhang M, Zhou TE, Yang ZF, Wen LQ, Chang JX. Changes of the
Immunological Barrier of Intestinal Mucosa in Rats With Sepsis. World J
Emerg Med (2010) 1:138–43. 22. Qiu D, Chu X, Hua L, Yang Y, Li K, Han Y, et al. Gpr174-Deficient Regulatory
T Cells Decrease Cytokine Storm in Septic Mice. Cell Death Dis (2019) 10:233. doi: 10.1038/s41419-019-1462-z 3. Stagg AJ. Intestinal Dendritic Cells in Health and Gut Inflammation. Front
Immunol (2018) 9:2883–3. doi: 10.3389/fimmu.2018.02883 23. Worbs T, Hammerschmidt SI, Förster R. Dendritic Cell Migration in Health
and Disease. Nat Rev Immunol (2017) 17:30–48. doi: 10.1038/nri.2016.116 4. Middel P, Raddatz D, Gunawan B, Haller F, Radzun HJ. Increased Number of 4. Middel P, Raddatz D, Gunawan B, Haller F, Radzun HJ. Increased Number of
Mature Dendritic Cells in Crohn's Disease: Evidence for a Chemokine
Mediated Retention Mechanism. Gut (2006) 55:220–7. doi: 10.1136/
gut.2004.063008 24. Woodruff MC, Heesters BA, Herndon CN, Groom JR, Thomas PG, Luster
AD, et al. Trans-Nodal Migration of Resident Dendritic Cells Into Medullary
Interfollicular Regions Initiates Immunity to Influenza Vaccine. J Exp Med
(2014) 211:1611–21. doi: 10.1084/jem.20132327 5. Baumgart DC, Metzke D, Guckelberger O, Pascher A, Grötzinger C,
Przesdzing I, et al. Aberrant Plasmacytoid Dendritic Cell Distribution and
Function in Patients With Crohn's Disease and Ulcerative Colitis. Clin Exp
Immunol (2011) 166:46–54. doi: 10.1111/j.1365-2249.2011.04439.x 25. Lutz MB, Kukutsch N, Ogilvie AL, Rössner S, Koch F, Romani N, et al. An
Advanced Culture Method for Generating Large Quantities of Highly Pure
Dendritic Cells From Mouse Bone Marrow. J Immunol Methods (1999)
223:77–92. doi: 10.1016/S0022-1759(98)00204-X 6. Dieleman LA, Ridwan BU, Tennyson GS, Beagley KW, Bucy RP, Elson CO. Dextran Sulfate Sodium-Induced Colitis Occurs in Severe Combined
Immunodeficient Mice. Gastroenterology (1994) 107:1643–52. doi: 10.1016/
0016-5085(94)90803-6 26. SUPPLEMENTARY MATERIAL In summary, our work revealed that the knockout of GPR174
could reduce intestinal inflammation and repair the epithelium
barrier by suppressing DCs maturation and naïve T cell
differentiation to alleviate DSS-induced colitis, which was
probably through inhibiting the TNF-a (NF-kB) pathway. The Supplementary Material for this article can be found online at:
https://www.frontiersin.org/articles/10.3389/fimmu.2022.841254/
full#supplementary-material May 2022 | Volume 13 | Article 841254 Frontiers in Immunology | www.frontiersin.org Wei et al. Gpr174 Influence Colitis via DCs REFERENCES Bruewer M, Luegering A, Kucharzik T, Parkos CA, Madara JL, Hopkins AM,
et al. Proinflammatory Cytokines Disrupt Epithelial Barrier Function by
Apoptosis-Independent Mechanisms. J Immunol (2003) 171:6164–72. doi:
10.4049/jimmunol.171.11.6164 19. Szekanecz Z, Koch AE. Successes and Failures of Chemokine-Pathway
Targeting in Rheumatoid Arthritis. Nat Rev Rheumatol (2016) 12:5–13. doi:
10.1038/nrrheum.2015.157 May 2022 | Volume 13 | Article 841254 Frontiers in Immunology | www.frontiersin.org 12 Wei et al. Gpr174 Influence Colitis via DCs 46. Rescigno M, Martino M, Sutherland CL, Gold MR, Ricciardi-Castagnoli P. Dendritic Cell Survival and Maturation are Regulated by Different Signaling
Pathways. J Exp Med (1998) 188:2175–80. doi: 10.1084/jem.188.11.2175 40. Schmitz H, Fromm M, Bentzel CJ, Scholz P, Detjen K, Mankertz J, et al. Tumor Necrosis Factor-Alpha (TNFalpha) Regulates the Epithelial Barrier in
the Human Intestinal Cell Line HT-29/B6. J Cell Sci (1999) 112(Pt 1):137–46. doi: 10.1242/jcs.112.1.137 Conflict of Interest Statement: The authors declare that the research was
conducted in the absence of any commercial or financial relationships that
could be construed as a potential conflict of interest. 41. Madara JL, Stafford J. Interferon-Gamma Directly Affects Barrier Function of
Cultured Intestinal Epithelial Monolayers. J Clin Invest (1989) 83:724–7. doi:
10.1172/JCI113938 42. Baumgart DC, Metzke D, Schmitz J, Scheffold A, Sturm A, Wiedenmann B,
et al. Patients With Active Inflammatory Bowel Disease Lack Immature
Peripheral Blood Plasmacytoid and Myeloid Dendritic Cells. Gut (2005)
54:228–36. doi: 10.1136/gut.2004.040360 Publisher’s Note: All claims expressed in this article are solely those of the authors
and do not necessarily represent those of their affiliated organizations, or those of
the publisher, the editors and the reviewers. Any product that may be evaluated in
this article, or claim that may be made by its manufacturer, is not guaranteed or
endorsed by the publisher. 43. Gad M, Kristensen NN, Kury E, Claesson MH. Characterization of T-Regulatory
Cells, Induced by Immature Dendritic Cells, Which Inhibit Enteroantigen-
Reactive Colitis-Inducing T-Cell Responses In Vitro and In Vivo. Immunology
(2004) 113:499–508. doi: 10.1111/j.1365-2567.2004.01977.x Copyright © 2022 Wei, Mu, Han, Chen, Kuang, Wu, Luo, Tong, Zhang, Yang and
Song. This is an open-access article distributed under the terms of the Creative
Commons Attribution License (CC BY). The use, distribution or reproduction in other
forums is permitted, provided the original author(s) and the copyright owner(s) are
credited and that the original publication in this journal is cited, in accordance with
accepted academic practice. May 2022 | Volume 13 | Article 841254 REFERENCES No use, distribution or reproduction is permitted which
does not comply with these terms. 44. Ungaro R, Mehandru S, Allen PB, Peyrin-Biroulet L, Colombel J-F. Ulcerative
Colitis. Lancet (London England) (2017) 389:1756–70. doi: 10.1016/S0140-
6736(16)32126-2 45. Berges C, Naujokat C, Tinapp S, Wieczorek H, Höh A, Sadeghi M, et al. A Cell
Line Model for the Differentiation of Human Dendritic Cells. Biochem
Biophys Res Commun (2005) 333:896–907. doi: 10.1016/j.bbrc.2005.05.171 May 2022 | Volume 13 | Article 841254 Frontiers in Immunology | www.frontiersin.org 13
|
https://openalex.org/W4386504070
|
https://journal.univpancasila.ac.id/index.php/ARTESIS/article/download/3764/1920
|
Indonesian
| null |
IMPLEMENTASI PERMIT TO WORK SYSTEM PADA PEKERJAAN KETINGGIAN DI AREA STEAM TURBINE BUILDING PROJECT PLTGU MUARA TAWAR BEKASI
|
Jurnal Artesis
| 2,022
|
cc-by
| 3,591
|
Diterima 10 Mei 2022, Disetujui 27 Mei 2022 Diterima 10 Mei 2022, Disetujui 27 Mei 2022 ABSTRAK Pekerjaan konstruksi adalah pekerjaan Penuh dengan aktivitas dengan risiko yang cukup tinggi, seperti mengangkat benda
berat, bekerja di ketinggian, dan Bekerja di ruang terbatas. Oleh karena itu, penerapan pada kesehatan dan keselamatan
kerja pekerjaan konstruksi (K3) khususnya pada izin kerja perusahaan (Permit To Work), sangat penting dan sesuai
peraturan perushaan yang dijalankan. Program sistem izin kerja atau Permit To Work adalah sistem otorisasi (pemberian
hak) dalam bentuk tertulis formal yang digunakan untuk mengontrol jenis pekerjaan yang berpotensi bahaya. Penelitian ini
bertujuan untuk mengetahui program, kesesuaian dan implementasi Permit To Work System pada Project PLTGU Muara
Tawar Bekasi. Dari data tersebut dilakukan analisis dengan metode wawancara, observasi serta data sekunder untuk
mendapatkan hasil kesesuaian pelaksanaan di lapangan dengan prosedur yang berlaku. Dari hasil analisis tersebut yang
didapat dari pengamatan dalam pelaksanaannya lokasi secara keseluruhan Proyek PLTGU Hutama Karya Muara Tawar
Bekasi telah menerapkan sistem izin kerja untuk bekerja di ketinggian yang sesuai Standar dengan Guidelines on Permit To
Work (P.T.W.) Systems Association of Oil & Gas Producers. Yang memiliki persentase penurunan bahaya pekerjaan yaitu
sekitar 100% yang membuktikan bahwa penerapan Permit To Work System pada Project PLTGU Muara Tawar terlaksana
dengan baik. Kata kunci: Sistem Izin Kerja, Administasi K3, Pekerjaan Di Ketinggian, Implementasi. Jurnal Artesis. Vol.2 (1): 74-79 Jurnal Artesis. Vol.2 (1): 74-79 Cahya Suci Wulandhari1, Akhmad Dofir1, Ayu Herzanita1
1Program Studi Teknik Sipil, Fakultas Teknik, Universitas Pancasila, Jakarta
E-mail: cahyasuciw98@gmail.com Diterima 10 Mei 2022, Disetujui 27 Mei 2022 (Implementation Of Permit To Work System On Height Works In The Steam Turbine Building Project Area
Of PLTGU Muara Tawar Bekasi) Cahya Suci Wulandhari1, Akhmad Dofir1, Ayu Herzanita1
1Program Studi Teknik Sipil, Fakultas Teknik, Universitas Pancasila, Jakarta
E-mail: cahyasuciw98@gmail.com Keywords: Permit To Work System, Safety Admisnitration, Working at height, Implementation. ketinggian [2]. Banyak masalah yang timbul ketika pekerja bekerja
di ketinggian misalnya tidak terdapat ijin kerja atau
Permit To Work di Ketinggian, bekerja tidak mematuhi
prosedur yang ada, alat kerja atau scaffolding yang tidak
aman
untuk
digunakan. Setelah
diindentifikasi
ditemukan beberapa faktor yaitu pekerjaan tidak
menggunakan full body harness dengan baik dan benar,
alat kerja (scaffolding) tidak sesuai dengan prosedur, dan
tidak ada izin pekerjaan atau Permit To Work System. Hal
ini
sangat
membahayakan
dikarenakan
dapat
menimbulkan kerugian bagi pekerja yang bersangkutan
dan perusahaan yang terkait. Pekerjaan konstruksi adalah pekerjaan yang padat
akan aktifitas dengan level risiko yang cukup tinggi,
misalnya pekerjaan pengangkatan benda-benda berat,
bekerja pada ketinggian, serta pekerjaan pada ruang
terbatas. Kecelakaan dalam sebuah pekerjaan konstruksi
sebagian besar disebabkan oleh kurangnya pengalaman
kerja pekerja, peralatan yang tidak layak pakai,
lingkungan kerja yang tidak aman, penggunaan peralatan
yang tidak sesuai spesifikasi, dan ketidakpedulian
karyawan
terhadap
perilaku. Keselamatan,
dan
manajemen perusahaan tidak sepenuhnya peduli dengan
metode kerja yang aman dan tidak aman. Dengan
mengambil keputusan yang tepat pada tahap desain
teknik, kecelakaan yang disebabkan oleh metode kerja
yang salah dapat dihindari, yang merupakan tanggung
jawab insinyur, dan untuk penyebab kecelakaan lainnya,
menjadi
tanggung
jawab
kontraktor
untuk
memperhatikan hal ini. Dalam setiap kegiatan di dunia
industri ataupun konstruksi tidak akan terlepas dari
suatu kondisi yang berpotensi terhadap timbulnya
bahaya, mulai dari kejadian hampir celaka (Nearmiss),
luka
ringan,
sedang,
berat
dan
fatal,
hingga
mengakibatkan kematian dan rusaknya peralatan
industri dan kerusakan lingkungan. Efek dari kecelakaan
tersebut tentu saja dapat menimbulkan kerugian, baik
bagi pihak perusahaan, karyawan dan masyarakat
sekitar. Maka
salah
satu
upaya
untuk
pencegahan
kecelakaan di suatu perusahaan Kontruksi adalah
dengan diberlakukannya Permit To Work System (PTW)
yaitu khususnya pekerjaan di ketinggian yang memiliki
resiko tinggi. Berdasarkan dengan latar belakang yang
ada, perlu dilakukan pengamatan dan observasi
lapangan mengenai Implementasi Permit To Work
System Pada Pekerjaan Ketinggian Di Area Steam Turbine
Building Project PLTGU Muara Tawar Bekasi. Lokasi Penelitian Lokasi penelitian pada Unit Pembangkitan Muara
Tawar sebuah pembangkit listrik tenaga gas uap (PLTGU)
yang dikelola oleh PT Pembangkitan Jawa-Bali di Jl. PLTGU
Muara Tawar No. 1, Segarajaya, Kec. Tarumajaya, Bekasi,
Jawa Barat 17212. ABSTRACT Construction work is a job full of activities with a fairly high risk, such as lifting heavy objects, working at heights, and working
in confined spaces. Therefore, the application of the occupational health and safety of construction work (K3), especially the
company's work permit (Permit To Work), is very important and is in accordance with the company's regulations. The work
permit system program or Permit To Work is a formal written authorization system that is used to control types of work that
are potentially hazardous. This study aims to determine the program, suitability and implementation of the Permit To Work
System on the Muara Tawar Bekasi PLTGU Project. From these data, analysis was carried out using interviews, observation and
secondary data to obtain results of conformity with the implementation in the field with applicable procedures. From the results
of the analysis obtained from observations in its implementation, the overall location of the PLTGU Hutama Karya Muara Tawar
Bekasi project has implemented a work permit system to work at a height that is in accordance with the Standards with the
Guidelines on Permit To Work (P.T.W.) Systems Association of Oil & Gas Producers. Which has a percentage of work hazard
reduction which is around 100% which proves that the application of the Permit To Work System in the Muara Tawar PLTGU
Project has been carried out well. Keywords: Permit To Work System, Safety Admisnitration, Working at height, Implementation. Implementasi Permit To Work System Pada Pekerjaan Ketinggian… Implementasi Permit To Work System Pada Pekerjaan Ketinggian…………………………………….…….... .Cahya, Dofir, Ayu ketinggian [2]. Berdasarkan latar belakang tersebut Adapun tujuan
penelitian ini yaitu, untuk mengetahui prosedur dari
pembuatan Permit To Work Pekerjaan Ketinggian
Cladding And Roofing Installation On Steam Turbine
Building berdasarkan workflow Permit To Work System
pada Project PLTGU Muara Tawar, untuk mengetahui
implementasi dan evaluasi dari Permit To Work System
Pekerjaan Di Ketinggian Area Steam Turbine Building
PLTGU Muara Tawar dengan Guidelines on Permit To
Work (P.T.W.) Systems Association of Oil & Gas Producers
[3], Untuk mengetahui persentase penurunan dari
pengaruh Permit To Work System Project PLTGU Muara
Tawar Bekasi dalam total pengurangan bahaya. Maka
dari
itu
penerapan
Kesehatan
dan
Keselamatan Kerja (K3) terutama pada Surat Ijin Kerja
(Permit To Work) dalam sebuah perusahaan sangat
penting dan sesuai dengan peraturan yang telah
ditetapkan. Hasil penelitian Aryanto menunjukkan angka
kecelakaan kerja menurun sejak diberlakukannya sistem
izin kerja [1]. Untuk memastikan keberhasilan industri,
aspek keselamatan berperan dalam meminimalkan
risiko bahaya di tempat kerja. Dalam hal ini keselamatan
kerja haruslah mendapat perhatian utama demi
berhasilnya program-program perusahaan dalam rangka
meningkatkan produktivitas bagi perusahaan salah
satunya yaitu dengan cara administrasi control atau izin
kerja. Keselamatan dan kesehatan kerja juga akan dapat
menciptakan keamanan dan kenyamanan kerja serta
mempunyai peranan penting dalam usaha mencegah dan
menanggulangi
adanya
resiko
kecelakaan,
serta
pengamanan aset perusahaan. c. Request For Work (RFW) 2. Pengawas Pekerjaan melengkapi formulir PTW bekerja
diketinggian 3. Pemberi
ijin
mereview
lembar
PTW
bekerja
diketinggian, metoda kerja, JSA, RFW 4. Melakukan identifikasi bahaya diarea sekitar bekerja
diketinggian gg
5. Bilamana pekerjaan diketinggian tidak aman untuk
dilakukan pemberi ijin melakukan mitigasi 6. Bila pekerjaan diketinggian aman untuk dilakukan
Pemberi
ijin
menandatangani
PTW
bekerja
diketinggian. PTW bekerja diketinggian dipampang
dilokasi dilokasi tempat pekerjaan berlangsung 7. Apabila
terdapat
perubahan
lingkup
pekerjaan,
pekerjaan dihentikan dan kembali ke poin satu Data Umum Proyek UP Muara Tawar merupakan Pembangkit Listrik
Tenaga Gas Uap (PLTGU) yang ter diri dari 5 blok. Blok 1
dan 5 merupakan blok siklus gabungan, sedangkan Blok 2,
3 dan 4 saat ini masih dalam siklus terbuka dan dalam
tahap tambahan pembangunan HRSG dan turbin uap. Saat
itu pembangunan blok 2, 3 dan 4 dilakukan oleh
kontraktor PT Hutama Karya. Pada Proyek PLTGU Muara Tawar yang melakukan
pekerjaan
konstruksi,
pekerjaan-pekerjaan
yang
memiliki banyak risiko tinggi. Jelas kiranya risiko-risiko
tersebut harus diminimalkan. Selain risiko kerugian
akibat rusaknya peralatan, hal lain yang sangat dihindari
adalah risiko kecelakaan kerja pada karyawan yang
sedang bekerja. Nama Proyek : PLTGU Muara Tawar Add on 2x650 MW
Lokasi Jalan : Muara
Tawar,
Desa
Segara
Jaya,
Kecamatan
Tarumajaya,
Kabupaten
Bekasi, Jawa Barat Nama Proyek : PLTGU Muara Tawar Add on 2x650 MW
Lokasi Jalan : Muara
Tawar,
Desa
Segara
Jaya,
Kecamatan
Tarumajaya,
Kabupaten
Bekasi, Jawa Barat Salah satu pekerjaan yang mengandung unsur
resiko tinggi yang dapat menimbulkan kerugian adalah
pekerjaan yang sangat berkaitan atau bekerja di
ketinggian
Working
at
Height. Pemerintah
Ketenagakerjaan telah mengatur peraturan mengenai
pekerjaan diketinggan yang disebutkan dalam Permen
No.9 tahun 2016 tentang kewajiban pengusaha dan atau
pengurus untuk menerapkan K3 dalam bekerja di J
Koordinat
: 06°05’11’’S 106°59’54’’E 75 Implementasi Permit To Work System Pada Pekerjaan Ketinggian…………………………………….……....…………....Cahya, Dofir, Ayu Implementasi Permit To Work System Pada Pekerjaan Ketinggian…… Luas Area
:
±40 Ha
Pemberi Karya : PT. Pembangkit Jawa Bali (PJB)
Sifat Kontraktor : EPC Fixed Lump sump & Unit Price
(Pilling Work)
Contraktor
: Doosan Heavy Industries & Hutama
Karya (persero)
Konsultan Review Engineering : Rekadaya Consultant
Konsultan Super visi Kontruksi : PLN Engineering
Cakupan Pekerjaan : 1. Procurement (FOB dan ex work)
2. Erection
3. Commissioning
Pemberi dana : 70 % by ECA; 30 % by owner Luas Area
:
±40 Ha
Pemberi Karya : PT. Pembangkit Jawa Bali (PJB)
Sifat Kontraktor : EPC Fixed Lump sump & Unit Price
(Pilling Work)
Contraktor
: Doosan Heavy Industries & Hutama
Karya (persero)
Konsultan Review Engineering : Rekadaya Consultant
Konsultan Super visi Kontruksi : PLN Engineering
Cakupan Pekerjaan : 1. Procurement (FOB dan ex work)
2. Erection
3. Commissioning
Pemberi dana : 70 % by ECA; 30 % by owner Gambar 2. Prosedur Permit To Work System Pada Project
PLTGU Muara Tawar Bekasi Implementasi dan Evaluasi Permit To Work System
Pada Pekerjaan di Ketinggian dengan Guidelines on
Permit To Work (P.T.W.) Systems Association of Oil & Gas
Producers 1993. Gambar 1. Bagan Alir Penelitian Berikut Implementasi dan evaluasi Permit To Work
System Pada Pekerjaan di Ketinggian dengan Guidelines on
Permit To Work (P.T.W.) Systems Association of Oil & Gas
Producers [3]: 1. Tahap Persiapan
Implementasi: 1. Tahap Persiapan
Implementasi: Gambar 1. Bagan Alir Penelitian a. Koordinasi
pihak
lapangan
kepada
pihak
engineering atas akan dimulainya pekerjaan agar
dirilisnya Request For Work Metodologi Penelitian Metode pelaksanaan penelitian ini yaitu dengan
beberapa tahapan sebagai berikut: Gambar 2. Prosedur Permit To Work System Pada Project
PLTGU Muara Tawar Bekasi beberapa tahapan sebagai berikut:
1. Metode Wawancara
2. Metode Observasi
3. Studi Litelatur
4. Pengolahan Data
5. Kesimpulan dan Saran
6. Selesai
Berikut bagan alur penelitian dalam penelitian ini: Berikut di bawah ini adalah alur atau urutan cara
pembuatan
PTW
bekerja
diketinggian
berdasarkan
workflow Permit To Work System pada Project PLTGU Muara
Tawar: 1. Pengawas pekerjaan mempersiapkan persyaratan
untuk pengajuan PTW. Ada tiga persyaratan PTW yang
wajib dilampirkan pada saat pengajuan PTW:
a. Metoda Kerja
b. Job Safety Analysis (JSA)
c. Request For Work (RFW) 1. Pengawas pekerjaan mempersiapkan persyaratan
untuk pengajuan PTW. Ada tiga persyaratan PTW yang
wajib dilampirkan pada saat pengajuan PTW: Berikut bagan alur penelitian dalam penelitian ini: Gambar 1. Bagan Alir Penelitian b. Job Safety Analysis (JSA) c. Request For Work (RFW) HASIL DAN PEMBAHASAN HASIL DAN PEMBAHASAN Analisa Prosedur Pembuatan Permit To Work System
Pada Pekerjaan di Ketinggian berdasarkan workflow
Permit To Work System pada Project PLTGU Muara Tawar b. Penentuan jenis pekerjaan b. Penentuan jenis pekerjaan c. Waktu
rencana
awal
pekerjaan
ketinggian
pemasangan cladding
d. Gas Tesing e. Pembuatan dan penandatanganan dokumen ijin
kerja 76 Implementasi Permit To Work System Pada Pekerjaan Ketinggian…………………………………….……....…………....Cahya, Dofir, Ayu Implementasi Permit To Work System Pada Pekerjaan Ketinggian… Implementasi Permit To Work System Pada Pekerjaan Ketinggian…………………………………….……....…………....Cahya, Dofir, Ayu .Cahya, Dofir, Ayu Evaluasi:
a. Kurangnya perencanaan dengan baik dalam
pembuatan isi permit yang menyebabkan ijin kerja
sedikit lebih lama di rilis
2. Tahap Kontrol Pelaksanaan
Implementasi:
a. Validasi dokumen ijin kerja
b. Serah terima dokumen ijin kerja dengan pihak
konstruksi lapangan
c. Tindakan HSE dalam menangani near miss
d. Pemantauan safety officer di lapangan atas alat dan
material yang digunakan
a. Dokumen ijin kerja banyak yang tidak on site
secara tepat waktu dikarenakan permit yang lama
di rilis
3. Tahap Pekerjaan Selesai
Implementasi:
a. Penutupan permit dan informasi kepada PTW
officer jika pekerjaan sudah selesai
b. 2 Inspeksi tempat pekerjaan dan housekeeping
Evaluasi:
Tidak adanya pengembalian dokumen ijin kerja kepada
PTW officer. a. Dokumen ijin kerja banyak yang tidak on site
secara tepat waktu dikarenakan permit yang lama
di rilis 3. Tahap Pekerjaan Selesai
Implementasi: 3. Tahap Pekerjaan Selesai
Implementasi: a. Penutupan permit dan informasi kepada PTW
officer jika pekerjaan sudah selesai Tabel 1. Implementasi dan Evaluasi Permit To Work PLTGU Muara dengan Guidelines on Permit To Work (P.T.W.) Systems
Association of Oil & Gas Producers [3]
Deskripsi Tahapan
Implementasi Permit To Work Ketinggian
PLTGU Muara Tawar Bekasi dengan
Guidelines on Permit To Work (P.T.W.)
Systems
Association of Oil & Gas Producers [3]
Evaluasi Permit To Work Ketinggian
PLTGU Muara Tawar Bekasi dengan
Guidelines on Permit To Work (P.T.W.)
Systems
Association of Oil & Gas Producers [3]
1. Tahapan Persiapan
1. Koordinasi pihak lapangan kepada
pihak engineering atas akan dimulainya
pekerjaan agar dirilisnya Request For
Work
2. Penentuan jenis pekerjaan
3. Waktu rencana awal pekerjaan
ketinggian pemasangan cladding
4. Gas Tesing
6. Pembuatan dan penandatanganan
dokumen ijin kerja
1. Kurangnya perencanaan dengan baik
dalam
pembuatan
isi
permit
yang
menyebabkan ijin kerja sedikit lebih lama
di rilis
2. Tahap Kontrol Pelaksanaan
Pekerjaan
1. Validasi dokumen ijin kerja
2. Serah terima dokumen ijin kerja
dengan pihak konstruksi lapangan
3. Tindakan HSE dalam menangani near
miss
4. HASIL DAN PEMBAHASAN Pada evaluasi Permit To Work System PLTGU Muara 77 Implementasi Permit To Work System Pada Pekerjaan Ketinggian…………………………………….……....…………....Cahya, Dofir, Ayu .Cahya, Dofir, Ayu Implementasi Permit To Work System Pada Pekerjaan Ketinggian Analisa Persentase Penurunan Dari Pengaruh Permit
To Work System PLTGU Muara Tawar Bekasi Dalam Total
Pengurangan Bahaya dengan workflow yang berlaku sudah berjalan dengan
baik dan sifatnya general atau umum, prosedurnya
adalah: a. Seorang
pekerja
mempersiapkan
persyaratan
untuk pengajuan PTW yang sudah di approve oleh
pihak yang berwenang. Persyaratan yang harus
disiapkan: Metoda Kerja, JSA dan RFW Gambar 3. Grafik Safety Matrix PLTGU Muara Tawar Bekasi
Periode Tahun 2021
0
20
40
60JanuariFebruariMaretAprilMeiJuniJuliAgustusSeptemberOktoberNovemberDesember
SAFETY MATRIX PLTGU MUARA TAWAR BEKASI
PERIODE TAHUN 2021
1 Fatality
2 Lost Workday Injuries (Lost Time Incident)
3 Resticted Work Injuries (Lost Work Day)
4 Medical Treatment Case
5 First Aid Case
6 Near Miss
7 Motor Vehicle Incident
8 Property Damaged
9 Enviroment Incident b. Seorang
pekerja
mengisi
form
PTW
dan
melampirkan persyaratan PTW 2. 2. Implementasi Permit To Work System bekerja di
ketinggian Project PLTGU Muara Tawar meliputi
Guidelines on Permit To Work (P.T.W.) Systems
Association of Oil & Gas Producers (1993): f
(
)
a. Tahapan Persiapan yaitu dengan koordinasi pihak
lapangan dan persiapan dokumen Permit To Work
(PTW) meliputi dokumen Metoda Kerja, Dokumen
JSA, dan Dokumen RFW yang sudah di approve oleh
pihan yang berwenang b. Tahap Kontrol Pelaksanaan Pekerjaaan yaitu
dengan pengecekan validasi Permit To Work System
yang ada di lapangan dan pengecekan alat bantu
dalam pekerjaan ketinggian di bangunan Steam
Turbine Building Gambar 3. Grafik Safety Matrix PLTGU Muara Tawar Bekasi
Periode Tahun 2021 Pada grafik dari Safety Matrix PLTGU Muara Tawar
Bekasi Periode Tahun 2021 di atas menunjukan adanya
penurunan pada Near Miss atau Hampir Celaka yaitu istilah
untuk suatu kejadian yang nyaris terjadi kecelakaan di
lapangan. Pada pekerjaan Ketinggian di Steam Turbine
Building dilakukan pada bulan September 2021 sampai
Desember 2021, yang memiliki tingkat near miss terendah
dan juga turun sebesar 13 orang sampai 0 orang hampir
celaka, hal ini merupakan keberhasilan dari ijin kerja dan
implementasi
dari
tahap
persiapan,
tahap
kontrol
pelaksanaan pekerjaan dan tahap pekerjaan selesai. Dimana
kontrol pekerjaan safety officer di lapangan dan inspeksi alat
dan scaffolder untuk pekerjaan ketinggian dilaksanakaan. c. HASIL DAN PEMBAHASAN Tahapan
Pekerjaan
selesai
yaitu
dilakukan
penutupan permit untuk menginformasikan bahwa
pekerjaan tersebut sudah selesai Evaluasi Permit To Work System bekerja di
ketinggian Project PLTGU Muara Tawar meliputi
Guidelines on Permit To Work (P.T.W.) Systems Association
of Oil & Gas Producers [3]: a. Pada tahapan persiapan kurangnya perencanaan
dengan baik dalam pembuatan isi permit yang
menyebabkan ijin kerja sedikit lebih lama di rilis b. Pada tahap kontrol pelaksanaan, dokumen ijin kerja
banyak yang tidak on site secara tepat waktu
dikarenakan permit yang lama di rilis Persentase penurunan jumlah orang yang mengalami near
miss pada Project PLTGU Muara Tawar Bekasi: c. Pada
tahap
pekerjaan
selesai
tidak
adanya
pengembalian dokumen ijin kerja kepada PTW officer Nilai Near Miss Januari 2021: 55 orang
Nilai Near Miss Desember 2021: 0 orang
𝐴=𝜋𝑟2 3. Persentase penurunan kecelakaan kerja yaitu near
miss (Hampir Celaka) pada pekerjaan Ketinggian di
Steam
Turbine
Building
dilakukan
pada
bulan
September 2021 sampai Desember 2021, yang
memiliki tingkat near miss terendah dan juga turun
sebesar 13 orang sampai 0 orang hampir celaka,
dengan nilai persentase penurunan yaitu 100% hal ini
merupakan
keberhasilan
dari
ijin
kerja
dan
implementasi mulai dari tahap persiapan, tahap
kontrol pelaksanaan pekerjaan dan tahap pekerjaan
selesai. Maka persentase :
(Nilai awal near miss 2021 – Nilai akhir near miss 2021)
𝑁𝑖𝑙𝑎𝑖 𝐴𝑤𝑎𝑙 𝑁𝑒𝑎𝑟 𝑀𝑖𝑠𝑠 2021
x 100%
(55 – 0)
55
x 100% = 100% Maka persentase :
(Nilai awal near miss 2021 – Nilai akhir near miss 2021)
𝑁𝑖𝑙𝑎𝑖 𝐴𝑤𝑎𝑙 𝑁𝑒𝑎𝑟 𝑀𝑖𝑠𝑠 2021
x 100% Maka persentase :
(Nilai awal near miss 2021 – Nilai akhir near miss 2021)
𝑁𝑖𝑙𝑎𝑖 𝐴𝑤𝑎𝑙 𝑁𝑒𝑎𝑟 𝑀𝑖𝑠𝑠 2021
x 100% (55 – 0)
55
x 100% = 100% (55 – 0)
55
x 100% = 100% Nilai persentase penurunan dari near miss adalah
100% dimana pada pekerjaan Project PLTGU Muara Tawar
dalam kurun waktu 12 bulan yaitu 100%. Hal ini merupakan
ketepatan selarasnya proses pekerjaan dari mulai tahap
persiapan, tahap pengontrolan pekerjaan dan tahap
pekerjaan selesai dalam surat ijin kerja/Permit To Work
System pada pekerjaan di project PLTGU Muara Tawar baik. UCAPAN TERIMA KASIH Penelitian ini dapat dilaksanakan dengan baik
berkat bantuan dari berbagai pihak. Untuk itu peneliti
mengucapkan terima kasih kepada PTW Officer Pada
Project PLTGU Muara Tawar Bekasi, Akhmad Dofir, Ir.,
M.T. dan Ayu Herzanita Y.,S.T.,M.T. selaku dosen
pembimbing, dan Ketua Jurusan Teknik Sipil Universitas
Pancasila yang telah memberikan kerja sama yang baik
dalam penelitian ini. HASIL DAN PEMBAHASAN Pemantauan safety officer di lapangan
atas alat dan material yang digunakan
1. Dokumen ijin kerja banyak yang tidak
on site secara tepat waktu dikarenakan
permit yang lama di rilis
3. Tahap Pekerjaan Selesai
1. Penutupan permit dan informasi
kepada PTW officer jika pekerjaan sudah
selesai
2. Inspeksi tempat pekerjaan dan
housekeeping
1. Tidak adanya pengembalian dokumen
ijin kerja kepada PTW officer an Evaluasi Permit To Work PLTGU Muara dengan Guidelines on Permit To Work (P.T.W.) Systems
l & Gas Producers [3] 1. Kurangnya perencanaan dengan baik
dalam
pembuatan
isi
permit
yang
menyebabkan ijin kerja sedikit lebih lama
di rilis 1. Dokumen ijin kerja banyak yang tidak
on site secara tepat waktu dikarenakan
permit yang lama di rilis 1. Dokumen ijin kerja banyak yang tidak
on site secara tepat waktu dikarenakan
permit yang lama di rilis 1. Tidak adanya pengembalian dokumen
ijin kerja kepada PTW officer 1. Tidak adanya pengembalian dokumen
ijin kerja kepada PTW officer Tawar perlu adanya perencanaan yang lebih baik dan
matang agar permit pekerjaan sudah selesai sebelum
pekerjaan dimulai, dikarenakan ijin kerja harus on site di
lapangan untuk memudahkan proses pemantauan oleh
safety officer. Bahwasanya di buku Guidelines on Permit To
Work (P.T.W.) Systems Association of Oil & Gas Producers
dijelaskan tentang proses tahapan perisapan, pengontrolan
pekerjaan, pekerjaan selesai dan siapa yang bertangguang
jawab dalam pengontrolan pekerjaan [3]. Dari hasil perbandingan antara implementasi dan
evaluasi Permit To Work (PTW) pekerjaan di ketinggian area
Steam Turbine Building yang ada di Project PLTGU Muara
Tawar Bekasi dengan Guidelines on Permit To Work (P.T.W.)
Systems Association of Oil & Gas Producers (1993) dapat
diambil kesimpulan bahwa implementasi yang ada di
Project PLTGU Muara Tawar telah sesuai dengan buku
Guidelines on Permit To Work (P.T.W.) Systems Association of
Oil & Gas Producers yang terbagi menjadi 3 bagian yaitu
dimulai dari proses tahapan perisapan, pengontrolan
pekerjaan, pekerjaan selesai dan siapa yang bertangguang
jawab dalam pengontrolan pekerjaan [3]. Tawar perlu adanya perencanaan yang lebih baik dan
matang agar permit pekerjaan sudah selesai sebelum
pekerjaan dimulai, dikarenakan ijin kerja harus on site di
lapangan untuk memudahkan proses pemantauan oleh
safety officer. Bahwasanya di buku Guidelines on Permit To
Work (P.T.W.) Systems Association of Oil & Gas Producers
dijelaskan tentang proses tahapan perisapan, pengontrolan
pekerjaan, pekerjaan selesai dan siapa yang bertangguang
jawab dalam pengontrolan pekerjaan [3]. KESIMPULAN Dari hasil penelitian dengan tema “Implementasi
Permit To Work System Pada Pekerjaan Ketinggian Di Area
Steam Turbine Building Project PLTGU Muara Tawar
Bekasi” dapat diambil kesimpulan: 1. Project PLTGU Muara Tawar Memiliki prosedur
pembuatan dalam Permit To Work System sesuai 78 Implementasi Permit To Work System Pada Pekerjaan Ketinggian…………………………………….……....…………....Cahya, Dofir, Ayu Implementasi Permit To Work System Pada Pekerjaan Ketinggian… Implementasi Permit To Work System Pada Pekerjaan Ketinggian…………………………………….……....…………....Cahya, Dofir, Ayu ..Cahya, Dofir, Ayu DAFTAR PUSTAKA [1]
Aryanto H, “Studi Pelaksanaan Sistem Ijin Kerja dan
Kejadian Kecelakaan Kerja (Studi Kasus di PT Petro
Oxo Nusantara Gresik),“ B.S. thesis, Fakultas
Kesehatan Masyarakat , Universitas Airlangga,
Gresik Jawa Timur, Indonesia, 2004. [2]
Peraturan Menteri Ketenagakerjaan Republik
Indonesia Nomor 9 Tahun 2016, “Keselamatan
Dan Kesehatan Kerja Dalam Pekerjaan Pada
Ketinggian,” Mar. 2016. [3]
Guidelines On Permit To Work (PTW) System,
“Internasional Association of Oil & Gas Producers ,
OGP,” 1993. [1]
Aryanto H, “Studi Pelaksanaan Sistem Ijin Kerja dan
Kejadian Kecelakaan Kerja (Studi Kasus di PT Petro
Oxo Nusantara Gresik),“ B.S. thesis, Fakultas
Kesehatan Masyarakat , Universitas Airlangga,
Gresik Jawa Timur, Indonesia, 2004. [2]
Peraturan Menteri Ketenagakerjaan Republik
Indonesia Nomor 9 Tahun 2016, “Keselamatan
Dan Kesehatan Kerja Dalam Pekerjaan Pada
Ketinggian,” Mar. 2016. gg
[3]
Guidelines On Permit To Work (PTW) System,
“Internasional Association of Oil & Gas Producers ,
OGP,” 1993. 79
|
https://openalex.org/W4380480427
|
https://ieeexplore.ieee.org/ielx7/6287639/10005208/10149324.pdf
|
English
| null |
Jamming Gripper-Inspired Soft Jig for Perceptive Parts Fixing
|
IEEE access
| 2,023
|
cc-by
| 11,423
|
TATSUYA SAKUMA
1, (Member, IEEE), TAKUYA KIYOKAWA
1,2, (Member, IEEE),
TAKAMITSU MATSUBARA
1, (Member, IEEE), JUN TAKAMATSU3, (Member, IEEE),
TAKAHIRO WADA
1, (Member, IEEE), AND TSUKASA OGASAWARA
4, (Life Member, IEEE)
1Division of Information Science, Graduate School of Science and Technology, Nara Institute of Science and Technology, Ikoma 630-0192, Japan
2Department of Systems Innovation, Graduate School of Engineering Science, Osaka University, Osaka 565-0871, Japan
3Microsoft, Redmond, WA 98052, USA
4Institute for Research Initiatives, Nara Institute of Science and Technology, Ikoma 630-0192, Japan
Corresponding author: Tatsuya Sakuma (sakuma.tatsuya.sn1@is.naist.jp) TATSUYA SAKUMA
1, (Member, IEEE), TAKUYA KIYOKAWA
1,2, (Member, IEEE),
TAKAMITSU MATSUBARA
1, (Member, IEEE), JUN TAKAMATSU3, (Member, IEEE),
TAKAHIRO WADA
1, (Member, IEEE), AND TSUKASA OGASAWARA
4, (Life Member, IEEE)
1Division of Information Science, Graduate School of Science and Technology, Nara Institute of Science and Technology, Ikoma 630-0192, Japan
2Department of Systems Innovation, Graduate School of Engineering Science, Osaka University, Osaka 565-0871, Japan
3Microsoft, Redmond, WA 98052, USA
4Institute for Research Initiatives, Nara Institute of Science and Technology, Ikoma 630-0192, Japan
Corresponding author: Tatsuya Sakuma (sakuma.tatsuya.sn1@is.naist.jp) TATSUYA SAKUMA
1, (Member, IEEE), TAKUYA KIYOKAWA
1,2, (Member, IEEE),
TAKAMITSU MATSUBARA
1, (Member, IEEE), JUN TAKAMATSU3, (Member, IEEE),
TAKAHIRO WADA
1, (Member, IEEE), AND TSUKASA OGASAWARA
4, (Life Member, IEEE)
1Division of Information Science, Graduate School of Science and Technology, Nara Institute of Science and Technology, Ikoma 630-0192, Japan
2Department of Systems Innovation, Graduate School of Engineering Science, Osaka University, Osaka 565-0871, Japan
3Microsoft, Redmond, WA 98052, USA
4Institute for Research Initiatives, Nara Institute of Science and Technology, Ikoma 630-0192, Japan
Corresponding author: Tatsuya Sakuma (sakuma.tatsuya.sn1@is.naist.jp) TATSUYA SAKUMA
1, (Member, IEEE), TAKUYA KIYOKAWA
1,2, (Member, IEEE),
TAKAMITSU MATSUBARA
1, (Member, IEEE), JUN TAKAMATSU3, (Member, IEEE),
TAKAHIRO WADA
1, (Member, IEEE), AND TSUKASA OGASAWARA
4, (Life Member, IEEE)
1Division of Information Science, Graduate School of Science and Technology, Nara Institute of Science and Technology, Ikoma 630-0192, Japan
2Department of Systems Innovation, Graduate School of Engineering Science, Osaka University, Osaka 565-0871, Japan
3Microsoft, Redmond, WA 98052, USA
4Institute for Research Initiatives, Nara Institute of Science and Technology, Ikoma 630-0192, Japan
Corresponding author: Tatsuya Sakuma (sakuma.tatsuya.sn1@is.naist.jp) ABSTRACT To achieve precise and time-efficient operation to ensure stable and reliable product assembly,
custom-made jigs have been developed for each assembly part. However, designing these jigs is a time-
consuming task. Therefore, this study proposes a jamming gripper-inspired flexible jig using a soft membrane
consisting of transparent beads and oil with an adjusted refractive index. The proposed jig is equipped with a
hydraulic drive system with visual control of the amount of transparent oil inside the jig. The proposed system
enables parts fixing by creating a jammed state while maintaining optical transparency, thereby facilitating
the visual sensing of the jig’s membrane from camera sensors embedded in the jig. Furthermore, a sensing
method was developed to perform a precise assembly without external sensors to estimate the pose of an
object based on the proposed point-to-function iterative closest point. TATSUYA SAKUMA
1, (Member, IEEE), TAKUYA KIYOKAWA
1,2, (Member, IEEE),
TAKAMITSU MATSUBARA
1, (Member, IEEE), JUN TAKAMATSU3, (Member, IEEE),
TAKAHIRO WADA
1, (Member, IEEE), AND TSUKASA OGASAWARA
4, (Life Member, IEEE)
1Division of Information Science, Graduate School of Science and Technology, Nara Institute of Science and Technology, Ikoma 630-0192, Japan
2Department of Systems Innovation, Graduate School of Engineering Science, Osaka University, Osaka 565-0871, Japan
3Microsoft, Redmond, WA 98052, USA
4Institute for Research Initiatives, Nara Institute of Science and Technology, Ikoma 630-0192, Japan
Corresponding author: Tatsuya Sakuma (sakuma.tatsuya.sn1@is.naist.jp) The results show that the proposed
system successfully estimates the orientation at a root mean square error of less than 4° when the position
is varied from 0 mm to 20 mm and the angle is varied from 0° to 20°. Furthermore, it can repeatedly fix the
object in any orientation from 0° to 20° at a position of less than 0.5 mm and an angle of less than 1.1°. INDEX TERMS Flexible manufacturing systems, robotic assembly, soft robotics, vision sensor FIGURE 1. Left: Soft jig with a thin and soft membrane and transparent
beads and oil. Right: Pose estimation result. his work is licensed under a Creative Commons Attribution 4.0 License. For more information, see https://creativecommons.org/licenses/by/4.0/ Received 22 May 2023, accepted 7 June 2023, date of publication 13 June 2023, date of current version 23 June 2023. Received 22 May 2023, accepted 7 June 2023, date of publication 13 June 2023, date of current version 23 June 2023. Digital Object Identifier 10.1109/ACCESS.2023.3285810 A. FLEXIBLE JIGS AND FIXTURESi Flexible jigs (fixtures) have been developed using various
strategies. However, reducing the uncertainty of the fixed
pose is an open issue in the previous strategies. Reconfigurable jigs [12], [13], [14] fix objects and
support their surfaces by rearranging or replacing multiple
components according to the shape of the object. The ability
to change a fixed configuration provides flexibility. Pin array
jigs [3], [4], [5] use a shape-memorable mechanism to fix
objects by constraining multiple points of an object with
pins. When fixing components using pins, a high degree
of freedom (DoF) facilitates flexibility. Flexible material
jigs [6], [15] use elastomeric materials to deform the surface
of the fixture depending on the object shapes to be fixed. In [6] and [15], parts fixing is accomplished by jamming
transition and vacuum mechanisms. This paper proposes a new two-in-one flexible jig (called
a soft jig) that provides flexibility to fix various objects and
the sensing ability to perceive the pose of a fixed object. The soft jig fixes an object by utilizing particle jamming,
which evacuates the inner fluid inside the jig; this technique
is famously used in a jamming gripper [9]. Furthermore, the
soft jig can estimate the pose of a fixed object by acquiring the
shape of the deformed membrane based on optical sensing. Because there is always a contact surface when fixing an
object, if the pose can be estimated using this surface, the
object-fixing and estimation can be completed using the jig
alone. Although flexible material jigs [6], [15] allow for versatile
fixing compared to other flexible fixtures, the precision of
the fixing in target poses is a disadvantage. One method for
overcoming this issue is to utilize tactile sensing to reduce the
uncertainty of a fixed pose. Based on our previous studies, we used transparent beads
and replaced the air used for the jamming grippers with
transparent oil that had a refractive index close to that of
the beads; this allows optical sensing under the jamming
mechanism [10], [11]. If the refractive indices of all materials
inside the jig are similar, the surface of the membrane can be
captured optically by the camera when light travels straight
through. Markers were painted on the inner surface of the
membrane, and two cameras were installed inside the jig to
convert the captured image into the shape of the membrane. A. FLEXIBLE JIGS AND FIXTURESi Furthermore, we used a hydraulic drive system with visual
control of the amount of transparent oil inside the jig. If the I. INTRODUCTION Comparison of contact points, flexibility, and sensing ability of
general-purpose jigs. In summary, our primary contributions are as follows:l In summary, our primary contributions are as follows:l • A design for a flexible assembly jig to achieve the object
pose estimation and parts fixing (Section III). design for a flexible assembly jig to achieve the ob because it can constrain the shape of the object. However,
if there are many contact points, the number of contact points
that must be measured to determine the constraining state of
an object increases, reducing the ease of sensing. pose estimation and parts fixing (Section III). • A hydraulic drive system with visual-based control of
the amount of transparent oil inside the jig, which
facilitates the creation of a jammed state for parts fixing
while maintaining optical transparency (Section IV-A).i To solve this trade-off problem, we investigated a method
for acquiring object orientation with a jamming jig with many
contact points. • A sensing method for estimating object poses fixed on
the jig by utilizing point-to-function ICP (Sections IV-B
and IV-C). Although it is possible to estimate the object pose fixed
on the jig using vision-based perception techniques [7], [8],
the direction other than the bottom surface that contacts the
jig should be unobstructed by sensors to allow the robot
to approach the part from various perspectives. Therefore,
incorporating an internal sensing system into the jig can
provide a more reliable solution that is accessible to the robot
and has sensing ability.l I. INTRODUCTION In robotic assembly processes, jigs play a critical role in
achieving precise assembly by fixing objects in desired
poses. Positioning the target object with a jig designed
based on the object shape facilitates precise and efficient
robotic assembly [1]. However, because the jig prepared for
a particular object is not applicable to other objects, the
jig must be redesigned each time the manufacturing object
changes, resulting in cost and time loss. This study aims to
address the issue by developing a general-purpose jig that can
achieve high-mix, low-volume production quickly and cost-
effectively. FIGURE 1. Left: Soft jig with a thin and soft membrane and transparent
beads and oil. Right: Pose estimation result. jig that fixes objects based on clamping force. General-
purpose jigs with improved object adaptability include pin-
array jigs [3], [4], [5] and jamming jigs [6]. Fig. 2 compares
the contact points, flexibility, and sensing ability of general-
purpose jigs. The general-purpose jig should be designed to position
and constrain the object even if there is a positioning error
in the robot [2]. A vise is a well-known general-purpose The jig has at least two contact points consisting of a
locator or a clamp that constrain the object. The more contact
points, the more versatile the object that can be handled The associate editor coordinating the review of this manuscript and
approving it for publication was Tao Wang
. 62187 62187 T. Sakuma et al.: Jamming Gripper-Inspired Soft Jig for Perceptive Parts Fixing FIGURE 2. Comparison of contact points, flexibility, and sensing ability of
general-purpose jigs. amount of oil discharged is not adequate when controlling
the pump, bubbles will be generated, and the membrane on
the jig will not be visible from the sensor as described in
Appendix A. This study estimates the object position and orientation
based on a proposed point-to-function iterative closest point
(ICP) that finds a transformation from the reconstructed
membrane surface to the bottom surface of the object model. More specifically, we detect the markers on the membrane,
estimate the marker positions, represent a curved surface
from the markers, and determine the object pose with the
transformation, as shown in Fig. 1. Using this approach,
we estimated the object’s orientation at a root mean square
error (RMSE) of less than 4° for positions from 0 mm to
20 mm and angles from 0° to 20°. FIGURE 2. B. TACTILE SENSING FOR SOFT ROBOTICS Because a high spatial resolution enables high object pose
estimation accuracy, we focus on camera-based methods. GelSight [16] is a well-known tactile sensor that acquires a
normal map for an object in contacting the sensor. The sensor
utilizes multiple light sources of different colors to obtain the
normal based on the photometric stereo. SoftBubble [17], [18] 62188 62188 VOLUME 11, 2023 T. Sakuma et al.: Jamming Gripper-Inspired Soft Jig for Perceptive Parts Fixing FIGURE 5. Mold (left) to fabricate the membrane (right) of the soft jig. The curved surface of the membrane can be formed by coating liquid
silicone material onto a rigid mold. FIGURE 3. Design of the proposed soft jig. Note that the oil flow piping is
simplified for clarity. FIGURE 5. Mold (left) to fabricate the membrane (right) of the soft jig. The curved surface of the membrane can be formed by coating liquid
silicone material onto a rigid mold. FIGURE 3. Design of the proposed soft jig. Note that the oil flow piping is
simplified for clarity. FIGURE 4. Appearance of the soft jig while being filled with oil, viewed
from the camera side. Although the beads are packed tightly inside the
membrane, the markers can be observed through the filling oil. • Two cameras to acquire the point cloud data around the
target object using triangulation The base plate size was 220mm × 220mm × 5mm, and the
height of the silicone membrane was approximately 30 mm
from the base plate. The pancake-like shape increases the
robot’s ability to reach the object from the top by preventing
the object from sinking deeper in a hemispherical shape. As shown in Fig. 4, the soft jig was filled with oil inside
the membrane, and the markers were observed even when the
membrane was filled with beads. FIGURE 4. Appearance of the soft jig while being filled with oil, viewed
from the camera side. Although the beads are packed tightly inside the
membrane, the markers can be observed through the filling oil. uses an air-inflated balloon to acquire the shape of the
object with a depth camera through the balloon membrane. Lin et al. [19] proposed a sensor that estimates the curvature
of a flexible material using the subtractive color-mixing
principle. A. SILICONE MEMBRANE A silicone membrane with multiple rounded markers was
formed using the mold shown in Fig. 5. The membrane
thickness was less than 1 mm (approximately 0.7 mm). To form a thin membrane that easily deformed to assume
the shape of the object, we employed a method for coating
liquid silicone rubber (Dragon Skin FX-Pro, Smooth-On) on
a mold surface instead of casting the material with several
molds. The mold had hemispherical indentations to create
markers, which in turn were filled with white silicone, which
was colored with a pigment (Silc Pig, Smooth-On) and mixed
with a thickener (THI-VEX, Smooth-On). The soft jig is designed to be larger than the sensing
grippers [10], [20], and the area of contact with the object
is increased by decreasing the size of the beads to be filled
to achieve a parts fixing performance suitable for the jig. These design changes were expected to reduce the internal
transparency and marker sensing accuracy in the previously
proposed method. Nevertheless, the use of passive stereo
with two cameras (in Section III) and the proposed image
processing system for blurred images (in Section IV-B)
helped overcome these problems to estimate the object pose. After the white silicone hardened, the black silicone
adhered to the mold surface after filling the indentations. Instead of using the original color of the silicone, it was
colored to allow easy detection of the marker during image
processing. The mold was printed using a 3D printer. We removed the layer lines on the mold surface using
polyester putty to allow easy removal of the hardened silicone
from the 3D-printed plastic mold. B. TACTILE SENSING FOR SOFT ROBOTICS In our previous work, we developed a sensing
jamming gripper that (1) grasps an object using transparent
beads and oil for jamming-based grasping and (2) acquires
the membrane shape using an embedded camera and markers
of known sizes attached to the membrane [10], [11]. TaTa [20]
is a soft gripper capable of sensing tactile information
by incorporating the techniques of GelSight [16] into our
previously developed jamming gripper [10]. A. OBJECT-FIXING PARTi To fix an object in the proposed jig, controlling the input and
output of the fluid inside the membrane is necessary. Table 1
shows the difference between pneumatic and hydraulic drives
for the jamming transition. Although RealSense D435 can be used to acquire point
cloud data based on the active stereo method, the function
cannot be used in this hardware configuration of the soft jig. This is because the infrared light projected from the camera
is reflected by the transparent base plate without reaching the
inner surface of the membrane. The advantage of the hydraulic-driven method is that it
can be combined with optical sensing. It is challenging to
perform optical sensing with the pneumatic-driven approach
because no object has the same refractive index as air. However, the hydraulic-driven method requires precise con-
trol to prevent excessive input–output that causes cavitation
and membrane breakage owing to the incompressibility
of oil. Because cavitation inhibits the passage of light,
it is prevented by precisely measuring the oil level of the
tank and controlling the amount of oil using a peristaltic
pump. To eliminate the influence of ambient light, the cameras
were surrounded with black plastic plates, and light-emitting
diode (LED) lights near the camera maintained a constant
brightness during sensing. B. GLASS BEADS The soft jig, shown in Fig. 3, comprises the following five key
components: If large beads and a thin membrane are used, the shape of the
bead will cause bumps on the surface of the membrane; as
a result, the contact between the membrane and the object
changes from surface contact to point contact, decreasing
the coefficient of friction between the object and the
membrane [21]. Therefore, beads covered with a silicone
membrane should be small in diameter to avoid deteriorating
the parts fixing performance. Furthermore, the beads must be • A silicone membrane with markers to capture the surface
deformation • Glass beads for jamming transition • A base plate cut out of transparent polyethylene
terephthalate • A pump to create a jammed state 62189 VOLUME 11, 2023 T. Sakuma et al.: Jamming Gripper-Inspired Soft Jig for Perceptive Parts Fixing TABLE 1. Fluid actuation methods for jamming transition. TABLE 1. Fluid actuation methods for jamming transition. FIGURE 6. System configuration of the soft jig. The blue and green lines
denote the oil flow tubes and electrical connections, respectively. FIGURE 6. System configuration of the soft jig. The blue and green lines
denote the oil flow tubes and electrical connections, respectively. transparent to make the inner markers visible to the cameras. Based on these requirements, we selected 1 mm glass beads
(FGB-20, Fuji Manufacturing). FIGURE 7. (a) Setup of the oil tank and the external camera. (b) Appearance of the oil level from the external camera side. Oil level per
pixel is approximately 10 µm. (c) Flowchart of the oil level measurement. D. CAMERA To acquire the deformation of the membrane, we used
two cameras embedded inside the jig to capture a pair of
images. The two cameras (RealSense D435, Intel) were
placed approximately 150 mm under the base plate with a
baseline of 30 mm. C. BASE PLATE The base plate must be transparent to make the membrane
markers visible. It cannot allow leakage of the oil and beads
inside and must allow only oil to be pumped through the
connected tube. To achieve this, the plate holds the silicone membrane and
tube connectors. Tube connectors with a filter were attached
to the underside of the base plate to pump the oil while
preventing the beads from flowing out. Because of the small
diameter of the beads, the filters must be finer, which reduces
the cross-sectional area A required to enter the tube and the
mass flow rate Q; this is indicated by the equation Q = ρAV,
where ρ is the oil density, A is the cross-sectional area, and V
is the velocity. In a production line, it is necessary to increase
Q to shorten the time. Eight tube connectors were attached to
the base plate to distribute the filter load. FIGURE 7. (a) Setup of the oil tank and the external camera. (b) Appearance of the oil level from the external camera side. Oil level per
pixel is approximately 10 µm. (c) Flowchart of the oil level measurement. IV. FIXING AND SENSING SYSTEMi To achieve a highly precise oil-level measurement of less
than 1 mm, which is unfeasible with commercially available
level sensors, such as capacitive, conductive, or float sensors,
a camera was used to capture the transparent oil tank from
the side with a high resolution. This method allows for Fig. 6 shows the system configuration of the soft jig. To achieve parts fixing and object-pose estimation, the
system comprises three parts: an object-fixing part, a mem-
brane shape-sensing part, and an object pose estimation
part. VOLUME 11, 2023 VOLUME 11, 2023 62190 T. Sakuma et al.: Jamming Gripper-Inspired Soft Jig for Perceptive Parts Fixing FIGURE 9. Flowchart of the proposed membrane shape-sensing. The
inputs were two paired images, and the output was the pose of the
object on the soft jig. FIGURE 10. Point-to-function ICP that finds a transformation from the
reconstructed membrane surface to the bottom surface of the object
model. FIGURE 9. Flowchart of the proposed membrane shape-sensing. The
inputs were two paired images, and the output was the pose of the
object on the soft jig. FIGURE 8. Appearance of markers on the inner surface of the soft jig
when viewed from the camera side. The markers were blurred, making it
challenging to determine their exact size; nevertheless, their positions
could be captured. The red circles superimposed on the image result from
marker detection by Laplacian of Gaussian (LoG). FIGURE 8. Appearance of markers on the inner surface of the soft jig
when viewed from the camera side. The markers were blurred, making it
challenging to determine their exact size; nevertheless, their positions
could be captured. The red circles superimposed on the image result from
marker detection by Laplacian of Gaussian (LoG). FIGURE 9. Flowchart of the proposed membrane shape-sensing. The
inputs were two paired images, and the output was the pose of the
object on the soft jig. FIGURE 9. Flowchart of the proposed membrane shape-sensing. The
inputs were two paired images, and the output was the pose of the
object on the soft jig. FIGURE 10. Point-to-function ICP that finds a transformation from the
reconstructed membrane surface to the bottom surface of the object
model. greater precision in oil-level control using a peristaltic pump. IV. FIXING AND SENSING SYSTEMi On the other hand, the method of sensing oil pressure is not
suitable for this application, as the pressure value may change
regardless of the oil level inside the jig, for example, when it
comes into contact with an object. Fig. 7 (a) shows the setup
of the oil tank and external camera, and Fig. 7 (b) shows the
actual captured image, where the yellow box represents the
area for creating a two-dimensional profile. Fig. 7 (c) shows
a flowchart of the oil-level measurement procedure. The oil
level per pixel in an image is approximately 10 µm. The liquid
surface is black because the oil rises on the wall surface by
capillary action and appears similar to a concave cylindrical
lens when viewed from the side. FIGURE 10. Point-to-function ICP that finds a transformation from the
reconstructed membrane surface to the bottom surface of the object
model. FIGURE 10. Point-to-function ICP that finds a transformation from the
reconstructed membrane surface to the bottom surface of the object
model. First, a two-dimensional profile was created by averaging
the pixel values in the yellow box in the horizontal
direction to analyze the intensity values of the image with
respect to the height direction of the tank. Second, the
minimum intensity candidate was detected to convert the
two-dimensional profile into an oil level measurement,
considering that the liquid surface viewed from the side
appeared black. To achieve sub-pixel accuracy in the oil level
measurement, a parabola function was fitted on measured
points around the minimum intensity candidate, and the
minimum value with the fitted function was determined. This
method enabled us to accurately measure the oil level in the
tank. FIGURE 11. Pre-processing to convert a computer-aided design (CAD)
model into points. (a) Original CAD model. The yellow surface is the
bottom when fixed. (b) Transformed CAD model in the fixed pose
coordinate system. (c) Generated points from the intersection of the z-up
vector and the bottom surface. FIGURE 11. Pre-processing to convert a computer-aided design (CAD)
model into points. (a) Original CAD model. The yellow surface is the
bottom when fixed. (b) Transformed CAD model in the fixed pose
coordinate system. (c) Generated points from the intersection of the z-up
vector and the bottom surface. Fig. 9 shows a flowchart of the proposed membrane shape-
sensing that uses acquired images. IV. FIXING AND SENSING SYSTEMi The system captures a
point cloud that represents the membrane shape from the
image and treats it as a surface to acquire the membrane shape
as a surface function. The acquired oil level was used to control the jamming
transition by driving the peristaltic pump until the target value
was reached. Since the peristaltic pump was driven by a
stepping motor, the system could control the amount of oil
in response to the pulse input in a closed loop, using the oil
level as a feedback parameter. LoG was used to accurately detect marker centers in
blurred images. Using the known variation in the marker size
in the image, σ of the Gaussian filter was determined, and the
center of the markers was calculated by extracting the local
maximum from the LoG image. R =
(
RxRy
if rotationally symmetric
RxRyRz
otherwise. Rx, Ry, Rz are the rotation matrices around the x-, y-,
and z-axes in the fixed coordinate system, respectively. R is
the combined rotation matrix, t is the translation vector, pi
(source), and qi (target) are the pairs of corresponding points,
ni is the normal vector of the target, and Np is the number
of corresponding points found in the ICP. Here, we utilize a
5-DoF metric when R is RxRy and a 6-DoF metric when R is
RxRyRz. C. OBJECT POSE ESTIMATION Fig. 10 shows the point-to-function ICP, the proposed
algorithm for estimating the orientation of a pushed object. Fig. 11 shows a point cloud of the bottom surface generated
using the CAD model as the source for the pose estimation. The coordinate system of the CAD model is adjusted to the
fixing pose, and the intersection points with the z-up vector
are calculated by shifting the xy-position of the z-up vector
at regular pitch intervals. The lowest value in the intersection
area is treated as the bottom surface point. We minimized this ε using the Levenberg–Marquard
algorithm [23] to search for pose parameters (R and t) and
treated the result as the pose of the object. p
The pose of the object is estimated by performing ICP
using a CAD point cloud and a differentiable surface function. In point-to-plane ICP, the source and target point clouds are
static, and the pair points of the source and target point
clouds must match. Conversely, point-to-function ICP and
the xy-position of all points in the source point cloud from
CAD were used as inputs for the surface function, and the
target point cloud (surface) was calculated from the function
for each iteration. Because the function is differentiable,
we calculated the target xyz points and the normals from the
derivative at each surface position. This makes it straight-
forward to calculate the point correspondence between the
source and target, even when the spatial resolutions of the
source and target point clouds differ. B. MEMBRANE SHAPE-SENSING Fig. 8 shows the markers observed in the oil. The markers
on the inner surface of the membrane were visible owing to
the refractive index-tuned oil, although light travels through
several beads to reach the membrane. Because the refractive
index was adjusted by the human eye while changing the
mixing ratio of paraffin and silicone oils, it did not perfectly
match the glass beads and appeared blurred in the camera
images. The centers of the markers detected in the left and right
images were acquired in integer pixel units. To acquire
the center with sub-pixel accuracy, a least-squares fitting is
performed to a quadratic function in the x- and y-directions,
respectively, using pixels in the neighborhood of the center. Subsequently, the marker centers were converted to the three-
dimensional point cloud based on the triangulation method. VOLUME 11, 2023 62191 T. Sakuma et al.: Jamming Gripper-Inspired Soft Jig for Perceptive Parts Fixing FIGURE 13. Three cylindrical objects with different diameters: 30 mm,
50 mm, and 80 mm. All objects have the same height of 30 mm. FIGURE 14. Illustration of the motion of the robot to press the object
against the jig. FIGURE 12. Experimental setup. (a) A manipulator is used to push the
object in a fixed pose. (b) Transformation relationships for each
coordinate. The transformation from the robot to the camera is computed
using model information and kinematics. The transformation from the
camera to the jig is computed using known camera parameters and
marker placement. (c) Four white circular markers at the ends of the
dotted arrows align the soft jig with the robot coordinates. FIGURE 13. Three cylindrical objects with different diameters: 30 mm,
50 mm, and 80 mm. All objects have the same height of 30 mm. FIGURE 13. Three cylindrical objects with different diameters: 30 mm,
50 mm, and 80 mm. All objects have the same height of 30 mm. FIGURE 13. Three cylindrical objects with different diameters: 30 mm,
50 mm, and 80 mm. All objects have the same height of 30 mm. FIGURE 13. Three cylindrical objects with different diameters: 30 mm,
50 mm, and 80 mm. All objects have the same height of 30 mm. FIGURE 12. Experimental setup. (a) A manipulator is used to push the
object in a fixed pose. (b) Transformation relationships for each
coordinate. B. MEMBRANE SHAPE-SENSING The transformation from the robot to the camera is computed
using model information and kinematics. The transformation from the
camera to the jig is computed using known camera parameters and
marker placement. (c) Four white circular markers at the ends of the
dotted arrows align the soft jig with the robot coordinates. FIGURE 12. Experimental setup. (a) A manipulator is used to push the
object in a fixed pose. (b) Transformation relationships for each
coordinate. The transformation from the robot to the camera is computed
using model information and kinematics. The transformation from the
camera to the jig is computed using known camera parameters and
marker placement. (c) Four white circular markers at the ends of the
dotted arrows align the soft jig with the robot coordinates. FIGURE 14. Illustration of the motion of the robot to press the object
against the jig. Here, the rotation around the z-axis in the objective
function ε for ICP can be fixed if the object is rotationally
symmetric (e.g., bearing). The objective function used in the
point-to-plane and point-to-function metric is expressed as The acquired point cloud is equivalent to the bottom
surface of the contacting object. Owing to the soft membrane,
the object shape appeared on the membrane as a continuous
curved surface smoothed from its original shape owing to
the uniform distribution of the fluid. When the object is
pressed against the membrane, the membrane enclosing the
fluid maintains its surface as a smooth curve. Because a point
cloud is a continuous curved surface, it can be defined as a
differentiable surface function using the radial basis function. This allows interpolation of the surface heights and normals
at any xy-position. ε =
Np
X
i=1
Rpi + t −qi
· ni
2
(1) (1) where R =
(
RxRy
if rotationally symmetric
RxRyRz
otherwise. R =
(
RxRy
if rotationally symmetric
RxRyRz
otherwise. A. JIG COORDINATE CALIBRATION The red and green lines denote the x- and y-tilts
acquired from the jig, respectively, whereas the blue and orange lines
denote the x- and y-tilt residuals from the ground truth, respectively. FIGURE 15. Tilt angles and residuals before calibration of cylinders with a
diameter of 50 mm. The red and green lines denote the x- and y-tilts
acquired from the jig, respectively, whereas the blue and orange lines
denote the x- and y-tilt residuals from the ground truth, respectively. FIGURE 17. Estimated tilt angles and its residuals after the calibration of
a cylinder with a diameter of 30 mm. FIGURE 18. Estimated tilt angles and their residuals after the calibration
of a cylinder with a diameter of 80 mm FIGURE 16. Estimated tilt angles and their residuals after calibration of a
cylinder with a diameter of 50 mm. FIGURE 18. Estimated tilt angles and their residuals after the calibration
of a cylinder with a diameter of 80 mm. FIGURE 16. Estimated tilt angles and their residuals after calibration of a
cylinder with a diameter of 50 mm. Next, we calculated three parameters (offsets Ox, Oy, and
scale S) for cylinders with diameters of 50 mm as calibration
results to obtain the correct value of the acquired tilt angle. Axj and Ayj can be regarded as pure sine waves in the yz- and
xz-planes, considering the tilt direction was changed by 360°
while maintaining the angle α constant in the radial direction
determined by the robot. Therefore, the offsets Ox, Oy, and
the scale S can be calculated using the following equations: Fig. 13 shows the cylindrical objects used for calibration. The robot grasps these objects and moves to the origin within
a horizontal orientation in the xy-plane of the jig coordinate
system, using the marker recognized by the camera. Fig. 14
shows a diagram of the motion of the robot pressing the
object. The robot presses the grasped object against the jig,
and the grasped object is tilted to specified angles α; however,
the tilt direction of the object was changed by 360° within a
period of 1° per second without using a pump to fix the object. A. JIG COORDINATE CALIBRATION We performed jig coordinate calibration to evaluate the pose
estimation method and convert the pose acquired from the jig
into a coordinate system for the robot. For calibration, we used a manipulator (LBR iiwa 14 R820,
KUKA) to precisely control the pose of the object pressed
against the jig. The robot and soft jig were arranged as
shown in Fig. 12. To transform the pose of the grasped object
from a robot coordinate system to a jig coordinate system,
a hand-eye camera and four circular markers on the jig were
used for coordinate alignment. To obtain the transformation,
we solved the perspective-n-point problem [24]. 62192 62192 VOLUME 11, 2023 VOLUME 11, 2023 T. Sakuma et al.: Jamming Gripper-Inspired Soft Jig for Perceptive Parts Fixing FIGURE 15. Tilt angles and residuals before calibration of cylinders with a
diameter of 50 mm. The red and green lines denote the x- and y-tilts
acquired from the jig, respectively, whereas the blue and orange lines
denote the x- and y-tilt residuals from the ground truth, respectively. FIGURE 16. Estimated tilt angles and their residuals after calibration of a
cylinder with a diameter of 50 mm. FIGURE 17. Estimated tilt angles and its residuals after the calibration of
a cylinder with a diameter of 30 mm. FIGURE 18. Estimated tilt angles and their residuals after the calibration
of a cylinder with a diameter of 80 mm. Next, we calculated three parameters (offsets Ox, Oy, and
scale S) for cylinders with diameters of 50 mm as calibration FIGURE 15. Tilt angles and residuals before calibration of cylinders with a
diameter of 50 mm. The red and green lines denote the x- and y-tilts
acquired from the jig, respectively, whereas the blue and orange lines
denote the x- and y-tilt residuals from the ground truth, respectively. FIGURE 16. Estimated tilt angles and their residuals after calibration of a
cylinder with a diameter of 50 mm. FIGURE 17. Estimated tilt angles and its residuals after the calibration of
a cylinder with a diameter of 30 mm. FIGURE 18. Estimated tilt angles and their residuals after the calibration
of a cylinder with a diameter of 80 mm. FIGURE 15. Tilt angles and residuals before calibration of cylinders with a
diameter of 50 mm. A. JIG COORDINATE CALIBRATION Tilt estimation results around the y-axis and the ground truth
measured by motion capture when the position is varied from 0 mm to
20 mm and the angle is varied from 0° to 20°. The data are plotted for
each pose for a period of 3 s. The red line denotes equal relation. Ns is the number of samples used for calibration. Ns is the number of samples used for calibration. The tilt angles of a cylinder with a diameter of 50 mm
acquired using plane estimation, ICP algorithms, and ground
truth used as a command value to the robot are shown in
Fig. 15. The tilt angles calibrated for cylinders with diameters
of 50 mm using these parameters are shown in Fig. 16. The
tilt angles calibrated for cylinders with diameters of 30 mm
and 80 mm, using the same parameters as those of 50 mm,
are shown in Fig. 17 and Fig. 18. The RMSE for the results
are listed on the right in Fig. 15–18. We observed that the planar estimation [25] and the
proposed method (5-DoF) have approximately the same
results and can be estimated with high accuracy at an RMSE
of less than 4° in Fig. 16–18. When the diameter of the object
was small, the number of markers that could be referenced
was small, and the pose estimation result became noisy. This
is evident from the considerable noise observed in the pose
estimation results at the diameter of 30 mm, where the point-
to-plane ICP method failed to produce accurate estimates. However, it is clear that the proposed method improves
estimation accuracy with an RMSE more than twice as large. In addition, the noise level varied depending on the estimated
degrees of freedom during the pose estimation. FIGURE 21. Tilt estimation results around the y-axis and the ground truth
measured by motion capture when the position is varied from 0 mm to
20 mm and the angle is varied from 0° to 20°. The data are plotted for
each pose for a period of 3 s. The red line denotes equal relation. FIGURE 22. Position estimation results in the x-direction and the ground
truth measured by motion capture when the position is varied from 0 mm
to 20 mm, and the angle is varied from 0° to 20°. The data are plotted for
each pose for a period of 3 s. A. JIG COORDINATE CALIBRATION Ox = 1
Ns
Ns
X
j=1
Axj,
(2)
Oy = 1
Ns
Ns
X
j=1
Ayj,
(3)
S =
2 × α
√
2 × (σx + σy)
,
(4) (2) While changing the pose of an object, we estimated the
object orientation from the jig using plane estimation [25],
a point-to-plane ICP algorithm (applied to a point cloud),
and the proposed point-to-function ICP algorithm (applied to
a function). Here, we apply ICP algorithms with the 6-DoF
and 5-DoF metrics (without rotation around the z-axis) while
considering the target objects to be cylindrical. (3) (4) where: The x-tilt Ax and y-tilt Ay were calculated using plane esti-
mation, the ICP algorithms, and ground truth by projecting
the normal vector of the object pose onto the yz- and xz-
planes and converting them into angles. σx =
v
u
u
u
t 1
Ns
Ns
X
j=1
(Axj −Ox)2,
(5) (5) VOLUME 11, 2023 T. Sakuma et al.: Jamming Gripper-Inspired Soft Jig for Perceptive Parts Fixing g. n
FIGURE 20. Setup to evaluate the estimation performance of the soft jig. A motion capture camera is mounted from above to capture a cylinder at
various orientations and positions on the jig. FIGURE 19. Variations of the positions that the robot pressed on the jig. The position is varied in 10 mm increments only in the x-axis direction in
the jig coordinate system. FIGURE 20. Setup to evaluate the estimation performance of the soft jig. FIGURE 19. Variations of the positions that the robot pressed on the jig. The position is varied in 10 mm increments only in the x-axis direction in
the jig coordinate system. FIGURE 19. Variations of the positions that the robot pressed on the jig. The position is varied in 10 mm increments only in the x-axis direction in
the jig coordinate system. FIGURE 20. Setup to evaluate the estimation performance of the soft jig. A motion capture camera is mounted from above to capture a cylinder at
various orientations and positions on the jig. σy =
v
u
u
u
t 1
Ns
Ns
X
j=1
(Ayj −Oy)2,
(6) σy =
v
u
u
u
t 1
Ns
Ns
X
j=1
(Ayj −Oy)2,
(6) (6) FIGURE 21. A. JIG COORDINATE CALIBRATION The red line denotes equal relation. At a diameter of 80 mm, the residuals were larger when the
tilt was large because when a large object is tilted and pressed
against the jig at a large angle, a part of the bottom surface of
the object lifts and loses contact with the jig, as described in
Appendix B. B. PUSHED OBJECT POSE ESTIMATION not a problem because such large positioning errors are
unlikely to occur in many cases under robot control of object
positions. The position estimation results in the x-direction are shown
in Fig. 22. In cases of planar estimation, the position cannot
be estimated; however, in the proposed method, the position
is also included in the output result. Although the deviations
from the ground truth occur at an RMSE of less than 6 mm
when the offset is between 0 mm and 10 mm, it is possible to
correct because R2 is higher than 0.92. C. PLACED OBJECT FIXING
We performed an object fixing experiment to clarify when FIGURE 23. Estimated tilt angles from pressed to fixed. FIGURE 24. Displacements from pressed to released. The minute
displacement after 0.5 s was caused by vibration when the robot control
it h d f
f
t
l t
iti
t
l FIGURE 23. Estimated tilt angles from pressed to fixed. FIGURE 24
Displacements from pressed to released The minute FIGURE 24. Displacements from pressed to released. The minute
displacement after 0.5 s was caused by vibration when the robot control
was switched from force control to position control. FIGURE 23. Estimated tilt angles from pressed to fixed. FIGURE 23. Estimated tilt angles from pressed to fixed. FIGURE 23. Estimated tilt angles from pressed to fixed. FIGURE 24. Displacements from pressed to released. The minute
displacement after 0.5 s was caused by vibration when the robot control
was switched from force control to position control. from the jig using plane estimation [25] and the proposed
point-to-function ICP algorithm (5-DoF).l Six reflective markers on the jig and four on the cylinder
were measured from the top using a motion capture system
(V120: Trio, OptiTrack), as shown in Fig. 20. The y-tilt
of the object determined by the jig and motion capture
system was calculated by projecting the normal vector of
the object pose onto the xz-planes and converting it into an
angle. not a problem because such large positioning errors are
unlikely to occur in many cases under robot control of object
positions. The position estimation results in the x-direction are shown
in Fig. 22. In cases of planar estimation, the position cannot
be estimated; however, in the proposed method, the position
is also included in the output result. B. PUSHED OBJECT POSE ESTIMATION Although the deviations
from the ground truth occur at an RMSE of less than 6 mm
when the offset is between 0 mm and 10 mm, it is possible to
correct because R2 is higher than 0.92. The results of the tilt estimation around the y-axis are
shown in Fig. 21. During planar estimation, angle errors
were included in the estimates as the object moved away
from the center. Additionally, when the angle of the object
was large, the position of the object changed when pressed
with a force of 5 N, indicating that errors occur even
when the object is pressed at the center. However, the
proposed method compensates for this by exploiting the
fact that the estimated angle values are roughly equivalent
to the ground truth at an RMSE of less than 4° and a
coefficient of determination R2 greater than 0.98, regardless
of the position when the offset is between 0 mm and
10 mm. Furthermore, we confirmed that an offset larger
than 20 mm includes estimation errors; however, this is B. PUSHED OBJECT POSE ESTIMATION We performed pose estimation at various positions and angles
to evaluate the pose estimation results when the pressed
position was not at the center. FIGURE 22. Position estimation results in the x-direction and the ground
truth measured by motion capture when the position is varied from 0 mm
to 20 mm, and the angle is varied from 0° to 20°. The data are plotted for
each pose for a period of 3 s. The red line denotes equal relation. The robot grasps a cylinder with a diameter of 50 mm and
moves the cylinder to the coordinates away from the center
in 10 mm increments, as shown in Fig. 19. This cylinder was
tilted around the y-axis by the robot to specific angles α
of 5°, 10°, 15°, and 20° from the center of the contacting
bottom. The contact with the jig was determined when the force applied to the wrist of the manipulator was changed
by 1 N or more. After contact detection, the robot pushed the
object with a force of 5 N; the jig estimated the object pose 62194 62194 VOLUME 11, 2023 T. Sakuma et al.: Jamming Gripper-Inspired Soft Jig for Perceptive Parts Fixing FIGURE 23. Estimated tilt angles from pressed to fixed. from the jig using plane estimation [25] and the proposed
point-to-function ICP algorithm (5-DoF). Six reflective markers on the jig and four on the cylinder
were measured from the top using a motion capture system
(V120: Trio, OptiTrack), as shown in Fig. 20. The y-tilt
of the object determined by the jig and motion capture
system was calculated by projecting the normal vector of
the object pose onto the xz-planes and converting it into an
angle. The results of the tilt estimation around the y-axis are
shown in Fig. 21. During planar estimation, angle errors
were included in the estimates as the object moved away
from the center. Additionally, when the angle of the object
was large, the position of the object changed when pressed
with a force of 5 N, indicating that errors occur even
when the object is pressed at the center However the
FIGURE 24. Displacements from pressed to released. The minute
displacement after 0.5 s was caused by vibration when the robot control
was switched from force control to position control. D. FIXING FORCE To evaluate the effectiveness of the proposed jig in securing
an object with a threaded hole, we measured the rotation angle
generated during screw-tightening using the jig. For the experiment, we used an electric screwdriver
(Bosch GO 2 Professional, Bosch) with various tightening
torques. The preselected torque mode (PTM) of the electric
screwdriver was adjusted using a torque-presetting ring. We measured the tightening torques for different PTMs
using a joint torque sensor, as shown in Fig. 26. We used
pre-measured PTM-1 to PTM-3 (0.25 Nm–1.03 Nm) in our
experiment. FIGURE 26. Measurement results of the torque exerted by the driver. The
joint torque of the manipulator was used for the measurement. The soft jig, screwdriver, and target object were arranged
as shown in Fig. 27 (a). We followed the screw-tightening
procedure before and after measuring the rotation angle of
the object, as depicted in Fig. 27 (b). The procedure involved
fixing the target object in place using the jig, grasping the
screwdriver, and tightening the screw. We used three types of
objects as targets: a circular, octagonal, and square cylinder
with a nut embedded in the center and markers for motion
capture. The angle of rotation of the object being measured is
shown in Fig. 27 (c) and was measured using motion capture
to calculate the rotation angle around the z-axis. sensing and registering the object CAD model to the
membrane shape.i From the results in Section V-A, we confirmed that point-
to-function ICP, which expresses the shape of the membrane
as a function and estimates the object’s pose, improves the
accuracy. This is because the correspondence of the points
can be obtained more appropriately than if the points were
used directly to estimate the object pose. We also confirmed
that changing the number of pose estimation parameters from
6-DoF to 5-DoF when using a cylindrical object improves the
accuracy, indicating the importance of limiting the DoF of the
estimation according to the object shape. The experiment results are shown in Fig. 28. The x-axis
of the graph shows the combination of object geometry and
PTM used in the experiments, and the y-axis shows the
amount of object rotation when screwing was performed. From the results in Section V-C, we confirmed that highly
accurate orientation estimation can be achieved using the
estimation result before fixation instead of that after fixation. C. PLACED OBJECT FIXINGi We performed an object-fixing experiment to clarify when
the pose should be estimated with the jig and to analyze the
fixed pose variability caused by supporting an object with a
flexible material instead of a rigid material. The setup and procedure for pushing the object were the
same as those described in Section V-B. In this experiment,
the object was pressed against the center of the jig only. After
the robot pushed the object with a force of 5 N, a peristaltic 62195 VOLUME 11, 2023 T. Sakuma et al.: Jamming Gripper-Inspired Soft Jig for Perceptive Parts Fixing FIGURE 25. Difference between pushed and released pose. The pushed
pose is the average of the 0–1 s timeframe in Fig. 24, and the released
pose is the average of the 2–3 s timeframe in Fig. 24. The difference in
position is the Euclidean distance between the two postures, and the
difference in angle is calculated from the angular difference between the
unit vectors when the unit vector is rotated by the two postures. FIGURE 26. Measurement results of the torque exerted by the driver. The
joint torque of the manipulator was used for the measurement. FIGURE 25. Difference between pushed and released pose. The pushed
pose is the average of the 0–1 s timeframe in Fig. 24, and the released
pose is the average of the 2–3 s timeframe in Fig. 24. The difference in
position is the Euclidean distance between the two postures, and the
diff
i
l
i
l
l t d f
th
l
diff
b t
th pump was used to fill the soft jig with oil; pumping stopped
when the liquid level limit acquired from the camera was
reached. Snapshots of the robot pushing an object and after parts
fixing are shown in Fig. 23. The angle was accurately
estimated when the object was pressed against it; however,
as the oil was pumped out for fixation, the accuracy of the
estimation decreased.i Snapshots showing the after parts fixing and the robot
releasing an object are displayed in Fig. 24. The results
illustrating the difference between the pushed and released
poses are shown in Fig. 25. The results show that the
pose of the object changed slightly when the gripper was
released; however, the amount was sufficiently small to
be compensated for by compliance with the remote center
compliance (RCC) devices. FIGURE 25. D. FIXING FORCE The pose estimation converges correctly because the oil
increases the shape difference between the object contact
region and the object non-contact region. We found that the circular and octagonal cylinders rotated
during screw-tightening, even with PTM-1, because their
shape provided little constraint. However, the square cylinder
was fixed and screw-tightened using PTM-1. When screw
tightening was performed using PTM-2 and PTM-3, the
object rotated significantly from PTM-3. C. PLACED OBJECT FIXINGi Difference between pushed and released pose. The pushed
pose is the average of the 0–1 s timeframe in Fig. 24, and the released
pose is the average of the 2–3 s timeframe in Fig. 24. The difference in
position is the Euclidean distance between the two postures, and the
difference in angle is calculated from the angular difference between the
unit vectors when the unit vector is rotated by the two postures. FIGURE 25. Difference between pushed and released pose. The pushed
pose is the average of the 0–1 s timeframe in Fig. 24, and the released
pose is the average of the 2–3 s timeframe in Fig. 24. The difference in
position is the Euclidean distance between the two postures, and the
difference in angle is calculated from the angular difference between the
unit vectors when the unit vector is rotated by the two postures. B. HARDWARE LIMITATIONS FOR SENSING Section V-A shows that there are limitations to the estimated
allowable tilt angle and size of the soft jig. To address these
limitations and improve performance, adjustments to the jig’s
diameter and the volume of the membrane interior may be
necessary. VII. CONCLUSION
Thi This paper proposed a novel soft jig inspired by jamming
grippers, which uses a soft membrane containing transparent
beads and oil with an adjusted refractive index to improve
the efficiency of robotic assembly processes. The soft jig
incorporates a hydraulic drive system for controlling the
amount of oil inside the membrane and fixes the object while
maintaining transparency inside the membrane. Additionally,
it is equipped with a sensing system that estimates the pose
of an object based on a point-to-function ICP. FIGURE 27. (a) Setup for the screw tightening experiment. (b) Procedure
for screw tightening using a jig. (c) Angle of the object after screw
tightening used for evaluation. The angle was calculated using motion
capture. FIGURE 27. (a) Setup for the screw tightening experiment. (b) Procedure
for screw tightening using a jig. (c) Angle of the object after screw
tightening used for evaluation. The angle was calculated using motion
capture. Our experiments validated the proposed jig’s ability
to estimate the orientation of cylindrical objects between
diameters of 30 mm and 80 mm by pressing objects at the
center of the jig, with an RMSE of less than 4°. Furthermore,
we demonstrated that the proposed method could estimate the
orientation and position even if the object is not pressed at the
center of the jig, provided that the amount of misalignment
is 10 mm or less. These results indicate that the proposed
method outperforms previously proposed sensing methods. Furthermore, in the object-fixing experiments with multiple
angles (from 0° to 20°), we confirmed that the jig maintained
the pose of the object at a position of less than 0.5 mm and an
angle of less than 1.1°. FIGURE 28. Rotation angle of the object during screw tightening. Each
measurement value was calculated as the pose after the screw tightening
procedure. FIGURE 28. Rotation angle of the object during screw tightening. Each
measurement value was calculated as the pose after the screw tightening
procedure. Combined with the object pose estimation, the jig allows
us to perform precise assembly operations by re-grasping
or changing the assembled object’s pose to the target pose. In other words, the soft jig is an active perception jig that
works with the assembly plan, unlike the conventional jigs
that only fix objects in a desired pose. APPENDIX A
DIFFICULTY OF ACTUATION In a pneumatic drive, the air is compressible and can be easily
driven by a vacuum pump or other devices. Conversely, in a
hydraulic drive, if oil is not output at the proper pressure
and/or volume, bubbles will be generated owing to cavitation
or insufficient sealing. This must be prevented because if
bubbles are generated, the markers on the membrane cannot
be recognized, as shown in Fig. 29. VII. CONCLUSION
Thi In the future, we intend
to improve the performance of the jig, integrate the proposed
jig and assembly operation planning, and work on general-
purpose assembly using robots. The current marker size of 5 mm and the distance between
the markers on the membrane of approximately 10 mm pose
accuracy challenges in object pose estimation. For instance,
a cylinder with a diameter of 30 mm has only about seven
reference markers. To enhance the accuracy, we propose
reducing the marker size and distance by half, which would
increase the number of reference points to 19. Alternatively, we could improve pose estimation accuracy
by introducing a mechanism that can change the membrane’s
color to visible or invisible, as shown in a previous study [26]. Another option would be to use multi-color illumination
techniques [16], [20] to minimize the uncertainty of the pose
estimation process. A. PRECISE ESTIMATION The soft jig performs object pose estimation for contacted
objects by acquiring the membrane shape using optical 62196 62196 VOLUME 11, 2023 T. Sakuma et al.: Jamming Gripper-Inspired Soft Jig for Perceptive Parts Fixing FIGURE 27. (a) Setup for the screw tightening experiment. (b) Procedure
for screw tightening using a jig. (c) Angle of the object after screw
tightening used for evaluation. The angle was calculated using motion
capture. FIGURE 28. Rotation angle of the object during screw tightening. Each
measurement value was calculated as the pose after the screw tightening
procedure. FIGURE 27. (a) Setup for the screw tightening experiment. (b) Procedure
for screw tightening using a jig. (c) Angle of the object after screw
tightening used for evaluation. The angle was calculated using motion
capture. the jig can be used for re-grasping to further improve
the gripper’s object grasping posture, as demonstrated in a
previous study [27]. However, the results in Section V-D reveal challenges in
constraining objects using the proposed jig. This is because
the jig relies on a form closure, which utilizes the shape of
the object to constrain it, rather than a force closure, which
applies a force to constrain the object. Nevertheless, it is
worth noting that the mechanical clutch was activated during
screw tightening, and the screw was tightened to the specified
torque. This can be attributed to the frictional effect of the
flexible membrane on the jig. C. HARDWARE LIMITATION FOR FIXING The results in Section V-C demonstrate that the proposed jig
successfully placed objects with high accuracy. Moreover, VOLUME 11, 2023 62197 T. Sakuma et al.: Jamming Gripper-Inspired Soft Jig for Perceptive Parts Fixing FIGURE 29. Marker visibility when air is introduced into the membrane. [6] T. Kiyokawa, T. Sakuma, J. Takamatsu, and T. Ogasawara, ‘‘Soft-jig-
driven assembly operations,’’ in Proc. IEEE Int. Conf. Robot. Autom. (ICRA), May 2021, pp. 3466–3472. [7] G. Du, K. Wang, S. Lian, and K. Zhao, ‘‘Vision-based robotic grasping
from object localization, object pose estimation to grasp estimation
for parallel grippers: A review,’’ Artif. Intell. Rev., vol. 54, no. 3,
pp. 1677–1734, Mar. 2021. [8] M. Q. Mohammed, K. L. Chung, and C. S. Chyi, ‘‘Review of deep
reinforcement learning-based object grasping: Techniques, open chal-
lenges, and recommendations,’’ IEEE Access, vol. 8, pp. 178450–178481,
2020. [9] E. Brown, N. Rodenberg, J. Amend, A. Mozeika, E. Steltz, M. R. Zakin,
H. Lipson, and H. M. Jaeger, ‘‘Universal robotic gripper based on the
jamming of granular material,’’ Proc. Nat. Acad. Sci. USA, vol. 107, no. 44,
pp. 18809–18814, Nov. 2010. [10] T. Sakuma, F. Von Drigalski, M. Ding, J. Takamatsu, and T. Ogasawara,
‘‘A universal gripper using optical sensing to acquire tactile information
and membrane deformation,’’ in Proc. IEEE/RSJ Int. Conf. Intell. Robots
Syst. (IROS), Oct. 2018, pp. 1–9. [11] T. Sakuma, E. Phillips, G. A. G. Ricardez, M. Ding, J. Takamatsu, and
T. Ogasawara, ‘‘A parallel gripper with a universal fingertip device using
optical sensing and jamming transition for maintaining stable grasps,’’
in Proc. IEEE/RSJ Int. Conf. Intell. Robots Syst. (IROS), Nov. 2019,
pp. 5814–5819. FIGURE 29. Marker visibility when air is introduced into the membrane. FIGURE 30. State of contact when a cylinder with a diameter of 80 mm is
tilted 20°. The membrane and cylinder are in contact in the green area but
not in the red area. [12] S. Lu, Z. Ahmad, M. Zoppi, X. Ding, D. Zlatanov, and R. Molfino, ‘‘Design
and testing of a highly reconfigurable fixture with lockable robotic arms,’’
J. Mech. Des., vol. 138, no. 8, pp. 1–8, Aug. 2016.i [13] R. Muller, M. Esser, and M. Vette, ‘‘Reconfigurable handling systems as
an enabler for large components in mass customized production,’’ J. Intell. Manuf., vol. 24, no. 5, pp. 977–990, Oct. 2013. [14] G. Levi, Y. Golan, and A. Shapiro, ‘‘SIMJig—Smart independent
minimalist jig,’’ IEEE Robot. Autom. Lett., vol. APPENDIX B
CONTACT STATE FOR A LARGE OBJECT CONTACT STATE FOR A LARGE OBJECT When a large object is tilted and pressed against the jig at
a large angle, part of the bottom surface of the object lifts
and loses contact with the jig, as shown in Fig. 30. Therefore,
the object’s pose cannot be accurately estimated from this
membrane shape. [19] X. Lin, L. Willemet, A. Bailleul, and M. Wiertlewski, ‘‘Curvature sensing
with a spherical tactile sensor using the color-interference of a marker
array,’’ in Proc. IEEE Int. Conf. Robot. Autom. (ICRA), May 2020,
pp. 603–609. [20] S. Li, X. Yin, C. Xia, L. Ye, X. Wang, and B. Liang, ‘‘TaTa: A universal
jamming gripper with high-quality tactile perception and its application
to underwater manipulation,’’ in Proc. Int. Conf. Robot. Autom. (ICRA),
May 2022, pp. 6151–6157. C. HARDWARE LIMITATION FOR FIXING 7, no. 2, pp. 3396–3403,
Apr. 2022. [15] A. Rubio-Mateos, M. Casuso, A. Rivero, E. Ukar, and A. Lamikiz,
‘‘Vibrations characterization in milling of low stiffness parts with a rubber-
based vacuum fixture,’’ Chin. J. Aeronaut., vol. 34, no. 6, pp. 54–66,
Jun. 2021. [16] W. Yuan, S. Dong, and E. Adelson, ‘‘GelSight: High-resolution robot
tactile sensors for estimating geometry and force,’’ Sensors, vol. 17, no. 12,
p. 2762, Nov. 2017. FIGURE 30. State of contact when a cylinder with a diameter of 80 mm is
tilted 20°. The membrane and cylinder are in contact in the green area but
not in the red area. [17] A. Alspach, K. Hashimoto, N. Kuppuswamy, and R. Tedrake, ‘‘Soft-
bubble: A highly compliant dense geometry tactile sensor for robot
manipulation,’’ in Proc. 2nd IEEE Int. Conf. Soft Robot. (RoboSoft),
Apr. 2019, pp. 597–604. [18] N. Kuppuswamy, A. Alspach, A. Uttamchandani, S. Creasey, T. Ikeda,
and R. Tedrake, ‘‘Soft-bubble grippers for robust and perceptive manipula-
tion,’’ in Proc. IEEE/RSJ Int. Conf. Intell. Robots Syst. (IROS), Oct. 2020,
pp. 9917–9924. REFERENCES [21] J. M. Gómez–Paccapelo,
A. A. Santarossa,
H. D. Bustos,
and
L. A. Pugnaloni, ‘‘Effect of the granular material on the maximum
holding force of a granular gripper,’’ Granular Matter, vol. 23, no. 1,
pp. 1–6, Feb. 2021. [1] A. Gameros, S. Lowth, D. Axinte, A. Nagy-Sochacki, O. Craig, and
H. R. Siller, ‘‘State-of-the-art in fixture systems for the manufacture and
assembly of rigid components: A review,’’ Int. J. Mach. Tools Manuf.,
vol. 123, pp. 1–21, Dec. 2017. [22] J. Amend, N. Cheng, S. Fakhouri, and B. Culley, ‘‘Soft robotics
commercialization: Jamming grippers from research to product,’’ Soft
Robot., vol. 3, no. 4, pp. 213–222, Dec. 2016. [2] H. Asada and A. By, ‘‘Kinematic analysis of workpart fixturing for flexible
assembly with automatically reconfigurable fixtures,’’ IEEE J. Robot. Autom., vol. JRA-1, no. 2, pp. 86–94, Jun. 1985. [23] J. J. More, ‘‘The Levenberg–Marquardt algorithm: Implementation
and theory,’’ in Numerical Analysis. Berlin, Germany: Springer, 1978,
pp. 105–116. [3] A. Mo and W. Zhang, ‘‘A novel universal gripper based on meshed pin
array,’’ Int. J. Adv. Robotic Syst., vol. 16, no. 2, pp. 1–12, 2019. [4] P. Shi, Z. Hu, K. Nagata, W. Wan, Y. Domae, and K. Harada, ‘‘Development
of a shape-memorable adaptive pin array fixture,’’ Adv. Robot., vol. 35,
no. 10, pp. 591–602, May 2021. [24] V. Lepetit, F. Moreno-Noguer, and P. Fua, ‘‘EPnP: An accurate O(n)
solution to the PnP problem,’’ Int. J. Comput. Vis., vol. 81, no. 2,
pp. 155–166, Feb. 2009. [5] J. W. Park, J. Park, H. Kim, N. Kim, and D. Y. Kim, ‘‘Assembly part
positioning on transformable pin array fixture by active pin maximization
and joining point alignment,’’ IEEE Trans. Autom. Sci. Eng., vol. 19, no. 2,
pp. 1047–1057, Apr. 2022. [25] T. Sakuma, T. Kiyokawa, J. Takamatsu, T. Wada, and T. Ogasawara,
‘‘Soft-jig: A flexible sensing jig for simultaneously fixing and estimating
orientation of assembly parts,’’ in Proc. Int. Conf. Robot. Autom. (ICRA),
May 2022, pp. 10945–10950. 62198 VOLUME 11, 2023 T. Sakuma et al.: Jamming Gripper-Inspired Soft Jig for Perceptive Parts Fixing TAKAHIRO WADA (Member, IEEE) received
the B.S. degree in mechanical engineering, the
M.S. degree in information science and systems
engineering, and the Ph.D. degree in robotics
from Ritsumeikan University, Japan, in 1994,
1996, and 1999, respectively. In 1999, he was a
Research Associate with Ritsumeikan University. REFERENCES In 2000, he joined as an Assistant Professor with
the Department of Intelligent Mechanical Systems
Engineering, Kagawa University, where he was
promoted as an Associate Professor, in 2003. In 2006 and 2007, he spent half
a year with the University of Michigan Transportation Research Institute,
Ann Arbor, as a Visiting Researcher. He joined as a Full Professor with the
College of Information Science and Engineering, Ritsumeikan University,
in 2012. Since 2021, he has been a Full Professor and the Director of
the Human Robotics Laboratory, Division of Information Science, Nara
Institute of Science and Technology (NAIST), Japan. His current research
interests include robotics, human–machine systems, and human modeling. He is a member of IEEE ITSS, IEEE SMC, IEEE RAS, the Robotics Society
of Japan, Society for Neuroscience, the Society of Automotive Engineers
of Japan, and the Society of Instrument and Control Engineers (SICE). He received the Best Paper Award from JSAE, in 2008 and 2011, and the
Outstanding Oral Presentation from the Society for Automotive Engineers
(SAE), in 2010. [26] F. R. Hogan, M. Jenkin, S. Rezaei-Shoshtari, Y. Girdhar, D. Meger,
and G. Dudek, ‘‘Seeing through your skin: Recognizing objects with a
novel visuotactile sensor,’’ in Proc. IEEE Winter Conf. Appl. Comput. Vis. (WACV), Jan. 2021, pp. 1217–1226. [27] Z. Hu, W. Wan, K. Koyama, and K. Harada, ‘‘Reducing uncertainty
using placement and regrasp planning on a triangular corner fixture,’’
IEEE Trans. Autom. Sci. Eng., early access, Jan. 10, 2023, doi:
10.1109/TASE.2023.3234047. TATSUYA SAKUMA (Member, IEEE) received
the B.E. degree from the Kobe City College of
Technology, Japan, and the M.E. degree from the
Nara Institute of Science and Technology, Japan,
in 2019. He is currently with Panasonic Hold-
ings Corporation. His research interests include
soft robotics and tactile sensors for dexterous
manipulations. TAKUYA KIYOKAWA (Member, IEEE) received
the Ph.D. degree in information science from
the Nara Institute of Science and Technology,
Japan, in 2021. Since 2021, he has been an
Assistant Professor with Osaka University, Japan,
and the Nara Institute of Science and Technology,
Japan. His current research interest includes agile
reconfigurable robot systems. TAKAMITSU MATSUBARA (Member, IEEE)
received the Ph.D. degree in information science
from the Nara Institute of Science and Technology,
Nara, Japan, in 2007. From 2005 to 2007, he was a
Research Fellow (DC1) of the Japan Society for
the Promotion of Science. REFERENCES From 2013 to 2014,
he was a Visiting Researcher with the Donders
Institute for Brain Cognition and Behavior, Rad-
boud University Nijmegen, Nijmegen, The Nether-
lands. He is currently a Professor with the Nara
Institute of Science and Technology and a Visiting Researcher with ATR,
Kyoto, Japan, and the AIST AI Center, Tokyo, Japan. His current research
interests include machine learning and control theory for robotics. TSUKASA OGASAWARA (Life Member, IEEE)
received the Ph.D. degree from The University
of Tokyo, Japan, in 1983. From 1983 to 1998,
he was with the Electrotechnical Laboratory, Min-
istry of International Trade and Industry, Japan. From 1993 to 1994, he was with the Institute
for Real-Time Computer Systems and Robotics,
University of Karlsruhe, Germany, as a Humboldt
Research Fellow. He joined the Nara Institute of
Science and Technology, Nara, Japan, in 1998,
where he has been with the Robotics Laboratory, until March 2021. He is
currently the Executive Director and the Vice President of the Nara Institute
of Science and Technology. His research interests include human–robot
interaction, dexterous manipulation, human modeling, and bio-inspired
robotics. JUN TAKAMATSU (Member, IEEE) received
the Ph.D. degree in computer science from
The
University
of
Tokyo,
Japan,
in
2004. From 2004 to 2008, he was with the Institute
of Industrial Science, The University of Tokyo. In 2007, he was a Visiting Researcher with
Microsoft Research Asia. From 2008 to 2020,
he was an Associate Professor with the Nara
Institute of Science and Technology, Japan. He was a Visitor with Carnegie Mellon University,
from 2012 to 2013. He was a Visiting Scientist with Microsoft, in 2018,
where he is currently a Senior Researcher of applied robotics. His current
research interests include robotics, including learning-from-observation, task
and motion planning, feasible motion analysis, 3D-shape modeling and
analysis, and physics-based vision. He is a member of RSJ. 62199 VOLUME 11, 2023
|
https://openalex.org/W4386923404
|
https://www.researchsquare.com/article/rs-3239830/latest.pdf
|
English
| null |
Efficacy of four indigenous Penicillium fungal isolates and Aspergillus fumigatus on Fusarium oxysporum f.sp. radicis lycopersici (FORL)
|
Research Square (Research Square)
| 2,023
|
cc-by
| 4,485
|
Efficacy of four indigenous Penicillium fungal
isolates and Aspergillus fumigatus on Fusarium
oxysporum f.sp. radicis lycopersici (FORL) Khosi Ramachela North West University
Mukondeleli Ndivhuwo Ramatsitsi
North West University
Galaletsang Petronella Segone
North West University Abstract In search for indigenous soil borne fungi with bio-control properties against plant pathogenic F. oxysporum f.sp. radicis-lycopersici (FORL) resulted in the isolation of five fungal species. These were
identified as: Penicillium commune-1 (MK660351.1), P. commune-2 (MK660335.1), Aspergillus
fumigatus (MN178806.1), P. concavorugulosum-1 (MK841454.1) and P. concavorugulosum-2
(MK841454.1). An in vitro dual culture plate experiment was carried out to assess radial growth
inhibition and growth competitiveness of potential antagonism on FORL. Analysis of the fungal bio-
agents’ interactions with FORL showed occurrence of various relationships such as antibiosis,
competition and mycoparasitism. High inhibition effect of P. concavorugulosum-2 (67.9%) and A. fumigatus (65.36%) against FORL was established. The other test isolates’ inhibition was not significant
as they ranged from 9–29%. A conclusion can be drawn that P. concavorugulosum-2 and A. fumigatus
contain potential bio-antagonistic properties that suppressed of growth of FORL. These fungal isolates
should be considered as good candidates for development as commercial bio-control agents as
individuals or in combination. However, these species’ suppressive properties should be evaluated under
natural soil conditions where various biotic and abiotic factors could influence the efficacy. Research Article License: This work is licensed under a Creative Commons Attribution 4.0 International License. d
ll Additional Declarations: No competing interests reported. Page 1/13 Page 1/13 1. Background Plant diseases continue to reduce crop yields and also affect crop quality worldwide causing global
threat to food security. Fusarium genus is one of the most important and prevalent wilt pathogens
infecting most agricultural crops [1; 2]. Fusarium oxysporum f.sp. radicis-lycopersici (FORL) is a
necrotrophic pathogen that causes diseases with global economic implications in crop production. The
disease causes significant damage to greenhouses, field crops and hydroponic cultures [3]. In an effort to manage Fusarium infection, several fungicides have previously been registered for use in
the control this problematic pathogen. These agro-chemicals have however been reported to have
negative effect on the environment and in some cases posing health risks to the consumers [4; 5]. Furthermore, emergence of resistant strains against these fungicides have been reported. For these
reasons, researchers are exploring for fungal isolates that can be used as biocontrol agents against
soilborne in vegetable crops. Previous studies have shown myco-parasitism potential of Penicillium and Aspergillus species in
inhibiting fungal pathogenic mycelial growth to up to 100 per cent Penicillium and Aspergillus species
are filamentous fungi that produce a range of secondary metabolites and mycotoxins that exerts certain
effects on plants, soil and other microbes [6–8]. Aflatoxins, gliotoxins and ochratoxins are some of
various toxins produced by Penicillium and Aspergillus species. The two genera have also been reported
to have antimicrobial effects as well as immune-suppressive properties on other micro-organisms. These species are of economic importance in agriculture because of their multifunctional properties that
may be beneficial as well as harmful to plants and other soil micro fauna [9]. For example, some
pathogenic effect on plants may be exerted by various Penicillium species through their metabolites Page 2/13 effect on plant physiological growth [10]. Fungal isolates with similar impacts on plants have been
studied and used in bio-logical control of weeds. There are other studies on Penicillium and Aspergillus
species that have exhibited their abilities to positively influence plant growth and development. These
effects included production of certain phytohormones and secondary metabolites [11]. These fungal
species inherently possess the potential to be developed as both plant growth enhancers and root
protection against soil borne pathogens. Fungal isolates can selectively adapt to various climatic and soil conditions, including varying
temperatures, pH levels, and salt levels [12]. 2.1 Pathogen isolation and inoculum preparation In order to test for the presence of fungal isolates, soil samples around plant roots with various cropping
systems were collected at North-West University, Molelwane research farm latitude of 25.810 S and
longitude of 25.630 E. Soil serial dilution plate method was used to carry out single spore isolation [15]. A soil suspension was created by vigorously mixing 1 g of rhizosphere soil sample with 9 mL of sterile
water. With the aim to plate the soil suspension on Potato Dextrose Agar (PDA) plates, the soil dilution
series was carried to obtain spore concentration of 10− 2 g/mL and 10− 3 g/mL. The plates infused with
spore suspension were incubated at 27°C and mycelia from single spore growths were pricked out and
re-cultured to obtain pure cultures of the respective isolates. 1. Background Given all of these considerations, it may be reckoned that
local fungal isolates would be well adapted to that the rhizosphere in a local environment, and
subsequently control target diseases better than the exotic/introduced isolates [13]. Efficacy of bio-
control agents entail use of a variety of microorganisms, ideally isolated from conditions similar to those
in which they must be applied [14]. The objective of the current study was to carry out in vitro analysis of the effect of the locally isolated
four Penicillium isolates, and A. fumigatus against FORL. IP =
× 100
−
X T1 (A0) −
−
X T1 (A1)
−
X T1 (A0) Where: Where: T1(A1) = Pathogen growing with respective isolate 2.3 Data analysis Analysis of variance was carried out to test the significance of the relationship/effect at 5% probability
level using GENSTAT software seventeenth edition (Copyright 2014, VSN International Ltd.). 2.2 Dual culture bio-assays An in-vitro experiment was carried out using each of the respective Penicillium species isolated from
different soil types, i.e., P. commune-1, P. commune-2, A. fumigatus, P. concavorugulosum-1 and P. concavorugulosum-2. Each of the respective fungal isolates was cultured in PDA and their mycelia cut
into 2-cm diameter discs and analyzed for their suppressive effect against FORL. This was carried out by
placing the respective test fungal isolates alongside the 2 cm diameter discs of FORL. The discs were
each placed 3 cm apart in each petri plate. Each respective plate made up an experimental unit. The
experimental treatment combinations were computed as follows: FT0, FT1, FT2, FT3, FT4 and FT5;
where FT0 = FORL, FT1 = P. commune-1; FT2 = P. commune-2; FT3 = A. fumigatus; FT4 = P. concavorugulosum-1 and FT5 = P. concavorugulosum-2 and this produced six treatment combinations. These treatment combinations were each replicated five times making a total of thirty experimental units concavorugulosum-1 and FT5 = P. concavorugulosum-2 and this produced six treatment combinations. These treatment combinations were each replicated five times making a total of thirty experimental units Page 3/13 Page 3/13 and these were laid out in a Completely Randomized Design (CRD) in an incubator set at 27 ± 2 ℃ for ten
days. and these were laid out in a Completely Randomized Design (CRD) in an incubator set at 27 ± 2 ℃ for ten
days. Mycelial growth assessment was carried out on a two-day interval for a period of ten days whereby the
diameter of FORL and respective Penicillium species and A. fumigatus were measured using a vernier
caliper. The correlation coefficient between the growth of pathogen and respective test antagonists was
determined using Pearson's correlation coefficient. Inhibition percentages (IP) were calculated using
growth assessment of i.e., pathogen growing alone (the control) and pathogen growing with isolate
respectively. The following formula was used calculate IP: 3.1.1 Mycelial growth rate assessments The mycelial growth rate of the respective isolates and FORL, and their ecological relationships
highlighted the range of suppressiveness of the respective isolates against FORL are shown in Fig. 1(a)
– 1(e). An inversely proportional relationship between P. commune-2 and FORL (r = − 0.198), and
between P. concavorugulosum-2 and FORL (r = − 0.429) were established. With a rapid growth rate of P. concavorugulosum-2, there was a corresponding decline of FORL [Fig. 1 (a)]. Aspergillus fumigatus
exhibited a continual increase on all the assessments whereas FORL followed a parabolic path with a
peak of 3,92 cm at assessment 3. A gradual decline was however established at assessment 4 and 5 for
FORL [Fig. 1 (b)]. Although A. fumigatus exhibited multiple suppressive characteristics against FORL, P. concavorugulosum-1 was not as strongly expressive. This is evidenced by the higher FORL growth rate
[Fig. 1 (c)]. Penicillium commune-1 exhibited a reduced growth rate against FORL with the result in the
rapid growth of the pathogen. Competition in favor of FORL was established between the species. This
resulted in subdued mycelial growth of P. commune-1 [Fig. 1 (d)]. A gradual decline of P. commune-2 in
mycelial growth was observed whereas a steep elevation was noted on FORL, expressing less
suppressiveness of P. commune-2 on FORL. [Fig. 1 (e)]. The mycelial growth rate of the respective isolates and FORL, and their ecological relationships
highlighted the range of suppressiveness of the respective isolates against FORL are shown in Fig. 1(a)
– 1(e). An inversely proportional relationship between P. commune-2 and FORL (r = − 0.198), and
between P. concavorugulosum-2 and FORL (r = − 0.429) were established. With a rapid growth rate of P. concavorugulosum-2, there was a corresponding decline of FORL [Fig. 1 (a)]. Aspergillus fumigatus
exhibited a continual increase on all the assessments whereas FORL followed a parabolic path with a
peak of 3,92 cm at assessment 3. A gradual decline was however established at assessment 4 and 5 for
FORL [Fig. 1 (b)]. Although A. fumigatus exhibited multiple suppressive characteristics against FORL, P. concavorugulosum-1 was not as strongly expressive. This is evidenced by the higher FORL growth rate
[Fig. 1 (c)]. Penicillium commune-1 exhibited a reduced growth rate against FORL with the result in the
rapid growth of the pathogen. Competition in favor of FORL was established between the species. This
resulted in subdued mycelial growth of P. commune-1 [Fig. 1 (d)]. fumigatus (MN178806.1), P. concavorugulosum-1 (MK841454.1) and P. concavorugulosum-2
(MK841454.1). fumigatus (MN178806.1), P. concavorugulosum-1 (MK841454.1) and P. concavorugulosum-2
(MK841454.1). 3.2 Dual antifungal relationship Mycelial discs diameter, mycelial form and various growth relationships of the respective species were
assessed at a two-day interval for a week. Rapid mycelial growth and multiple suppressive
characteristics i.e., competitiveness, antibiosis and myco-parasitism were established for P. concavorugulosum-2 against FORL. Penicillium concavorugulosum-2 rapidly grew around and over FORL
while slowly reducing its growth and size. Multiple suppressive effects were also exhibited by A. fumigatus against FORL where strong competition, antibiosis and myco-parasitic effect were observed. A. fumigatus rapidly grew around FORL subsequently covering it. Slow growth of P. concavorugulosum-1
was established as it grew away from FORL, exhibiting signs of antibiosis between fungal isolate and
FORL. Notably, antibiosis effect of FORL against P. commune-1 was evident. The decline of P. commune-
1 mycelial growth when compared to FORL was indicative of its restricted growth. 3.1.1 Mycelial growth rate assessments A gradual decline of P. commune-2 in
mycelial growth was observed whereas a steep elevation was noted on FORL, expressing less
suppressiveness of P. commune-2 on FORL. [Fig. 1 (e)]. 3.1.2 Correlation co-efficient Correlation analysis determines the statistical relationship or association and strength between two
variables. Statistical relationships of the test isolate against FORL were therefore determined and are
exhibited in Fig. 2(a) – 1(e). Although there was an evident correlation between the respective test
isolates and FORL albeit relatively weak, there was notable inhibition effects on FORL due to the
respective test fungal isolates. 3.1 Indigenous fungal isolates Macro-morphological variations according to color, shape, size and structure of each pure isolate were
characterized following standard procedure [16]. Isolates were initially grouped into their respective
genera following morphological variations. Microscopic morphological characteristics of spores and
fruiting bodies, conidiophores, hyphae and conidia were used to separate isolates according to species. On the basis of macro- and micro-morphological traits, five indigenous fungal isolates were all initially
identified as Penicillium species, but later, one isolate was confirmed as Aspergillus species by DNA
molecular technique. This highlights the similarity across genera and the value of suing DNA technique. Following molecular sequencing of the internal transcribed spacer (ITS) gene area (containing ITS1 and
ITS2 sub-regions), which was used to support morphological identification, identities of fungal isolates
were verified and confirmed as follows: P. commune-1 (MK660351.1), P. commune-2 (MK660335.1), A. Page 4/13 fumigatus (MN178806.1), P. concavorugulosum-1 (MK841454.1) and P. concavorugulosum-2
(
) fumigatus (MN178806.1), P. concavorugulosum-1 (MK841454.1) and P. concavorugulosum-2
(MK841454.1). 4. Discussion The in vitro results exhibited various relationships including competition, antibiosis and mycoparasitism
of the respective test isolates against FORL. This fungus-on-fungus interaction is most likely attributed
to the release of labile chemicals by respective fungi that affect the development and survival of the
microorganisms with which they interact [18; 19]. The inhibitory action of A. fumigatus and P. concavorugulosum-2 on FORL was established to be due to mycoparasitism, an interaction that occurs
when one fungal species parasitizes another [20]. Gliotoxin, a pleiotropic mycotoxin that is produced by
A. fumigatus has been reported to cause cytotoxicity in other fungal isolates, including A. flavus, A. oryzae and F. graminearum [21]. This could have caused the rapid degradation and reduced growth of
FORL. Myco-parasitism established on P. concavorugulosum-2 against FORL could be because FORL
released extrolites that acted as a nutritional source for P. concavorugulosum-2 [22]. Aspergillus fumigatus and P. concavorugulosum-2 demonstrated rapid growth and a mycoparasitism
mechanism against FORL as their mycelia expressed a degradation effect. This was observed with
mycelia of these respective isolates overlaying FORL mycelia. The two test isolates could probably have
catabolized metabolites produced by FORL to decrease its growth and survival rate. Chatterjee et al. [23]
reported that species-specific metabolite catabolism by the producer or an organism in the same growth
environment is possible. The good performance of P. concavorugulosum-2 against FORL is indicative of
its potential use as a biocontrol agent against the pathogen. The Penicillium species could be used as a
single agent or be used in combination with other Penicillium isolates such as P. oxalicum that also
showed a suppressive effect against Fusarium [24]. It is expected that a synergistic effect of collective
activity against FORL of the combined fungal species could be generated. Further studies should
therefore be undertaken to determine whether such synergistic effect could be achieved. The antibiosis interaction i.e., growth inhibition zone of P. commune-2 against FORL that was established
(Fig. 1 (e)) was possible because of the production of metabolites that can weaken and reduce
competitiveness and growth of other microorganisms. For example, Perincherry et al. [23] reported on
the production of toxic secondary metabolites by Fusarium species during various interactions with
plants or other microbes. These metabolites can therefore influence/weaken the growth of other
microbial species. 3.1.3 Inhibition percentage Page 5/13 Page 5/13 Percentage inhibitions of the test isolates against FORL are graphically presented in Fig. 3 where P. concavorugulosum-2 (67.9%) and A. fumigatus (65.36%) exhibited the highest inhibition percentage
against FORL. The lowest inhibition was noted with P. commune-1 (9% and 29%.) against FORL. Inhibition percentage may be explained as the measure of inhibitory influence by subtracting inhibited
growth from normal growth of the same object and dividing by the normal growth of the same object
[17]. 4. Discussion Fusarium species have been reported to produce mycotoxins such as zearalenone,
deoxynivalenol, fumonisins and trichothecenes that can invade cell membranes, alter microbial growth
patterns and increase pathogen multiplication [23]. Therefore, this might have given FORL growth and
multiplication advantage over P. commune-2. Furthermore, the growth inhibition zone observed during Page 6/13 the incubation stage may be explained by the production of inhibitory metabolites by FORL that led to the
growth-free zone, allowing both species to grow but giving FOR competitive growth [24]. There was an observed strong variation of the relationship between test isolates and FORL, ranging from
weakly negative to moderate positive associations. Correlation analysis between A. fumigatus and FORL
showed a very weak positive relationship where r = 0.062 [Fig. 2(c)] compared to moderate correlation
value between P. concavorugulosum-2 and FORL that showed a moderate negative relationship of r =
0.43 [Fig. 2 (e)]. This means P. concavorugulosum-2 expressed a range of fungal suppressive properties
i.e., competitive, antibiotic and myco-parasitic properties. A similar expression was established between
A. fumigatus and FORL albeit at a weaker level [Fig. 2(c)]. The moderate positive association between P. commune-1 and FORL showed no indication of pathogen suppressiveness [Fig. 2(a)]. There was no
evident relationship between P. commune-1 and FORL on all the five assessments. There was, however, a
relatively weak negative association of P. commune-2 and FORL indicating a decline of P. commune-2,
and an increase of FORL. − The negative correlation established between P. concavorugulosum-2 and FORL is indicative of the
inhibition ability of the test fungus against the pathogen. The inhibition characteristics of a test isolate
exerted on FORL that was calculated and expressed in percentages showed P. concavorugulosum-2 had
the highest inhibitory effect against FORL (67.9%) followed by A. fumigatus (65.36%) (Fig. 3). The other
test isolates’ inhibition was not very significant as they ranged from 9–29%. The results demonstrated
that the indigenous isolate, P. concavorugulosum-2, significantly inhibited FORL development on culture
media; demonstrating that the test isolate may have had a direct effect, such as competition and
mycoparasitism. 5. Conclusion A conclusion can be drawn that P. concavorugulosum-2 and A. fumigatus contain potential antagonistic
properties that had suppressive effect on the growth of FORL. These fungal isolates should be
considered as good candidates for development of a commercial biocontrol agent. However, these
species’ suppressive properties should be evaluated for their efficacy against the various F. oxysporum
biotypes under different soil types. In-vivo studies will determine their influence in the presence of
various other soil microorganisms. They should also be investigated for their phyto-pathological and
ecological biosafety status. A. fumigatus was however reported to produce mycotoxins that are harmful
to human health when exposure to high amounts occurs. Caution on handling A. fumigatus in future
studies is therefore strongly advisable. Acknowledgements None. Author contributions GP and K Conceptualized the study. GP set up the methodology, collected and analyzed data. GP wrote
initial draft, K and MN wrote revisions. All authors read and approved the final manuscript. Consent for publication All data generated or analyzed during this study are included in this published article. The authors declare that they have no competing interests. This work was supported by Food Security and Safety Niche Research Entity. Abbreviations FORL
F. oxysporum f.sp. radicis-lycopersici
PDA FORL
F. oxysporum f.sp. radicis-lycopersici
PDA FORL
F. oxysporum f.sp. radicis-lycopersici
PDA Page 7/13 Potato Dextrose Agar
CRD
Completely Randomized Design
ITS
Internal Transcribed Spacer
Declarations
Ethics approval and consent to participate
Not applicable
Consent for publication
Not applicable
Availability of data and materials
All data generated or analyzed during this study are included in this published article. Competing interest
The authors declare that they have no competing interests. Funding
This work was supported by Food Security and Safety Niche Research Entity. Author contributions
GP and K Conceptualized the study. GP set up the methodology, collected and analyzed
initial draft, K and MN wrote revisions. All authors read and approved the final manuscrip
Acknowledgements
None. References
1. Çolak A, Biçici M. PCR detection of Fusarium oxysporum f. sp. radicis-lycopersici an
oxysporum f. sp. lycopersici of tomato in protected tomato-growing areas of the eas
M di
i
f T
k
T
k J A
i F
2013 37 457 67 d i 10 3906/
1 Potato Dextrose Agar
CRD
Completely Randomized Design
ITS
Internal Transcribed Spacer
Declarations
Ethics approval and consent to participate
Not applicable
Consent for publication
Not applicable
Availability of data and materials
All data generated or analyzed during this study are include
Competing interest
The authors declare that they have no competing interests
Funding
This work was supported by Food Security and Safety Nich
Author contributions
GP and K Conceptualized the study. GP set up the methodo
initial draft, K and MN wrote revisions. All authors read and
Acknowledgements
None. References
1 Çolak A Biçici M PCR detection of Fusarium oxysporu Potato Dextrose Agar
CRD
Completely Randomized Design
ITS
Internal Transcribed Spacer References 1. Çolak A, Biçici M. PCR detection of Fusarium oxysporum f. sp. radicis-lycopersici and races of F. oxysporum f. sp. lycopersici of tomato in protected tomato-growing areas of the eastern
Mediterranean region of Turkey. Turk J Agric For. 2013;37:457-67. doi:10.3906/tar-1203-71. 1. Çolak A, Biçici M. PCR detection of Fusarium oxysporum f. sp. radicis-lycopersici and races of F. oxysporum f. sp. lycopersici of tomato in protected tomato-growing areas of the eastern
Mediterranean region of Turkey. Turk J Agric For. 2013;37:457-67. doi:10.3906/tar-1203-71. 2. Bodah ET. Root rot diseases in plants: A review of common causal agents and management
strategies. Agric Res Technol. 2017;5:555661. doi:10.19080/ARTOAJ.2017.05.555661. 2. Bodah ET. Root rot diseases in plants: A review of common causal agents and management
strategies. Agric Res Technol. 2017;5:555661. doi:10.19080/ARTOAJ.2017.05.555661. Page 8/13 Page 8/13 3. Mcgovern, R. Management of tomato diseases caused by Fusarium oxysporum. Crop Prot. 2015;73:78-92. doi:10.1016/j.cropro.2015.02.021. 4. Özkara A, Akyıl D, Konuk M. Pesticides, environmental pollution, and health. In: Larramendy ML,
Soloneski S, editors. Environmental health risk-hazardous factors to living species. IntechOpen. 2016. 5. Kumari D, John, S. Safety and occupational health hazards of agricultural workers handling
pesticides: A case study. Advances in health and environment safety: Select proceedings of HSFEA
2016. Springer; 2018. p 75-82. 6. Navale V, Vamkudoth KR, Ajmera S, Dhuri V. Aspergillus derived mycotoxins in food and the
environment: Prevalence, detection, and toxicity. Toxicol Rep. 2021;8:1008-30. doi:10.1016/j.toxrep.2021.04.013. 7. Dass RS, Mythili KR, Thorat P, Suresh AJ, Mahata PK. The biological implications of fungi as agents
of mycotoxigenicity and potential therapeutics in medicine. In: Singh I, Rajpal VR, Navi SS, editors. Fungal resources for sustainable economy: Current status and future perspectives. Singapore:
Springer Nature Singapore; 2023. p 433-58. 8. El Hajj Assaf C, Zetina-Serrano C, Tahtah N, Khoury AE, Atoui A, Oswald IP, Puel O, Lorber S. Regulation of secondary metabolism in the Penicillium genus. Int J Mol Sci. 2020;21. agar D, Yu J, Ehrlich KC. Toxins of filamentous fungi. Chem Immunol. 2002;8 10. Shuping D, Eloff J. The use of plants to protect plants and food against fungal pathogens: A review. Afr J Tradit Complement Altern Med. 2017;14:120-7. doi:10.21010/ajtcam.v14i4.14. 11. Kumar A, Asthana M, Gupta A, Nigam D, Mahajan S. Secondary metabolism and antimicrobial
metabolites of Penicillium. In: Gupta VK, Rodriguez-Couto S, editors. New and future developments
in microbial biotechnology and bioengineering. Amsterdam: Elsevier; 2018. p 47-68. 12. References Pang KL, Chiang MW, Guo SY, Shih CY, Dahms HU, Hwang JS, Cha HJ. Growth study under combined
effects of temperature, pH and salinity and transcriptome analysis revealed adaptations of
Aspergillus terreus NTOU4989 to the extreme conditions at Kueishan Island hydrothermal vent field,
Taiwan. PLoS One. 2020;15:e0233621. doi:10.1371/journal.pone.0233621. 13. Monfort E, Lopez-Llorca LV, Jansson HB, Salinas J. In vitro soil receptivity assays to egg-parasitic
nematophagous fungi. Mycol Progress. 2006;5:18-23. doi:10.1007/s11557-005-0002-7. 13. Monfort E, Lopez-Llorca LV, Jansson HB, Salinas J. In vitro soil receptivity assays to egg-parasitic
nematophagous fungi. Mycol Progress. 2006;5:18-23. doi:10.1007/s11557-005-0002-7. 14. Collinge DB, Jørgensen HJ, Latz MA, Manzotti A, Ntana F, Rojas EC, Jensen B. Searching for novel
fungal biological control agents for plant disease control among endophytes. In: Hodkinson TR,
Doohan FM, Saunders MJ, Murphy BR, editors. Endophytes for a growing world. United Kingdom;
Cambridge University Press; 2019. p 25. 14. Collinge DB, Jørgensen HJ, Latz MA, Manzotti A, Ntana F, Rojas EC, Jensen B. Searching for novel
fungal biological control agents for plant disease control among endophytes. In: Hodkinson TR,
Doohan FM, Saunders MJ, Murphy BR, editors. Endophytes for a growing world. United Kingdom;
Cambridge University Press; 2019. p 25. 15. Pradhan N, Sukla L. Solubilization of inorganic phosphates by fungi isolated from agriculture soil. Afr J Biotechnol. 2006;5. 15. Pradhan N, Sukla L. Solubilization of inorganic phosphates by fungi isolated from agriculture soil. Afr J Biotechnol. 2006;5. 16. Wang GL, Xiong F, Que F, Xu ZS, Wang F, Xiong AS. Morphological characteristics, anatomical
structure, and gene expression: Novel insights into gibberellin biosynthesis and perception during
carrot growth and development. Hortic Res. 2015;2:15028. doi:10.1038/hortres.2015.28. 16. Wang GL, Xiong F, Que F, Xu ZS, Wang F, Xiong AS. Morphological characteristics, anatomical
structure, and gene expression: Novel insights into gibberellin biosynthesis and perception during
carrot growth and development. Hortic Res. 2015;2:15028. doi:10.1038/hortres.2015.28. Page 9/13 17. Wan Harun WHA, Nordin A, Abdul Razak F. Determination of the percentage inhibition of diameter
growth (PIDG) of Piper betle crude aqueous extract against oral Candida species. J Med Plant Res. 2011;5:878-84. 18. Zeilinger-Migsich S, Mukherjee PK. Fungus-fungus interactions. Open Mycol.J. 2014;8:27. doi:10.2174/1874437001408010027. 19. Schmidt R, Cordovez V, De Boer W, Raaijmakers J, Garbeva P. Volatile affairs in microbial
interactions. ISME J. 2015;9:2329-35. doi:10.1038/ismej.2015.42. 20. Qualhato TF, Lopes FAC, Steindorff AS, Brandao RS, Jesuino RSA, Ulhoa CJ. Mycoparasitism studies
of Trichoderma species against three phytopathogenic fungi: evaluation of antagonism and
hydrolytic enzyme production. References Biotechnol Lett. 2013;35:1461-8. doi:10.1007/s10529-013-1225-3. 21. Carberry S, Molloy E, Hammel S, Grainne O, Jones GW, Kavanagh K, Doyle S. Gliotoxin effects on
fungal growth: Mechanisms and exploitation. Fungal Genet Biol. 2012;49:302-12. doi:10.1016/j.fgb.2012.02.003. 22. Lebeau J, Petit1 T, Dufossé L, Caro S. Putative metabolic pathway for the bioproduction of bikaverin
and intermediates thereof in the wild Fusarium oxysporum LCP531 strain. AMB Expr. 2019;9:186. doi:10.1186/s13568-019-0912-4. 22. Lebeau J, Petit1 T, Dufossé L, Caro S. Putative metabolic pathway for the bioproduction of bikaverin
and intermediates thereof in the wild Fusarium oxysporum LCP531 strain. AMB Expr. 2019;9:186. doi:10.1186/s13568-019-0912-4. 23. Chatterjee S, Kuang Y, Splivallo R, Chatterjee P, Karlovsky P. Interactions among filamentous fungi
Aspergillus niger, Fusarium verticillioides and Clonostachys rosea: Fungal biomass, diversity of
secreted metabolites and fumonisin production. BMC Microbiol. 2016;16:83. doi:10.1186/s12866-
016-0698-3. 23. Chatterjee S, Kuang Y, Splivallo R, Chatterjee P, Karlovsky P. Interactions among filamentous fungi
Aspergillus niger, Fusarium verticillioides and Clonostachys rosea: Fungal biomass, diversity of
secreted metabolites and fumonisin production. BMC Microbiol. 2016;16:83. doi:10.1186/s12866-
016-0698-3. 24. Perincherry L, Lalak-Kańczugowska J, Stępień Ł. Fusarium-produced mycotoxins in plant-pathogen
interactions. Toxins. 2019;11:664. doi:10.3390/toxins11110664. 24. Perincherry L, Lalak-Kańczugowska J, Stępień Ł. Fusarium-produced mycotoxins in plant-pathogen
interactions. Toxins. 2019;11:664. doi:10.3390/toxins11110664. 25. De Cal AGLR, Garcia-Lepe R, Melgarejo P. Induced resistance by Penicillium oxalicum against
Fusarium oxysporum f. sp. lycopersici: Histological studies of infected and induced tomato stems. Phytopathol. 2000; 90:260-268. doi:10.1094/PHYTO.2000.90.3.260. 25. De Cal AGLR, Garcia-Lepe R, Melgarejo P. Induced resistance by Penicillium oxalicum against
Fusarium oxysporum f. sp. lycopersici: Histological studies of infected and induced tomato stems. Phytopathol. 2000; 90:260-268. doi:10.1094/PHYTO.2000.90.3.260. 26. Perrone G, Susca A. Penicillium species and their associated mycotoxins. In: Moretti A, Susca A,
editors. Mycotoxigenic fungi. Methods in molecular biology. New York: Humana Press; 2017. p 107-
19. 26. Perrone G, Susca A. Penicillium species and their associated mycotoxins. In: Moretti A, Susca A,
editors. Mycotoxigenic fungi. Methods in molecular biology. New York: Humana Press; 2017. p 107-
19. Figures Page 10/13 Figure 1 See image above for figure legend Page 11/13 Figure 2
See image above for figure legend Figure 2 Figure 2 See image above for figure legend Page 12/13 Figure 3 Inhibition percentages of four Penicilliumisolates and Aspergillus fumigatus against FORL. Page 13/13
|
https://openalex.org/W3006961534
|
https://www.researchsquare.com/article/rs-5486/v1.pdf
|
English
| null |
Using the anterior capsule of the hip joint to protect the tensor fascia lata muscle during direct anterior total hip arthroplasty: a randomized prospective trial
|
BMC musculoskeletal disorders
| 2,020
|
cc-by
| 6,299
|
Gongyin Zhao Gongyin Zhao
The affiliated changzhou no.2 people's hospital of nanjing medical university
Ruixia Zhu
The affiliated changzhou no.2 people's hospital of nanjing medical university
Shijie Jiang
The affiliated changzhou no.2 people's hospital of nanjing medical university
Nanwei Xu
The affiliated changzhou no.2 people's hospital of nanjing medical university
Hongwei Bao
jingjiang people's hospital
Yuji Wang
(
yujiwang@sohu.com
)
Changzhou No.2 People's Hospital The affiliated changzhou no.2 people's hospital of nanjing medical university
Ruixia Zhu The affiliated changzhou no.2 people's hospital of nanjing medical university
Shijie Jiang Yuji Wang
(
yujiwang@sohu.com
)
Changzhou No.2 People's Hospital Using the anterior capsule of the hip joint to protect
the tensor fascia lata muscle during direct anterior
total hip arthroplasty: a randomized prospective trial Using the anterior capsule of the hip joint to protect
the tensor fascia lata muscle during direct anterior
total hip arthroplasty: a randomized prospective trial Research article Keywords: direct anterior approach, total hip arthroplasty, tensor fasciae lata muscle, capsular, protect,
trauma, muscle cross-sectional area, fatty atrophy
Posted Date: September 25th, 2019
DOI: https://doi.org/10.21203/rs.2.14815/v1
License:
This work is licensed under a Creative Commons Attribution 4.0 International
License. Read Full License Keywords: direct anterior approach, total hip arthroplasty, tensor fasciae lata muscle, capsular, protect,
trauma, muscle cross-sectional area, fatty atrophy Posted Date: September 25th, 2019 License:
This work is licensed under a Creative Commons Attribution 4.0 International
License. Read Full License Version of Record: A version of this preprint was published on January 11th, 2020. See the published
version at https://doi.org/10.1186/s12891-019-3035-9. Page 1/17 Abstract Background The direct anterior approach for total hip arthroplasty (THA) has specific advantages, but
injury to the tensor fasciae lata muscle (TFLM) remains a concern. This injury in part negates some of
the advantages of the intermuscular approach, because injury of the muscle fibers of the TFLM can lead
to less satisfactory clinical results. Thus, in this study, we propose an intraoperative method to protect the
TFLM and demonstrate its feasibility. Methods Fifty-six patients undergoing THA by the direct anterior approach were divided randomly into
two groups. In group A, the TFLM was protected by an autogenous tissue “pad” created from the anterior
capsule of the joint that was rotated to protect the TFLM from direct contact with the retractors needed to
expose the joint space. In group B, the operation was carried out in the typical fashion with no other
protection of the TFLM except the attempt by the surgeons to consciously avoid compressive injury of
the TFLM. We evaluated the magnitude of changes in the muscle cross-sectional area (MSCA) and fatty
atrophy (FA) early and postoperatively between groups by magnetic resonance imaging. The differences
in blood hemoglobin and serum levels of myoglobin, lactate dehydrogenase (LDH), and creatine
phosphokinase (CPK) were compared at 8, 24, and 48 h, postoperatively. Postoperative drainage volume
and pain evaluated by visual analogue scores (VAS) were compared between the two groups. Results LDH, CPK and myoglobin in group B were significantly higher than those in group A at 8, 24, and
48 hours after the surgery. (p<0.05) Compared to the group A, the decrease of hemoglobin in group B
displayed significantly at 24 and 48 hours after surgery. (P<0.05) The significantly increased MSCA and
FA of TFLM were demonstrated in group B. The PDV and VAS in group B were significantly higher
compared with those in group A. (P<0.05) Conclusions Using the anterior capsule of the hip joint as an autogenous, protective capsular tissue pad
to limit the trauma to the TFLM during a direct anterior approach to THA is an effective method to protect
the TFLM and improve the clinical effect. Introduction By forming the lateral wall of the surgical wound, the muscle fibers of the tensor fascia lata muscle
(TFLM) are often traumatized due to mechanical retraction by surgical instruments. While preparing the
acetabulum for the prosthesis, the TFLM is often damaged by the handle of the acetabular reamer. In
addition, muscle fibers of the TFLM can be injured by the edge of the retractor and blades of the saw
during the process of elevating the proximal femur and implanting the femoral prosthesis. Therefore,
injury of the TFLM is one of the most common complications of a DAA for THA. Inadvertent injury to the
muscle can increase bleeding and postoperative pain, lead to abnormalities in gait, and overall poorer
recovery for patients. Meanwhile, excessive stimulation of muscles causes irreversible injuries, such as
compensatory thickening, hypertrophy, and lipogenesis of muscle fibers. [13, 14, 16, 17] Therefore,
avoiding secondary injury to the periarticular muscles caused during THA operations via a DAA is an
increasing concern for orthopedic surgeons. Some surgeons have used protective agents, such as gauze
or plastic sheets, to prevent injury to the TFLM, some have used curved abdominal retractors, and others
have used nothing but conscious intraoperative awareness of the potential for injury to these muscles. These methods have not proven satisfactory, however, because of their complexity and lack of effective
protection of the muscle. After the iliofemoral ligament and the anterior articular capsule be revealed in a DAA, many surgeons
open the anterior capsule in a "T" or transverse fashion to expose the hip joint and later close the capsule
by suture or not close the capsule after the arthroplasty; both closing or not closing the capsule are widely
accepted. [5, 18-22] Although some surgeons believe that the anterior capsule is not important, we have
found a new application for this structure. We hypothesized that we could protect the TFLM by using the
anterior articular capsule as a barrier between the TFLM and the retractors needed for exposure of the
joint for the arthroplasty. This randomized, controlled prospective study was designed to determine if this maneuver would protect
the TFLM during a DAA approach to hip arthroplasty. Introduction The direct anterior approach (DAA) is considered a more minimally invasive operative approach for total
hip arthroplasty (THA). This approach is favored by surgeons, because it utilizes the intermuscular plane,
thereby avoiding the need to transect any of the periarticular muscles. Moreover, this approach is favored
because it provides adequate exposure of the acetabular side. [1-5] Several studies have reported that
DAA decreases the duration of hospital stay and increases the percentage of patients discharged home
due to the lesser operative trauma which permits a more rapid recovery. [5-9] Whether DAA is actually
superior to other surgical approaches, however, remains controversial. Due to some concerns of less
adequate exposure caused by difficulty elevating of proximal femur, some complications can occur
during the operation, such as contusion of periarticular muscles, injury to the anterolateral femoral
cutaneous nerve, and fractures around the prosthesis. [10-15] Page 2/17 Page 2/17 Page 2/17 By forming the lateral wall of the surgical wound, the muscle fibers of the tensor fascia lata muscle
(TFLM) are often traumatized due to mechanical retraction by surgical instruments. While preparing the
acetabulum for the prosthesis, the TFLM is often damaged by the handle of the acetabular reamer. In
addition, muscle fibers of the TFLM can be injured by the edge of the retractor and blades of the saw
during the process of elevating the proximal femur and implanting the femoral prosthesis. Therefore,
injury of the TFLM is one of the most common complications of a DAA for THA. Inadvertent injury to the
muscle can increase bleeding and postoperative pain, lead to abnormalities in gait, and overall poorer
recovery for patients. Meanwhile, excessive stimulation of muscles causes irreversible injuries, such as
compensatory thickening, hypertrophy, and lipogenesis of muscle fibers. [13, 14, 16, 17] Therefore,
avoiding secondary injury to the periarticular muscles caused during THA operations via a DAA is an
increasing concern for orthopedic surgeons. Some surgeons have used protective agents, such as gauze
or plastic sheets, to prevent injury to the TFLM, some have used curved abdominal retractors, and others
have used nothing but conscious intraoperative awareness of the potential for injury to these muscles. These methods have not proven satisfactory, however, because of their complexity and lack of effective
protection of the muscle. Introduction By examining the release of muscle-specific
enzymes as a marker of the degree of muscle damage and using high-quality images obtained by
magnetic resonance imaging (MRI) of the periarticular muscles to determine the muscle cross-sectional
area (MCSA) and fatty atrophy (FA), we show that this technique helps to minimize trauma, thereby
protecting the integrity of the TFLM during the DAA to THA. Subjects From March 2018 to March 2019, 76 primary THAs were performed using the DAA by a single surgeon. The 76 patients, who ranged in age from 56 to 103 years old, suffered from femoral neck fractures and
underwent an uncemented THA. Of these patients, those patients older than 75 years (n=8) were
excluded. Patients with hepatic or renal failure, coagulation abnormalities, thrombosis, myocardial
infarction, cerebrovascular accidents within the previous 6 months, or long-term use of anticoagulants
were excluded (n=6). Patients who had surgery lasting more than 90 min or had an unexpected condition, Page 3/17 Page 3/17 such as an intraoperative fracture, were also excluded (n=4). In addition, patients who were unable to
undergo an MRI because of claustrophobia and who could not be evaluated effectively by MRI were
excluded (n=2). In the end, 56 patients were enrolled in the study, ranging in age from 56 to 75 years old. These patients
were randomly divided into 2 groups according to the time of admission. Patients with singular
admission were involved in group A, and patients with even days were involved in group B. The patients
were assigned to groups by a supervisor nurse who didn’t know the surgery situation. In Group A, the
TFLM was covered by the anterior capsular pad which was rotated to cover the surface of the TFLM,
while in group B, the surgeons were reminded to try their best to remember to minimize tension on,
traction, or undue retraction of the TFLM during the operation. Baseline demographic information of the
patients is shown in Table 1, which confirms the comparability of the patients. The approval for this
study was obtained from the Institutional Review Boards of the first authors’ affiliated institutions. Surgical technique The TFLM was protected only by the surgeon’s conscious effort to avoid undue trauma to the
muscle during the exposure of the joint to allow adequate visualization for the THA, without the use of
special instruments or devices. In both groups, acetabular and femoral prostheses were implanted in an identical way. Each patient
received a non-cemented femoral stem, a press-fit acetabular cup, and a 32-mm ceramic head and
ceramic lining. The prosthesis came from a single manufacturer (Waldemar Link Gmbh & Co, Hamburg,
Germany). The capsule was then re-sutured closed after the arthroplasty was completed. No other
structures were injured during the operation, and all 56 patients had their THA completed within 90 min. Drainage tubes were routinely placed after the operation to allow wound drainage for 48 h. Antibiotics
were administered perioperatively for 48 h, and nadroparin calcium as prophylaxis against deep vein
thrombosis was administered within 6 h postoperatively. After the patients were fully wake, they were
encouraged to walk immediately with the aid of a walker. At 8, 24, and 48 h after surgery, blood hemoglobin (Hb), serum levels of creatine kinase (CK), lactate
dehydrogenase (LDH), and myoglobin (MYO), as well as postoperative drainage from the wound drains
[13], were recorded and later compared between the two groups. In addition, we used a VAS scale to score
pain at 8, 24, and 48 h after surgery. No patient developed infection, facture, or other complications, and none of the patients in ether group
required any allogeneic blood transfusions during the perioperative period. All patients were discharged
from the hospital within one week, with the shortest discharge 3 days after surgery and the longest 7
days after the THA. The relevant clinical data were collected by a single observer blinded to the patient
group. An MRI was obtained 4 weeks after surgery and compared with the preoperative MRI. Imaging was
performed on a 3.0-T MR tomograph (Achieva 3.0 T TX, Philips, Netherlands) according to a standard
protocol. The T2-weighted, cross-sectional images were obtained at the level of the lesser trochanter. The
location of the MRI cross-section is shown in Fig. 2. By fully identifying the surrounding muscle tissue,
the TFLM was located and its cross-sectional area and degree of fatty atrophy were measured. The TFLM
was identified in the images by its elliptical shape and its position anterolateral to the femur and lateral to
the rectus femoris muscle. Surgical technique All patients underwent operation under general anesthesia within 5 days of femoral neck fracture. During
anesthesia, the same proportion of muscle relaxants were given according to body weight. Intravenous
tranexamic acid (1 g) was administered to patients in each group. All patients also received autologous
blood transfusion whenever possible during the process. All operations were performed by the same
group of surgeons. The patients were placed in a supine position using a regular table. A typical incision was made over the
medial margin of the TFLM. After the location of the TFLM was confirmed, the overlying fascia was
incised. After identifying the space between the TFLM and rectus femoris muscle, the TFLM was
retracted laterally. After transecting the ascending branch of the lateral circumflex femoral artery, the
anterior capsule was exposed satisfactorily. On the medial side of the capsule, the iliocapsularis muscle lies posterior to the rectus femoris muscle
and is attached to the anterior capsule. The iliocapsularis is an elongated muscle that is attached to the
full length of the anterior capsule. Though rarely reported, some surgeons have suggested that the
iliocapsularis muscle stabilizes the femoral head by contracting within the dysplastic acetabulum. [19,
27, 28] The iliocapsularis muscle serves as an anatomic marker separating the joint capsule of the hip
from the hazardous area posteriorly, which contains the arteries and nerves that can be injured by
operating on the superomedial aspect of the iliocapsularis muscle. Keeping the iliocapsularis muscle as the medial boundary of the joint capsule, we dissected along the
lateral side of the iliocapsularis, being careful to detach the muscle from the joint capsule to expose the
maximum possible area of the anterior articular capsule. After incising the articular capsule, the now
separated anterior capsule was folded outward as a layer of soft tissue acting as a “pad” covering the
outer musculature and sutured to the skin at the edge of the incision with silk suture material. With this
maneuver, the retractor and the handle of the reamer were separated by the “capsular pad” from direct Page 4/17 contact with the TFLM, while still allowing the exposure of the hip joint for the THA. The specific steps are
described in Fig. 1. In group B, the anterior capsule was incised in the middle and retracted to the sides to expose the hip
joint. Results From March 2018 to March 2019, 76 patients scheduled for a primary THA with a single surgeon were
screened and assessed for eligibility. Twenty of the patients were excluded for failing to meet the
selection criteria. After randomization, 56 patients (28 patients in group A; 28 patients in group B) that
were included in the statistical analysis of our trial. (Fig.4.) There were no statistically significant differences in the demographics or the results of preoperative
laboratory examinations between the two groups. In contrast, at 8, 24, and 48 hours postoperatively,
serum levels of MYO, CPK, and LDH were significantly greater in group B than in group A (Fig 5a, 5b, 5c;
p<0.05 each). When evaluating the decrease in blood Hb at 8 hours postoperatively, there was a trend in
group B having a greater decrease in Hb content (p=0.0637); however, the decrease in blood Hb became
greater in group B at 24 and 48 hours postoperatively (Fig 5d, p=0.0039, p<0.0001). When evaluating postoperative pain, the VAS scores were higher in group B than in group A at 8, 24 and
48 hours postoperatively (Fig 5e; p=0.0016, p<0.0001, p=0.0475). The amount of wound drainage in
group A was also less than in group B (Fig 5f; p<0.001). These results are summarized in Fig. 5. According to the comparative analysis of the MRI imaging data from the two groups, the degree of fatty
atrophy of the TFLM as a measure of muscle injury was significantly less in group A than that of group B
(Fig 6a; p=0.0134), as was the volume of the TFLM cross sectional area (Fig 6b; p<0.001). Surgical technique As a measure of pathologic change, fatty atrophy was first identified and
classified by Daniel in 1994. The classification system quantitated fatty changes in the rotator cuff into 5
grades: Grade 0, no fat; Grade 1, scant amounts of fat; Grade 2, fat visible but less than muscle; Grade 3,
muscle and fat in equal amounts; and Grade 4, fatty streaks present more than muscle. Florian Engelken
modified the classification to give the percentage of fatty tissue increments of 25% in 2014. [29, 30] We
adopted the classification method of Daniel et al. and divided the fatty atrophy into 5 grades. The ellipse Page 5/17 Page 5/17 of the TFLM was measured along its long axis and short axis to obtain the muscle cross-sectional area
from the formula. These observations and measurements of the MRI imaging are shown in Fig. 3. Statistical analysis We used a Student’s paired t test to compare the differences in the clinical and imaging data with or
without use of the capsular pad and a Kruskal Wallace test for wound drainage. P values <0.05 were
considered statistically significant. The data were analyzed by GraphPad Prism 5 software. Analysis of data In order to determine the effects of the two types of operative attempts to protect the TFLM, we measured
the mean level of blood Hb and serum muscle-specific enzymes, as well as the mean changes in fatty
atrophy and cross-sectional areas, within groups before and after the THA. For wound drainage, because
the data were not normally distributed, we calculated the median drainage (with interquartile ranges as
well as actual ranges). All images were analyzed by an observer who was blinded to both groups. Discussion The intermuscular plane is utilized in the DAA for THA; however, due to the more limited exposure of the
hip joint, in many cases the necessary retraction of the TFLM to expose the joint adequately to allow a Page 6/17 safe and reliable THA can lead to potentially serious tissue injury of the periarticular muscles, including
the TFLM, gluteus medius, and rectus femoris muscles. This type of muscle injury can lead to more
postoperative pain and delayed physical rehabilitation. [2, 3, 31] We used the anterior capsule as a
protective autogenous “capsular tissue pad” to try to minimize collateral damage, especially to the TFLM
during the operation. While this maneuver cannot protect all of the periarticular tissues that might be
traumatized during a THA, the use of this autogenous capsular tissue pad appears to protect the TFLM
substantially. Using nearby tissues to protect important structures is a common practice in surgery. This capsular tissue
pad does not require extra preparation and has enough bulk and strength to protect the muscle from the
retractor and instruments used for the THA. The synovial layer of the capsule is smooth and moist, and
the fibrous layer is dense and firm. As the liner of the joint, the capsule pad is both convenient and strong. After the detached articular capsule was reversed and stitched to the lateral skin, the hip joint was
exposed more sufficiently. The key point in using the capsular tissue pad is to identify the iliocapsularis
muscle and transect the anteromedial capsule along the length of the iliocapsularis muscle, thereby
exposing the largest area of articular tissue pad. The iliocapsularis, which is attached to the anterolateral
aspect of the capsule, is rarely mentioned in the anatomy of the hip. In this study, we show what we
believe to be an important role of the iliocapsularis muscle. [19, 27, 28, 32-34] We used objective measures of hematologic parameters of blood loss, serum levels of circulating,
muscle-specific enzymes, and quantitative MRI imaging to determine the extent of muscle damage. As
indicators of muscle damage, MYO, CPK, and LDH were used to determine the extent of cell muscle injury. [14, 17, 23, 25] Among these data, CPK showed the greatest relative differences in the three postoperative
time periods, while the differences in LDH and MYO were less impressive though still statistically
significant. Conclusions In conclusion, the use of a “protective “ articular capsulular tissue pad effectively decreased operative
injury to the TFLM, decreased bleeding and postoperative pain. This method has its limitaitons. Rather
than simply cutting open or excising the capsule, the need to carefully strip the capsule from its
attachment to the iliocapsularis muscle requires more manipulation and takes more time. Because of the
small number of patients in this study and the short follow-up time, the long-term effects of such
methods need to be further observed. Another limitation of this method is that because of the size and
position of the released anterior capsule, the proximal wound cannot be covered by the capsular tisue
pad, and therefore the gluteus minimus and gluteus medius muscles cannot be protected. Fianlly,
whether too much dissetion of the iliocapuslaris could lead to severe consequences still needs to be
determined. List Of Abbreviations THA total hip arthroplasty; TFLM tensor fasciae lata muscle; DAA direct anterior approach; MRI magnetic
resonance imaging; MSCA muscle cross-sectional area; FA fatty atrophy; Hb hemoglobin; MYO myoglobin; LDH lactate dehydrogenase; CPK creatine
phosphokinase; VAS visual analogue scores Discussion We postulate that these somewhat delayed changes in group B are related to the time
sequence of anabolism and catabolism of some proteins. Alterations in Hb indicate the amount of blood
lost, which can indicate soft tissue damage indirectly. The decrease in hemoglobin became more
pronounced and the difference between the two groups more obvious over time, which may be related to
“invisible blood loss” into the periarticular tissues after the THA. Consistent with this delay was the lack
of a statistically significant difference in the VAS pain scores at 8 hours postoperatively; however, the
difference became statistically significant at 24 and 48 hours postoperatively. These findings suggest
that using the capsular tissue pad can improve patient experience. The absence of effective protection of
the TFLM results in damage to the muscle, which can cause more postoperative bleeding, increase
postoperative drainage, and may also cause a delayed recovery during rehabilitation. MRI was used to visualize muscle injury. Comparing the two groups of patients, we found that the fatty
atrophy was more severe in the group not protected with the capsular tissue pad. In addition, the CSA of
TFLM became larger, indicative of muscular edema and possibly inflammation. Some studies have
mentioned that there is increased fatty atrophy with greater injury, associated with a decrease in muscle
CSA. [14, 26, 29] We believe this inconsistency is caused by the different timing of post-operative MRI
examinations. We performed our MRIs four weeks after THA, but according the relevant literature, the Page 7/17 Page 7/17 earliest reports of evaluation of hip CSA were 3 months after surgery; this is probably after the edema we
noted had resolved, followed by evidence of muscle atrophy. earliest reports of evaluation of hip CSA were 3 months after surgery; this is probably after the edema we
noted had resolved, followed by evidence of muscle atrophy. Ethics approval and consent to participate: The approval for this study was obtained from the Institutional Review Board committees of the affiliated
Changzhou No.2 people’s hospital of Nanjing medical university. All patients were enrolled voluntarily and
signed informed consent forms. Availability of data and material: Not applicable Consent for publication: Not applicable Acknowledgements: Not applicable Authors' contributions: YJW designed the study, GYZ, HWB, RXZ and SJJ performed the research, RXZ and GYZ analyzed data,
GYZ wrote the paper. The remaining authors contributed to refining the ideas, carrying out additional
analyses and finalizing this paper. All authors read and approved the final manuscript. CONSORT statement This study follows the CONSORT principle. The check list is attached in another file named “CONSORT
CHECKLIST”. Funding: Social development fund of Jiangsu province, BE2015632 to Y.W. Social development fund of Jiangsu province, BE2015632 to Y.W. The fund provided financial support for patient follow-up and post-operative MR review, as well as
assistance in data collection, statistics and language polishing. Competing interests: The authors declare that they have no competing interests Page 8/17 References 10: p. 12. 10. Petis, S., et al., Surgical approach in primary total hip arthroplasty: anatomy, technique and clinical
outcomes. Canadian Journal of Surgery, 2015. 58(2): p. 128-139. 10. Petis, S., et al., Surgical approach in primary total hip arthroplasty: anatomy, technique and clinical
outcomes. Canadian Journal of Surgery, 2015. 58(2): p. 128-139. 11. de Steiger, R.N., M. Lorimer, and M. Solomon, What is the learning curve for the anterior approach for
total hip arthroplasty? Clin Orthop Relat Res, 2015. 473(12): p. 3860-6. 11. de Steiger, R.N., M. Lorimer, and M. Solomon, What is the learning curve for the anterior approach for
total hip arthroplasty? Clin Orthop Relat Res, 2015. 473(12): p. 3860-6. 12. Rivera, F., et al., Risk of stem undersizing with direct anterior approach for total hip arthroplasty. Hip
Int, 2016. 26(3): p. 249-53. 13. Nistor, D.V., et al., Transitioning to the direct anterior approach in total hip arthroplasty. Is it a true
muscle sparing approach when performed by a low volume hip replacement surgeon? Int Orthop,
2017. 41(11): p. 2245-2252. 14. Kawasaki, M., et al., Muscle Damage After Total Hip Arthroplasty Through the Direct Anterior
Approach for Developmental Dysplasia of the Hip. J Arthroplasty, 2017. 32(8): p. 2466-2473. 15. Barton, C. and P.R. Kim, Complications of the direct anterior approach for total hip arthroplasty. Orthop Clin North Am, 2009. 40(3): p. 371-5. 16. De Anta-Díaz, B., et al., No differences between direct anterior and lateral approach for primary total
hip arthroplasty related to muscle damage or functional outcome. International Orthopaedics, 2016. 40(10): p. 2025-2030. 17. Unis, D.B., et al., Postoperative Changes In The Tensor Fascia Lata Muscle After Using The Modified
Anterolateral Approach For Total Hip Arthroplasty. The Journal of Arthroplasty, 2013. 28(4): p. 663-
665. 18. Dewar DC., et al., The relative contribution of the medial and lateral femoral circumflex arteries to the
vascularity of the head and neck of the femur: a quantitative MRI-based assessment. Bone joint
Journal, 2016. 98-B(12): p. 1582-1588. 19. Cooper, H.J., B.L. Walters, and J.A. Rodriguez, Anatomy of the hip capsule and pericapsular
structures: A cadaveric study. Clin Anat, 2015. 28(5): p. 665-71. 20. Sasaki, M., et al., Relationship between the hip joint capsule and piriformis tendon in a simulation of
the modified Watson-Jones anterolateral approach in THA cadaver study. Clin Anat, 2013. 26(5): p. 610-3. 21. References 1. Post ZD., et al., Direct anterior approach for total hip arthroplasty: indications, technique, and results
J Am Acad Orthop Surg. 2014.22(9): p. 595-603 1. Post ZD., et al., Direct anterior approach for total hip arthroplasty: indications, technique, and results. J Am Acad Orthop Surg. 2014.22(9): p. 595-603 2. Manrique, J., et al., Direct anterior approach for revision total hip arthroplasty. Ann Transl Med, 2014. 2(10): p. 100. 2. Manrique, J., et al., Direct anterior approach for revision total hip arthroplasty. Ann Transl Med, 2014. 2(10): p. 100. 3. Horne, P.H. and S.A. Olson, Direct anterior approach for total hip arthroplasty using the fracture table. Curr Rev Musculoskelet Med, 2011. 4(3): p. 139-45. 3. Horne, P.H. and S.A. Olson, Direct anterior approach for total hip arthroplasty using the fracture table. Curr Rev Musculoskelet Med, 2011. 4(3): p. 139-45. 4. Chen, A.F., et al., Higher Acetabular Anteversion in Direct Anterior Total Hip Arthroplasty: A
Retrospective Case-Control Study. HSS J, 2016. 12(3): p. 240-244. 4. Chen, A.F., et al., Higher Acetabular Anteversion in Direct Anterior Total Hip Arthroplasty: A
Retrospective Case-Control Study. HSS J, 2016. 12(3): p. 240-244. 5. Connolly, K.P., Direct anterior total hip arthroplasty: Literature review of variations in surgical
technique. World Journal of Orthopedics, 2016. 7(1): p. 38. 6. Van Den Eeden, Y.N.T., B.J.G. De Turck, and F.M.C. Van Den Eeden, 24 hours stay after hip
replacement. Acta Orthopaedica, 2016. 88(1): p. 24-28. 6. Van Den Eeden, Y.N.T., B.J.G. De Turck, and F.M.C. Van Den Eeden, 24 hours stay after hip
replacement. Acta Orthopaedica, 2016. 88(1): p. 24-28. 7. Siguier, T., M. Siguier, and B. Brumpt, Mini-incision anterior approach does not increase dislocation
rate: a study of 1037 total hip replacements. Clin Orthop Relat Res, 2004(426): p. 164-73. 7. Siguier, T., M. Siguier, and B. Brumpt, Mini-incision anterior approach does not increase dislocation
rate: a study of 1037 total hip replacements. Clin Orthop Relat Res, 2004(426): p. 164-73. 8. Soderquist, M.C., R. Scully, and A.S. Unger, Acetabular Placement Accuracy With the Direct Anterior
Approach Freehand Technique. J Arthroplasty, 2017. 32(9): p. 2748-2754. 8. Soderquist, M.C., R. Scully, and A.S. Unger, Acetabular Placement Accuracy With the Direct Anterior
Approach Freehand Technique. J Arthroplasty, 2017. 32(9): p. 2748-2754. Page 9/17 9. Thurig, G., et al., Safety of total hip arthroplasty for femoral neck fractures using the direct anterior
approach: a retrospective observational study in 86 elderly patients. Patient Saf Surg, 2016. 9. Thurig, G., et al., Safety of total hip arthroplasty for femoral neck fractures using the direct anterior
approach: a retrospective observational study in 86 elderly patients. Patient Saf Surg, 2016. 10: p.
12. References 304(6): P.78-83. 31. Lanting, B.A., et al., Bikini versus traditional incision direct anterior approach: is there any difference
in soft tissue damage? Hip Int, 2017. 27(4): p. 397-400. 31. Lanting, B.A., et al., Bikini versus traditional incision direct anterior approach: is there any difference
in soft tissue damage? Hip Int, 2017. 27(4): p. 397-400. 32. Retchford TH., et al., Can local muscles augment stability in the hip? A narrative literature review. J
Musculoskelet Neuronal Interact, 2013. 13(1): p. 1-12. 32. Retchford TH., et al., Can local muscles augment stability in the hip? A narrative literature review. J
Musculoskelet Neuronal Interact, 2013. 13(1): p. 1-12. 33. Haefeli, P.C., et al., An increased iliocapsularis-to-rectus-femoris ratio is suggestive for instability in
borderline hips. Clin Orthop Relat Res, 2015. 473(12): p. 3725-34. 33. Haefeli, P.C., et al., An increased iliocapsularis-to-rectus-femoris ratio is suggestive for instability in
borderline hips. Clin Orthop Relat Res, 2015. 473(12): p. 3725-34. 34. Ricci, V. and L. Özçakar, Ultrasound Imaging for Anterior Hip Pain: Hypertrophic Bursitis between the
Direct Tendon of the Rectus Femoris and the Iliocapsularis Muscle. Pm&R, 2019. 34. Ricci, V. and L. Özçakar, Ultrasound Imaging for Anterior Hip Pain: Hypertrophic Bursitis between the
Direct Tendon of the Rectus Femoris and the Iliocapsularis Muscle. Pm&R, 2019. References Kanda, A., et al., Preservation of the articular capsule and short lateral rotator in direct anterior
approach to total hip arthroplasty. Eur J Orthop Surg Traumatol, 2018. 28(6): p. 1111-1116. 22. Benad, K., et al., Technique to treat iliopsoas irritation after total hip replacement: Thickening of
articular hip capsule through an abridged direct anterior approach. Orthop Traumatol Surg Res, 2015. 101(8): p. 973-6. 23. Cui, T. and M.S. Jiang, Myoglobin A79G polymorphism association with exercise-induced skeletal
muscle damage. Genet Mol Res, 2016. 15(2). 23. Cui, T. and M.S. Jiang, Myoglobin A79G polymorphism association with exercise-induced skeletal
muscle damage. Genet Mol Res, 2016. 15(2). Page 10/17 24. Rebalka IA., Hawke TJ., Potential biomarkers of skeleta muscle damage. Biomark Med. 2014. 8(3): p. 375-8. 25. Sorichter S., et al., Skeletal muscle injury induced by eccentric muscle action: muscle proteins as
markers of muscle fiber injury. Exerc Immunol Rev. 1999 5. p:5-21. 25. Sorichter S., et al., Skeletal muscle injury induced by eccentric muscle action: muscle proteins as
markers of muscle fiber injury. Exerc Immunol Rev. 1999 5. p:5-21. 26. Muller, M., et al., Evidence of reduced muscle trauma through a minimally invasive anterolateral
approach by means of MRI. Clin Orthop Relat Res, 2010. 468(12): p. 3192-200. 26. Muller, M., et al., Evidence of reduced muscle trauma through a minimally invasive anterolateral
approach by means of MRI. Clin Orthop Relat Res, 2010. 468(12): p. 3192-200. 27. Walters, B.L., J.H. Cooper, and J.A. Rodriguez, New findings in hip capsular anatomy: dimensions of
capsular thickness and pericapsular contributions. Arthroscopy, 2014. 30(10): p. 1235-45. 28. Babst, D., et al., The Iliocapsularis Muscle: An Important Stabilizer in the Dysplastic Hip. Clinical
Orthopaedics and Related Research®, 2010. 469(6): p. 1728-1734. 29. Engelken, F., et al., Assessment of fatty degeneration of the gluteal muscles in patients with THA
using MRI: reliability and accuracy of the Goutallier and quartile classification systems. J
Arthroplasty, 2014. 29(1): p. 149-53. 29. Engelken, F., et al., Assessment of fatty degeneration of the gluteal muscles in patients with THA
using MRI: reliability and accuracy of the Goutallier and quartile classification systems. J
Arthroplasty, 2014. 29(1): p. 149-53. 30. Goutallier D., et al., Fattu muscle degeneration in cuff ruptures. Pre- and postoperative evaluation by
CT scan. Clin Orthop Relat Res, 1994. 304(6): P.78-83. 30. Goutallier D., et al., Fattu muscle degeneration in cuff ruptures. Pre- and postoperative evaluation by
CT scan. Clin Orthop Relat Res, 1994. Table Group A
Group B
t-test
P-value
Age (y)
67(56)
70(5)
1.993
0.056
Male (%)
10 (36%)
8 (29%)
-
-
Female (%)
18 (64%)
20 (71%)
-
-
BMI (kg/m2)
23.6 (2.4)
23.2 (2.3)
0.759
0.454
Table I. Demographics of patients in group A (TFLA Coverage) and group B (No
Coverage)
Values are presented as means (±
SD)
Figures Figure 1
The stars represent the anterior capsule in A and B, and in C and D the stars represent the capsule pad
that is prepared from the anterior capsule. The white arrows indicate the iliocapsularis, which can be
separated from the capsule along the dotted line in B. The detached capsule is flipped over the TFLM in
C. The capsule pad is spaced between the retractor and the TFLM in D. Figure 1 The stars represent the anterior capsule in A and B, and in C and D the stars represent the capsule pad
that is prepared from the anterior capsule. The white arrows indicate the iliocapsularis, which can be
separated from the capsule along the dotted line in B. The detached capsule is flipped over the TFLM in
C. The capsule pad is spaced between the retractor and the TFLM in D. Page 12/17 Page 12/17 Page 12/17 Figure 2
The measured plane we selected is shown with the dotted line on the coronal position in A B sh Figure 2 The measured plane we selected is shown with the dotted line on the coronal position in A. B shows
anatomic features before measurement. TFL indicates the TFLM, Rf represents the rectus femoris, and
Sa represents the rectus femoris sartorius. The measured plane we selected is shown with the dotted line on the coronal position in A. B shows
anatomic features before measurement. TFL indicates the TFLM, Rf represents the rectus femoris, and
Sa represents the rectus femoris sartorius. Page 13/17 Page 13/17 Figure 3
A and B show typical measurements of CSA and fatty atrophy of the TFLM before and after surgery in
group A. Measurements for group B are shown in C and D. Photographs of the TFLM after arthroplasty
corresponding to the left MRI image are shown in E and F. The TFLM was protected with the capsule pad
as shown in C. In group B, evidence of substantial damage to the TFLM is shown. Figure 3 Figure 3 A and B show typical measurements of CSA and fatty atrophy of the TFLM before and after surgery in
group A. Measurements for group B are shown in C and D. Photographs of the TFLM after arthroplasty
corresponding to the left MRI image are shown in E and F. The TFLM was protected with the capsule pad
as shown in C. In group B, evidence of substantial damage to the TFLM is shown. A and B show typical measurements of CSA and fatty atrophy of the TFLM before and after surgery in
group A. Measurements for group B are shown in C and D. Photographs of the TFLM after arthroplasty
corresponding to the left MRI image are shown in E and F. The TFLM was protected with the capsule pad
as shown in C. Figure 1 In group B, evidence of substantial damage to the TFLM is shown. Page 14/17 Page 14/17 Figure 4
Flow diagram of patient enrollment and randomization. Figure 6 Comparison of differences in CSA and fatty atrophy of the TFLM before and after surgery between the
two groups. ‘Fatty Atrophy’ represents the difference in grades of fatty atrophy before and after operation. ‘CSA’ is the difference between preoperative and postoperative CSA of TFLM. *, * p<0.05, ** p<0.01,
***p<0.001. CONSORTChecklist.pdf Figure 4 Flow diagram of patient enrollment and randomization. Flow diagram of patient enrollment and randomization. Page 15/17 Page 15/17 Figure 5
Comparison of laboratory examination results, postoperative pain scores, and postoperative drainage
volumes between the two groups of patients at pre-operation and three time periods post-operation. *
p<0.05, ** p<0.01, ***p<0.001. Figure 5 Figure 5 Figure 5 Comparison of laboratory examination results, postoperative pain scores, and postoperative drainage
volumes between the two groups of patients at pre-operation and three time periods post-operation. *
p<0.05, ** p<0.01, ***p<0.001. Page 16/17 Figure 6
Comparison of differences in CSA and fatty atrophy of the TFLM before and after surgery between the
two groups. ‘Fatty Atrophy’ represents the difference in grades of fatty atrophy before and after operation. ‘CSA’ is the difference between preoperative and postoperative CSA of TFLM. *, * p<0.05, ** p<0.01,
***p<0.001. Figure 6 Figure 6 Supplementary Files This is a list of supplementary files associated with this preprint. Click to download. CONSORTChecklist.pdf Page 17/17 Page 17/17 Page 17/17
|
https://openalex.org/W4385548330
|
https://journal.formosapublisher.org/index.php/ijbae/article/download/5261/5159
|
English
| null |
The Influence of Compensation, Work Environment, and Workload on Turnover Intention Phenomena at Royal Orchid Garden Hotel
|
International Journal of Business and Applied Economics
| 2,023
|
cc-by
| 5,369
|
International Journal of Business and Applied Economics (IJBAE)
Vol. 2, No. 4, 2023: 2963-6124 The Influence of Compensation, Work Environment, and
Workload on Turnover Intention Phenomena at Royal Orchid
Garden Hotel Sahala Manalu1, Felik Sad Windu Wisnu Broto2, Ameilia Magdalena
Kristiyani3*
Universitas Ma Chung Malang
Corresponding Author: Ameilia Magdalena Kristiyani lchris0107@gmail.com Sahala Manalu1, Felik Sad Windu Wisnu Broto2, Ameilia Magdalena
Kristiyani3*
Universitas Ma Chung Malang
Corresponding Author: Ameilia Magdalena Kristiyani lchris0107@gmail.com Universitas Ma Chung Malang
Corresponding Author: Ameilia Magdalena Kristiyani lchris0107@gmail.com A B S T R A C T
Turnover intention refers to the employee's
voluntary or involuntary intention to leave the
job. There are several variables that influence
the level of intention to leave: compensation,
work environment, and workload. The purpose
of this study is to determine how compensation,
work environment, and workload influence
retirement intentions at the Royal Orchid
Garden Hotel. The population and sample for
this study are employees of the Royal Orchid
Garden Hotel, and the sampling method used is
nonprobability sampling. The number of study
samples is 38, and data are collected by
distributing questionnaires. A data analysis
technique using multiple regression As a result,
it
was
found
that
compensation,
work
environment,
and
workload
all
affected
turnover intention. Work environment has a
positive and significant impact on turnover
intentions, whereas compensation and workload
have a negative impact on turnover intentions. A R T I C L E I N F O
Keywords: Compensation,
Work Environment,
Workload, Turnover Intention
Received : 30, May
Revised : 28, June
Accepted: 27, July A R T I C L E I N F O
Keywords: Compensation,
Work Environment,
Workload, Turnover Intention INTRODUCTION A company is an institution that aims for the purpose of earning a profit. Service-based companies are those that make money by providing non-physical
things in the form of services. Human resource is one of the most significant
components in the service industry. Human resource is company assets that can
provide value or contribute to the company's growth. Human resource is an
aspect of a business that must be managed correctly and effectively to be able to
minimize expenses. Employees or workers are one of the functions of human
resource in a company. According to Khomariah et al., (2020), one of the problems related to
employees is the high turnover rate. Turnover or the phenomenon of employees
leaving a company is a concrete manifestation of turnover intention. This can be
a serious problem for the company, especially if the person leaving has special
talents, skills, or holds an important position within the company. High
turnover intention has various negative consequences for a firm, including
more expense on recruiting and training new employees, a decline in the
quantity of skilled workers, and diminished stability and consistency of services
provided by the company itself (Pranata and Utama, 2019). p
y
p
y
(
)
Employees quit companies for a variety of reasons, one of which being the
work environment. The work environment plays an essential role in developing
a positive atmosphere at work. Aside from the work environment, another
factor that affects turnover intention is workload (Mobley, 2011). Workload is a
factor that affects employee stress levels in a company. Workload is frequently
underestimated by companies, despite the fact that workload can have an
impact on turnover intention. Workloads that are excessively heavy and not
matched by balanced compensation will lead employees to seek out more
profitable alternative jobs. Providing compensation is one way to improve employee welfare. A
company must take consideration for compensation in accordance with each
worker's duties. Compensation based on each employee's task is also effective
in improving their work performance. There are several forms of compensation
provided by the company, including salaries, wages, allowances, incentives,
promotions, and many others. p
y
Royal Orchid Garden Hotel is a hotel and resort service provider in Batu
City, East Java Province, Indonesia. The hotel has been in operation for around
31 years and has been granted a four-star rating. (
DOI: https://doi.org/10.55927/ijbae.v2i4.5261
ISSN-E: 2963-6124
https://journal.formosapublisher.org/index.php/ijbae (
DOI: https://doi.org/10.55927/ijbae.v2i4.5261
ISSN-E: 2963-6124
https://journal.formosapublisher.org/index.php/ijbae 569 Manalu, Broto, Kristiyani INTRODUCTION Royal Orchid Garden Hotel
employs roughly 87 people, including 48 permanent staff and 39 casual
employees split among seven departments (Human Resource Management,
Sales and Marketing, Front Office, Engineering, Food and Beverage,
Accounting, and Housekeeping). Human resources are the most important
aspect that must be preserved and improved in a company that provides
services. However, according to data provided by Royal Orchid Garden Hotel,
the company noticed a rather high turnover rate of 11.29% within one year. The
following data relates to the turnover intention rate of Royal Orchid Garden
Hotel: 570 Table 1. Turnover Intention Rate
Month
Year
Number of
Employees
In
Out
Turnover Rate
(%)
April
2022
35
0
0
0%
May
2022
34
0
1
2,94%
June
2022
36
2
0
0%
July
2022
36
0
0
0%
Agugust 2022
35
0
1
2,86%
September 2022
36
1
0
0%
October
2022
36
0
0
0%
November 2022
35
0
1
2,86%
Desember 2022
38
3
0
0%
January
2023
39
1
0
0%
February 2023
38
0
1
2,63%
March
2023
38
0
0
0%
Total
Avg: 36
7
4
11,29%
Source: Company Data (2023) According to Pranata and Utama (2019), turnover intention is said to be
normal if it has a percentage between 5%-10% in one year. If the percentage
amount is over 10%, the turnover intention is said to be high. According to the
calculations in the table above, the turnover intention rate of Royal Orchid
Garden Hotel employees is high at 11.29%. The researcher chose the Royal
Orchid Garden Hotel as the target of study since it is one of the ancient hotels
that still exists and competes with new modern hotels. Furthermore, the
researcher observed the high turnover intention of employees in this company
for around two months during the internship. As a result, the researcher choose
the research title “The Influence of Compensation, Work Environment, and
Workload on Turnover Intention Phenomena at Royal Orchid Garden Hotel.”
The purpose of this study is to analyze the effects of compensation, work
environment, and workload on turnover intentions at the Royal Orchid Garden
Hotel. This study measured three variables at the same time, which are
compensation, work environment, and workload. The aspect of uniqueness in
this research is statistical tests leading to results that contradict previous
research. Both previous studies show that workload has positive and significant
impact on turnover intention. INTRODUCTION On the other hand, in this study, workload has a
negative and significant impact on turnover intention. Therefore, this study is
expected to contribute further to research on turnover intention. Work Environment The working environment includes all factors that directly or indirectly
affect a company and can affect employee performance and satisfaction. Sihaloho and Siregar (2019) classify work environments into two types:
The first is physical work environment. The physical work environment
includes all the physical conditions of an company that can affect employee
performance, such as: buildings, work equipment, transportation equipment,
and others. The physical work environment has several indicators: lighting,
temperature and airflow, quietness, cleanliness, facilities, infrastructure,
workplace conditions, etc. The second is non-physical work environment. Non-
physical work environment includes all invisible work environment within an
company that can affect employee performancesuch as: workplace atmosphere
and employee relationships. There are several indicators of the non-physical
work environment, including: relationships between coworkers, relationships
with superiors, the responsibilities each employee has, workflow, and so on. Syauqi et al. (2020) conducted a previous study to assess the impact of
work environment on turnover intentions at Putra Mustika Prima, Bandung
City. Their study found that work environment had positive and significant
impact on turnover intention. Khomariah et al. (2020) also conducted a study
assess the impact of work environment on turnover intention for Efrata
Retailindo Company. Their study found that work environment had positive
and significant impact on turnover intention. g
p
H2: Work environment has a positive impact on turnover intention. Compensation According to Waskito and Putri (2022), compensation is a reward given to
employees for their contributions, whether in the form of energy, thoughts, or
other performance. There are two broad categories of compensation indicators:
financial compensation and non-financial compensation. Wages, salaries,
allowances, and incentives are all examples of financial compensation supplied 571 Manalu, Broto, Kristiyani by companies to their employees (Andreani et al., 2020). Meanwhile, non-
financial compensation refers to non-monetary compensation received by
employees, such as job training, promotion, and so on. The amount of compensation can be affected by a variety of things. First,
consider the level of education. The greater one's education, the greater one's
level of professionalism. Second, work experience. The more work experience
someone has, the more proficient she or he will become in the field she or he
works in. Third, company's position. The higher person's position, the more
income they receive. Nugroho and Darmawati (2018) conducted a previous study at Circle K
Indonesia Utama Company, Yogyakarta Branch to discover the impact of
compensation on turnover intention. They found that compensation had
negative and significant impact on turnover intention. Another study
conducted by Rukhviyanti and Susanti (2020), aimed to determine the effect of
compensation on turnover intention at Kwanduk Worldwide Company located
in Cikalong Wetan. Their research found that compensation had negative and
significant impact on turnover intention. Based on previous research results, we
can form a hypothesis in the following sentence: H1: Compensation has a negative impact on turnover intention. Workload Workload is a work capacity that individuals must perform within a
specific time frame and utilizing their talents and potential. Workload is 572 International Journal of Business and Applied Economics (IJBAE)
Vol. 2, No. 4, 2023: 2963-6124 divided into two categories by Juhnisa and Fitria (2020), physical workload and
mental workload. Physical workload is defined as the amount of work that can
be assessed by a person's physical strength when they are doing their job. Mental workload is a workload that cannot be measured by physical strength,
because it involves cognitive (thoughts, actions, and feelings) and emotional. According to Wahyuningsih et al. (2021) found factors that affect workload
include: internal factors (age, gender, weight, health condition) and external
factors (work environment, work conditions, difficulty of work, working hours,
break times, and responsibilities). There are several indicators used to measure workload consisting of
physical demand (a measure of the capacity of physical activity required by
employees in doing their jobs), mental demand (the ability of individuals to
realize their surroundings through the five senses), temporal demand (a
measure of the quantity of pressure related to time and felt by employees while
doing their jobs), effort (the effort made by employees, both physically and
mentally with the aim of achieving employee performance levels), frustration
level (a measure used to measure the level of security, satisfaction, and comfort
felt by employees), and performance (a measure of employee success in doing
their jobs well and satisfactorily). Previous research by Ernawati et al. (2018) intend to find out the impact of
workload on turnover intention at Mirota Batik Company, Yogyakarta. They
discovered that workload had a positive and significant effect on turnover
intention in their study. Another study being undertaken by Fitriantini et al. (2020) intend to find out the impact of workload on turnover intention at
Mataram Regional Hospital. They discovered that workload has a positive and
significant impact on turnover intention in their study. H3: Workload has a positive impact on turnover intention. Research Model Misno et al. (2021) define a research model as an overarching image or
representation of the variables raised and hypothesis investigated in a study. The research model seeks to simplify the study into a more understandable
format. Generally, research models take the shape of diagrams or charts that
describe the relationship between variables. A study model linked to the topic
at hand is as follows: 573 Manalu, Broto, Kristiyani
Figure 1. Research Model Manalu, Broto, Kristiyani Figure 1. Research Model The chart above shows the relationship between three independent
factors, compensation (X1), work environment (X2), and workload (X3) with the
dependent variable, turnover intentions (Y). Compensation (X1), work
environment (X2), and workload (X3) are variables that can influence or change
turnover intentions (Y). Validity Test According to (Setyawan, 2022), a validity test is a measurement that is
regarded to be valid if it can measure what is to be tested. According to
Soegiyono (2019), data is declared valid if the significance is < 0.05 or has R
value > R table. R table in this research is 0.312. Here are the results of the
validity tests performed: p
Table 2. Validity Test
Variable
Item R Value R Table Description
Compensation
(X1)
X1.1
0.637
0.312
Valid
X1.2
0.611
0.312
Valid
X1.3
0.793
0.312
Valid
X1.4
0.644
0.312
Valid
X1.5
0.753
0.312
Valid
Work Environment (X2)
X2.1
0.829
0.312
Valid
X2.2
0.818
0.312
Valid
X2.3
0.661
0.312
Valid
X2.4
0.726
0.312
Valid
Workload
(X3)
X3.1
0.501
0.312
Valid
X3.2
0.664
0.312
Valid
X3.3
0.681
0.312
Valid
X3.4
0.792
0.312
Valid
X3.5
0.633
0.312
Valid
Turnover Intention (Y)
Y.1
0.819
0.312
Valid
Y.2
0.761
0.312
Valid
Y.3
0.882
0.312
Valid
Y.4
0.867
0.312
Valid
Y.5
0.801
0.312
Valid
Source: SPSS Data Processing Results (2023) Table 2 shows that the compensation, work environment, and workload
variables have computed R value > R table. This indicates the research
instrument's statement items met the data validity requirements. METHODOLOGY This study uses descriptive quantitative research with a causal approach. According to Mulyadi (2019), quantitative descriptive research is a study that
can provide answers to research issues in the form of numerical data through
statistical processing and is used to analyze and clarify an event or phenomena
by describing the variables related to the topic under study. The causality
approach is a research method that seeks to discover the relationship between
one variable and another (Sekaran and Bougie, 2018). The data sources used in this research are primary data. Rahman (2021)
defines primary data as information collected directly from its source without
any involvement of intermediaries. Primary data can be collected through a
variety of methods, including questionnaires, interviews, polls, focus group
discussions, and observations. A questionnaire will be used for data collection
in this study. The survey was distributed using Google Forms. In this study, the
Likert scale is used as a measurement tool and to assess an individual's or a
group's opinions, perceptions, and attitudes toward a phenomenon (Suwandi et
al., 2018). Soegiyono (2019) defined a population as a generalization consisting of
research subjects or subjects with specific characteristics and traits determined
by researchers for the purpose of conducting research and drawing conclusions. The study population consisted of 87 employees of the Royal Orchid Garden
Hotel. A sample is an element of the population under study (Arikunto, 2017). Sampling technique is required to calculate the number of samples. The
sampling method used in this study is non-probability with sample size of 38
employees of Royal Orchid Garden Hotel. 574 International Journal of Business and Applied Economics (IJBAE)
Vol. 2, No. 4, 2023: 2963-6124 International Journal of Business and Applied Economics (IJBAE)
Vol. 2, No. 4, 2023: 2963-6124 Reliability Test Reliability test determines whether the measurement equipment used is
reliable and remains consistent after repeated measurements. Variables are
considered reliable if Cronbach's alpha > 0.6. Below is the result of reliability
test for the four variables in this study: 575 Manalu, Broto, Kristiyani Table 3. Reliability Test
Variable
Cronbach's Alpha
Description
Compensation
0.709
Reliable
Work Environment
0.753
Reliable
Workload
0.648
Reliable
Turnover Intention
0.869
Reliable
Source: SPSS Data Processing Results (2023) The reliability test result in Table 3 shows that Cronbach’s Alpha for the
variables compensation, work environment, and workload are greater than
0.60, it can be concluded that all variables are reliable. Normality Test y
Normality test determines whether residual variables in a regression
model have normal or near-normal distribution values (Suganda and Cahyadi,
2020). By looking at Asimp. Sig. (two-tailed), the normality test can be
performed using Kolmogorov-Smirnov test. Data is distributed periodically
during Asymp. Sign (two-sided) > 0.05. The result of the normality test is
shown in the table below: Table 4. Normality Test Table 4. Normality Test Table 4. Normality Test
One-Sample Kolmogorov-Smirnov Test
Compensation
Work
Environtment Workload
Turnover
Intention
N
38
38
38
38
Normal Parametersa
Mean
14.87
13.66
16.63
13.34
Std. Deviation
2.782
2.989
2.149
3.505
Most Extreme Differences Absolute
.118
.200
.133
.153
Positive
.118
.100
.086
.094
Negative
-.105
-.200
-.133
-.153
Kolmogorov-Smirnov Z
.724
1.230
.818
.946
Asymp. Sig. (2-tailed)
.670
.097
.516
.333
a. Test distribution is Normal. Source: SPSS Data Processing Results (2023) Table 4. Normality Test
One-Sample Kolmogorov-Smirnov Test Based on the data shown above it can be concluded that all the data in
this study have two-tailed Asymp.sig values greater than 0.05 and all the data
are normally distributed. 576 International Journal of Business and Applied Economics (IJBAE)
Vol. 2, No. 4, 2023: 2963-6124 International Journal of Business and Applied Economics (IJBAE)
Vol. 2, No. 4, 2023: 2963-6124 Figure 2. Normality Test with Normal Probability Plot
Source: SPSS Data Processing Results (2023) Figure 2. Normality Test with Normal Probability Plot
Source: SPSS Data Processing Results (2023) Figure 2 shows the results of the normality test using normal probability
plot technique. The test result shows that the points are scattered around the
diagonal line and follow the direction of the line. This means that the data are
normally distributed. Figure 3. Histogram Normality Test
Source: SPSS Data Processing Results (2023) Figure 3. Histogram Normality Test
Source: SPSS Data Processing Results (2023) The result of normality test using the histogram method is shown in
Figure 3. Because the data distribution is bell-shaped, we can conclude that the
data in this study is normally distributed. Heteroscedasticity Test Heteroscedasticity test is used to determine if there is a difference in
variance between the regression model and the residual data. Glejser's test is
used to test for heteroscedasticity, and values of Sig. > 0.05 indicates that the
regression does not contain heteroscedasticity. The test result is shown in the
table below: Table 6. Heteroscedasticity Test
Coefficientsa
Model
Unstandardized
Coefficients
Standardized
Coefficients
t
Sig. B
Std. Error
Beta
1
(Constant)
5.181
2.255
2.298
.028
Compensation
-.143
.104
-.245
-1.381
.176
Work Environment
.097
.104
.178
.935
.356
Workload
-.128
.140
-.169
-.909
.370
a. Dependent Variable: ABS_RES
Source: SPSS Data Processing Results (2023) Table 6. Heteroscedasticity Test The significance (hereinafter abbreviated as sig.) of the three variables
above reveals that each independent variable has a value > 0.05, implying that
in this regression model there are no issues related to heteroscedasticity. Multicollinearity Test The multicollinearity test determines whether a regression model detects
correlations between independent variables. Use to detect the presence or
absence of multicollinearity in a regression model when tolerance > 0.01 or VIF
value < . 10. The result of the multicollinearity test is shown in the table below: 577 Manalu, Broto, Kristiyani Table 5. Multicollinearity Test
Coefficientsa
Model
Collinearity Statistics
Tolerance
VIF
Compensation
.852
1.173
Work Environment
.739
1.353
Workload
.779
1.284
a. Dependent Variable: Turnover_Intention
Source: SPSS Data Processing Results (2023) Table 5. Multicollinearity Test
Coefficientsa
Model
Collinearity Statistics
Tolerance
VIF
Compensation
.852
1.173
Work Environment
.739
1.353
Workload
.779
1.284
a. Dependent Variable: Turnover_Intention
Source: SPSS Data Processing Results (2023) According to the multicollinearity test results in Table 5, the tolerance
coefficients for variables compensation, work environment, and workload >
0.01 and VIF values < 10. These result shows that the regression model does not
produce symptoms of multicollinearity. Multiple Linear Regression Test In testing research hypothesis, researchers used multiple linear regression
to test the effect of independent variables, consisting of compensation, work
environment, and workload, while in dependent variable section there is
turnover intention. The following is multiple linear regression model in this
research: 578 Table 7. Multiple Linear Regression Test
Coefficientsa
Model
Unstandardized
Coefficients
Standardized
Coefficients
t
Sig. B
Std. Error
Beta
(Constant)
.465
4.068
.114 .910
Compensation
.331
.187
.262
1.767 .086
Work Environment
.510
.187
.435
2.727 .010
Workload
.060
.253
.037
.236 .814
a. Dependent Variable: Turnover_Intention
Source: SPSS Data Processing Results (2023) Based on the data provided in Table 7, the multiple linear regression
model equation is as follows: (1)
Description:
α
: Constant
Y
: Turnover Intention
: Conpensation
: Work Environment
: Workload
: Error Term (1) p
α
: Constant Multiple linear regression is tested at multiple stages, including the F
test, coefficient of determination test, and T test. F Test F-test is used to simultaneously determine the influence of independent
factors on the dependent variable at the same time. If F-value > F-table, or value
of Sig < 0.05, the independent variable has a significant influence on the
dependent variable. The result of F-test in this study is: Table 8. F Test
ANOVAb
Model
Sum of Squares df Mean Square
F
Sig. Regression
164.120
3
54.707
6.404 .001a
Residual
290.433
34
8.542
Total
454.553
37
Source: SPSS Data Processing Results (2023) 579 Manalu, Broto, Kristiyani Manalu, Broto, Kristiyani Based on the data in Table 8, it can be concluded that compensation, work
environment, and workload simultaneously have a significant influence on
turnover intention phenomenon at Royal Orchid Garden Hotel (F-value > F-
table, and Sig. < 0.05). Determination Coefficient Test The coefficient of determination test is used to measure the extent to
which the power of the independent variable explains the variance of the
dependent variable. The result of the coefficient of determination test in this
study is as follows: Table 9. Determination Coefficient Test
Model Summaryb
Model R R Square Adjusted R Square Std. Error of the Estimate
1
.601a
.361
.305
2.923
Source: SPSS Data Processing Results (2023) Table 9. Determination Coefficient Test Based on data from Table 9, it is known that adjusted R square value is
0.305, so it can be interpreted that the independent variables consisting of
compensation, work environment, and workload are able to influence the
dependent variable of turnover intention by 30.5%. The remaining 69.5% is
impacted by other variables. a. If thei Sig valuei is > 0,05, thein
is acceipteid and
is reijeicteid. Work Environment Impact on Turnover Intention The work environment has positive and significant effect on the
phenomenon of turnover intention. Based on on-site observation, work
environment at the Royal Orchid Garden Hotel is slighty messy and damp. This
makes employees feel uncomfortable doing their job. At the Royal Orchid
Garden Hotel, the higher number of inconveniences in the work environment,
the higher turnover intention rate. Compensation Impact on Turnover Intention Compensation Impact on Turnover Intention From the description of the materials and the result of statistical test, it is
clear that compensation has negative and significant effect on the phenomenon
of turnover intention. According to observations and interviews conducted, it is
known that the company has not maximized compensation regularly and
thoroughly. This decreases staff motivation and has an impact on the increase
in turnover intention at Royal Orchid Garden Hotel. T Test T-test is used to determine the significance level of each independent
variable relative to the dependent variable. If the computed T value > T table or
Sig < 0.05, then
is rejected and
is accepted. T-test result for this study is: Table 10. T Test
Coefficientsa
Model
Unstandardized Coefficients Standardized Coefficients
t
Sig. B
Std. Error
Beta
(Constant)
.465
4.068
.114 .910
Compensation
.331
.187
.262
1.767 .086
Work Environment
.510
.187
.435
2.727 .010
Workload
.060
.253
.037
.236 .814
a. Dependent Variable: Turnover Intention
Source: SPSS Data Processing Results (2023) The test results of acceptance or rejection of
can be determined by
comparing the significance T value (Sig.) with the determined probability level
of 0.05. The test criteria to explain the influence between each variable is as
follows: 580 International Journal of Business and Applied Economics (IJBAE)
Vol. 2, No. 4, 2023: 2963-6124 b. If thei Sig valuei is < 0.05, thein
is reijeicteid and
is acceipteid. From the result above, it can be concluded as follows: First, compensation
calculation shows that T value < T table and Sig. > 0.05, it can be concluded that
the compensation variable has negative and significant effect on turnover
intention. Second, work environment calculation shows that T value > T table
and Sig. < 0.05, it can be concluded that work environment variable has positive
and significant effect on turnover intention. Third, workload calculation shows
that T value < T table and Sig. > 0.05, it can be concluded that workload
variable has negative and significant effect on turnover intention. Workload Impact on Turnover Intention Workload
has
negative
and
significant
impact
on
turnover
intention. Based on the research and observation conducted, it is known that
the workload at Royal Orchid Garden Hotel is still in a reasonable and normal
level. Employees are assigned duties based on their ability. Workload is
balanced with work facilities, making employees feel comfortable. CONCLUSIONS Based on the results of this research, we can draw the following
conclusions: 1. From the distribution result of the questionnaire, 65.79% of the
respondents are male, 36.84% are 20 to 25 years old, 50% have working
period less than 5 years, and 68.42% have high school or vocational
school degree. g
2. Based on the result of regression model's F-test, we can conclude that
compensation, work environment, and workload all have significant
impact on turnover intention phenomenon at the Royal Orchid Garden
Hotel (Sig. 0.001 < 0.05). 581 Manalu, Broto, Kristiyani 3. Based on T-test result of regression model, compensation has negative
and significant effect on turnover intention phenomenon at Royal Orchid
Garden Hotel. The higher the compensation of Royal Orchid Garden
Hotel employees, the lower number of turnover intention phenomena (T-
value 1.767 < T table 2.032 and Sig. 0.086 > 0.05). 3. Based on T-test result of regression model, compensation has negative
and significant effect on turnover intention phenomenon at Royal Orchid
Garden Hotel. The higher the compensation of Royal Orchid Garden
Hotel employees, the lower number of turnover intention phenomena (T-
value 1.767 < T table 2.032 and Sig. 0.086 > 0.05). g
4. 2. Work environment has positive and significant impact on turnover
phenomenon at Royal Orchid Garden Hotel. The higher number of
unfavorable working conditions at Royal Orchid Garden Hotel, the
higher turnover intention number (T-value 2.727 > T-table 2.032 and Sig. 0.010 < 0.05). )
5. Workload has negative and significant impact on turnover intention
phenomenon at Royal Orchid Garden Hotel. It also indicates that no
matter how much workload employees have, if it is balanced with
appropriate compensation, facilities, or other positive elements, it will
not increase the intention rate of turnover (T value 0.236 < T table and
Sig. 0.814 > 0.05). Based on the analysis results and conclusion, here are some suggestions
or relevant parties: Based on the analysis results and conclusion, here are some suggestions
or relevant parties: p
1. For further research, it is hoped that other variables that have not been
used in this study can be added, such as job stress, competitors, age, job
satisfaction, company policies, corporate culture, and so on. ADVANCED RESEARCH There are several limitations of this study which will be explained below: There are several limitations of this study which will be explained belo y
1. The independent variables compensation, work environment, and
workload can explain only 30.5% of the dependent variable (turnover
intention), with the remaining 69.5% influenced by other variables and
factors. Additional factors or variables that may influence turnover
intention are job stress, competitors, age, job satisfaction, business
policy, company culture, and so on. y
y
2. A small number of respondents. Because the number of permanent
employees at the Royal Orchid Garden Hotel are just 38, the researcher
decided to use a sample approach called non-probability sampling
with saturation sampling to gather accurate data. CONCLUSIONS Not only
that, but future researchers can search for research objects with more
samples than this study, so that the research result can explain overall
condition of the entire population, especially employees in a company. p p
p
y
p
y
p
y
2. For the company, this study is expected to be a major concern for the
hospitality company Royal Orchid Garden Hotel to pay more attention
to the well-being of its employees and reduce the company's high
turnover rate. Employee well-being is a central aspect of a company's
success. ACKNOWLEDGMENT The researcher would like to thank the Royal Orchid Garden Hotel
employees for their contributions to the success of this study. Special thanks to 582 International Journal of Business and Applied Economics (IJBAE)
Vol. 2, No. 4, 2023: 2963-6124 my family and lecturer who always encourage, support, and motivate me to
complete my research, including my completion of this thesis report and this
journal. Finally, the researcher apologizes for any errors in the writing. I hope
that this research can help to further research and knowledge. REFERENCES Andreani, J. L. R., Sulistiyani, E., & Azizah. (2020). Kompensasi Finansial dan
Non Finansial serta Pengaruhnya terhadap Kinerja Karyawan. Jurnal
Orbith, 16(1), 16–26. ( )
Arikunto, S. (2017). Prosedur Penelitian Suatu Pendekatan Praktik (S. Arikunto
(ed.); 16th ed.). Rineka Cipta. Ernawati, Prayekti, & Herawati, J. (2018). Pengaruh Beban Kerja terhadap
Turnover Intention melalui Stres Kerja sebagai Variabel Intervening (Studi
pada Mirota Batik Yogyakarta). Jurnal UMKM Dewantara Vol., 1(1), 29–38. https://core.ac.uk/download/pdf/196255896.pdf Fitriantini, R., Agusdin, & Nurmayanti, S. (2020). Pengaruh Beban Kerja,
Kepuasan Kerja dan Stres Kerja terhadap Turnover Intention Tenaga
Kesehatan Berstatus Kontrak di RSUD Kota Mataram. Jurnal Distribusi,
8(1), 23–38. Juhnisa, E., & Fitria, Y. (2020). Pengaruh Beban Kerja terhadap Burnout
Karyawan pada PT PLN (Persero) dengan Dukungan Sosial sebagai
Variabel Pemediasi. Jurnal Kajian Manajemen Dan Wirausaha, 2(4), 168. https://doi.org/10.24036/jkmw02100350 Khomariah, E., Pawenang, S., & Hadi, S. (2020). Turnover Intention PT. Efrata
Retailindo Ditinjau dari Beban Kerja, Lingkungan Kerja dan Kepuasan
Kerja. Business Management Analysis Journal (BMAJ), 3(1), 35–45. https://doi.org/10.24176/bmaj.v3i1.4429 Misno, A., Mulyapradana, A., Tajibu, M. J., Saputra, N., Aziza, N., & Anita, T. L. (2021). Fundamentals of Social Research. In Gcaindo (Ed.), Fundamentals of
Social Research: Methods, Processes and Applications (1st ed.). Diandra Kreatif. Mulyadi, M. (2019). Penelitian Kuantitatif dan Kualitatif serta Pemikiran Dasar
Menggabungkannya. Jurnal Studi Komunikasi Dan Media, 15(1), 128–138. Nugroho, T. T., & Darmawati, A. (2018). Pengaruh Kompensasi dan
Lingkungan Kerja terhadap Turnover Intention Pramuniaga PT Circle K
Indonesia Utama Cabang Yogyakarta. Jurnal Ilmu Manajemen, 15(2), 101–
108. Pranata, I. G. N., & Utama, I. W. M. (2019). Pengaruh Iklim Organisasi terhadap
Turnover Intention dengan Kepuasan Kerja sebagai Variabel Mediasi. E-
Jurnal Manajemen, 8(1), 7486–7518. Rahman, J. H. (2021). Jenis jenis data penelitian. Jurnal Teknik Pengumpulan Data
Dalam Rancangan Penelitian, 4(2), 33–41. Sekaran, U., & Bougie, R. (2018). Metode Penelitian untuk Bisnis: Pendekatan
Pengembangan Keahlian (R. Bougie (ed.); 6th ed.). Salemba Empat. Setyawan, D. A. (2022). Uji Validitas dan Reliabilitas Instrumen Pengumpulan
Data Menggunakan SPSS. In D. A. Setyawan (Ed.), Buku Petunjuk Praktikum 583 Manalu, Broto, Kristiyani Uji Validitas dan Reliabilitas Instrumen Pengumpulan Data (Menggunakan
Aplikasi SPSS) (1st ed., Issue July, p. 12). Politeknik Kesehatan Surakarta. Uji Validitas dan Reliabilitas Instrumen Pengumpulan Data (Menggunakan
Aplikasi SPSS) (1st ed., Issue July, p. 12). Politeknik Kesehatan Surakarta. Sihaloho, R. D., & Siregar, H. (2019). Pengaruh Lingkungan Kerja Terhadap
Kinerja Karyawan pada PT Super Setia Sagita Medan. Jurnal Ilmiah Socio
Secretum,
9(2),
273–281. https://jurnal.darmaagung.ac.id/index.php/socio/article/view/413/406 https://jurnal.darmaagung.ac.id/index.php/socio/article/view/413/406
Soegiyono. (2019). REFERENCES Metode Penelitian Kuantitatif, Kualitatif dan R&D (21st ed.). Alfabeta. Soegiyono. (2019). Metode Penelitian Kuantitatif, Kualitatif dan R&D (21st ed.). Alfabeta. Suganda, R., & Cahyadi, R. T. (2020). Modul Praktikum Statistika (R. Suganda &
R. T. Cahyadi (eds.); 1st ed.). Fakultas Ekonomi dan Bisnis Universitas Ma
Chung. Suwandi, E., Imansyah, F., & Dasril. (2018). Analisis Tingkat Kepuasan
Menggunakan Skala Likert pada Layanan Speedy yang Bermigrasi ke
Indihome. Jurnal Teknik Elektro, 2(3), 11. Syauqi, A., Abdurrahman, D., & Frendika, R. (2020). Pengaruh Beban Kerja dan
Lingkungan Kerja terhadap Turnover Intention pada PT. Putra Mustika
Prima Bandung. Jurnal Prosiding Manajemen, 6(1), 284–288. Wahyuningsih, S., Maulana, A., & Ligita, T. (2021). Faktor-faktor yang
Mempengaruhi Beban Kerja Perawat dalam Memberikan Asuhan
Keperawatan di Ruang Rawat Inap: Literature Review. Jurnal Keperawatan,
6(2), 1–8. Waskito, M., & Putri, A. R. (2022). Pengaruh Kompensasi dan Kepuasan Kerja
terhadap Turnover Intention pada Karyawan Office Pt Cipta Nugraha
Contrindo. KINERJA
Jurnal
Ekonomi
Dan
Bisnis,
4(1),
111–122. https://doi.org/10.34005/kinerja.v4i1.1799 584
|
https://openalex.org/W3155503043
|
https://journals.plos.org/plosone/article/file?id=10.1371/journal.pone.0248978&type=printable
|
English
| null |
Humeral elevation workspace during daily life of adults with spinal cord injury who use a manual wheelchair compared to age and sex matched able-bodied controls
|
PloS one
| 2,021
|
public-domain
| 9,769
|
PLOS ONE PLOS ONE RESEARCH ARTICLE Humeral elevation workspace during daily life
of adults with spinal cord injury who use a
manual wheelchair compared to age and sex
matched able-bodied controls Brianna M. GoodwinID1,2, Stephen M. Cain3, Meegan G. Van Straaten1,2,4,
Emma Fortune1,2, Omid JahanianID1,2, Melissa M. B. Morrow1,2* 1 Division of Health Care Delivery Research, Mayo Clinic, Rochester, MN, United States of America,
2 Robert D. and Patricia E. Kern Center for the Science of Health Care Delivery, Mayo Clinic, Rochester, MN,
United States of America, 3 Department of Mechanical Engineering, University of Michigan, Ann Arbor, MI,
United States of America, 4 Department of Physical Medicine and Rehabilitation, Mayo Clinic, Rochester,
MN, United States of America a1111111111
a1111111111
a1111111111
a1111111111
a1111111111 a1111111111
a1111111111
a1111111111
a1111111111
a1111111111 * morrow.melissa@mayo.edu * morrow.melissa@mayo.edu Abstract OPEN ACCESS
Citation: Goodwin BM, Cain SM, Van Straaten MG,
Fortune E, Jahanian O, Morrow MMB (2021)
Humeral elevation workspace during daily life of
adults with spinal cord injury who use a manual
wheelchair compared to age and sex matched able-
bodied controls. PLoS ONE 16(4): e0248978. https://doi.org/10.1371/journal.pone.0248978 Shoulder pain and pathology are extremely common for individuals with spinal cord injuries
(SCI) who use manual wheelchairs (MWC). Although risky humeral kinematics have been
measured during wheelchair-based activities performed in the lab, little is known about arm
kinematics in the free-living environment. The purpose of this study was to measure the
humeral elevation workspace throughout a typical day for individuals with SCI who use a
MWC and matched able-bodied controls. Thirty-four individuals with SCI who use a MWC
(42.7±12.7 years of age, 28 males/6 females, C6-L1) and 34 age-and sex-matched controls
were enrolled. Participants wore three inertial measurement units (IMU) on their upper arms
and torso for one to two days. Humeral elevation angles were estimated and the percentage
of time individuals spent in five elevation bins (0–30˚, 30–60˚, 60–90˚, 90–120˚, and 120–
180˚) were calculated. For both arms, the SCI cohort spent a significantly lower percentage
of the day in 0–30˚ of humeral elevation (Dominant: SCI = 15.7±12.6%, Control = 32.1
±15.6%, p<0.0001; Non-Dominant: SCI = 21.9±17.8%, Control = 34.3±15.5%, p = 0.001)
and a significantly higher percentage of time in elevations associated with tendon compres-
sion (30–60˚ of humeral elevation, Dominant: SCI = 62.8±14.4%, Control = 49.9.1±13.0%,
p<0.0001; Non-Dominant: SCI = 58.8±14.9%, Control = 48.3±13.6%, p = 0.003) than con-
trols. The increased percentage of time individuals with SCI spent in elevations associated
with tendon compression may contribute to increased shoulder pathology. Characterizing
the humeral elevation workspace utilized throughout a typical day may help in understand-
ing the increased prevalence of shoulder pain and pathology in individuals with SCI who use
MWCs. Editor: Yih-Kuen Jan, University of Illinois at
Urbana-Champaign, UNITED STATES Introduction Shoulder pain is the most common site of musculoskeletal pain in adults with spinal cord inju-
ries (SCI) who use manual wheelchairs (MWC) and its existence can significantly limit a per-
son’s functional abilities [1]. Shoulder pain is reported in 37–70% of individuals with SCI who
use a MWC [2–7]. This differs vastly from the 2.9% of the general able-bodied population who
experience shoulder pain [8]. Although shoulder pain can develop any time after SCI, it is
most commonly developed within the first five years [9] and often lasts longer than one year
[3]. Of the MWC users who experience pain, up to 93% have pathological signs on MRI [10],
most commonly in the supraspinatus tendon [11]. In general, non-traumatic supraspinatus tendon tears in the shoulder have been thought to
be caused by a combination of intrinsic and extrinsic factors [12]. However, these effects can
be exacerbated by overuse [13]. One extrinsic factor is the narrowing of the subacromial space
which causes compression of the supraspinatus tendons under the coracoacomial arch, and is
hypothesized to lead to increased tendon pathology and pain [14,15]. Individualized musculo-
skeletal models utilizing MRI have estimated the risk of supraspinatus tendon compression
through various humeral planes and elevations [16]. The magnitude of glenohumeral elevation
was the greatest kinematic predictor of tendon compression risk, followed by the specific
plane of elevation. The supraspinatus tendon had the greatest risk of compression at humer-
othoracic elevations angles between 30–60˚ [15]. Biplane fluoroscopic imaging of the shoulder
joint during dynamic motion has shown similar results and demonstrated that at higher
humeral elevations, as the humeral head rotates posteriorly, the supraspinatus tendon may no
longer be under the coracoacromial arch, therefore, not at risk of compression [17]. Under-
standing where tendon compression risk occurs can provide insights when interpreting the
humeral elevation workspace of activities of daily living. MWC propulsion, transfers, and other wheelchair-based activities of daily living have been
investigated in laboratory environments to characterize the upper extremity kinematics that
pose a risk for shoulder tendon compression from a reduction in subacromial space [18–20]. Although in-laboratory data provide accurate quantifications of how MWC users utilize their
arms to complete specific activities, it is unable to quantify the exposure to postures known to
reduce subacromial space in daily-living. Editor: Yih-Kuen Jan, University of Illinois at
Urbana-Champaign, UNITED STATES Received: August 13, 2020
Accepted: March 8, 2021
Published: April 23, 2021 Copyright: This is an open access article, free of all
copyright, and may be freely reproduced,
distributed, transmitted, modified, built upon, or
otherwise used by anyone for any lawful purpose. The work is made available under the Creative
Commons CC0 public domain dedication. Data Availability Statement: All relevant data are
within the manuscript and its Supporting
Information files. Funding: National Institutes of Health (grant no. R01 HD84423-01, MMM), Mayo Clinic Robert D. and Patricia E. Kern Center for the Science of
Health Care Delivery, and National Center for
Advancing Translational Sciences (UL1
TR002377). 1 / 17 PLOS ONE | https://doi.org/10.1371/journal.pone.0248978
April 23, 2021 PLOS ONE Humeral elevation workspace during daily life of manual wheelchair users compared to controls Competing interests: The authors have declared
that no competing interests exist. Competing interests: The authors have declared
that no competing interests exist. Competing interests: The authors have declared
that no competing interests exist. PLOS ONE | https://doi.org/10.1371/journal.pone.0248978
April 23, 2021 Introduction To understand the daily exposure to shoulder motion
and potential shoulder tendon compression, inertial measurement units (IMUs) can be used
to measure the angular velocity and acceleration of body segments throughout an entire day in
environments of daily living. IMU-based methods for quantifying shoulder movement show
good agreement with position-based motion capture and have been used to quantify shoulder
elevation angles; however, a limited number of studies have applied these methods to free-liv-
ing full-day collections [21–26]. To the best of our knowledge no study has utilized these meth-
ods to understand the humeral elevation workspace of MWC users throughout an entire day. The purpose of this study was to use IMUs to measure the humeral elevation workspace
throughout a typical day for individuals with SCI who use a MWC and compare it to matched
able-bodied controls. Comparison to controls allows for understanding of how humeral eleva-
tion exposure during daily life differs when the option to use the lower extremities for weight
bearing and mobility is removed. This study also aimed to understand the effects of years of
MWC use, pain, sex, and level of SCI on the humeral elevation workspace. Due to the
increased prevalence of shoulder pain and pathology in MWC users compared to able-bodied
controls [11] and the potential role that humeral elevation has on shoulder tendon compres-
sion [15], we hypothesized that MWC users would utilize a different humeral elevation work-
space than able-bodied adults. Specifically, we hypothesized individuals with SCI would spend
a higher percentage of time at elevation angles previously associated with tendon compression 2 / 17 PLOS ONE | https://doi.org/10.1371/journal.pone.0248978
April 23, 2021 PLOS ONE Humeral elevation workspace during daily life of manual wheelchair users compared to controls risk [27]. Understanding the humeral elevation workspace of individuals with SCI may con-
tribute to understanding why increased levels of shoulder pain and pathology occur for this
population. Participant enrollment This study was approved by the Mayo Clinic Institutional Review Board. Individuals with an
SCI who used a MWC as their main mode of mobility were recruited through querying medi-
cal records and care providers of local clinics. Sex- and age- (±2.5 years) matched able-bodied
controls were recruited through email distribution lists and classified ads. Participants for both
cohorts were considered for inclusion in the study if they were between 18–70 years of age and
had functional range of motion at both shoulders. Functional range of motion was defined as
active humeral thoracic flexion, abduction of at least 150˚ and the ability of the participant to
touch the opposite shoulder, the back of his/her neck and his/her low back. Prior to accrual to
the study a licensed physical therapist performed a screening physical exam to confirm inclu-
sion and exclusion criteria listed above. This study is part of a larger longitudinal study that fol-
lows rotator cuff pathology progression over time via magnetic resonance imaging (MRI). Therefore, participants were also excluded if they self-reported a previous diagnosis of com-
plete supraspinatus tendon tear or they were withdrawn from the study if a complete tear was
seen during the first MRI. Participants with SCI who had unilateral supraspinatus complete
tears were still eligible to be followed for the contralateral shoulder. Additionally, participants
in both cohorts were excluded if there were conditions/factors which might have hindered
protocol adherence and controls were also excluded if they had any musculoskeletal or neuro-
logical disorder which might have impacted shoulder health or changed the individual’s ability
to walk independently. PLOS ONE | https://doi.org/10.1371/journal.pone.0248978
April 23, 2021 Questionnaires and IMU instrumentation Upon enrollment, participants attended an in-lab visit. A licensed physical therapist screened
participants for eligibility and informed consent was obtained. Participants self-reported their
hand dominance and were asked if they had pain in either or both shoulders. To assess the
presence of shoulder pain, the physical therapist asked the participant if they experience any
shoulder pain in either or both shoulders during their daily life. The therapist clarified with the
participant that the pain can come and go, is muscular or joint pain and not nerve pain, and
may happen before or after certain activities. All participants from both cohorts completed the
Disabilities of the Arm, Shoulder, and Hand (DASH) questionnaire [28] for both right and left
arms. The DASH is a measure of physical function and symptoms, and is not specific to the
shoulder, but rather the whole arm is considered when responding to the questions. It encom-
passes 30 questions which ask individuals to rate their difficulty, pain, and satisfaction when
accomplishing specific tasks on a 5 point scale. Scores range from 0–100, with 0 indicating no
difficulty and 100 indicating the most difficulty, pain, and dissatisfaction. The DASH has been
shown to be reliable and to have high validity [29]. Additionally, the SCI cohort filled out the
Wheelchair User’s Shoulder Pain Index (WUSPI) for both the right and left shoulders. To
complete the WUSPI, participants were asked to rate their shoulder pain when completing 15
tasks on a visual analog scale between “no pain” and “worst pain ever experienced” [30]. Possi-
ble scores ranged from 0 (no pain) and 150 (worst pain ever experienced in all categories). The
WUSPI is valid and reliable for this population [31]. Although we acknowledge that the DASH
and WUSPI were designed to be filled out once, as part of a larger study, both surveys were
filled out for both arms to evaluate pain and function as it related to each arm. 3 / 17 PLOS ONE | https://doi.org/10.1371/journal.pone.0248978
April 23, 2021 PLOS ONE Humeral elevation workspace during daily life of manual wheelchair users compared to controls Participants were given three wireless IMUs (Emerald or Opal, APDM, Inc., Portland, OR). Each IMU contained a 3-axis accelerometer (±200 g), 3-axis gyroscope (±2000˚/s), and 3-axis
magnetometer (±8 Gauss). The three IMUs remained synchronized via a proprietary wireless
protocol, recorded data at 128 Hz and saved the data to internal storage. Questionnaires and IMU instrumentation In order to maximize
the consistency of IMU placement and functional calibration movements across participants,
written handouts, video guides and in-person instruction were provided. Participants were
instructed to wear one IMU on each lateral upper arm and one on the anterior of the torso;
IMUs were secured on the body with elastic and Velcro straps. Each IMU was labeled with the
wear location (left arm, right arm, or torso) and an arrow indicating the proper mounting ori-
entation. Participants were instructed to wear the sensors during the entire length of two typi-
cal days, excluding bathing and swimming, and take them off before going to bed. Both
cohorts were asked to perform their regular daily routines; participants in the control cohort
did not use MWCs. Upon donning the sensors for a day, participants performed a set of func-
tional calibration postures (Fig 1, Appendix A, the individual in this manuscript has given
written informed consent (as outlined in PLOS consent form) to publish these case details). Due to the collection of multiple days of data, participants were responsible for charging the
IMUs overnight using a provided charging station. After the data collection, participants
returned the sensors with a pre-paid mailer or in person to the study staff. Data processing Data were downloaded through Motion Studio (APDM, Inc., Portland, OR) and outputs
included estimates of the orientations of each IMU relative to an inertial frame (Fig 2). The Fig 1. Functional calibration used to align IMU’s with the body. Postures included static upright neutral posture with upper
arms resting against the thorax (posture 1), static and dynamic arm t-pose/movement (shoulder abduction = 90˚, posture 2),
static and dynamic flexion pose/movement (shoulder flexion = 90˚, posture 3), dynamic flexion and extension of the torso
(movement 4), and simulated wheelchair use or walking (movement 5). Postures 2 and 3 were completed for both the right
and left arms separately. (Note: The individual pictured is a co-author who is able-bodied). Fig 1. Functional calibration used to align IMU’s with the body. Postures included static upright neutral posture with upper
arms resting against the thorax (posture 1), static and dynamic arm t-pose/movement (shoulder abduction = 90˚, posture 2),
static and dynamic flexion pose/movement (shoulder flexion = 90˚, posture 3), dynamic flexion and extension of the torso
(movement 4), and simulated wheelchair use or walking (movement 5). Postures 2 and 3 were completed for both the right
and left arms separately. (Note: The individual pictured is a co-author who is able-bodied). Fig 1. Functional calibration used to align IMU’s with the body. Postures included static upright neutral posture with upper
arms resting against the thorax (posture 1), static and dynamic arm t-pose/movement (shoulder abduction = 90˚, posture 2),
static and dynamic flexion pose/movement (shoulder flexion = 90˚, posture 3), dynamic flexion and extension of the torso
(movement 4), and simulated wheelchair use or walking (movement 5). Postures 2 and 3 were completed for both the right
and left arms separately. (Note: The individual pictured is a co-author who is able-bodied). https://doi.org/10.1371/journal.pone.0248978.g001 https://doi.org/10.1371/journal.pone.0248978.g001 4 / 17 PLOS ONE | https://doi.org/10.1371/journal.pone.0248978
April 23, 2021 PLOS ONE Humeral elevation workspace during daily life of manual wheelchair users compared to controls Fig 2. Data processing workflow. This included data collected, IMU orientation, IMU alignment to arm through
calibration postures, representative data, and humeral elevation bins. The percentage of time spent in each bin was
calculated and used for analysis. (Note: The individual pictured is a co-author who is able-bodied). https://doi.org/10.1371/journal.pone.0248978.g002 Fig 2. Data processing workflow. This included data collected, IMU orientation, IMU alignment to arm through
calibration postures, representative data, and humeral elevation bins. Data processing The percentage of time spent in each bin was
calculated and used for analysis. (Note: The individual pictured is a co-author who is able-bodied). https://doi.org/10.1371/journal.pone.0248978.g002 https://doi.org/10.1371/journal.pone.0248978.g002 orientation estimates were derived from the combined acceleration and angular velocity data,
rather than only the acceleration data. While researchers have used IMU-measured accelera-
tion only to estimate arm orientation [32,33], there are known limitations to this approach
[34], namely, the challenge of separating the measured acceleration into gravitational and
body caused components. Algorithms that use both measured acceleration and angular 5 / 17 PLOS ONE | https://doi.org/10.1371/journal.pone.0248978
April 23, 2021 PLOS ONE Humeral elevation workspace during daily life of manual wheelchair users compared to controls velocity to estimate IMU orientation or attitude (orientation relative to gravity) are well under-
stood and are critical in strapdown inertial navigation [35]. These algorithms integrate the
angular velocity signal to estimate orientation during periods with high dynamics (significant
body acceleration) and use the acceleration signal to update or correct the orientation during
periods with low dynamics (measured acceleration close to the acceleration of gravity). Fur-
ther, these algorithms take different forms and have been proven to be highly accurate for esti-
mating attitude [36–39]. Additionally, the orientation estimates were calculated without
magnetometer data due to the unknown and likely non-uniform magnetic fields present
throughout field data collections. While the orientation algorithm used by APDM is proprie-
tary, sensor fusion methods (e.g. Kalman filters) used to estimate IMU orientation from raw
sensor data are well understood and well documented in the literature [21,35,36]. Custom
MATLAB (Mathworks, Natick, MA) code was written to calculate orientations of anatomical
axes relative to IMU-fixed reference frames using data collecting during each participant’s
functional calibration postures and movements (Fig 1; Appendix A). Orientation of a given
body segment (upper arm or thorax) in an inertial (world) reference frame was then estimated
using the orientation of the IMU and the orientation of the anatomical axes relative to the
IMU-fixed reference frame (Appendix B). Humeral elevation and thorax deviation angles
were defined as the angle between the long axis of the body segment (defined from the func-
tion calibration) and vertical; these angles are only dependent on the estimated direction of
gravity relative to the body segment and, therefore, are drift-free metrics for quantifying body
segment motions. Data processing 6 / 17 PLOS ONE | https://doi.org/10.1371/journal.pone.0248978
April 23, 2021 PLOS ONE Humeral elevation workspace during daily life of manual wheelchair users compared to controls Table 1. Participant demographics. SCI
Control
P-value
Sample size
34
34
-
Age
42.7 +/- 12.7 (22.6–
63.3)
42.6 +/- 12.5 (24.3–
61.0)
Sex
28 males/6 females
28 males/6 females
-
Self-reported weight (kg)
80.7 +/- 17.2 (54.0–
145.1)
81.6 +/- 17.5 (56.7–
149.7)
0.822
Self-reported height (cm)
177.4 +/- 7.6
(160.0–195.6)
178.4 +/- 9.5
(160.0–205.7)
0.417
Injury Level
Cervical (C6-C8)
7
-
-
High/mid thoracic (T1-T8)
16
Low thoracic/lumbar (T9-L1)
11
Years of manual wheelchair use (years)
11.5 +/- 10.7 (0.5–
36.0)
-
-
Dominant arm
27 right/7 left
32 right/2 left
0.374
DASH (dominant arm)
15.2 +/- 17.5 (0–
71.7)
1.3 +/- 2.9 (0–15)
<0.0001
DASH (non-dominant arm)
13.6 +/- 14.2 (0–
51.7)
1.1 +/- 3.3 (0–15)
<0.0001
WUSPI (dominant arm)
12.7 +/- 20.4 (0–
71.2)
-
-
WUSPI (non-dominant arm)
10.8 +/- 16.2 (0–
71.6)
-
-
Self-reported shoulder pain (Number of participants
reporting pain, % of cohort)
26 (76%)
9 (26%)
<0.0001
https://doi.org/10.1371/journal.pone.0248978.t001 Table 1. Participant demographics. Periods of non-wear were determined using methods from Lugade and colleagues (2014)
[49] and were excluded from data analysis. Data were also excluded from analysis if the func-
tional calibration postures were not completed properly or if at least eight hours of data were
not collected after the elimination of non-wear time. Data were included if one or two com-
plete days were collected; if two days were included all data were combined before the calcula-
tion of the percent of time in humeral elevation bins. Data processing The calculated humeral elevation angles range between 0–180˚, with 0˚ indi-
cating the arm was down and perfectly aligned with gravity and 180˚ indicating the arm was
raised overhead and aligned with gravity. These methods have previously been validated in
unpublished data where five individuals with SCI performed 10 reaching tasks. The absolute
error and percent of error when compared to the gold standard (electromagnetic system) were
-0.06±1.12˚ and -1.44±1.28%, respectively, for the range of motion. The absolute error and
percent of error for the maximum elevation achieved during each reach were 2.59±2.47˚ and
2.04±2.47%, respectively. It is important to note that humerothoracic elevation angles and elevation planes relative to
the thorax were not calculated as these calculations require accounting for relative drift
between the orientation estimates of arm and torso IMUs. While the attitude estimates are
accurate and do not drift, the yaw or heading estimates, which describe the rotation angle or
direction of a body segment about a vertical axis, do drift, making accurate calculation of
humerothoraic angles over long periods of time difficult [24]. This difficulty is best illustrated
by the fact that studies that use IMUs to quantify shoulder motion during long periods in the
real world either do not calculate shoulder angles [32,40–42] or acknowledge the limitations of
the methodology [24]. Other work [25,43] claims to accurately calculate shoulder angle of ele-
vation but not plane of elevation; however, shoulder angle of elevation cannot be calculated
accurately without the plane of elevation [44]. Correcting the drift between sensors about verti-
cal is an active research area and requires a joint specific approach [45,46]. Therefore, in our
analysis, data in which the thorax deviation angle was more than 30˚ were eliminated in order
to allow humeral elevation angles to be interpreted similarly to humerothoracic elevation
angles; 30˚ was selected based on an unpublished sensitivity analysis performed during a prior
study. The percentage of daily wear time each participant spent in five humeral elevation bins were
calculated (0–30˚, 30–60˚, 60–90˚, 90–120˚, and 120–180˚). The bin sizes were chosen as a
means to combine three theories: 1) a painful arc of motion occurs between 60–120˚ of arm
abduction [47], 2) Rapid Upper Limb Assessment (RULA) which bins risky arm postures
between 0–20˚, 20–45˚, 45–90˚, and >90˚ [48], and 3) the subacromial risk area of 30–60˚ [15]. PLOS ONE | https://doi.org/10.1371/journal.pone.0248978
April 23, 2021 Results Thirty-four participants with SCI who used a MWC, and 34 age (±2.5 years) and sex matched,
able-bodied adults were enrolled (Table 1). There were no statistical differences between the
cohort’s self-reported weight, height, and dominant hand. Excluded data One control participant was ineligible for the study due to a self-reported complete supraspi-
natus tear which was confirmed with the medical records of an MRI. Seven pairs of data were
excluded from the analysis due to exclusion criteria (Fig 3). Data were collected for an average
(SD) of 11.4(2.1) and 11.9(1.3) hours for the SCI and control cohorts, respectively. Addition-
ally, on average 18.3(14.0) and 28.0(10.3) percent of the day was excluded because the trunk
was at or over 30˚ for the SCI and control cohorts, respectively. Statistical analysis Between cohort differences for the demographics data were assessed using paired t-tests for
the continuous variables (weight, height, and DASH), McNemar’s test for the presence of
shoulder pain and Fisher’s Exact test for hand dominance. Multivariate analyses of variance
(MANOVA) were used to test for the main effect of cohort on time spent in each humeral
elevation bin of both the dominant and non-dominant side (α = 0.05). Similarly, within
each cohort, MANOVA was also used to test main effects of sex, age, and arm function
(DASH) on time in bins. Finally, withing the MWC user cohort, MANOVA was used to test
main effects of shoulder pain (WUSPI), level of SCI, and years of MWC use. Linear regres-
sion analysis was used to test the strength of the relationship between the time spent in each
humeral elevation bin with age and years of MWC use for the MWC cohort. Within
cohorts, analysis of variance (ANOVA) was used to test the effect of humeral elevation bin
for both arms. When significant main effects were observed, post hoc paired t-tests were
performed. A Bonferroni correction factor was used to adjust the alpha level from 0.05 to
0.01 due to comparisons across five bins. PLOS ONE | https://doi.org/10.1371/journal.pone.0248978
April 23, 2021 7 / 17 PLOS ONE Humeral elevation workspace during daily life of manual wheelchair users compared to controls Humeral elevation workspace There was a main effect of cohort across humeral elevation bins on both dominant and non-
dominant sides (p<0.0001). Additionally, there was a main effect of humeral elevation bin for
both cohorts and arms (dominant: p<0.0001, non-dominant: p = 0.005, Fig 4). Individuals Fig 3. Data exclusion processes. Data were excluded if either SCI or control did not collect data, a minimum of 8
hours of data were not collected, or one sensor malfunctioned. Data were included in analysis if one or two days of
data were collected. https://doi.org/10.1371/journal.pone.0248978.g003 Fig 3. Data exclusion processes. Data were excluded if either SCI or control did not collect data, a minimum of 8
hours of data were not collected, or one sensor malfunctioned. Data were included in analysis if one or two days of
data were collected. 8 / 17 PLOS ONE | https://doi.org/10.1371/journal.pone.0248978
April 23, 2021 PLOS ONE Humeral elevation workspace during daily life of manual wheelchair users compared to controls with SCI spent significantly more time in 30–60˚ of humeral elevation than all other elevations
Fig 4. The percentage of time in each humeral elevation. A) The average percentage of time throughout a typical day individuals in the
SCI and control cohorts spent in 0–30˚, 30–60˚, 60–90˚, 90–120˚, and 120–180˚ of humeral elevation for their dominant and non-
dominant sides. B) Percentage of time throughout a typical day individuals in the SCI and control cohorts spent in 0–30˚, 30–60˚, 60–90˚,
90–120˚, and 120–180˚ of humeral elevation for their dominant arm (top) and their non-dominant arm (bottom). For each boxplot the
central line (red) represents the median, the edges of the box are the 25th and 75th percentiles, and the error bars extend the most extreme
data points not considered outliers and, the outliers are denoted by red +. indicates p < 0.0001 and indicates p < 0.005. https://doi.org/10.1371/journal.pone.0248978.g004 Fig 4. The percentage of time in each humeral elevation. A) The average percentage of time throughout a typical day individuals in the
SCI and control cohorts spent in 0–30˚, 30–60˚, 60–90˚, 90–120˚, and 120–180˚ of humeral elevation for their dominant and non-
dominant sides. B) Percentage of time throughout a typical day individuals in the SCI and control cohorts spent in 0–30˚, 30–60˚, 60–90˚,
90–120˚, and 120–180˚ of humeral elevation for their dominant arm (top) and their non-dominant arm (bottom). PLOS ONE PLOS ONE Table 2. The average (SD) percentage of the day individuals with SCI and matched able-bodied controls spent in five humeral elevation bins throughout one or two
days. Bin
SCI Dominant Arm
Percentage (%)
Control Dominant Arm
Percentage (%)
P-Value
SCI Non-Dominant Arm
Percentage (%)
Control Non-Dominant Arm
Percentage (%)
P-Value
0–30˚
15.7 (12.6)
32.1 (15.6)
<0.0001
21.9 (17.8)
34.3 (15.5)
0.001
30–60˚
62.8 (14.4)
49.9 (13.0)
<0.0001
58.8 (14.9)
48.3 (13.6)
0.003
60–90˚
18.4 (11.0)
16.2 (9.6)
0.410
17.7 (14.8)
15.6 (6.2)
0.430
90–120˚
2.8 (5.3)
1.4 (1.0)
0.145
1.4 (1.4)
1.6 (1.2)
0.589
120˚-
180˚
0.2 (0.4)
0.4 (0.8)
0.320
0.2 (0.5)
0.3 (0.3)
0.430
https://doi.org/10.1371/journal.pone.0248978.t002 Table 2. The average (SD) percentage of the day individuals with SCI and matched able-bodied controls spent in five
d h SCI and matched able-bodied controls spent in five humeral elevation bins throughout one or two https://doi.org/10.1371/journal.pone.0248978.t002 on their dominant and non-dominant sides, respectively. The controls also spent the greatest
amount of daily wear time in this elevation bin at 50% and 48% on their dominant and non-
dominant arm respectively, which was significantly lower than the SCI cohort for both arms
(dominant: p < .0001, non-dominant: p = 0.003, Table 2). For the SCI cohort, the second largest percentage of time was spent in 60–90˚ of humeral
elevation (approximately 20% of their day for both arms). Controls spent their second largest
percentage of time in 0–30˚ of elevation for both arms, which was significantly higher than the
amount of time the SCI cohort spent in this elevation bin (p<0.001). Individuals with SCI
spent comparable amounts of time in 0–30˚ and 60–90˚ of elevation, while controls spent sig-
nificantly more time in 0–30˚ than 60–90˚ of humeral elevation on the dominant (p<0.001)
and non-dominant (p<0.0001) sides. On average, participants in both cohorts spent less than 3% of their day (<25 minutes) in
elevations over 90˚ for both arms. There were no significant differences between cohorts for
the 60–90˚ 90–120˚and >120˚ humeral elevation bins or between dominant and non-domi-
nant arms for each cohort and each elevation bin. Pain and arm function Pain, measured by the WUSPI (in the SCI cohort) and arm function measured by the DASH
(in both both cohorts), did not have a significant effect on the percentage of time an individual
spent in any humeral elevation bins for both dominant and non-dominant arms. Sex, age, injury level, and years of MWC use There were no main effects of sex (Table 3), age (Table 4), injury level (Table 3), or years of
MWC Use (Table 4) on either arm. PLOS ONE | https://doi.org/10.1371/journal.pone.0248978
April 23, 2021 Humeral elevation workspace For each boxplot the
central line (red) represents the median, the edges of the box are the 25th and 75th percentiles, and the error bars extend the most extreme
data points not considered outliers and, the outliers are denoted by red +. indicates p < 0.0001 and indicates p < 0.005. https://doi.org/10.1371/journal.pone.0248978.g004 https://doi.org/10.1371/journal.pone.0248978.g004 with SCI spent significantly more time in 30–60˚ of humeral elevation than all other elevations
bins on both their dominant and non-dominant sides (p<0.001, Table 2). The SCI cohort
spent 63% and 59% of their daily wear time (approximately 7 hours per day) at these elevations PLOS ONE | https://doi.org/10.1371/journal.pone.0248978
April 23, 2021 9 / 17 Humeral elevation workspace during daily life of manual wheelchair users compared to controls Discussion This study aimed to understand the humeral elevation workspace utilized throughout a typical
day by individuals with SCI who use a MWC. These results were compared to a matched able-
bodied control cohort to better understand factors which may contributed to a higher rate of
both pain and tendon pathology associated with years of MWC use [11]. Both individuals with
SCI and controls spent the majority of their day (~80%) in elevation angles between 0 and 60˚. However, individuals with SCI spent significantly more time in humeral elevations previously
found to be associated with supraspinatus tendon compression (30–60˚) than controls
[27,50,51]. There was no evidence of the effect of injury level, years of MWC use, age, or sex
on the humeral elevation workspace for individuals with SCI. With the growing capabilities of wearable technology, many SCI-specific algorithms have
been created and validated to accompany and enhance data captured in a lab setting [52]. Many PLOS ONE | https://doi.org/10.1371/journal.pone.0248978
April 23, 2021 10 / 17 PLOS ONE Humeral elevation workspace during daily life of manual wheelchair users compared to controls Table 3. The percentage of time individuals spent in humeral elevation bins based on their injury level and sex. Injury Level
Cervical
High/mid thoracic
Low thoracic/lumbar
P-Value
Dominant Arm (% of the day)
0–30˚
20.3 ± 12.5
15.5 ± 14.0
13.1 ± 11.4
0.521
30–60˚
57.4 ± 11.0
63.0 ± 15.9
66.0 ± 15.1
0.493
60–90˚
18.0 ± 9.7
18.4 ± 12.4
18.8 ± 11.2
0.988
90–120˚
3.7 ± 5.2
3.0 ± 6.9
1.8 ± 2.9
0.769
120–180˚
0.6 ± 0.8
0.1 ± 0.2
0.2 ± 0.2
0.050
Non-Dominant Arm (% of the day)
0–30˚
24.7 ± 16.9
20.6 ± 21.1
22.1 ± 15.0
0.888
30–60˚
56.9 ± 11.3
59.4 ± 18.6
59.0 ± 12.6
0.937
60–90˚
16.5 ± 9.1
18.9 ± 19. 16.6 ± 11.4
0.910
90–120˚
1.6 ± 1.2
1.0 ± 1.0
1.9 ± 1.9
0.275
120–180˚
0.3 ± 0.4
0.06 ± 0.09
0.4 ± 0.8
0.266
Sex
Male
Female
P—Value
Dominant Arm (% of the day)
0–30˚
14.4 ± 12. Discussion PLOS ONE | https://doi.org/10.1371/journal.pone.0248978
April 23, 2021 11 / 17 PLOS ONE Humeral elevation workspace during daily life of manual wheelchair users compared to controls elevation angles or overuse of the arms of MWC users. The data presented in the current study
supplements data collected in a laboratory setting and other free-living MWC use metrics by
providing lengths of exposure to risky postures in the free-living environment. Recently it has been suggested that compression of the supraspinatus tendon occurs at low
elevation angles. Giphart, et al. [17] suggested that subacromial impingement syndrome occurs
below 70˚ of humeral elevation and the minimum distance between the footprint of the
supraspinatus tendon and greater tuberosity occurred between 36˚ and 65˚ of humeral eleva-
tion during forward flexion. Additionally, using individualized bone models (from MRI) and
group averaged kinematics, Lawrence, et al. [15] used musculoskeletal simulation models to
suggest the minimum distance between the coracoacomial arch and supraspinatus tendon area
occurred at 42˚ of humerothoracic elevation. Our results show that individuals with SCI who
use a MWC spent significantly more time than controls in a similar range of humeral eleva-
tions (30–60˚). This difference could be in part due to differences in the arm elevation work-
space during mobility. During MWC propulsion the humeral elevation is approximately 25 to
55˚ at a self-selected speed [55,56]; however, during walking, the humeral elevation angles
required are much lower [57]. The difference in humeral elevation during mobility likely is
not the only contributor to this increase, as MWC users move about 3 km less than able-bod-
ied individuals and only spend a small amount of their day actually propelling themselves; esti-
mates range from 16 to 54 minutes per day [53,58]. Another contributing factor to this
discrepancy may be wheelchair setup; for example, MWC users may not place their arms in a
neutral resting position of 0–30˚ due to the location of their arm rest. It will be important in
future studies to understand whether a large proportion of the time in 30–60˚ of humeral ele-
vation is actually due to MWC users resting on their arm rests. In addition to the humeral ele-
vation workspace differing during propulsion for MWC users and walking for able-bodied
individuals, the loading of the shoulder is also different during these two tasks and likely con-
tributes to the increase in pathology in MWC users. Discussion 21.9 ± 15.3
0.199
30–60˚
62.8 ± 14.4
63.2 ± 17.4
0.950
60–90˚
19.4 ± 11.7
13.8 ± 8.0
0.267
90–120˚
3.2 ± 5.9
1.0 ± 0.8
0.381
120–180˚
0.2 ± 0.5
0.2 ± 0.3
0.814
Non-Dominant Arm (% of the day)
0–30˚
20.4 ± 18.3
29.0 ± 16.2
0.294
30–60˚
59.1 ± 15.6
57.4 ± 14.2
0.814
60–90˚
18.8 ± 16.1
12.2 ± 7.6
0.337
90–120˚
1.5 ± 1.5
1.2 ± 0.8
0.691
120–180˚
0.2 ± 0.5
0.1 ± 0.1
0.565
https://doi.org/10.1371/journal.pone.0248978.t003 Table 3. The percentage of time individuals spent in humeral elevation bins based on their injury level and sex. Injury Level of the studies using wearable technology to understand movement of MWC users have focused
specifically on wheelchair propulsion and use [53,54], with less focus on understanding humeral of the studies using wearable technology to understand movement of MWC users have focused
specifically on wheelchair propulsion and use [53,54], with less focus on understanding humeral Table 4. Linear regression results for the percentage of time individuals spent in humeral elevation bins based on
age and years of MWC use. Age
Years of MWC use
R2
P-Value
R2
P-Value
Dominant Arm
0–30˚
0.104
0.06
0.009
0.60
30–60˚
0.037
0.28
0.000
0.97
60–90˚
0.013
0.41
0.008
0.62
90–120˚
0.000
0.94
0.002
0.82
120–180˚
0.002
0.79
0.047
0.22
Non-Dominant Arm
0–30˚
0.027
0.35
0.001
0.84
30–60˚
0.002
0.81
0.009
0.58
60–90˚
0.025
0.37
0.007
0.65
90–120˚
0.000
0.88
0.054
0.19
120–180˚
0.049
0.21
0.023
0.39
https://doi.org/10.1371/journal.pone.0248978.t004
PLOS ONE | https://doi.org/10.1371/journal.pone.0248978
April 23, 2021
11 / 17 Table 4. Linear regression results for the percentage of time individuals spent in humeral elevation bins based on
age and years of MWC use. Age
Years of MWC use
R2
P-Value
R2
P-Value
Dominant Arm
0–30˚
0.104
0.06
0.009
0.60
30–60˚
0.037
0.28
0.000
0.97
60–90˚
0.013
0.41
0.008
0.62
90–120˚
0.000
0.94
0.002
0.82
120–180˚
0.002
0.79
0.047
0.22
Non-Dominant Arm
0–30˚
0.027
0.35
0.001
0.84
30–60˚
0.002
0.81
0.009
0.58
60–90˚
0.025
0.37
0.007
0.65
90–120˚
0.000
0.88
0.054
0.19
120–180˚
0.049
0.21
0.023
0.39
https://doi.org/10.1371/journal.pone.0248978.t004 Table 4. Linear regression results for the percentage of time individuals spent in humeral elevation bins based on
age and years of MWC use. Discussion Further, additional data collections and
analyses are needed to fully understand the clinical implications of the differences in humeral
elevation between the wheelchair users and able-bodied control group. Capturing a holistic view of an individual’s exposure to potentially risky humeral elevation
is dependent on many factors including occupation and activities performed throughout a
day. A study looking at 10 able-bodied elderly adults using only accelerometry data found that
less than 4% of an individual’s day was spent in elevations above 90˚, with the average elevation
angle occurring at 40˚ [43]. These results are very similar to the data presented in the current
study for both cohorts; about 3% of the day was spent in elevations over 90˚. Previous reports
have suggested that extended periods of time in overhead motion may be the cause of
increased shoulder pain. Our results paired with the most recent modeling and imaging data
may suggest that injury to the supraspinatus tendon due to tendon compression of the SCI
cohort also occurs in-part due to increased time between 30–60˚ of humeral elevations. Fur-
ther, pain in higher elevation angles may be caused by other mechanisms [15]. Continuing to
map this workspace for individuals with SCI who use a MWC while they perform specific
tasks (i.e., propulsion or transfers) may help us to further understand daily risk exposures and
the contribution of specific tasks. Multiple challenges exist when using unsupervised real-world IMU data. First, accounting
for and correcting the drift of IMU-based body segment orientation estimates is a common
challenge in understanding the relative orientation of body segments (i.e. joint angles), espe-
cially for extended data collections (see excellent discussion in [24]). The current algorithms
utilized in this study do not take the plane of motion into account; 30˚ of humeral elevation in
front of the body, to the side, or behind would all be interpreted as 30˚ of humeral elevation
and are indistinguishable. While we could have used the orientation estimates to calculate PLOS ONE | https://doi.org/10.1371/journal.pone.0248978
April 23, 2021 12 / 17 PLOS ONE Humeral elevation workspace during daily life of manual wheelchair users compared to controls humerothoracic angle of elevation and plane of elevation, we know that those calculations
would contain errors from the heading/yaw drift. Heading drift directly affects the plane of ele-
vation and accurately calculating the plane of elevation is critical to accurately calculating
humerothoracic angle of elevation. Discussion Therefore, the data presented here only used the angle of
the humerus relative to vertical (humeral elevation angle) and not the trunk (humerothoracic
angle). This was compensated for by eliminating humeral elevation time points where the
trunk angle was at or over 30˚ of tilt; participants may have been leaning over or lying down. On average about 10% more data were eliminated from the control data sets than the SCI data
sites, indicating the controls had more variability and movement of their trunk than the SCI
cohort. Even with these limitations, the methods used in this study to estimate sensor orienta-
tion and humeral elevation are more accurate than other methods using only acceleration
data, especially during movements with high dynamics [36–39]. Since this study included a limited number of participants with a cervical level SCI who use
manual wheelchairs, these results should not be generalized to individuals with a cervical level
SCI who use power wheelchairs. Power wheelchairs often have the ability to recline or tilt in
space. Individuals with a higher level cervical SCI may also routinely sit in a more reclined
position to increase their stability and compensate for lack of trunk control. If persons with
higher level cervical SCIs perform a large amount of arm movement in these reclined or tilted
positions, then a large portion of their daily routine data would be eliminated by the 30˚ trunk
tilt threshold that was used in this study. If this is the case, a different approach to study arm
use in this population would be needed. There are limitations with the data presented in this study to consider. Previous studies
have found that up to four days of data collection are needed to represent propulsion trends
consistent throughout a MWC user’s daily life [59]. Only one or two days of data were col-
lected for participants in this study due to participant availability and adherence to the proto-
col. We attempted to compensate for this by asking participants to wear the sensors on ‘typical
days;’ however, we did not account for the day of the week or whether it was a workday or not
in our analysis. Further, our analysis represents the data from the full day and does not account
for the distribution of the humeral elevation angles at specific segments of the day such as the
morning, afternoon or evening. Discussion The calibration protocol used in this study enabled us to deter-
mine humeral elevations without an in-lab calibration. As participants performed the calibra-
tion protocol unsupervised, it’s possible that there could be errors induced by incorrect neutral
and 90˚ calibration postures. The data presented here were binned into 30˚ ranges below 120˚
of humeral elevation; however, creating bins with different boundaries may affect the results. Appendix C shows the average percent of time in 10˚ bins. Additionally, there are other factors
beyond humeral elevation that contribute to shoulder injury in the SCI population including
scapular motion, shoulder muscle strength, and increased load on the shoulder due to MWC
propulsion, body transfers, and repetitive motion. Loading of the shoulder although not mea-
sured in this study, has an important role in the increased pathology and pain for MWC users. This analysis does not report rotator cuff pathology and how it relates to differences in daily
humeral elevations between the cohorts. Shoulder tendon pathology from MRI is part of a
larger longitudinal study that follows rotator cuff pathology via MRI over time, and future
reports will provide meaningful information about humeral elevations and associations with
pathology progression. Supporting information S1 Appendix. Defining sensor-to-segment alignment matrices. (DOCX) S1 Appendix. Defining sensor-to-segment alignment matrices. (DOCX) S2 Appendix. Calculating humeral elevation thorax deviation angles. (DOCX) S3 Appendix. The distribution of the percentage of time the SCI and control cohort spent
in 10˚ bins for both the dominant and non-dominant arms. (TIF) S4 Appendix. Data underlying the humeral elevation plots for each participant in each
humeral elevation bin for dominant (dom) and nondominant (nondom) arms. (S) indicates
SCI participant and (C) indicates control participant. (XLSX) Conclusions This study aimed to understand the humeral elevation workspace throughout a typical day of
individuals with SCI who use a MWC and compare it to the workspace of age- and sex- 13 / 17 PLOS ONE | https://doi.org/10.1371/journal.pone.0248978
April 23, 2021 PLOS ONE Humeral elevation workspace during daily life of manual wheelchair users compared to controls matched controls. Our data suggest that individuals with SCI who use a MWC may spend
more time in a potentially risky humeral elevation range (30–60˚) than the controls. The find-
ings from this study do not support an effect of age, sex, pain, injury level, or years since injury
on the humeral elevation workspace for adults with SCI who use a MWC. Future work should
expand the understanding of loading of the upper extremity during daily life and characterize
more in-depth information about shoulder workspace and activities of daily living across
injury levels and groups with and without pain and pathology. Author Contributions Conceptualization: Brianna M. Goodwin, Stephen M. Cain, Meegan G. Van Straaten, Emma
Fortune. Formal analysis: Brianna M. Goodwin, Stephen M. Cain, Omid Jahanian, Melissa M. B. Morrow. Funding acquisition: Meegan G. Van Straaten, Emma Fortune, Melissa M. B. Morrow. Investigation: Melissa M. B. Morrow. Methodology: Brianna M. Goodwin, Stephen M. Cain, Melissa M. B. Morrow. Project administration: Brianna M. Goodwin, Meegan G. Van Straaten, Melissa M. B. Morrow. Software: Brianna M. Goodwin, Stephen M. Cain. Visualization: Brianna M. Goodwin. Visualization: Brianna M. Goodwin. Writing – original draft: Brianna M. Goodwin, Stephen M. Cain. Writing – original draft: Brianna M. Goodwin, Stephen M. Cain. Writing – review & editing: Brianna M. Goodwin, Stephen M. Cain, Meegan G. Van Straaten,
Emma Fortune, Omid Jahanian, Melissa M. B. Morrow. References M., Bruckner T., Munzinger J., et al., "Prevalence of rotator
cuff tear in paraplegic patients compared with controls," JBJS, vol. 92, pp. 23–30, 2010. https://doi.org/
10.2106/JBJS.H.01373 PMID: 20048092 12. Seitz A. L., McClure P. W., Finucane S., Boardman N. D. III, and Michener L. A., "Mechanisms of rotator
cuff tendinopathy: intrinsic, extrinsic, or both?," Clinical biomechanics, vol. 26, pp. 1–12, 2011. https://
doi.org/10.1016/j.clinbiomech.2010.08.001 PMID: 20846766 13. Carpenter J. E., Flanagan C. L., Thomopoulos S., Yian E. H., and Soslowsky L. J., "The effects of over-
use combined with intrinsic or extrinsic alterations in an animal model of rotator cuff tendinosis," The
American journal of sports medicine, vol. 26, pp. 801–807, 1998. https://doi.org/10.1177/
03635465980260061101 PMID: 9850782 14. Braman J. P., Zhao K. D., Lawrence R. L., Harrison A. K., and Ludewig P. M., "Shoulder impingement
revisited: evolution of diagnostic understanding in orthopedic surgery and physical therapy," Medical &
biological engineering & computing, vol. 52, pp. 211–219, 2014. https://doi.org/10.1007/s11517-013-
1074-1 PMID: 23572144 15. Lawrence R. L., Schlangen D. M., Schneider K. A., Schoenecker J., Senger A. L., Starr W. C., et al.,
"Effect of glenohumeral elevation on subacromial supraspinatus compression risk during simulated
reaching," Journal of Orthopaedic Research, vol. 35, pp. 2329–2337, 2017. https://doi.org/10.1002/jor. 23515 PMID: 28071815 16. Lawrence R. L., Sessions W. C., Jensen M. C., Staker J. L., Eid A., Breighner R., et al., "The effect of
glenohumeral plane of elevation on supraspinatus subacromial proximity," Journal of Biomechanics,
vol. 79, pp. 147–154, 2018. https://doi.org/10.1016/j.jbiomech.2018.08.005 PMID: 30172354 17. Giphart J. E., van der Meijden O. A., and Millett P. J., "The effects of arm elevation on the 3-dimensional
acromiohumeral distance: a biplane fluoroscopy study with normative data," Journal of shoulder and
elbow surgery, vol. 21, pp. 1593–1600, 2012. https://doi.org/10.1016/j.jse.2011.11.023 PMID: 22361718 18. Koontz A. M., Cooper R. A., Boninger M. L., Souza A. L., and Fay B. T., "Shoulder kinematics and kinet-
ics during two speeds of wheelchair propulsion," Journal of rehabilitation research and development,
vol. 39, pp. 635–650, 2002. PMID: 17943666 19. Morrow M. M., Kaufman K. R., and An K. N., "Scapula kinematics and associated impingement risk in
manual wheelchair users during propulsion and a weight relief lift," Clinical biomechanics, vol. 26, pp. 352–7, May 2011. https://doi.org/10.1016/j.clinbiomech.2010.12.001 PMID: 21216055 20. Requejo P., Mulroy S., Haubert L. References 1. Cooper R. A., Boninger M. L., and Robertson R. N., "Repetitive Strain Injury Among Manual Wheelchair
Users," Team Rehab Report, vol. 9, pp. 35–38, 1998. 2. Curtis K. A., Drysdale G. A., Lanza R. D., Kolber M., Vitolo R. S., and West R., "Shoulder pain in wheel-
chair users with tetraplegia and paraplegia," Archives of physical medicine and rehabilitation, vol. 80, pp. 453–457, 1999. https://doi.org/10.1016/s0003-9993(99)90285-x PMID: 10206610 PLOS ONE | https://doi.org/10.1371/journal.pone.0248978
April 23, 2021 14 / 17 PLOS ONE Humeral elevation workspace during daily life of manual wheelchair users compared to controls 3. Alm M., Saraste H., and Norrbrink C., "Shoulder pain in persons with thoracic spinal cord injury: preva-
lence and characteristics," Journal of rehabilitation medicine, vol. 40, pp. 277–283, 2008. https://doi. org/10.2340/16501977-0173 PMID: 18382823 4. Samuelsson K., Tropp H., and Gerdle B., "Shoulder pain and its consequences in paraplegic spinal
cord-injured, wheelchair users," Spinal cord, vol. 42, p. 41, 2004. https://doi.org/10.1038/sj.sc.3101490
PMID: 14713943 5. Dalyan M., Cardenas D., and Gerard B., "Upper extremity pain after spinal cord injury," Spinal cord, vol. 37, pp. 191–195, 1999. https://doi.org/10.1038/sj.sc.3100802 PMID: 10213328 6. Divanoglou A., Augutis M., Sveinsson T., Hultling C., and Levi R., "Self-reported health problems and
prioritized goals in community-dwelling individuals with spinal cord injury in Sweden," Journal of rehabili-
tation medicine, vol. 50, pp. 872–878, 2018. https://doi.org/10.2340/16501977-2383 PMID: 30225513 7. Dyson-Hudson T. A. and Kirshblum S. C., "Shoulder pain in chronic spinal cord injury, part 1: epidemiol-
ogy, etiology, and pathomechanics," ed: Taylor & Francis, 2004. https://doi.org/10.1080/10790268. 2004.11753724 PMID: 15156931 8. Greving K., Dorrestijn O., Winters J., Groenhof F., Van der Meer K., Stevens M., et al., "Incidence, preva-
lence, and consultation rates of shoulder complaints in general practice," Scandinavian journal of rheu-
matology, vol. 41, pp. 150–155, 2012. https://doi.org/10.3109/03009742.2011.605390 PMID: 21936616 9. Sie I. H., Waters R. L., Adkins R. H., and Gellman H., "Upper extremity pain in the postrehabilitation spi-
nal cord injured patient," Archives of physical medicine and rehabilitation, vol. 73, pp. 44–48, 1992. PMID: 1729973 10. Giner-Pascual M., Alcanyis-Alberola M., Gonza´lez L. M., Aguilar-Rodriguez M., and Querol F., "Shoul-
der pain in cases of spinal injury: influence of the position of the wheelchair seat," International Journal
of Rehabilitation Research, vol. 34, pp. 282–289, 2011. https://doi.org/10.1097/MRR. 0b013e32834a8fd9 PMID: 21971486 11. Akbar M., Balean G., Brunner M., Seyler T. References G., Tarasuk V., and Bombardier C., "Measuring the
wole or the parts?: Validity, reliability, and responsiveness of the disabilities of the arm, shoulder and
hand outcome measure in different regions of the upper extremity," Journal of Hand Therapy, vol. 14, pp. 128–142, 2001. PMID: 11382253 30. Curtis K., Roach K., Applegate E. B., Amar T., Benbow C., Genecco T., et al., "Development of the
wheelchair user’s shoulder pain index (WUSPI)," Spinal Cord, vol. 33, p. 290, 1995. 31. Curtis K., Roach K., Applegate E., Amar T., Benbow C., Genecco T., et al., "Reliability and validity of
the wheelchair user’s shoulder pain index (WUSPI)," Spinal Cord, vol. 33, p. 595, 1995. 32. Coley B., Jolles B. M., Farron A., and Aminian K., "Arm position during daily activity," Gait & Posture,
vol. 28, pp. 581–587, 2008. https://doi.org/10.1016/j.gaitpost.2008.04.014 PMID: 18547809 33. Amasay T., Zodrow K., Kincl L., Hess J., and Karduna A., "Validation of tri-axial accelerometer for the cal-
culation of elevation angles," International Journal of Industrial Ergonomics, vol. 39, pp. 783–789, 2009. 34. Amasay T., Latteri M., and Karduna A. R., "In vivo measurement of humeral elevation angles and expo-
sure using a triaxial accelerometer," Human factors, vol. 52, pp. 616–626, 2010. https://doi.org/10. 1177/0018720810386951 PMID: 21284365 35. Savage P. G., "Strapdown inertial navigation integration algorithm design part 1: Attitude algorithms,"
Journal of guidance, control, and dynamics, vol. 21, pp. 19–28, 1998. 36. Sabatini A. M., "Quaternion-based extended Kalman filter for determining orientation by inertial and
magnetic sensing," IEEE transactions on Biomedical Engineering, vol. 53, pp. 1346–1356, 2006. https://doi.org/10.1109/TBME.2006.875664 PMID: 16830938 37. Madgwick S., "An efficient orientation filter for inertial and inertial/magnetic sensor arrays," Report x-io
and University of Bristol (UK), vol. 25, pp. 113–118, 2010. 38. McGinnis R. S., Cain S. M., Tao S., Whiteside D., Goulet G. C., Gardner E. C., et al., "Accuracy of
femur angles estimated by IMUs during clinical procedures used to diagnose femoroacetabular
impingement," IEEE Transactions on Biomedical Engineering, vol. 62, pp. 1503–1513, 2015. https://
doi.org/10.1109/TBME.2015.2392758 PMID: 25608299 39. R. S. McGinnis, S. M. Cain, S. P. Davidson, R. V. Vitali, S. G. McLean, and N. Perkins, "Validation of
complementary filter based IMU data fusion for tracking torso angle and rifle orientation," in ASME Inter-
national Mechanical Engineering Congress and Exposition, 2014, p. V003T03A052. 40. Coley B., Jolles B. References L., Newsam C., Gronley J., and Perry J., "Evidence-based strategies
to preserve shoulder function in manual wheelchair users with spinal cord injury," Topics in Spinal Cord
Injury Rehabilitation, vol. 13, pp. 86–119, 2008. 21. El-Gohary M. and McNames J., "Shoulder and elbow joint angle tracking with inertial sensors," IEEE
Transactions on Biomedical Engineering, vol. 59, pp. 2635–2641, 2012. https://doi.org/10.1109/TBME. 2012.2208750 PMID: 22911538 PLOS ONE | https://doi.org/10.1371/journal.pone.0248978
April 23, 2021 15 / 17 PLOS ONE Humeral elevation workspace during daily life of manual wheelchair users compared to controls 22. Bouvier B., Duprey S., Claudon L., Dumas R., and Savescu A., "Upper Limb Kinematics Using Inertial
and Magnetic Sensors: Comparison of Sensor-to-Segment Calibrations," Sensors, vol. 15, pp. 18813–
33, 2015. https://doi.org/10.3390/s150818813 PMID: 26263993 23. Cutti A. G., Giovanardi A., Rocchi L., Davalli A., and Sacchetti R., "Ambulatory measurement of shoul-
der and elbow kinematics through inertial and magnetic sensors," Med Biol Eng Comput, vol. 46, pp. 169–78, Feb 2008. https://doi.org/10.1007/s11517-007-0296-5 PMID: 18087742 24. Kirking B., El-Gohary M., and Kwon Y., "The feasibility of shoulder motion tracking during activities of
daily living using inertial measurement units," Gait & posture, vol. 49, pp. 47–53, 2016. https://doi.org/
10.1016/j.gaitpost.2016.06.008 PMID: 27371783 25. Chapman R. M., Torchia M. T., Bell J.-E., and Van Citters D. W., "Continuously monitoring shoulder
motion after total shoulder arthroplasty: maximum elevation and time spent above 90˚ of elevation are
critical metrics to monitor," Journal of shoulder and elbow surgery, 2019. https://doi.org/10.1016/j.jse. 2019.01.003 PMID: 30956145 26. Langohr G. D. G., Haverstock J. P., Johnson J. A., and Athwal G. S., "Comparing daily shoulder motion
and frequency after anatomic and reverse shoulder arthroplasty," Journal of shoulder and elbow sur-
gery, vol. 27, pp. 325–332, 2018. https://doi.org/10.1016/j.jse.2017.09.023 PMID: 29133073 27. Lawrence R. L., Braman J. P., and Ludewig P. M., "Shoulder kinematics impact subacromial proximi-
ties: a review of the literature," Braz J Phys Ther, vol. 24, pp. 219–230, May—Jun 2020. https://doi.org/
10.1016/j.bjpt.2019.07.009 PMID: 31377124 28. Hudak P. L., Amadio P. C., Bombardier C., Beaton D., Cole D., Davis A., et al., "Development of an
upper extremity outcome measure: the DASH (disabilities of the arm, shoulder, and head)," American
journal of industrial medicine, vol. 29, pp. 602–608, 1996. https://doi.org/10.1002/(SICI)1097-0274
(199606)29:6<602::AID-AJIM4>3.0.CO;2-L PMID: 8773720 29. Beaton D. E., Katz J. N., Fossel A. H., Wright J. References M., Farron A., and Aminian K., "Detection of the movement of the humerus during
daily activity," Medical & biological engineering & computing, vol. 47, pp. 467–474, 2009. https://doi. org/10.1007/s11517-009-0464-x PMID: 19277750 41. Granzow R. F., Schall M. C. Jr, Smidt M. F., Chen H., Fethke N. B., and Huangfu R., "Characterizing
exposure to physical risk factors among reforestation hand planters in the Southeastern United States,"
Applied Ergonomics, vol. 66, pp. 1–8, 2018. https://doi.org/10.1016/j.apergo.2017.07.013 PMID:
28958420 PLOS ONE | https://doi.org/10.1371/journal.pone.0248978
April 23, 2021 16 / 17 PLOS ONE Humeral elevation workspace during daily life of manual wheelchair users compared to controls 42. Schall M. C. Jr, Fethke N. B., and Chen H., "Working postures and physical activity among registered
nurses," Applied ergonomics, vol. 54, pp. 243–250, 2016. https://doi.org/10.1016/j.apergo.2016.01. 008 PMID: 26851483 43. Chapman R. M., Torchia M. T., Bell J.-E., and Van Citters D. W., "Assessing shoulder biomechanics of
healthy elderly individuals during activities of daily living using inertial measurement units: high maxi-
mum elevation is achievable but rarely used," Journal of biomechanical engineering, vol. 141, p. 041001, 2019. https://doi.org/10.1115/1.4042433 PMID: 30758509 44. An K. N., Browne A., Korinek S., Tanaka S., and Morrey B., "Three-dimensional kinematics of gleno-
humeral elevation," Journal of Orthopaedic Research, vol. 9, pp. 143–149, 1991. https://doi.org/10. 1002/jor.1100090117 PMID: 1984044 45. T. Seel, T. Schauer, and J. Raisch, "Joint axis and position estimation from inertial measurement data
by exploiting kinematic constraints," in 2012 IEEE International Conference on Control Applications,
2012, pp. 45–49. 46. Vitali R., Cain S., McGinnis R., Zaferiou A., Ojeda L., Davidson S., et al., "Method for estimating three-
dimensional knee rotations using two inertial measurement units: Validation with a coordinate measure-
ment machine," Sensors, vol. 17, p. 1970, 2017. https://doi.org/10.3390/s17091970 PMID: 28846613 47. Kessel L. and Watson M., "The painful arc syndrome. Clinical classification as a guide to management,"
The Journal of bone and joint surgery. British volume, vol. 59, pp. 166–172, 1977. https://doi.org/10. 1302/0301-620X.59B2.873977 PMID: 873977 48. McAtamney L. and Corlett E. N., "RULA: a survey method for the investigation of work-related upper
limb disorders," Applied ergonomics, vol. 24, pp. 91–99, 1993. https://doi.org/10.1016/0003-6870(93)
90080-s PMID: 15676903 49. Lugade V., Fortune E., Morrow M., and Kaufman K., "Validity of using tri-axial accelerometers to mea-
sure human movement—Part I: Posture and movement detection," Med Eng Phys, vol. 36, pp. 169–
76, Feb 2014. https://doi.org/10.1016/j.medengphy.2013.06.005 PMID: 23899533 50. Giphart J. E., van der Meijden O. References A., and Millett P. J., "The effects of arm elevation on the 3-dimensional
acromiohumeral distance: a biplane fluoroscopy study with normative data," J Shoulder Elbow Surg,
vol. 21, pp. 1593–600, Nov 2012. https://doi.org/10.1016/j.jse.2011.11.023 PMID: 22361718 51. Lawrence R. L., Schlangen D. M., Schneider K. A., Schoenecker J., Senger A. L., Starr W. C., et al.,
"Effect of glenohumeral elevation on subacromial supraspinatus compression risk during simulated
reaching," J Orthop Res, vol. 35, pp. 2329–2337, Oct 2017. https://doi.org/10.1002/jor.23515 PMID:
28071815 52. Goodwin B. M., Fortune E., Van Straaten M. G., and Morrow M. M., "Outcome Measures of Free-Living
Activity in Spinal Cord Injury Rehabilitation," Current Physical Medicine and Rehabilitation Reports, pp. 1–6, 2019. https://doi.org/10.1007/s40141-019-00228-5 PMID: 31406630 53. Sonenblum S. E., Sprigle S., and Lopez R. A., "Manual wheelchair use: bouts of mobility in everyday
life," Rehabilitation research and practice, vol. 2012, 2012. 54. Tolerico M. L., Ding D., Cooper R. A., and Spaeth D. M., "Assessing mobility characteristics and activity
levels of manual wheelchair users," Journal of rehabilitation research and development, vol. 44, p. 561,
2007. https://doi.org/10.1682/jrrd.2006.02.0017 PMID: 18247253 55. Collinger J. L., Boninger M. L., Koontz A. M., Price R., Sisto S. A., Tolerico M. L., et al., "Shoulder bio-
mechanics during the push phase of wheelchair propulsion: a multisite study of persons with paraple-
gia," Archives of physical medicine and rehabilitation, vol. 89, pp. 667–676, 2008. https://doi.org/10. 1016/j.apmr.2007.09.052 PMID: 18373997 56. Rao S. S., Bontrager E. L., Gronley J. K., Newsam C. J., and Perry J., "Three-dimensional kinematics of
wheelchair propulsion," IEEE Transactions on Rehabilitation Engineering., vol. 4, pp. 152–60, 1996. https://doi.org/10.1109/86.536770 PMID: 8800218 57. Collins S. H., Adamczyk P. G., and Kuo A. D., "Dynamic arm swinging in human walking," Proceedings
of the Royal Society B: Biological Sciences, vol. 276, pp. 3679–3688, 2009. https://doi.org/10.1098/
rspb.2009.0664 PMID: 19640879 58. Fortune E., Cloud-Biebl B. A., Madansingh S. I., Ngufor C. G., Van Straaten M. G., Goodwin B. M.,
et al., "Estimation of manual wheelchair-based activities in the free-living environment using a neural
network model with inertial body-worn sensors," Journal of electromyography and kinesiology, 2019. https://doi.org/10.1016/j.jelekin.2019.07.007 PMID: 31353200 59. Schneider S., Popp W. L., Brogioli M., Albisser U., Demko´ L., Debecker I., et al., "Reliability of wear-
able-sensor-derived measures of physical activity in wheelchair-dependent spinal cord injured
patients," Frontiers in neurology, vol. 9, p. 1039, 2018. https://doi.org/10.3389/fneur.2018.01039 PMID:
30619026 17 / 17 PLOS ONE | https://doi.org/10.1371/journal.pone.0248978
April 23, 2021
|
W4392601021.txt
|
https://bmcpublichealth.biomedcentral.com/counter/pdf/10.1186/s12889-024-18178-6
|
en
|
Boundary violations and university teachers’ well-being during mandatory telework: Recovery’s role and gender differences
|
BMC public health
| 2,024
|
cc-by
| 6,960
|
Mascarenhas et al. BMC Public Health
(2024) 24:747
https://doi.org/10.1186/s12889-024-18178-6
BMC Public Health
Open Access
RESEARCH
Boundary violations and university teachers’
well-being during mandatory telework:
Recovery’s role and gender differences
Madalena Mascarenhas1*, Vânia Sofia Carvalho1*, Cleide Fátima Moretto2 and Maria José Chambel1
Abstract
Background This study aimed to explore the role of psychological detachment from work in the relationship of
boundary violations and flourishing, as well as gender differences among university teachers during mandatory
telework. We developed and tested a moderate mediation model where psychological detachment was the
explanatory mechanism of the relationship between boundary violations with flourishing and using gender as the
moderating variable.
Methods A cross-sectional study was conducted with a sample of 921 Brazilian university teachers (mean age 44
years, 681 women and 240 men) during mandatory telework. Multigroup analysis and moderate mediation were
performed using Mplus 7.2.
Results Psychological detachment mediated the relationship between boundary violations (in both directions) and
flourishing and work-to-family violations were more harmful to women’ recovery instead family-to-work violations
were more harmful to men’ recovery, among university teachers during mandatory telework.
Conclusion By focusing on boundary violations in the context of mandatory telework, the study sheds light on the
impact of blurred boundaries between work and personal life. This contributes both literature on work-life balance
and literature recovery. Moreover, it helps to understand a crisis setting of remote work. Further, the study’s findings
regarding gender differences highlight how men and women may experience and cope with boundary violations
differently during mandatory telework, supporting future specific interventions across genders.
Keywords Recovery, Psychological detachment, Boundary violations, Well-being, Flourishing, Gender equality,
Telework, Work-family interface
*Correspondence:
Madalena Mascarenhas
madalenamascarenhas@edu.ulisboa.pt
Vânia Sofia Carvalho
vscarvalho@psicologia.ulisboa.pt
1
Faculdade de Psicologia, CICPSI, Universidade de Lisboa, Alameda da
Universidade, 1649-013 Lisboa, Portugal
2
Universidade de Passo Fundo, 99052-900 Passo Fundo, Brasil
© The Author(s) 2024. Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use,
sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and
the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this
article are included in the article’s Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included
in the article’s Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will
need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by/4.0/. The
Creative Commons Public Domain Dedication waiver (http://creativecommons.org/publicdomain/zero/1.0/) applies to the data made available
in this article, unless otherwise stated in a credit line to the data.
Mascarenhas et al. BMC Public Health
(2024) 24:747
Background
Telework is a flexible form of work that promotes the
articulation between work and family, its imposition
caused by the COVID-19 pandemic, has had negative
consequences on this interaction and on well-being [1–
7]. Although university teachers already carried out some
tasks from home (e.g., preparing lessons or correcting
tests), the lockdown forced them to carry out all educational activities remotely (e.g., lessons and meetings
with colleagues or parents), without having the time and
resources to prepare to do it [8, 9]. It was a major challenge that required effort from these professionals, as
teaching in a distant format requires a range of information technology and different pedagogical skills that university teachers may not have (e.g., tech proficiency; to
provide feedback in an online environment; active learning activities) [10]. It is also important to mention that
the social distancing during the lockdown had a negative
psychosocial impact, namely a reduction in opportunities
for leisure activities [11].
According to boundary theory [12, 13] there are
boundaries that separate different domains (e.g., work
and family). Telework makes these boundaries more
permeable and transitions (i.e., sudden role changes)
are more frequent [12, 14–17]. Boundary violations are
a specific form of transition, and it are characterized by
the unwanted intrusion of one domain into the other,
which consumes resources (cognitive and temporal), during or after the interruption [18, 19]. These violations
can occur from work-to-family (e.g., answer a phone call
from the boss after shift work) or family-to-work (e.g.,
help a child perform schoolwork during shift work) and
they increased exponentially in the context of mandatory telework during COVID-19 with a negative impact
on the well-being [12, 20–23]. The mandatory lockdown
did not affect men and women equally: women lost more
jobs, reduced, or fragmented their working hours and
had more disruptions due to an increase in childcare and
other domestic responsibilities (e.g., housework or eldercare) [24–27].
It is known that recovery, in special the psychological
detachment from work, was greatly impaired by telework [28], and boundary violations were one of the most
harmful aspects, as they decreased the time and availability required for recovery. All workers need to recover,
which means, to replenish their resources, by psychologically distancing themselves from work (i.e., avoiding work-related thoughts or behaviours, and engaging
in leisure activities) [29–32]. For recovery to occur, it is
necessary to ensure that people do not engage in workrelated activities in their free time [33]. According to the
stressor-detachment model [34], work stressors do not
allow the employees to distance from work when they are
not working, and this happens because stressors trigger
Page 2 of 9
a highly negative response (e.g., ruminant thoughts or
actions to decrease stressor) which don’t allow the person to psychological detach from the stressor. Boundary
violations can be stressors because they are unscheduled interruptions that incur costs for the person, and
they decrease the chances of recovery (either because
they take time or because they keep work/family mentally present) [17, 35, 36]. For this reason, we intend to
understand how boundary violations compromised the
recovery of university teachers in lockdown, especially
because university teachers have great difficulty in recovering [37, 38]. Furthermore, psychological detachment
from work is important for well-being and flourishing
[39, 40]. Flourishing is an indicator of subjective wellbeing and it is characterized by a positive emotional
state, a high satisfaction with life and it is the major
manifestation of mental health [41–43]. Boundary violations are harmful to flourishing because they imply the
consumption of resources necessary for flourishing and
cause negative emotions [20, 44–46]. We may understand
why psychological detachment has this positive effect,
as it is a protective factor against the negative effects of
work stressors and decreases ruminant thoughts about
work [34, 39], and, according to the stressor-detachment
model, psychological detachment from work mediates
the relationship between work stressors and well-being
[34].
Thus, in the present study, it is expected to find a mediation whereas boundary violations (in both directions)
will inhibit distancing from work, which, in turn, will
hinder the flourishing of telecommuting university teachers during the period of confinement due to COVID-19.
Going beyond, we expect that the relationship between
boundary violations and psychological detachment may
be different between men and women. According to
role theory [47], a person devotes more resources to the
role that is most relevant to him or her, investing more
effort and resources in maintaining the boundaries of
that domain [47, 48]. Specifically, the family role is more
salient in highly feminine people (i.e., the traits often
associated with the female gender) and is tendentially
more valued by women, while men tend to value the work
role more [47, 48]. The fact that women have more family responsibilities and housework than men means that
women should have more family-to-work violations [49]
and telework women should have less effective recovery
than men [50]. During mandatory telework, women had
to provide more family support (e.g., elderly care), which
led them to perceive more stress and was detrimental to
their work [24, 25, 51]. As work plays a central role for
men, they tend to develop a series of strategies to protect
this domain, preventing the family from invading and
damaging it [52].Although the impact on men’s work has
not been as great as the impact on women’s work, many
Mascarenhas et al. BMC Public Health
(2024) 24:747
fathers have taken on more childcare responsibilities and
there have also been more opportunities for work interruptions as the whole family has been at home [50]. It
is psychologically damaging when the role that is most
important to the person is threatened [53] (e.g., violated
from another domain), which leads us expect that violations of work in the family are more detrimental to women’s recovery and that violations of family at work are
more detrimental to men’s recovery.
According to the stressor-detachment model [34], work
stressors impair psychological detachment, which, in
turn mediates the relationship between work stressors
and well-being. According to role theory [47], a person
invests more resources in the role that is most relevant
to his/her gender (i.e., the family role tends to be valued
more by women, whereas men tend to value the work
role) [47, 48]. Boundary violations interfere with the psychological detachment, which, in turn, is fundamental to
explain flourishing [39, 40] and this may be different for
men and women. This led us to the final research model,
represented in Fig. 1.
Thus, we propose a model of gender-moderated mediation of psychological detachment in the relationship
between boundary violations and flourishing, and suggest
the following study hypotheses:
H1a Violations form work-to-family are negatively associated with recovery, namely, psychological detachment
from work.
H1b Violations from family-to-work are negatively associated with recovery, namely, psychological detachment
from work.
H2a The relationship between work-to-family violations
and flourishing is mediated by recovery, namely, psychological detachment from work.
Fig. 1 Representation of the research model
Page 3 of 9
H2b The relationship between family-to-work violations
and flourishing is mediated by recovery, namely, psychological detachment from work.
H3a The relationship between violations of work-to-family boundaries and recovery is moderated by gender, with
this relationship being stronger in women than in men.
H3b The relationship between violations of family-work
boundaries and recovery is moderated by gender, with
this relationship being stronger in men than in women.
Methods
Data
Data were collected from a sample of university teachers in June 2020 in the state of Rio Grande do Sul, Brazil,
during the period of mandatory confinement imposed
by COVID-19. After obtaining the approval of the ethics
committees of the Faculty of Passo Fundo (Brazil) (May
28th, 2020) and of the Faculty of Psychology, University
of Lisbon (Portugal) (April 16th, 2020), the study was
subsequently presented to the Union of Private Education Teachers of Rio Grande do Sul– Sinpro/RS, which
randomly distributed the survey among 1500 members.
Of these, 921 completed the questionnaires, for a 61.4%
return rate. The distribution of the study and survey
was conducted by email and consisted of the following
components: explanation of the purpose of the study;
guarantee of the anonymity of the participants; access
to a link to the questionnaire through the Survey Monkey platform; and the availability of the results obtained.
Informed consent was obtained in writing on the first
page of the questionnaire. Participants who agreed to
participate in the study proceeded to answer the questionnaire. Each participant was assigned an individual
code to be entered when filling out the questionnaire,
which would later allow them to obtain their results in
terms of measures of stress and malaise. These individual
results were included in the overall results.
Mascarenhas et al. BMC Public Health
(2024) 24:747
Page 4 of 9
Participants
Gender
The participants were Brazilian university professors
who were in mandatory telecommuting. The mean age
was 44 years, and most participants were female (73.9%),
married (72.3%) and had children (65.7%). A significant
percentage had spouses who were also telecommuting and had children under 10 years of age (43.7% and
45.2%, respectively). None of the participants had previous experience with telecommuting, but the sample had
a mean of 13.5 years of experience as teachers. More
than half worked fulltime (53.5%), and the majority had
an open-ended contract (86.6%) and did not perform any
management or coordination role (80.1%).
Participants were asked to indicate the gender with which
they identified (0 = male; 1 = female).
Instruments
The scales were adapted to Brazilian Portuguese by one
of the authors and then verified in a pretest that was
applied to 10 Brazilian university teachers. All scales are
5-point Likert scales ranging from “strongly disagree” (1)
to “strongly agree” (5).
Boundary violations
Boundary violations were assessed using the scale of
Hunter et al. [54], the original scale has an α = 0.95 in
both directions (i.e., violations from work-to-family and
from family-to-work), which, although not yet validated
for the Portuguese population, has been previously used
in a previous study in Portugal, with an α = 0.95 in violations from work-to-family and an α = 0.91 in violations
from family-to-work α = 0.84 [20]. The scale allows for
the evaluation of violations of borders in both directions (i.e., family-to-work and work-to-family), with each
direction being evaluated through 3 items for family-towork direction with an α = 0.79 (e.g., “A family member
has interrupted my work more than I wanted”), and workto-family direction, with an α = 0.88 (e.g., “My work has
interrupted my personal/family life more than I wanted
to”).
Psychological detachment from work
Psychological detachment from work was assessed with
3 items (e.g., “I have had times when I distanced myself
from work”) with an α = 0.90, from the adaptation to Portuguese with an α = 0.76 [55] of the Recovery Experiences
Questionnaire with an α = 0.90 [31].
Flourishing
Flourishing was assessed using the Portuguese adaptation by Silva and Caetano, with an α = 0.83 [56] of the
flourishing scale of Diener et al., with an α = 0.80 [57]. For
the present study, 8 items were used that evaluated the
self-perception of success in relevant areas of the subject’s life (e.g., “I lead a life with purpose and meaning”),
with an α = 0.89.
Control variables
The age of the children was used as a control variable, as
it has an impact on adaptation to mandatory telecommuting (1 = under 1 year old; 2 = between 1 and 5 years
old; 3 = between 6 and 10 years old; 4 = between 11 and
15 years old; 5 = between 15 and 18 years old; 6 = over 18
years old).
Data analysis
Data were analyzed with SPSS 20.0 (SPSS, 2011) and
Mplus 7.0 [58]. First, we performed descriptive variable
and correlation analyses with SPSS. We used Confirmatory Factor Analysis to evaluate the latent structure of
the measures included in the study by Mplus 7.0. We follow the recommendation of Podsakoff et al. [59] to test
the common error variance method, applying Harman’s
single-factor test. Then, the best-fit model was used to
evaluate the mediation model and invariance across genders. To test Hypotheses 1 and 2, that is, the mediation
model, we compared models based on chi-square difference tests and on other fit indices: the standardized root
mean square (SRMR), the incremental fit index (IFI), the
Bentler comparative fit index (CFI) and the root mean
square error of approximation (RMSEA). Regarding CFI
and IFI, values greater than 0.90 represent a good model
fit, and for SRMR and RMSEA, values less than 0.07 indicate a good model fit.
To test Hypotheses 3, we performed MULTIGROUP
analysis. The following steps were conducted: (1) an
unconstrained multiple-group model across gender, with
all free parameters (baseline model), was investigated in
the first step; (2) a constrained multiple-group model,
where all structural paths were constrained to be equal
across groups; and (3) semirestricted models that constrained the strength of different paths successively to be
equal in view of inspecting invariance across the subsamples and compared these with the constraint-free models through the chi-square differences. These analyses
were conducted using maximum likelihood ratio (MLR).
Referring to the asymmetrical distribution between men
and women, the MLR estimator was used (maximum
likelihood parameter estimates with standard errors and
a chi-square test statistic that are robust to nonnormality). Thus, the comparison between two nested models
was tested through the significance of the difference in
the chi-square value using the MLR estimator and the
Satorra–Bentler scaled (mean-adjusted) chi-square.
Mascarenhas et al. BMC Public Health
(2024) 24:747
Page 5 of 9
Table 1 Means, correlations and standard deviations, n = 921
1
2
3
4
5
6
VWF
VFW
PDFW
Flourishing
Gender
AYS
Mean
3.13
2.6
2.8
3.9
DP
1.17
1.03
1
0.66
3.9
1.7
Skewness
− 0.25
0.33
− 0.2
− 0.85
2.1
0.80
Kurtosis
− 0.91
− 0.53
− 0.92
1.8
9.71
2.3
1
2
3
4
5
0.63**
− 0.48**
− 0.31**
− 0.07*
− 0.21**
0.28**
− 0.26**
− 0.01
− 0.29**
0.26**
0.12**
0.17**
− 0.09**
0.14**
− 0.02
Notes: ** p <.01, * p <.05, VWF = violations from work to family; VFW = violations from family to work; WFB = work–family balance; PDFW = psychological detachment
from work; Gender = 1 = male, 2 = female: AYS = age younger son
Table 2 Model fit indices for multigroup structural equation modeling
Models
Baseline model
Constrained multiple-group
Semi-restricted model 1a
Semi-restricted model 1b
χ2(df)
767.41, p <.01 (302)
785,019, p <.01 (315)
773,180, p <.01 (303)
771,567, p <.01 (303)
CFI
0.94
0.94
0.94
0.94
TLI
0.93
0.93
0.93
0.93
RMSEA
0.07
0.06
0.06
0.06
SRMR
0.06
0.06
0.06
0.07
Δχ 2(df)
15.48, p >.05 (13)1
6.87, p <.01 (1)2
6.13, p <.01 (1)3
Notes: baseline model = unrestricted model across gender; constrained multi-goup = a model with all structural paths were constrained to be equal across men and
women; Semi-restricted model 1a = model with the relationship between violations from work to family and psychological detachment from work constrained;
Semi-restricted model 1b = model with the relationship between family-to-work violations and psychological detachment from work constrained; df = degrees of
freedom, CFI = Comparative Fit Index, TLI = Tucker–Lewis Index, RMSEA = Root Mean Square Error Of Approximation, SRMR = standardized Root Mean Square Residual;
Δχ 2 = χ2—difference tests; 1difference between the baseline and fully constrained models; 2difference between the baseline model and the semi-restricted Model
1a; 3difference between the baseline model and the semi-restricted Model 1b
Results
The measurement model was constructed, including
the four distinct latent constructs (work to family, family to work violations, psychological detachment from
work and flourishing). The fit indices for this model
showed very good fit, χ 2 (112) = 544.57, p <.01; CFI = 96;
TLI = 0.95; SRMR = 0.05; RMSEA = 0.07. Then, a onefactor model was performed. The one-factor model
presented poor fit to the data χ 2 (118) = 4607.89, p <.01;
IFC = 54; TLI = 0.47; SRMR = 0.15; RMSEA = 0.20. Thus,
our theoretical model was better than the one-factor
model, Δχ 2 (6) = 4063.32, p <.01, and confirmed the construct validity of the measurement model.
from work, was also supported since this relationship
was negative and significant (β= − 0.26, p <.01).
H2 stated that psychological detachment from work
was the mediator between work-family violations and
flourishing. First, was observed that the relationship
between psychological detachment from work and flourishing was positive and significant (β = 0.27, p <.01). Second, both indirect effects were significant (violations
from work to family (b = − 0.10, p <.01); violations from
family to work (b = − 0.07, p <.01). However, the relationship between work-family violations and flourishing was
nonsignificant (β= − 0.03, p >.01). Taken together, H2a
was supported, and H2b was partially supported because
the direct relationship between violations from family to
work and flourishing was significant with the presence of
a mediator (psychological detachment from work), showing that it was a partial mediation.
Regarding the control variable, the son’s age has a significant effect on psychological detachment from work
(b = 0.19, p <.05).
Structural equation model
Multigroup Structural Equation Modeling (MGSEM)
To test Hypotheses 1 and 2, we performed a structural
equation model in MPLUS. This model presented a very
good fit (χ 2 (138) = 614.57, p <.01; CFI = 95; TLI = 0.94;
SRMR = 0.05; RMSEA = 0.06). Regarding H1a, where we
postulated that violations from work to the family are
negatively associated with psychological detachment
from work, the data indicate that the relationship was
negative and significant (β= − 0.38, p <.01); thus, H1a was
supported. H1b, where violations from family to work
are negatively associated with psychological detachment
The fit indices for the unrestricted (baseline) model
indicated a good fit (χ 2 (302) = 767.41, p <.01; CFI = 94;
TLI = 0.93; SRMR = 0.06; RMSEA = 0.07) and are
described in Table 2. Following step 2, we tested the
fully constrained model, and the fit indices are χ 2
(315) = 785,019, p <.01; CFI = 94; TLI = 0.93; SRMR = 0.06;
RMSEA = 0.07. The chi-square difference between
the unconstrained and fully constrained models was
nonsignificant Δχ 2 (13) = 15.48, p >.05. Despite this
Descriptive variables and correlations
By means of a correlation matrix (r), the variables were
observed to correlate with each other (p <.001) and
are described in Table 1. All variables included in our
hypotheses are correlated in the expected direction.
Confirmatory factor analysis
Mascarenhas et al. BMC Public Health
(2024) 24:747
nonsignificant difference, we pursued testing potential
differences in the different paths.
The semirestricted Model 1a, where the relationship
between violations from work to family and psychological detachment from work was constrained, presents a
good fit (χ 2 (303) = 773,180, p <.01; CFI = 94; TLI = 0.93;
SRMR = 0.06; RMSEA = 0.06). The difference between
the unrestricted model and the semirestricted Model 1a
was Δχ 2 (1) = 6.87, p <.01 and confirmed the difference
between men and women in this path. The semirestricted
Model 1b, where the relationship between family-towork violations and psychological detachment from work
was constrained, presents a good fit (χ 2 (303) = 771,567,
p <.01; CFI = 94; TLI = 0.93; SRMR = 0.07; RMSEA = 0.06).
The difference between the unrestricted model and the
semirestricted Model 1b was Δχ 2 (1) = 6.13, p <.01 and
confirmed the difference between men and women in
this path.
Overall, male and female differences were revealed in
the relationship between work-to-family and familyto-work violations and psychological detachment from
work. Specifically, violations from work to family presented a negative and significant relationship with psychological detachment from work for women (β= − 0.43,
p <.01), while for men, this relationship was nonsignificant (β = − 0.02, p >.01). Furthermore, despite violations
from family to work present a negative and significant
relationship for women (β= − 0.20, p <.01) for men this
relationship is stronger (β = − 0.62, p <.05). Thus, H3a and
H3b were supported.
Regarding control variables, we observe that son’s age
maintains a negative and significant effect (β = − 0.25,
p <.01) on women’s psychological detachment, but for
men, the effect was not significant (β = 0.00).
Discussion
Based on the boundary theory [12, 13], it was observed,
as expected, that boundary violations impair recovery,
and that recovery mediates the relationship between violations and flourishing. Based on role theory [47], was
also observed that violations from work-to-family hinder
recovery more for women compared to men, and violations from family-to-work affect recovery more intensely
for men.
As expected, our study indicates that boundary violations are an impediment to the recovery of university
teachers who were telecommuting in mandatory confinement during COVID-19. This is an important fact, as it
can be an example of the additional difficulties that arise
in situations of health crisis. When university teachers
had their professional activities interrupted by the family, it affected their recovery (e.g., they had to help their
children with a school activity, they probably had to
extend their working hours to complete all their tasks)
Page 6 of 9
[38]. Similarly, when university teachers saw their family
lives disturbed by work (e.g., interrupting a family task to
answer a call from a colleague) they probably kept thinking about work, which hindered their recovery or ability
to psychologically distance themselves from work.
Thus, boundary violations can be considered to impair
psychological detachment, which is most likely because
they keep work mentally present or because they reduce
the time allocated to leisure and rest activities that allow
recovery from the expenditure of resources [22, 32, 33].
Other studies have shown that university teachers tend
to keep their work psychologically present and to work
outside working hours, thus presenting great difficulty
in recovering [37]. As expected, it was also found that
boundary violations are stressors that affected the wellbeing of telecommuting university teachers during mandatory confinement [33, 45, 46]. More interestingly, the
results of this study also supported the assumption that
the absence of recovery was the explanatory mechanism of the effect of boundary violations on well-being,
because they consume resources and do not allow teleworkers to recover, which does not allow for the replacement of resources and negatively affect well-being [34].
However, we found differences according to the direction of the violation: in work-to-family violations, the
relationship with flourishing is fully explained by recovery, but in family-to-work violations, in addition to this
mediating effect of recovery, there is a direct relationship with flourishing. This difference may have occurred
because the data collection occurred during confinement, a period in which the whole family was forced
to stay at home, making this the stage for school, work
and play activities, increasing stress, namely boundaries
violations, especially involving the family at work [7, 22,
48]. Thus, university teachers felt that their professional
activity and the quality of their teaching were affected
(e.g., when they established a more distant relationship
with students, had less ability to prepare classes or lacked
mastery in the use of technological tools), and they felt
less involvement and flourishing [41, 42].
Based on role theory [47], we observed in university
teachers that violations of work in the family are more
harmful to the recovery of female teachers than to male.
When the family role of women is violated, they tend to
invest more in family life to mitigate the negative effects
of this violation, which does not allow them time to
recover. These results agree with Hartig et al. [50], who
emphasized that women have greater difficulty recovering when they work in the space dedicated to the family, because women attach more importance to their
family role. Thus, the fact that women mostly assumed
the family role increased the number of violations from
work-to-family, which adversely affected their recovery
[20, 22]. The weaker relationship for men results from the
Mascarenhas et al. BMC Public Health
(2024) 24:747
centrality of work role [47] and people who have work as
a central role do not feel the effects of work-family violations as much [52].
Additionally, as expected, our results showed that family-work violations impair recovery, especially in the case
of male university teachers. People tend to develop strategies to protect the domain that is most relevant to them
[13], thus men try to prevent the family from interfering
and damaging their work role. However, during mandatory telework, family violations at work were very frequent and people had been great difficulty in maintaining
their work preservation strategies and great difficulty to
disconnect from work, particularly those that have work
as their central role.
Our study has some limitations that should be considered. First, the sample collection was performed during
the first mandatory confinement, which facilitated the
occurrence of boundaries violations. Thus, these data
should be interpreted with caution. Despite that, data
research conducted during the COVID-19 pandemic
serves as a valuable resource for informing future crisis
preparedness, response, and recovery efforts. By leveraging the implications of this study namely the effects of
boundary violations on psychological detachment and
well-being, the global community can better prepare for
and mitigate the impact of future crises. Second, there
was no consideration of other non-binary options, which
may have had an impact on some participants’ response
options. Future studies should include this option. Third,
because this was a cross-sectional study, we were unable
to assess causal relationships among the variables. Therefore, we recommend conducting a longitudinal study to
assess causal relationships. Fourth, information on the
composition of the original sample was not available to
us. Fifth, we don’t know if the data was collected during the evaluation period. Future studies with university
professors should take this into account. Sixth, the study
uses a sample consisting of people of the same nationality, so it will be relevant to explore samples from other
countries and cultures, which may differ from this sample
in gender issues and the relationship between work and
family. At last, it is also important to note that the participants’ technological or pedagogical skills to perform
remotely tasks were not controlled. This variable could
influence the relationship between variables included in
this research. Thus, for future studies involving teleworkers and, specifically, university teachers, we recommend
collecting participants’ technological and pedagogical
skills. Future investigations should also test this model in
other professions, as well as in no mandatory telework, to
compare the results.
Although this study was conducted during mandatory
confinement, our results suggest ways in which telecommuting should be managed to maintain or promote the
Page 7 of 9
well-being of teleworkers. Considering that telework is
a form of work that will prevail in the future and that it
facilitates violations of borders, teleworkers should use
strategies that prevent the occurrence of violations, such
as turning off notifications from work devices during the
time dedicated to the family, limiting the period during
work hours in which they can be contacted, or after the
occurrence of a violation, communicate it immediately to
prevent repetition [18]. To psychologically detachment
themselves from work, teleworkers may use strategies
such as complying with work hours, seeking to have an
exclusive physical space for work, using breaks as a time
to disconnect and seeking to define a routine that allows
them to mentally disconnect from work, such as changing clothes and going for a walk. In addition, this study
highlighted the differences between genders in the repercussions of border violations on the recovery of teleworkers. Considering that telecommuting can intensify the
differences between men and women in the work-family
relationship, it is important that organizations implement strategies that promote gender equality. Finally, we
emphasize the social need for a more equitable distribution of family support between genders.
Conclusion
In conclusion, this study shows that boundary violations
are detrimental to the well-being of university teacher
during mandatory telework, and the results of this study
also support the assumption that the lack of recovery is
the explanatory mechanism for the effect of boundary
violations on well-being. Finally, this study also showed
that work boundary violations are more detrimental to
the recovery of female university teachers, whereas work
boundary violations are more detrimental to the recovery
of male university teachers.
Author contributions
MM wrote the main manuscript text. VC, MJC, and MM are involved in the
design of the work. VC was involved in the analysis and MJC with their
revision. CM was involved in the data collection and its interpretation.
Funding
This work received Portuguese national funding from FCT– Fundação para a
Ciência e a Tecnologia, I.P, through the project Ref. PTDC/PSI-GER/32367/2017
and through the Research Center for Psychological Science of the Faculty of
Psychology, University of Lisbon (UIDB/04527/2020; UIDP/04527/2020).
Data availability
The datasets generated and/or analysed during the current study are available
in the Figshare repository, [https://doi.org/10.6084/m9.figshare.24243307].
Declarations
Ethics approval and consent to participate
Ethics committees of the Faculty of Passo Fundo (Brazil) and of the Faculty
of Psychology, University of Lisbon (Portugal) approved this study. Informed
consent was obtained in writing on the first page of the questionnaire.
Participants who agreed to participate in the study proceeded to answer the
questionnaire.
Mascarenhas et al. BMC Public Health
(2024) 24:747
Consent for publication
Not applicable.
Competing interests
The authors declare no competing interests.
Received: 18 September 2023 / Accepted: 21 February 2024
References
1. Evans S, Alkan E, Bhangoo JK, Tenenbaum H, Ng-Knight T. Effects of the
COVID-19 lockdown on mental health, wellbeing, sleep, and alcohol use
in a UK student sample. Psychiatry Res. 2021;298. https://doi.org/10.1016/j.
psychres.2021.113819.
2. Platts K, Breckon J, Marshall H. Enforced home-working under lockdown
and its impact on employee wellbeing: a cross-sectional study. BMC Public
Health. 2022;22(199):1–13. https://doi.org/10.1186/s12889-022-12630-1.
3. Escudero-Castillo I, Mato-Díaz FJ, Rodriguez-Alvarez A, Furloughs. Teleworking and other Work situations during the COVID-19 lockdown: impact on
Mental Well-Being. Int J Environ Res Public Health. 2021;18:2898. https://doi.
org/10.3390/ijerph18062898.
4. Allen T, Johnson R, Kiuburz K, Shockley K. Work-family conflict and flexible
work arrangements: deconstructing flexibility. Pers Psychol. 2013;66:345–76.
https://doi.org/10.1111/peps.12012.
5. Kniffin KM, Narayanan J, Anseel F, Antonakis J, Ashford SP, Bakker AB,
Bamberger P, Bapuji H, Bhave DP, Choi VK, Creary SJ, Demerouti E, Flynn FJ,
Gelfand MJ, Greer LL, Johns G, Kesebir S, Klein PG, Lee SY, Ozcelik H, Petriglieri
JL, Rothbard NP, Rudolph CW, Shaw JD, Sirola N, Wanberg CR, Whillans A,
Wilmot MP, Vugt MV. COVID-19 and the Workplace: implications, issues, and
insights for Future Research and Action. Am Psychol. 2021;76:63. https://doi.
org/10.1037/amp0000716. Advance online publication.
6. Oakman J, Kinsman N, Stuckey R, Graham M, Weale V. A rapid review
of mental and physical health effects of working at home: how do we
optimise health? BMC Public Health. 2020;20:1825. https://doi.org/10.1186/
s12889-020-09875-z.
7. MacIntyre PD, Gregersen T, Mercer S. Language teachers’ coping strategies
during the Covid-19 Conversion to Online Teaching: correlations with stress,
wellbeing and negative emotions. System. 2020;94:102352. https://doi.
org/10.1016/j.system.2020.102352.
8. González A, Fernandez MA, Pino-Yancovic M, Madrid R. Teaching in the
pandemic: reconceptualizing Chilean educators’ professionalism now and for
the future. J Prof Capital Community. 2020;5:265–72. https://doi.org/10.1108/
JPCC-06-2020-0043.
9. Hollweck T, Doucet A. Pracademics in the pandemic: pedagogies and
professionalism. J Prof Capital Community. 2020;5:295–305. DOI: 0.1108/
JPCC-06-2020-0038.
10. UNESCO, UNICEF, and the World Bank. (2020). What Have We Learnt?
Overview of Findings from a Survey of Ministries of Education on National
Responses to COVID-19. https://openknowledge.worldbank.org/handle/10986/34700. Accessed 16 Jan 2021.
11. Chiesa V, Antony G, Wismar M, Rechel B. COVID-19 pandemic: health impact
of staying at home, social distancing and ‘lockdown’ measures-a systematic
review of systematic reviews. J Public Health (Oxf ). 2021;43:462–81. https://
doi.org/10.1093/pubmed/fdab102.
12. Ashforth BE, Kreiner GE, Fugate M. All in a day’s work: boundaries and
micro role transitions. Acad Manag Rev. 2000;25:472–91. https://doi.
org/10.2307/259305.
13. Clark SC. Work-family border theory: a new theory of work/family balance.
Hum Relat. 2000;53:747–70. https://doi.org/10.1177/0018726700536001.
14. Anderson D, Kelliher C. Flexible working and engagement: the
importance of choice. Strategic HR Rev. 2009;8:13–8. https://doi.
org/10.1108/14754390910937530.
15. Beauregard TA, Basile KA, Canonico E, Telework. Outcomes and facilitators for
employees. In Handbook of Technology and Employee Behavior; Landers, R.N.,
Ed.; Cambridge University Press: Cambridge, UK, 2019; pp. 511–543.ISBN.
16. G T.D. Telework and the navigation of work-home boundaries. Organ Dyn.
2021;50:100822. https://doi.org/10.1016/j.orgdyn.2020.100822.
17. Kinnunen U, Rantanen J, Bloom J, Mauno S, Feldt T, Korpela K. The role of
Work-Nonwork Boundary Management in work stress recovery. Int J Stress
Manage. 2016;23:99–123. https://doi.org/10.1037/a0039730.
Page 8 of 9
18. Kreiner GE, Hollensbe EC, Sheep ML. Balancing borders and bridges: negotiating the work-home interface via boundary work tactics. Acad Manag J.
2009;52:704–30. https://doi.org/10.5465/AMJ.2009.43669916.
19. Leroy S, Schmidt AM. The effect of regulatory focus on attention residue
and performance during interruptions. Organ Behav Hum Decis Process.
2016;137:218–35. https://doi.org/10.1016/j.obhdp.2016.07.006.
20. Carvalho VS, Santos A, Ribeiro MT, Chambel MJ, Please. Do not interrupt
me: work–family balance and segmentation behavior as mediators of
Boundary violations and teleworkers’ Burnout and Flourishing. Sustainability.
2021;13:7339. 10.3390/ su13137339.
21. Haun VC, Remmel C, Haun S. Boundary management and recovery when
working from home: the moderating roles of segmentation preference and
availability demands. German J Hum Resource Manage. 2022;36:270–99.
https://doi.org/10.1177/23970022221079048.
22. Leroy S, Schmidt AM, Madjar N. Working from Home during COVID-19: a
study of the Interruption Landscape. J Appl Psychol. 2021;106:1448–65.
https://doi.org/10.1037/apl0000972.
23. Perry SJ, Carlson DS, Kacmar MK, Wan MM, Thompson MJ. Interruptions in
remote work: a resourcebased model of work and family stress. J Bus Psychol.
2022;23:1–19. https://doi.org/10.1007/s10869-022-09842-y.
24. Alon T, Doepke M, Olmstead-Rumsey J, Tertilt M. The Impact of COVID-19 on
Gender Equality. NBER Working Paper No. 26947. Natl. Bur. Econ. Res. 2020,
26947; https://doi.org/10.3386/w26947.
25. Collins C, Landivar LC, Ruppanner L, Scarborough WJ. COVID-19 and the
gender gap in work hours. Gend Work Organ. 2021;28:101–12. https://doi.
org/10.1111/gwao.12506.
26. Carli LL, Women. Gender equality and COVID-19. Gend Manage. 2020;35:647–
55. https://doi.org/10.1108/GM-07-2020-0236.
27. Hjálmsdóttir A, Bjarnadóttir VS. I have turned into a foreman here at home:
families and work–life balance in times of COVID-19 in a gender equality paradise. Gend Work Organ. 2021;28:268–83. https://doi.org/10.1111/
gwao.12552.
28. Charalampous M, Grant CA, Tramontano C, Michailidis E. Systematically
reviewing remote e-workers’ well-being at work: a multidimensional
approach. Eur J Work Organizational Psychol. 2019;28:51–73. https://doi.org/1
0.1080/1359432X.2018.1541886.
29. Dettmers J. Characteristics of extended availability for work: the role of
demands and resources. Int J Stress Manage. 2016;23:276–97. https://doi.
org/10.1037/str0000014.
30. Sonnentag S, Bayer UV. Switching off mentally: predictors and consequences
of psychological detachment from work during offjob time. J Occup Health
Psychol. 2005;10:393–414. https://doi.org/10.1037/1076-8998.10.4.393.
31. Sonnentag S, Fritz C. The recovery experience questionnaire: development
and validation of a measure for assessing recuperation and unwinding from work. J Occup Health Psychol. 2007;12:204–21. https://doi.
org/10.1037/1076-8998.12.3.204.
32. Sonnentag S, Kuttler I, Fritz C. Job stressors, emotional exhaustion, and
need for recovery: a multi-source study on the benefits of psychological detachment. J Vocat Behav. 2010;76:355–65. https://doi.org/10.1016/j.
jvb.2009.06.005.
33. Volman FE, Bakker AB, Xanthopoulou D. Recovery at home and performance
at work: a diary study on self–family facilitation. Eur J Work Organizational
Psychol. 2013;22:218–34. https://doi.org/10.1080/1359432X.2011.648375.
34. Sonnentag S, Fritz C. Recovery from job stress: the stressor-detachment
model as an integrative framework. J Organizational Behav. 2015;36:72–S103.
https://doi.org/10.1002/job.1924.
35. Carlson DS, Kacmar MK, Zivnuska S, Ferguson M. Do the benefits of familyto-work transitions come at too great a cost? J Occup Health Psychol.
2015;20:161–71. https://doi.org/10.1037/a0038279.
36. Pachler D, Kuonath A, Specht J, Kennecke S, Agthe M, Frey D. Workflow
interruptions and employee work outcomes: the moderating role of polychronicity. J Occup Health Psychol. 2018;23:417–27. https://doi.org/10.1037/
ocp0000094.
37. Cropley M, Millward LJ. Job strain and work rumination about work issues
during leisure time: a diary study. Eur J Work Organizational Psychol.
2003;12:195–207. https://doi.org/10.1080/13594320344000093.
38. Mojza E, Lorenz C, Sonnentag S, Binnewies C. Daily recovery experiences:
the role of volunteer work during leisure time. J Occup Health Psychol.
2010;15:60–74. https://doi.org/10.1037/a0017983.
39. Weigelt O, Gierer P, Syrek CJ. My mind is working overtime—towards an integrative perspective of psychological detachment, work-related rumination,
Mascarenhas et al. BMC Public Health
40.
41.
42.
43.
44.
45.
46.
47.
48.
49.
50.
51.
(2024) 24:747
and work reflection. Int J Environ Res Public Health. 2019;16:2987. https://doi.
org/10.3390/ijerph16162987.
Wendsche J. A. Lohmann-Haislah 2017 A Meta-analysis on antecedents
and outcomes of detachment from work. Front Psychol 7 2072 https://doi.
org/10.3389/fpsyg.2016.02072.
Keyes CS. The mental health continuum: from languishing to flourishing in
life. J Health Social Behav. 2002;43:207–22. https://doi.org/10.2307/3090197.
Keyes CS. Promoting and protecting mental health as flourishing: a
complementary stretegy for improving national mental health. Am Psychol.
2007;62:95–108. https://doi.org/10.1037/0003-066X.62.2.95.
Rautenbach CL. Flourishing of employees in a fast moving consumable
goods environment. PhD, Doctor degree of Philosophy in Industrial Psychology, North-West University, 2015.
Diedericks E, Rothmann S. Flourishing of information technology professionals: the role of work engagement and job satisfaction. J Psychol Afr.
2013;23:225–33. https://doi.org/10.1080/14330237.2013.10820618.
Hunter EM, Clark MA, Carlson DS. Violating work-family boundaries: reactions
to interruptions at work and home. J Manag. 2019;45:1284–308. https://doi.
org/10.1177/0149206317702221.
Smit BW, Maloney PW, Maertz CP Jr, Montag-Smit T. Out of sight, out
of mind? How and when cognitive role transition episodes influence
employee performance. Hum Relat. 2016;69:2141–68. https://doi.
org/10.1177/0018726716636204.
Pleck JH. The work-family role system. Soc Probl. 1977;24:417–27. https://doi.
org/10.2307/800135.
Powell GN, Greenhaus JH. Sex, gender, and the work-to-family interface: exploring negative and positive interdependencies. Acad Manag J.
2010;53:513–34. https://doi.org/10.1177/0149206309350774.
Russo M, Ollier-Malaterre A, Kossek EE, Ohana M. Boundary management
permeability and relationship satisfaction in dual-earner couples: the asymmetrical gender effect. Front Psychol. 2018;9:1723. https://doi.org/10.3389/
fpsyg.2018.01723.
Hartig T, Kylin C, Johansson G. The telework tradeoff: stress mitigation vs.
constrained restoration. Appl Psychology: Int Rev. 2007;56:231–53. https://
doi.org/10.1111/j.1464-0597.2006.00252.x.
Voitsidis P, Nikopoulou VA, Holeva V, Parlapani E, Sereslis K, Tsipropoulou V,
Karamouzi P, Giazkoulidou A, Tsopaneli N, Diakogiannis I. The mediating role
Page 9 of 9
52.
53.
54.
55.
56.
57.
58.
59.
of fear of COVID-19 in the relationship between intolerance of uncertainty
and depression. Psychol Psychotherapy: Theory Res Pract. 2020;94:884–93.
https://doi.org/10.1111/papt.12315.
Sanz-Vergel AI, Demerouti E, Bakker AE, Moreno-Jiménez B. Daily detachment
from work and home: the moderating effect of role salience. Hum Relat.
2011;64:775–99. https://doi.org/10.1177/0018726710393368.
Thoits PA. On merging, identity, theory, and stress research. Social Psychol Q.
1991;54:101–12. https://doi.org/10.2307/2786929.
Hunter EM, Clark M, Carlson DS. Violating work-family boundaries: reactions
to interruptions at work and home. J Manag. 2017;45:1284–308. https://doi.
org/10.1177/0149206317702221.
Araújo MF, Pinheiro M. (2013). (2012, 12 e 13 junho). Recovery Experiences
Questionnaire. Adaptação para a população portuguesa In F. Lobo, editor, I
Congresso Internacional de Psicologia do Trabalho e das Organizações: Trabalho,
Riscos Psicossociais e Saúde: Conceptualização, Diagnóstico e Intervenção (pp.
1–10). http://hdl.handle.net/10400.14/17034.
Silva AJ, Caetano A. Validation of the flourishing scale and scale of positive
and negative experience in Portugal. Soc Indic Res. 2013;110:469–78. https://
doi.org/10.1007/s11205-011-9938-y.
Diener E, Wirtz D, Tov W, Kim-Prieto C, Choi D, Oishi S, Biswas-Diener R. New
measures of well-being: flourishing and positive and negative feelings. Soc
Indic Res. 2010;39:247–66. https://doi.org/10.1007/s11205-009-9493-y.
Muthén LK, Muthén BO. Mplus user’s guide. 7th ed. Los Angeles, CA, USA:
Muthén and Muthén; 2015.
Podsakoff PM, MacKenzie SB, Podsakoff NP, Lee JY. Common method
biases in behavioral research: a critical review of the literature and
recommended remedies. J Appl Psychol. 2003;88:879–903. https://doi.
org/10.1037/0021-9010.88.5.879.
Publisher’s Note
Springer Nature remains neutral with regard to jurisdictional claims in
published maps and institutional affiliations.
|
|
https://openalex.org/W2753776141
|
https://europepmc.org/articles/pmc5610841?pdf=render
|
English
| null |
The Neuroprotective Effects of SIRT1 on NMDA-Induced Excitotoxicity
|
Oxidative medicine and cellular longevity
| 2,017
|
cc-by
| 9,284
|
Research Article
The Neuroprotective Effects of SIRT1 on
NMDA-Induced Excitotoxicity Xiaorong Yang,1 Peipei Si,1,2 Huaping Qin,1 Litian Yin,1 Liang-Jun Yan,3 and Ce Zhang1
1National Key Disciplines, Key Laboratory for Cellular Physiology of Ministry of Education, Department of Neurobiology,
Shanxi Medical University, No. 56 Xin Jian South Road, Taiyuan, 030001 Shanxi, China Xiaorong Yang,1 Peipei Si,1,2 Huaping Qin,1 Litian Yin,1 Liang-Jun Yan,3 and Ce Zhang1
1National Key Disciplines, Key Laboratory for Cellular Physiology of Ministry of Education, Department of Neurobiology,
Shanxi Medical University, No. 56 Xin Jian South Road, Taiyuan, 030001 Shanxi, China
2Key Laboratory of Neurology of Hebei Province, The Second Hospital of Hebei Medical University, Shijiazhuang,
050071 Hebei, China
3Department of Pharmaceutical Sciences, UNT System College of Pharmacy, University of North Texas Health Science Center,
Fort Worth, TX 76107, USA 3Department of Pharmaceutical Sciences, UNT System College of Pharmacy, University of North Texas Health Science Center
Fort Worth TX 76107 USA 3Department of Pharmaceutical Sciences, UNT System College of Pharmacy, University of North Texas Health Science Center
Fort Worth, TX 76107, USA Correspondence should be addressed to Xiaorong Yang; rong_522522@163.com and Ce Zhang; cezh2002@yahoo.com Received 29 March 2017; Revised 22 June 2017; Accepted 5 July 2017; Published 1 September 2017 Academic Editor: Javier Egea Academic Editor: Javier Egea Copyright © 2017 Xiaorong Yang et al. This is an open access article distributed under the Creative Commons Attribution License,
which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited. Silent information regulator 1 (SIRT1), an NAD+-dependent deacetylase, is involved in the regulation of gene transcription,
energy metabolism, and cellular aging and has become an important therapeutic target across a range of diseases. Recent
research has demonstrated that SIRT1 possesses neuroprotective effects; however, it is unknown whether it protects
neurons from NMDA-mediated neurotoxicity. In the present study, by activation of SIRT1 using resveratrol (RSV) in
cultured cortical neurons or by overexpression of SIRT1 in SH-SY5Y cell, we aimed to evaluate the roles of SIRT1 in
NMDA-induced excitotoxicity. Our results showed that RSV or overexpression of SIRT1 elicited inhibitory effects on
NMDA-induced excitotoxicity including a decrease in cell viability, an increase in lactate dehydrogenase (LDH) release,
and a decrease in the number of living cells as measured by CCK-8 assay, LDH test, and Calcein-AM and PI double
staining. RSV or overexpression of SIRT1 significantly improved SIRT1 deacetylase activity in the excitotoxicity model. Hindawi
Oxidative Medicine and Cellular Longevity
Volume 2017, Article ID 2823454, 11 pages
https://doi.org/10.1155/2017/2823454 Hindawi
Oxidative Medicine and Cellular Longevity
Volume 2017, Article ID 2823454, 11 pages
https://doi.org/10.1155/2017/2823454 Research Article
The Neuroprotective Effects of SIRT1 on
NMDA-Induced Excitotoxicity Further study suggests that overexpression of SIRT1 partly suppressed an NMDA-induced increase in p53 acetylation. These results indicate that SIRT1 activation by either RSV or overexpression of SIRT1 can exert neuroprotective effects
partly by inhibiting p53 acetylation in NMDA-induced neurotoxicity. 2. Experimental Procedures 2.6. Lactate Dehydrogenase (LDH) Assay. LDH is released
from cells into a culture medium upon cell lysis. The cells
were plated in 24-well plates. At 24 h after NMDA exposure,
the supernatant was collected to measure LDH release
according to the manufacturer’s instructions. 2.1. Reagents. Neurobasal/B27, DMEM/F-12, and fetal
bovine serum (FBS) were purchased from Gibco-BRL (Grand
Island, NY, USA). Lipofectamine 2000 transfection reagents
were obtained from life technologies (St. Louis, MO, USA). Poly-D-lysine (MW 150,000–300,000), trypsin, arabinoside
cytosine, Calcein-AM, propidium iodide (PI), RSV, Sirtinol,
NMDA, MK-801, and SIRT1 assay kit were all purchased
from Sigma-Aldrich (St. Louis, MO, USA). The Cell Count-
ing Kit-8 (CCK-8) was from Dojindo, and the kit of LDH
was from Njjcbio. The polyclonal antibody to SIRT1 was
from Santa Cruz Biotechnology (Santa Cruz, CA, USA). Two polyclonal antibodies to p53 and Ace-p53 were obtained
from Cell Signaling Technology (Beverly, MA, USA). 2.7. Calcein-AM and PI Staining. Calcein-AM solution
(20 μM) was added to coverslips and the cells were incubated
at 37°C for 30min. PI solution was added and the cells were
incubated at 37°C for 5 min. The cells were examined by
using confocal microscope (Olympus, FV-1000) at the exci-
tation wave of 490 nm and emission wave of 515 nm. 2.8. SIRT1 Deacetylase Activity Assay. To measure SIRT1
activity, the protein was extracted from cells. The enzyme
activity of SIRT1 was measured using a SIRT1 assay kit
(CS1040; Sigma-Aldrich) based on the fleur de Lys-SIRT1
substrate peptide. The fluorescence intensity was measured
with a microplate reader (Packard, Meridien, MS), and the
excitation wavelength was 365 nm, and the emission wave-
length was 460 nm. 2.2. Cell Culture. Primary cortical cells were isolated from 1–
3-day-old Wistar rats and were cultured as previously
described [10]. In brief, cortical neurons from rats anesthe-
tized with ketamine (intraperitoneal injection, 100 mg/kg,
3 min) were dissected and digested in 0.025% trypsin,
followed by centrifugation at 800 g for 5 min. Cells were
resuspended in neurobasal/B27 medium and cultured at
37°C in 5% CO2. Arabinoside cytosine (10 μM) was added
after 24 h in vitro to inhibit non-neuronal cell growth. Exper-
iments were performed after 10–12 days in vitro. 2.9. Quantitative Real-Time PCR (qRT-PCR). Total RNA
from SH-SY5Y cell was isolated using TRIzol reagent (Invi-
trogen Life Technologies, Carlsbad, CA, USA) according to
the manufacturer’s instruction. Reverse transcription was
performed with High-Capacity cDNA Archive Kit (Applied
Biosystem). Oxidative Medicine and Cellular Longevity Midi Kit (Omega, GA, USA). The SH-SY5Y cells were seeded
into plates at a density of 1 × 105, and after 24 h, the plasmids
were transfected into the cells with a Lipofectamine 2000
Transfection Reagent. Midi Kit (Omega, GA, USA). The SH-SY5Y cells were seeded
into plates at a density of 1 × 105, and after 24 h, the plasmids
were transfected into the cells with a Lipofectamine 2000
Transfection Reagent. NMDA-induced excitotoxicity is a useful tool to evaluate
neurotoxicity in isolated cells and is a good model of nerve
injury that mimics closely the situation in vivo [9]. The present study was designed to investigate the neuro-
protection of SIRT1 in NMDA-induced excitotoxicity by
activation of SIRT1 using resveratrol (RSV) in cultured corti-
cal neurons or by overexpression of SIRT1 in the SH-SY5Y
cell line. The neuroprotective role of SIRT1 activity in vitro
may be useful for the development of new treatments for
central nervous system disorders. 2.5. Cell Viability Assay. Cells were seeded in 96-well plates,
and cell viability was assayed 24 h after NMDA exposure. Administration of 10 μL cck-8 solution into each well was
performed followed by incubation at 37°C for 2 h. Absor-
bance at 490 nm was measured using a microplate reader
(Packard, Meridien, MS). 1. Introduction SIRT1 deacetylates p53, PGC-1α, and NF-κB to prevent
many pathogenic processes. However, it remains unknown
whether SIRT1 protects neurons from NMDA-mediated
neurotoxicity in different excitotoxic insult models. Silent information regulator 1 (SIRT1), an NAD+-dependent
deacetylase, is known to deacetylate histone and nonhis-
tone proteins such as transcription factors. It participates
in a variety of physiopathological processes such as health
maintenance in development, gametogenesis, homeostasis,
longevity, and several neurodegenerative diseases as well
as age-related disorders [1–5]. Recently, the neuroprotec-
tive effects of SIRT1 have attracted great interest. It has
been found that SIRT1 could be upregulated to antagonize
neuronal injury in different animal models, such as cere-
bral ischemia, Alzheimer’s disease (AD), and Huntington’s
disease (HD) [6]. It has also been demonstrated that Glutamate is a primary excitatory amino acid neuro-
transmitter and activation of glutamate receptors including
NMDA receptor plays crucial roles in the central nervous
system. However, overactivation of NMDA receptor may
cause intracellular calcium overload, leading to an enzymatic
cascade of events resulting ultimately in cell death known as
excitotoxicity [7]. A wide range of acute and chronic brain
injury diseases, such as stroke/ischemia and epilepsy, and
certain neurodegenerative disorders have been linked to
NMDA receptor-mediated excitotoxicity [8]. Therefore, Oxidative Medicine and Cellular Longevity 2 2. Experimental Procedures qRT-PCR primers were synthesized by the
software of Primer Premier according to the following
sequences: β-actin (forward primer: 5′-TCATCACCATTGG
CAATGAG-3′, reverse primer: 5′-CACTGTGTTGGCGTA
CAGGT-3′); SIRT1: (forward primer: 5′-GCCTCACATG
CAAGCTCTAGTGACT-3′, reverse primer: 5′-ACTCAGG
TGGAG GTATTGTTTCCG-3′). qRT-PCR was performed
using StepOne Real-Time PCR Detection System (ABI) and
SYBR premix EX taqII (Takara). The human neuroblastoma SH-SY5Y cell, obtained from
the Chinese Academy of Sciences Institute of Cell Resource
Center, Shanghai, China, was maintained under a DMEM/
F12 medium with 10% FBS in 5% CO2 incubator. They were
washed by PBS buffer before adding 0.25% Trypsin-EDTA,
followed by incubation for 5 min at room temperature. Then,
the cells were detached, resuspended in medium, counted,
and seeded into plates at the density of 1× 105. 2.3. NMDA Treatment. After overnight incubation allowing
the cells to reach 80% confluency, cells were treated with
NMDA-containing Mg2+-free Locke’s buffer for 2 h. RSV
was added to cultures 12h prior to NMDA induction. Sirtinol was added 2 h before NMDA treatment. MK-801
and NMDA was simultaneously added to Mg2+-free Locke’s
buffer in the NMDA + MK-801 group. Control cells were
incubated with drug-free Mg2+-free Locke’s buffer and grown
at 37°C in an atmosphere containing 5% CO2. 2.10. Western Blot Analysis. The SH-SY5Y cells were col-
lected at 24 h after exposure to NMDA. Then, cells were lysed
in a lysis buffer (10 mM Tris-HCl (pH 7.4), 1 mM EDTA, and
1% Triton X-100). Cleared cell lysates were obtained after
centrifugation at 10000 ×g for 30min at 4°C. After measure-
ment of protein concentration using a BCA Protein Assay kit,
cell lysates (30 ~ 50 μg/lane) were subjected to SDS-PAGE,
and separated proteins were electrotransferred to nitro-
cellulose membranes. The membranes were washed in
Tris-buffered saline (TBS) containing 0.1% Tween 20 and
3% bovine serum albumin (BSA). The membranes were 2.4. Transfection of SIRT1. The expression vector expressing
human wild-type SIRT1 (WT-SIRT1) and the dominant-
negative form of human SIRT1 (DN-SIRT1) was constructed
by Genecopoeia. 2. Experimental Procedures Films
scanned and a densitometric analysis of the bands was
ormed with AlphaEase image analysis software. Statistical Analysis. The data were expressed as means ±
M. of at least three independent experiments. One-way
ysis of variance (ANOVA) with Bonferroni post hoc test
used for statistical comparisons. P < 0 05 was considered
e significant. Results
Neuroprotective Effects of RSV on NMDA-Induced
totoxicity in Primary Neurons
. Effects of RSV on NMDA-Induced Decrease in Cell
bility. Our previous study showed that the optimal excito-
city was induced 24h after NMDA (100 μM) exposure
2 h in primary cortical neurons. Figure 1(a) shows that
DA-induced cell viability decreased by 51.97% as
pared to the control in primary neurons (P < 0 05). Pre-
ment with five dosages (10 μM, 25 μM, 50μM, 75μM,
100 μM) of RSV, a potent SIRT1 activator, showed
cell viability was increased by 20.43% (P < 0 05),
2% (P < 0 05), 17.78% (P < 0 05), 11.85% (P < 0 05),
LDH viability (U/mgPro)
0
Control
NMDA
MK + NMDA
RSV + NMDA
RSV + Sirtinol + NMDA
DMSO
20
40
60
80
100
120
140
⁎
#
&
n = 6 per group
Figure 2: Effects of RSV on NMDA-induced LDH release in
primary neurons. RSV (25 μM) reduced NMDA-induced LDH
release (P < 0 05), and Sirtinol (10 μM) abolished the role of
RSV (P < 0 05). RSV: resveratrol; MK: MK-801. Each value
represents the mean ± S.E.M. of six independent experiments. ∗P < 0 05 versus the control group, #P < 0 05 versus the NMDA
group. &P < 0 05 versus the RSV + NMDA group. LDH viability (U/mgPro)
0
Control
NMDA
MK + NMDA
RSV + NMDA
RSV + Sirtinol + NMDA
DMSO
20
40
60
80
100
120
140
⁎
#
&
n = 6 per group
Figure 2: Effects of RSV on NMDA-induced LDH release in
primary neurons. RSV (25 μM) reduced NMDA-induced LDH
release (P < 0 05), and Sirtinol (10 μM) abolished the role of
RSV (P < 0 05). RSV: resveratrol; MK: MK-801. Each value
represents the mean ± S.E.M. of six independent experiments. ∗P < 0 05 versus the control group, #P < 0 05 versus the NMDA
group. &P < 0 05 versus the RSV + NMDA group. 2. Experimental Procedures incubated overnight at 4°C in TBS containing 3% BSA and
one of the following primary antibodies: SIRT1 (1 : 100),
p53 (1 : 1000), and Ace-p53 (1 : 1000). Subsequently, the
labeled proteins were incubated with an HRP-conjugated
anti-rabbit IgG (1 : 10,000) for 2 h. Blots were developed with
the ECL chemiluminescence system and were captured on
autoradiographic films (Kodak Image Station 440). Films
were scanned and a densitometric analysis of the bands was
performed with AlphaEase image analysis software. 2.11. Statistical Analysis. The data were expressed as means ±
S.E.M. of at least three independent experiments. One-way
analysis of variance (ANOVA) with Bonferroni post hoc test
was used for statistical comparisons. P < 0 05 was considered
to be significant. 2. Experimental Procedures The plasmids were extracted with a Plasmid Oxidative Medicine and Cellular Longevity 3 3 Control
MK
RSV (10 휇M)
RSV (25 휇M)
RSV (50 휇M)
RSV (75 휇M)
RSV (100 휇M)
DMSO
Cell viability (%)
0
20
40
60
80
100
120
⁎
#
#
#
#
#
NMDA
n = 6 per group
(a)
Cell viability (%)
0
Control
NMDA
RSV + NMDA
RSV + Sirtinol + NMDA
20
40
60
80
100
120
⁎
#
&
n = 6 per group
(b)
Figure 1: Effects of RSV on NMDA-induced decrease in cell viability in primary neurons. (a) Pretreatment of RSV (10 μM, 25 μM, 50 μM,
and 75 μM) improved cell viability compared with the NMDA treatment group (P < 0 05). (b) RSV (25 μM) significantly reversed NMDA-
induced decrease in cell viability (P < 0 05), and Sirtinol (10 μM) inhibited the effect of RSV (P < 0 05). RSV: resveratrol; MK: MK-801. Each
value represents the mean ± S.E.M. of six independent experiments. ∗P < 0 05 versus the control group, #P < 0 05 versus the NMDA
group, &P < 0 05 versus the RSV + NMDA group. (a) Figure 1: Effects of RSV on NMDA-induced decrease in cell viability in primary neurons. (a) Pretreatment of RSV (10 μM, 25 μM, 50 μM,
and 75 μM) improved cell viability compared with the NMDA treatment group (P < 0 05). (b) RSV (25 μM) significantly reversed NMDA-
induced decrease in cell viability (P < 0 05), and Sirtinol (10 μM) inhibited the effect of RSV (P < 0 05). RSV: resveratrol; MK: MK-801. Each
value represents the mean ± S.E.M. of six independent experiments. ∗P < 0 05 versus the control group, #P < 0 05 versus the NMDA
group, &P < 0 05 versus the RSV + NMDA group. LDH viability (U/mgPro)
0
Control
NMDA
MK + NMDA
RSV + NMDA
SV + Sirtinol + NMDA
DMSO
20
40
60
80
100
120
140
⁎
#
&
n = 6 per group bated overnight at 4°C in TBS containing 3% BSA and
of the following primary antibodies: SIRT1 (1 : 100),
(1 : 1000), and Ace-p53 (1 : 1000). Subsequently, the
ed proteins were incubated with an HRP-conjugated
rabbit IgG (1 : 10,000) for 2 h. Blots were developed with
ECL chemiluminescence system and were captured on
radiographic films (Kodak Image Station 440). Figure 2: Effects of RSV on NMDA-induced LDH release in
primary neurons. RSV (25 μM) reduced NMDA-induced LDH
release (P < 0 05), and Sirtinol (10 μM) abolished the role of
RSV (P < 0 05). RSV: resveratrol; MK: MK-801. Each value
represents the mean ± S.E.M. of six independent experiments.
∗P < 0 05 versus the control group, #P < 0 05 versus the NMDA
group. &P < 0 05 versus the RSV + NMDA group. 3. Results 3.1. Neuroprotective Effects of RSV on NMDA-Induced
Excitotoxicity in Primary Neurons 3.1.1. Effects of RSV on NMDA-Induced Decrease in Cell
Viability. Our previous study showed that the optimal excito-
toxicity was induced 24h after NMDA (100 μM) exposure
for 2 h in primary cortical neurons. Figure 1(a) shows that
NMDA-induced cell viability decreased by 51.97% as
compared to the control in primary neurons (P < 0 05). Pre-
treatment with five dosages (10 μM, 25 μM, 50μM, 75μM,
and 100 μM) of RSV, a potent SIRT1 activator, showed
that cell viability was increased by 20.43% (P < 0 05),
31.92% (P < 0 05), 17.78% (P < 0 05), 11.85% (P < 0 05), Figure 2: Effects of RSV on NMDA-induced LDH release in
primary neurons. RSV (25 μM) reduced NMDA-induced LDH
release (P < 0 05), and Sirtinol (10 μM) abolished the role of
RSV (P < 0 05). RSV: resveratrol; MK: MK-801. Each value
represents the mean ± S.E.M. of six independent experiments. ∗P < 0 05 versus the control group, #P < 0 05 versus the NMDA
group. &P < 0 05 versus the RSV + NMDA group. Oxidative Medicine and Cellular Longevity Calcein-AM
PI
Merge
Bright feld
Control
NMDA
MK + NMDA
RSV + NMDA
RSV + Sritinol + NMDA
DMSO
(a)
Survival rate (%)
0
Control
NMDA
MK + NMDA
RSV + NMDA
RSV + Sirtinol + NMDA
DMSO
20
40
60
80
100
120
⁎
#
&
n = 3 per group
(b)
cts of RSV on NMDA-induced decrease in the number of living cells in primary neurons. (a) Represent
uppression of RSV (25 μM) on NMDA-induced decrease of living cells (P < 0 05), which was abo
05). Living cells were stained by Calcein-AM (green), and dead cells were stained by PI (red). (b) B
RSV: resveratrol; MK: MK-801. Each value represents the mean ± S.E.M. of three independent exper
trol group, #P < 0 05 versus the NMDA group. &P < 0 05 versus the RSV + NMDA group. 3. Results NMDA (100 μM, 500 μM,
and 1000 μM) decreased cell viability at 6 h, 12 h, and 24 h after
NMDA exposure for 2 h (P < 0 05). Only at 12 h and 24 h after
NMDA (10 μM), treatment was significant (P < 0 05). Numbers
represent the percentage of the living cells normalized to the
control. Each
value
represents
the
mean ± S.E.M. of
six
independent experiments. ∗P < 0 05 versus the control group. Figure 4: Effects of RSV on NMDA-induced decrease of SIRT1
deacetylase activity in primary neurons. RSV (25 μM) significantly
ameliorated SIRT1 activity reduced by NMDA (P < 0 05), and
Sirtinol (10 μM) abolished the effect of RSV (P < 0 05). MK: MK-
801. Each value represents the mean ± S.E.M. of six independent
experiments. ∗P < 0 05 versus the control group, #P < 0 05 versus
the NMDA group, &P < 0 05 versus the RSV + NMDA group. Figure 4: Effects of RSV on NMDA-induced decrease of SIRT1
deacetylase activity in primary neurons. RSV (25 μM) significantly
ameliorated SIRT1 activity reduced by NMDA (P < 0 05), and
Sirtinol (10 μM) abolished the effect of RSV (P < 0 05). MK: MK-
801. Each value represents the mean ± S.E.M. of six independent
experiments. ∗P < 0 05 versus the control group, #P < 0 05 versus
the NMDA group, &P < 0 05 versus the RSV + NMDA group. 3.1.3. Effects of RSV on NMDA-Induced Decrease in the
Number of Living Cells. Exposure to NMDA resulted in a
significant decrease in the cell survival rate estimated by
Calcein-AM and PI staining (Figure 3). After treatment with
NMDA, the cell survival rate decreased by 55.98% as com-
pared to that of the control group (P < 0 05). MK-801
(10 μM) completely inhibited NMDA-induced decrease of
living cells. Administration of RSV (25 μM) significantly
increased the cell survival rate by 18.99% as compared to
the NMDA-treated group (P < 0 05) and pretreatment with
Sirtinol (10 μM) abolished the role of RSV (P < 0 05). and 0.37% (P > 0 05), respectively, when compared to that
of the NMDA-treated group (Figure 1(a)). In the following
experiments, RSV was administered at the concentration
of 25μM based on the significant protective effect observed
in the 25μM RSV group. 3. Results Calcein-AM
PI
Merge
Bright feld
Control
NMDA
MK + NMDA
RSV + NMDA
RSV + Sritinol + NMDA
DMSO
(a) Bright feld Merge PI (a)
Survival rate (%)
0
Control
NMDA
MK + NMDA
RSV + NMDA
RSV + Sirtinol + NMDA
DMSO
20
40
60
80
100
120
⁎
#
&
n = 3 per group
(b) Figure 3: Effects of RSV on NMDA-induced decrease in the number of living cells in primary neurons. (a) Representative micrographs
showing the suppression of RSV (25 μM) on NMDA-induced decrease of living cells (P < 0 05), which was abolished by Sirtinol
(10 μM) (P < 0 05). Living cells were stained by Calcein-AM (green), and dead cells were stained by PI (red). (b) Bar graph of mean
of living cells. RSV: resveratrol; MK: MK-801. Each value represents the mean ± S.E.M. of three independent experiments. ∗P < 0 05
versus the control group, #P < 0 05 versus the NMDA group. &P < 0 05 versus the RSV + NMDA group. Oxidative Medicine and Cellular Longevity 5 SIRT1 activity (%)
0
Control
NMDA
MK + NMDA
RSV + NMDA
RSV + Sirtinol + NMDA
DMSO
20
40
60
80
100
120
⁎
#
&
n = 6 per group
Figure 4: Effects of RSV on NMDA-induced decrease of SIRT1
deacetylase activity in primary neurons. RSV (25 μM) significantly
ameliorated SIRT1 activity reduced by NMDA (P < 0 05), and
Sirtinol (10 μM) abolished the effect of RSV (P < 0 05). MK: MK-
801. Each value represents the mean ± S.E.M. of six independent
experiments. ∗P < 0 05 versus the control group, #P < 0 05 versus
the NMDA group, &P < 0 05 versus the RSV + NMDA group. SIRT1 activity (%)
0
Control
NMDA
MK + NMDA
RSV + NMDA
RSV + Sirtinol + NMDA
DMSO
20
40
60
80
100
120
⁎
#
&
n = 6 per group 0
Control
10 휇M
100 휇M
500 휇M
1000 휇M
NMDA
30
60
⁎
⁎
⁎
⁎
⁎
⁎
⁎
⁎
⁎
⁎⁎
Cell viability (%)
n = 6 per group
6 h
12 h
24 h
90
120 n = 6 per group Figure 5: Effects of different concentrations (10–1000 μM) of
NMDA on cell viability in SH-SY5Y cell. 3. Results MK-801 (10μM) did not alter cell
viability and yielded a value nearly equivalent to that of the
control group (P > 0 05). These data support the notion that
the toxic effects were induced by NMDA. Similarly,
DMSO as a NMDA vehicle had no significant effects on
cell viability (P > 0 05). As shown in Figure 1(b), pretreatment with RSV (25 μM)
significantly increased the viability of primary neurons com-
pared to that of the NMDA group (P < 0 05). However, acom-
bination of Sirtinol (10 μM) and RSV (25 μM) did not affect
the NMDA-induced decrease in cell viability (P > 0 05), sug-
gesting that Sirtinol, a specific inhibitor of SIRT1, blocked
theprotective effectofRSVonNMDA-induced excitotoxicity. Pretreatment of Sirtinol (10 μM) alone does not affect cell
survival in primary neurons (data not shown). 3.1.4. Effects of RSV on NMDA-Induced Decrease in SIRT1
Deacetylase Activity. As shown in Figure 4, NMDA greatly
reduced the SIRT1 activity (P < 0 05), which was inhibited
by MK-801. Pretreatment with RSV significantly ameliorated
SIRT1 activity reduced by NMDA (P < 0 05), while Sirtinol
abolished the effect of RSV (P < 0 05). There was no differ-
ence in SIRT1 deacetylase activity between the DMSO group
and the control group (P > 0 05). 3.1.2. Effects of RSV on NMDA-Induced LDH Release. After
treatment of NMDA, LDH levels rose by 116.03% compared
with those of the control group (P < 0 05, Figure 2). NMDA-
induced LDH release was fully blocked by MK-801 (10 μM). Administration of RSV (25 μM) reduced NMDA-induced
LDH release by 27.34% (P < 0 05), and pretreatment with
Sirtinol (10 μM) abolished the role of RSV (P < 0 05). There
was no difference in LDH release between the DMSO group
and the control group (P > 0 05). 3.2. Neuroprotective Effects of SIRT1 on NMDA-Induced
Excitotoxicity in the SH-SY5Y Cell Line 3.2.1. NMDA-Induced Decrease in Cell Viability. To better
characterize NMDA-induced neuronal insults of the SH-
SY5Y cell line, the administration of NMDA at different con-
centrations (10μM, 100 μM, 500 μM, and 1000 μM) and the
cell viability were measured 6 h, 12 h, and 24 h after NMDA
exposure for 2 h. 3. Results Results showed that 500 μM NMDA Oxidative Medicine and Cellular Longevity 6 ell viability by 49.26% (P < 0 05) 12 h after
osure (Figure 5), which were used as an insult-
del for further experiments. In SH-SY5Y cell,
d
ll i bili
hi h
ld b
i
d
3.2.2. Overexpression of SIRT1 Increased the Levels of S
mRNA
and
Protein
in
NMDA-Induced
Excitotox
NMDA significantly decreased the level of SIRT1 m
(Fi
6( ))
d SIRT1
i
(Fi
6(b))
h
SIRT1 mRNA expression
(fold versus control)
0
Control
NMDA
Vector + NMDA
WT-SIRT1 + NMDA
WT-SIRT1
DN-SIRT1 + NMDA
DN-SIRT1
1
2
3
4
5
⁎
⁎
#
#
⁎
⁎
&
n = 3 per group
▲
(a)
Control
NMDA
Vector + NMDA
WT-SIRT1 + NMDA
WT-SIRT1
DN-SIRT1 + NMDA
DN-SIRT1
SIRT1
훽-Actin
SIRT1 protein expression
(fold versus control)
0.0
0.5
1.0
1.5
2.0
2.5
3.0
3.5
⁎
#
⁎
⁎
⁎
#
&
n = 3 per group
▲
(b)
erexpression of SIRT1 (WT-SIRT1 or DN-SIRT1) increased the levels of SIRT1 mRNA and protein reduced by N
SH-SY5Y cell (P < 0 05). (a) Quantitative representations of SIRT1 mRNA by bar graph. (b) Western blot probed
ainst SIRT1 and β-actin (upper panel) and quantitative representations of SIRT1 protein expression by bar graph (
alue represents the mean ± S.E.M. of three independent experiments. ∗P < 0 05 versus the control group. #P < 0 05 versu
A group, &P < 0 05 versus the WT-SIRT1 + NMDA group, ▲P < 0 05 versus the DN-SIRT1 + NMDA group. g Figure 6: Overexpression of SIRT1 (WT-SIRT1 or DN-SIRT1) increased the levels of SIRT1 mRNA and protein reduced by NMDA
(500 μM) in SH-SY5Y cell (P < 0 05). (a) Quantitative representations of SIRT1 mRNA by bar graph. (b) Western blot probed with
antibodies against SIRT1 and β-actin (upper panel) and quantitative representations of SIRT1 protein expression by bar graph (lower
panel). Each value represents the mean ± S.E.M. of three independent experiments. ∗P < 0 05 versus the control group. #P < 0 05 versus the
vector + NMDA group, &P < 0 05 versus the WT-SIRT1 + NMDA group, ▲P < 0 05 versus the DN-SIRT1 + NMDA group. decreased cell viability by 49.26% (P < 0 05) 12 h after
NMDA exposure (Figure 5), which were used as an insult-
induced model for further experiments. 3. Results Effects of SIRT1 Overexpression on p53 Acetylation in
NMDA-Induced Excitotoxicity. Figure 8 shows that NMDA
induced acetylation of p53 and the level of acetylated
p53 (Ace-p53) was significantly higher (36.60%) than
that of the control group (P < 0 05). WT-SIRT1 overex-
pression partially inhibited NMDA-stimulated p53 acety-
lation (P < 0 05), and DN-SIRT1 overexpression had no
effect on Ace-p53 increased by NMDA (P > 0 05). The
total levels of p53 were virtually unchanged under all
of these experimental conditions. SIRT1 deacetylase activity (%)
0
Control
NMDA
Vector + NMDA
WT-SIRT1 + NMDA
WT-SIRT1
DN-SIRT1 + NMDA
DN-SIRT1
20
40
60
80
100
120
140
160
180
⁎
⁎
#
&
⁎
⁎
n = 6 per group 3.2.5. Effects of SIRT1 Overexpression on the Cell Viability
Reduced by NMDA. Figure 9 shows that WT-SIRT1 overex-
pression reversed NMDA-induced decrease in cell viability
(P < 0 05), while DN-SIRT1 overexpression did not affect
cell viability in the NMDA group (P > 0 05). WT-SIRT1
overexpression alone did not affect cell viability (P > 0 05),
whereas DN-SIRT1 overexpression alone reduced cell viabil-
ity (P < 0 05). There was a difference between the WT-
SIRT1+ NMDA and WT-SIRT1 groups (P < 0 05); however,
there was no difference between the DN-SIRT1 +NMDA
and DN-SIRT1 groups (P > 0 05). Figure 7: Effects of SIRT1 overexpression on the deacetylase
activity in NMDA-induced excitotoxicity of SH-SY5Y cell. WT-
SIRT1 overexpression significantly reversed the deacetylase activity
decreased by NMDA (P < 0 05), and DN-SIRT1 overexpression
had no effect (P > 0 05). Each value represents the mean ± S.E.M. of six independent experiments. ∗P < 0 05 versus the control group. #P < 0 05 versus the vector + NMDA group, &P < 0 05 versus the
WT-SIRT1 + NMDA group. 3.2.6. Effects of SIRT1 Overexpression on NMDA-Induced
LDH Release. As shown in Figure 10, WT-SIRT1 overex-
pression reduced NMDA-induced LDH release by 24.26%
(P < 0 05). Whereas, DN-SIRT1 overexpression did not
protect against NMDA-mediated LDH release (P > 0 05). The effects of WT-SIRT1 or DN-SIRT1 overexpression
alone on LDH release were completely consistent with
those of WT-SIRT1 or DN-SIRT1 overexpression alone on
cell viability. 3.2.6. Effects of SIRT1 Overexpression on NMDA-Induced
LDH Release. As shown in Figure 10, WT-SIRT1 overex-
pression reduced NMDA-induced LDH release by 24.26%
(P < 0 05). 3. Results Whereas, DN-SIRT1 overexpression did not
protect against NMDA-mediated LDH release (P > 0 05). The effects of WT-SIRT1 or DN-SIRT1 overexpression
alone on LDH release were completely consistent with
those of WT-SIRT1 or DN-SIRT1 overexpression alone on
cell viability. DN-SIRT1 overexpression restored the NMDA-induced
decrease in the levels of SIRT1 mRNA and protein compared
to those of the control (P < 0 05); however, WT-SIRT1/DN-
SIRT1-overexpressing cells without NMDA treatment exhib-
ited about a 2 ~ 3-fold increase in SIRT1 mRNA and protein
when compared with those of the control group (P < 0 05). The level of SIRT1 mRNA and protein showed a great differ-
ence between the WT-SIRT1 + NMDA group and the WT-
SIRT1 group and also between the DN-SIRT1 + NMDA
and DN-SIRT1 groups (P < 0 05), indicating that the trans-
fection efficiency may be downregulated by NMDA. There
was no difference between the NMDA group and the
NMDA + vector group (P > 0 05). 3.2.7. Effects of SIRT1 Overexpression on the Number of Living
Cells Reduced by NMDA. Calcein-AM and PI staining results
(Figure 11) showed that NMDA resulted in a significant
decrease in the number of living cells, which was inhibited
by WT-SIRT1 overexpression (P < 0 05). While DN-SIRT1
overexpression has no effect on the number of survival cells
when compared with the NMDA group (P > 0 05). The
effects of WT-SIRT1 or DN-SIRT1 overexpression alone on
cell survival showed similar results as those of the data
described above. 3. Results SIRT1 activity showed a great difference between the W
SIRT1 +NMDA group and the WT-SIRT1 group (P < 0
There was no difference between the DN-SIRT1 + NM
and DN SIRT1 group and also between the NMDA gr
SIRT1 deacetylase activity (%)
0
Control
NMDA
Vector + NMDA
WT-SIRT1 + NMDA
WT-SIRT1
DN-SIRT1 + NMDA
DN-SIRT1
20
40
60
80
100
120
140
160
180
⁎
⁎
#
&
⁎
⁎
n = 6 per group
Figure 7: Effects of SIRT1 overexpression on the deacety
activity in NMDA-induced excitotoxicity of SH-SY5Y cell. W
SIRT1 overexpression significantly reversed the deacetylase act
decreased by NMDA (P < 0 05), and DN-SIRT1 overexpres
had no effect (P > 0 05). Each value represents the mean ± S.E
of six independent experiments. ∗P < 0 05 versus the control gr
#P < 0 05 versus the vector + NMDA group, &P < 0 05 versus
WT-SIRT1 + NMDA group. Oxidative Medicine and Cellular Longevity SIRT1 deacetylase activity (%)
0
Control
NMDA
Vector + NMDA
WT-SIRT1 + NMDA
WT-SIRT1
DN-SIRT1 + NMDA
DN-SIRT1
20
40
60
80
100
120
140
160
180
⁎
⁎
#
&
⁎
⁎
n = 6 per group
Figure 7: Effects of SIRT1 overexpression on the deacetylase
activity in NMDA-induced excitotoxicity of SH-SY5Y cell. WT-
SIRT1 overexpression significantly reversed the deacetylase activity
decreased by NMDA (P < 0 05), and DN-SIRT1 overexpression
had no effect (P > 0 05). Each value represents the mean ± S.E.M. of six independent experiments. ∗P < 0 05 versus the control group. #P < 0 05 versus the vector + NMDA group, &P < 0 05 versus the
WT-SIRT1 + NMDA group. SIRT1 deacetylase activity (%)
0
Control
NMDA
Vector + NMDA
WT-SIRT1 + NMDA
WT-SIRT1
DN-SIRT1 + NMDA
DN-SIRT1
20
40
60
80
100
120
140
160
180
⁎
⁎
#
&
⁎
⁎
n = 6 per group
Figure 7: Effects of SIRT1 overexpression on the deacetylase
activity in NMDA-induced excitotoxicity of SH-SY5Y cell. WT-
SIRT1 overexpression significantly reversed the deacetylase activity
decreased by NMDA (P < 0 05), and DN-SIRT1 overexpression
had no effect (P > 0 05). Each value represents the mean ± S.E.M. of six independent experiments. ∗P < 0 05 versus the control group. #P < 0 05 versus the vector + NMDA group, &P < 0 05 versus the
WT-SIRT1 + NMDA group. 3.2.4. 3. Results In SH-SY5Y cell,
NMDA decreased cell viability, which could be antagonized
by MK-801 (data not shown). decreased cell viability by 49.26% (P < 0 05) 12 h after
NMDA exposure (Figure 5), which were used as an insult-
induced model for further experiments. In SH-SY5Y cell,
NMDA decreased cell viability, which could be antagonized
by MK-801 (data not shown). 3.2.2. Overexpression of SIRT1 Increased the Levels of SIRT1
mRNA
and
Protein
in
NMDA-Induced
Excitotoxicity. NMDA significantly decreased the level of SIRT1 mRNA
(Figure 6(a)) and SIRT1 protein (Figure 6(b)) when com-
pared with that of the control group (P < 0 05). WT-SIRT1/ 3.2.2. Overexpression of SIRT1 Increased the Levels of SIRT1
mRNA
and
Protein
in
NMDA-Induced
Excitotoxicity. NMDA significantly decreased the level of SIRT1 mRNA
(Figure 6(a)) and SIRT1 protein (Figure 6(b)) when com-
pared with that of the control group (P < 0 05). WT-SIRT1/ 7 Oxidative Medicine and Cellular Longevity DN-SIRT1 overexpression restored the NMDA-indu
decrease in the levels of SIRT1 mRNA and protein compa
to those of the control (P < 0 05); however, WT-SIRT1/D
SIRT1-overexpressing cells without NMDA treatment exh
ited about a 2 ~ 3-fold increase in SIRT1 mRNA and pro
when compared with those of the control group (P < 0
The level of SIRT1 mRNA and protein showed a great diff
ence between the WT-SIRT1 + NMDA group and the W
SIRT1 group and also between the DN-SIRT1 + NM
and DN-SIRT1 groups (P < 0 05), indicating that the tra
fection efficiency may be downregulated by NMDA. Th
was no difference between the NMDA group and
NMDA + vector group (P > 0 05). 3.2.3. Effects of SIRT1 Overexpression on the Deacety
Activity in NMDA-Induced Excitotoxicity. As shown
Figure 7, NMDA inhibited SIRT1 deacetylase acti
(P < 0 05). WT-SIRT1 overexpression after exposure
NMDA reversed the deacetylase activity decreased by NM
(P < 0 05), whereas DN-SIRT1 overexpression with NM
administration had no effect (P > 0 05). Compared with
of the control group, WT-SIRT1 overexpression itself sig
icantly increased SIRT1 activity (P < 0 05), and DN-SIR
overexpression itself reduced the activity (P < 0 05). 4. Discussion WT-SIRT1 overexpression
partially inhibited NMDA-stimulated p53 acetylation (P < 0 05), and DN-SIRT1 overexpression had no effect (P > 0 05). (a) Western blot
probed with antibodies against p53 and Ace-p53. (b) Quantitative representations of Ace-p53 by bar graph. Each value represents the
mean ± S.E.M. of three independent experiments. ∗P < 0 05 versus the control group, #P < 0 05 versus the NMDA group. LDH viability (U/L)
0
Control
NMDA
WT-SIRT1 + NMDA
WT-SIRT1
DN-SIRT1 + NMDA
DN-SIRT1
100
200
300
400
500
600
⁎
#
&
⁎
n = 6 per group
Figure 10: Effects of SIRT1 overexpression on NMDA-induced the
amount of LDH release the in SH-SY5Y cell line. WT-SIRT1
overexpression reduced NMDA-induced LDH release (P < 0 05),
and DN-SIRT1 overexpression did not protect against NMDA-
mediated LDH release (P > 0 05). Each value represents the mean ±
S.E.M. of six independent experiments. ∗P < 0 05 versus the control
group, #P < 0 05 versus the NMDA group, &P < 0 05 versus the
WT-SIRT1 + NMDA group. protection against disorders of the nervous system, for
example, brain ischemia reperfusion injury [11], Alzhei-
mer’s disease, Parkinson’s disease [12], and traumatic
CNS
injury
[13]. We
also
found
that
Sirtinol,
a
pharmacological inhibitor of SIRT1, abolished the protec-
tion of RSV against NMDA-mediated nerve injury, indi-
cating that the neuroprotective role of RSV is possibly
achieved by activation of SIRT1. It has been shown that
Cell viability (%)
0
Control
NMDA
WT-SIRT1 + NMDA
WT-SIRT1
DN-SIRT1 + NMDA
DN-SIRT1
20
40
60
80
100
120
⁎
#
&
⁎
n = 6 per group
Figure 9: Effects of SIRT1 overexpression on cell viability reduced
by NMDA in the SH-SY5Y cell line. WT-SIRT1 overexpression
reversed NMDA-induced decrease in cell viability (P < 0 05), and
DN-SIRT1 overexpression did not affect cell viability in the
NMDA group (P > 0 05). Each value represents the mean ± S.E.M. of six independent experiments. ∗P < 0 05 versus the control
group, #P < 0 05 versus the NMDA group, &P < 0 05 versus the
WT-SIRT1 + NMDA group. LDH viability (U/L)
0
Control
NMDA
WT-SIRT1 + NMDA
WT-SIRT1
DN-SIRT1 + NMDA
DN-SIRT1
100
200
300
400
500
600
⁎
#
&
⁎
n = 6 per group
Figure 10: Effects of SIRT1 overexpression on NMDA-induced the
amount of LDH release the in SH-SY5Y cell line. 4. Discussion 3.2.3. Effects of SIRT1 Overexpression on the Deacetylase
Activity in NMDA-Induced Excitotoxicity. As shown in
Figure 7, NMDA inhibited SIRT1 deacetylase activity
(P < 0 05). WT-SIRT1 overexpression after exposure to
NMDA reversed the deacetylase activity decreased by NMDA
(P < 0 05), whereas DN-SIRT1 overexpression with NMDA
administration had no effect (P > 0 05). Compared with that
of the control group, WT-SIRT1 overexpression itself signif-
icantly increased SIRT1 activity (P < 0 05), and DN-SIRT1
overexpression itself reduced the activity (P < 0 05). The
SIRT1 activity showed a great difference between the WT-
SIRT1 +NMDA group and the WT-SIRT1 group (P < 0 05). There was no difference between the DN-SIRT1 + NMDA
and DN-SIRT1 group and also between the NMDA group
and the NMDA + vector group (P > 0 05). The present study provided the following three important
findings. First, activation of SIRT1 or overexpression of
SIRT1 protected against NMDA-mediated excitotoxicity; sec-
ond, the neuroprotective effects of SIRT1 on NMDA-induced
excitotoxicity were attributed to its deacetylase activity; and
third, inhibition of p53 acetylation might be one of the mech-
anisms underlying SIRT1-mediated neuroprotection. In this study, we found that either preincubation of corti-
cal neurons with RSV or overexpression of WT-SIRT1 in the
SH-SY5Y cell line prevented NMDA-induced excitotoxicity
including a decrease in cell viability, an increase in LDH
release, and an increase in cell death, suggesting that SIRT1
has neuroprotection in NMDA-induced excitotoxicity. As
has been reported, activation of SIRT1 using RSV has 8 8 Oxidative Medicine and Cellular Longevity Ace-p53
p53
(a)
Ace-p53/p53 (%)
0
Control
NMDA
WT-SIRT1 + NMDA
DN-SIRT1 + NMDA
20
40
60
80
100
120
⁎
#
n = 3 per group
(b)
Figure 8: Effects of SIRT1 overexpression on p53 acetylation in NMDA-induced excitotoxicity of SH-SY5Y cell. WT-SIRT1 overexpression
partially inhibited NMDA-stimulated p53 acetylation (P < 0 05), and DN-SIRT1 overexpression had no effect (P > 0 05). (a) Western blot
probed with antibodies against p53 and Ace-p53. (b) Quantitative representations of Ace-p53 by bar graph. Each value represents the
mean ± S.E.M. of three independent experiments. ∗P < 0 05 versus the control group, #P < 0 05 versus the NMDA group. 8
Oxidative Medicine and Cellular Longevity (a) (b) Figure 8: Effects of SIRT1 overexpression on p53 acetylation in NMDA-induced excitotoxicity of SH-SY5Y cell. 4. Discussion WT-SIRT1
overexpression reduced NMDA-induced LDH release (P < 0 05),
and DN-SIRT1 overexpression did not protect against NMDA-
mediated LDH release (P > 0 05). Each value represents the mean ±
S.E.M. of six independent experiments. ∗P < 0 05 versus the control
group, #P < 0 05 versus the NMDA group, &P < 0 05 versus the
WT-SIRT1 + NMDA group. Figure 9: Effects of SIRT1 overexpression on cell viability reduced
by NMDA in the SH-SY5Y cell line. WT-SIRT1 overexpression
reversed NMDA-induced decrease in cell viability (P < 0 05), and
DN-SIRT1 overexpression did not affect cell viability in the
NMDA group (P > 0 05). Each value represents the mean ± S.E.M. of six independent experiments. ∗P < 0 05 versus the control
group, #P < 0 05 versus the NMDA group, &P < 0 05 versus the
WT-SIRT1 + NMDA group. Figure 9: Effects of SIRT1 overexpression on cell viability reduced
by NMDA in the SH-SY5Y cell line. WT-SIRT1 overexpression
reversed NMDA-induced decrease in cell viability (P < 0 05), and
DN-SIRT1 overexpression did not affect cell viability in the
NMDA group (P > 0 05). Each value represents the mean ± S.E.M. of six independent experiments. ∗P < 0 05 versus the control
group, #P < 0 05 versus the NMDA group, &P < 0 05 versus the
WT-SIRT1 + NMDA group. Figure 10: Effects of SIRT1 overexpression on NMDA-induced the
amount of LDH release the in SH-SY5Y cell line. WT-SIRT1
overexpression reduced NMDA-induced LDH release (P < 0 05),
and DN-SIRT1 overexpression did not protect against NMDA-
mediated LDH release (P > 0 05). Each value represents the mean ±
S.E.M. of six independent experiments. ∗P < 0 05 versus the control
group, #P < 0 05 versus the NMDA group, &P < 0 05 versus the
WT-SIRT1 + NMDA group. pharmacological inhibitor of SIRT1, abolished the protec-
tion of RSV against NMDA-mediated nerve injury, indi-
cating that the neuroprotective role of RSV is possibly
achieved by activation of SIRT1. It has been shown that protection against disorders of the nervous system, for
example, brain ischemia reperfusion injury [11], Alzhei-
mer’s disease, Parkinson’s disease [12], and traumatic
CNS
injury
[13]. 4. Discussion We
also
found
that
Sirtinol,
a 9 9 Oxidative Medicine and Cellular Longevity Control
NMDA
WT-SIRT1 + NMDA
WT-SIRT1
DN-SIRT1 + NMDA
DN-SIRT1
Calcein-AM
Bright feld
Merge
PI
(a)
Survival rate (%)
0
Control
NMDA
WT-SIRT1 + NMDA
WT-SIRT1
DN-SIRT1 + NMDA
DN-SIRT1
20
40
60
80
100
120
#
⁎
&
⁎
n = 3 per group
(b)
Figure 11: Effects of SIRT1 overexpression on the number of living cells reduced by NMDA in the SH-SY5Y cell line. (a) Repr
micrographs showing the suppression of WT-SIRT1 overexpression on NMDA-induced decrease of living cells (P < 0 05) and no
DN-SIRT1 overexpression on the number of survival cells in the NMDA group (P > 0 05). Living cells were stained by Ca
(green), and dead cells were stained by PI (red). (b) Bar graph of mean of living cells. Each value represents the mean ± S.E.M
independent experiments. ∗P < 0 05 versus the control group, #P < 0 05 versus the NMDA group, &P < 0 05 versus the WT
NMDA Control
NMDA
WT-SIRT1 + NMDA
WT-SIRT1
DN-SIRT1 + NMDA
DN-SIRT1
Calcein-AM
Bright feld
Merge
PI
(a) PI WT-SIRT1 + NMDA (a)
Survival rate (%)
0
Control
NMDA
WT-SIRT1 + NMDA
WT-SIRT1
DN-SIRT1 + NMDA
DN-SIRT1
20
40
60
80
100
120
#
⁎
&
⁎
n = 3 per group
(b) Figure 11: Effects of SIRT1 overexpression on the number of living cells reduced by NMDA in the SH-SY5Y cell line. (a) Representative
micrographs showing the suppression of WT-SIRT1 overexpression on NMDA-induced decrease of living cells (P < 0 05) and no effect of
DN-SIRT1 overexpression on the number of survival cells in the NMDA group (P > 0 05). Living cells were stained by Calcein-AM
(green), and dead cells were stained by PI (red). (b) Bar graph of mean of living cells. Each value represents the mean ± S.E.M. of three
independent experiments. ∗P < 0 05 versus the control group, #P < 0 05 versus the NMDA group, &P < 0 05 versus the WT-SIRT1 +
NMDA group. Figure 11: Effects of SIRT1 overexpression on the number of living cells reduced by NMDA in the SH-SY5Y cell line. Oxidative Medicine and Cellular Longevity 10 function, reduces brain atrophy, and attenuates mutant-
HTT-mediated metabolic abnormalities in a mouse model of
Huntington’s disease [17]. Overexpression of SIRT1 protein
in neurons protects against experimental autoimmune
encephalomyelitis through activation of multiple SIRT1
targets [18]. In addition to the findings in support of the
protective effects of SIRT1 on neurodegeneration, there are
also contradictory studies reporting the opposite effect. In this
respect, it was shown that SIRT1 inhibition reduces IGF-1/
IRS-2/Ras/ERK1/2 signalling and protects neurons [19]. activation and Ace-p53 induces apoptosis and is involved
in neuronal death [26, 27]. Together, these experiments
demonstrate that deacetylation of p53 is at least in part
required for SIRT1-mediated neuroprotection in the exci-
totoxicity model. SIRT1 is an endogenous neuroprotective factor and
mediates protection through different pathways. The mech-
anisms of the neurotoxic effects of NMDA are very complex
including calcium overload, oxidative stress, mitochondrial
dysfunction, cell necrosis, and apoptosis [28]. Nonetheless,
our results suggest that NMDA may inhibit the activity of
SIRT1 and weakens the protective effect of SIRT1. Subse-
quent experimental observation confirmed this speculation,
because SIRT1 activation by RSV or overexpression of
SIRT1 ameliorates NMDA-induced neurotoxicity and exerts
the neuroprotection. g
g
p
Further observation shows that RSV significantly amelio-
rated NMDA-reduced SIRT1 deacetylase activity in primary
neurons, and this amelioration was prevented when SIRT1
activity was inhibited by Sirtinol. Therefore, it raises the pos-
sibility that the deacetylase activity is required for SIRT1’s
neuroprotection in the excitotoxicity model. In addition, we
observed
that
overexpression
of
WT-SIRT1
reversed
NMDA-induced reduction of SIRT1 mRNA, SIRT1 protein
level, and SIRT1 deacetylase activity and inhibition of
NMDA-induced insults of SH-SY5Y cell. However, overex-
pression of DN-SIRT1 increased the levels of SIRT1 mRNA
and protein reduced by NMDA but had no effect on
NMDA-induced decrease in the deacetylase activity and also
did not inhibit subsequent excitotoxic cell death. These
results clearly indicated that SIRT1 deacetylase activity is cru-
cial to the neuroprotective effects of SIRT1 in NMDA-
induced insults. A previous work by a number of other labo-
ratories has also established that RSV potentiates SIRT1
activity and provides neuroprotection in recurrent stroke
models [20], stress resistance, and prosurvival effects [21]. The deacetylase-deficient SIRT1 (H363Y) completely elimi-
nated the protective effects of SIRT1 in HD models [17]. Acknowledgments Additionally, we found that overexpression of WT-
SIRT1 significantly inhibited NMDA-induced p53 acetyla-
tion and subsequent neurotoxicity. However, DN-SIRT1
overexpression has no such effect. The findings suggest that
SIRT1 might provide potent neuroprotection against NMDA
insult through regulating p53 acetylation. As a deacetylase,
SIRT1 is known to deacetylate and modulate the activity of
key transcription factors, such as P53, NF-κB, PGC-1α,
LKB1, TSC2, HSF1, and other substrates, which participate
in the adjustment of the process of a variety of injuries. The
available evidence indicated that SIRT1 reduces the activity
of p53 by removing these acetyl groups that inhibits apopto-
sis and promotes cell survival [24, 25]. In this experiment, we
observed that NMDA induced p53 acetylation which may be
one of the mechanisms of inducing neuronal death via
apoptosis. Acetylation is thought to be a key event for p53 This work was supported by the National Natural Science
Foundation of China (no. 31000481, no. 31171023, and no. 81601167), the Natural Science Foundation of Shanxi Prov-
ince, China (no. 2011011040-2), the Fund for Shanxi Key
Subjects Construction (FSKSC) and the Students’ Innovation
and Entrepreneurship Training Program of Shanxi Province
(no. 2015141). Oxidative Medicine and Cellular Longevity Modulation of sirtuin activity has been shown to impact the
course of several aggregate-forming neurodegenerative dis-
orders including Alzheimer’s disease, Parkinson’s disease,
Huntington’s disease, amyotrophic lateral sclerosis, and spi-
nal and bulbar muscular atrophy [22]. The above evidences
and our results support that SIRT1 deacetylase activity is crit-
ical to its neuroprotection. But there are different opinions
about SIRT1 on neuronal survival that SIRT1-mediated neu-
roprotection is independent of its deacetylase activity, and
this mechanism might involve interactions between SIRT1
and other apoptosis-regulatory proteins [23]. In summary, a growing body of evidence has confirmed
the neuroprotective effects of SIRT1. The finding of the
present study suggests that SIRT1 might be a therapeutic
target for certain neurological diseases related to NMDA-
mediated excitotoxicity. Authors’ Contributions Xiaorong Yang and Peipei Si have contributed equally to
this work. Conflicts of Interest There are no conflicts of interest to declare. Abbreviations SIRT1:
Silent information regulator 1
RSV:
Resveratrol
LDH:
Lactate dehydrogenase
FBS:
Fetal bovine serum
PI:
Propidium iodide
Ace-p53: Acetylated p53
AD:
Alzheimer’s disease
HD:
Huntington’s disease. SIRT1:
Silent information regulator 1
RSV:
Resveratrol
LDH:
Lactate dehydrogenase
FBS:
Fetal bovine serum
PI:
Propidium iodide
Ace-p53: Acetylated p53
AD:
Alzheimer’s disease
HD:
Huntington’s disease. 4. Discussion (a) Representative
micrographs showing the suppression of WT-SIRT1 overexpression on NMDA-induced decrease of living cells (P < 0 05) and no effect of
DN-SIRT1 overexpression on the number of survival cells in the NMDA group (P > 0 05). Living cells were stained by Calcein-AM
(green), and dead cells were stained by PI (red). (b) Bar graph of mean of living cells. Each value represents the mean ± S.E.M. of three
independent experiments. ∗P < 0 05 versus the control group, #P < 0 05 versus the NMDA group, &P < 0 05 versus the WT-SIRT1 +
NMDA group. neuronal insults in in vivo and in vitro models of Waller-
ian degeneration [15, 16]. Further evidence demonstrates
that SIRT1 overexpression can also play a protective role
in a variety of in vivo and in vitro models of nerve
injury. Overexpression
of
SIRT1
improves
motor RSV
ameliorates
motor
neuron
degeneration
and
improves survival mainly through increasing the expres-
sion of SIRT1 in the SOD1G93A mouse model of amyo-
trophic
lateral
sclerosis
[14]. Inhibition
of
SIRT1
increased axonal injury and activation of SIRT1 prevented Oxidative Medicine and Cellular Longevity References Andrechek, and H. Wang,
“NMDA receptor-mediated excitotoxicity depends on the
coactivation of synaptic and extrasynaptic receptors,” Cell
Death & Disease, vol. 4, article e560, 2013. [24] T. S. Anekonda and P. H. Reddy, “Neuronal protection by sir-
tuins in Alzheimer’s disease,” Journal of Neurochemistry,
vol. 96, no. 2, pp. 305–313, 2006. [25] A. F. Paraiso, K. L. Mendes, and S. H. Santos, “Brain activation
of SIRT1: role in neuropathology,” Molecular Neurobiology,
vol. 48, no. 3, pp. 681–689, 2013. [10] Y. Li, X. Yang, C. Ma, J. Qiao, and C. Zhang, “Necroptosis con-
tributes to the NMDA-induced excitotoxicity in rat’s cultured
cortical neurons,” Neuroscience Letters, vol. 447, no. 2-3,
pp. 120–123, 2008. [26] H. Vaghefiand K. E. Neet, “Deacetylation of p53 after nerve
growth factor treatment in PC12 cells as a post-translational
modification mechanism of neurotrophin-induced tumor
suppressor activation,” Oncogene, vol. 23, no. 49, pp. 8078–
8087, 2004. [11] G. A. Abdel-Aleem, E. F. Khaleel, D. G. Mostafa, and L. K. Elberier, “Neuroprotective effect of resveratrol against brain
ischemia reperfusion injury in rats entails reduction of DJ-1
protein expression and activation of PI3K/Akt/GSK3b survival
pathway,” Archives of Physiology and Biochemistry, vol. 122,
no. 4, pp. 200–213, 2016. [27] K. Hasegawa and K. Yoshikawa, “Necdin regulates p53
acetylation via Sirtuin1 to modulate DNA damage response
in cortical neurons,” The Journal of Neuroscience, vol. 28,
no. 35, pp. 8772–8784, 2008. [12] M. Caruana, R. Cauchi, and N. Vassallo, “Putative role of
red wine polyphenols against brain pathology in Alzheimer’s
and Parkinson’s disease,” Frontiers in Nutrition, vol. 3, p. 31,
2016. [28] H. Prentice, J. P. Modi, and J. Y. Wu, “Mechanisms of neu-
ronal protection against excitotoxicity, endoplasmic reticu-
lum stress, and mitochondrial dysfunction in stroke and
neurodegenerative diseases,” Oxidative Medicine and Cellular
Longevity, vol. 2015, Article ID 964518, 7 pages, 2015. [13] M. S. Lopez, R. J. Dempsey, and R. Vemuganti, “Resveratrol
neuroprotection in stroke and traumatic CNS injury,” Neuro-
chemistry International, vol. 89, pp. 75–82, 2015. [14] L. Song, L. Chen, X. Zhang, J. Li, and W. Le, “Resveratrol ame-
liorates motor neuron degeneration and improves survival in
SOD1(G93A) mouse model of amyotrophic lateral sclerosis,”
BioMed Research International, vol. 2014, Article ID 483501,
10 pages, 2014. [15] T. Araki, Y. Sasaki, and J. Milbrandt, “Increased nuclear NAD
biosynthesis and SIRT1 activation prevent axonal degenera-
tion,” Science, vol. 305, no. 5686, pp. 1010–1013, 2004. [16] A. Calliari, N. Bobba, C. References [1] H. S. Ghosh, “The anti-aging, metabolism potential of SIRT1,”
Current Opinion in Investigational Drugs, vol. 9, no. 10,
pp. 1095–1102, 2008. [2] M. W. McBurney, X. Yang, K. Jardine et al., “The mam-
malian SIR2α protein has a role in embryogenesis and Oxidative Medicine and Cellular Longevity 11 gametogenesis,” Molecular and Cellular Biology, vol. 23, no. 1,
pp. 38–54, 2003. multiple SIRT1 targets,” Journal of Immunology, vol. 190,
no. 9, pp. 4595–4607, 2013. [3] G. Asher, D. Gatfield, M. Stratmann et al., “SIRT1 regulates
circadian clock gene expression through PER2 deacetylation,”
Cell, vol. 134, no. 2, pp. 317–328, 2008. [19] Y. Li, W. Xu, M. B. MW, and V. D. Longo, “SirT1 inhibition
reduces
IGF-I/IRS-2/Ras/ERK1/2
signaling
and
protects
neurons,” Cell Metabolism, vol. 8, no. 1, pp. 38–48, 2008. [20] L. M. Wang, Y. J. Wang, M. Cui et al., “A dietary polyphenol
resveratrol acts to provide neuroprotection in recurrent
stroke models by regulating AMPK and SIRT1 signaling,
thereby reducing energy requirements during ischemia,”
The European Journal of Neuroscience, vol. 37, no. 10,
pp. 1669–1681, 2013. [4] J. N. Feige and J. Auwerx, “Transcriptional targets of sirtuins
in the coordination of mammalian physiology,” Current
Opinion in Cell Biology, vol. 20, no. 3, pp. 303–309, 2008. [5] S. Lavu, O. Boss, P. J. Elliott, and P. D. Lambert, “Sirtuins—
novel therapeutic targets to treat age-associated diseases,”
Nature Reviews. Drug Discovery, vol. 7, no. 10, pp. 841–853,
2008. [21] B. L. Tang, “Resveratrol is neuroprotective because it is not a
direct activator of Sirt1-a hypothesis,” Brain Research Bulletin,
vol. 81, no. 4-5, pp. 359–361, 2010. [6] B. L. Tang and C. E. Chua, “SIRT1 and neuronal diseases,”
Molecular Aspects of Medicine, vol. 29, no. 3, pp. 187–200,
2008. [22] A. Z. Herskovits and L. Guarente, “Sirtuin deacetylases in
neurodegenerative diseases of aging,” Cell Research, vol. 23,
no. 6, pp. 746–758, 2013. [7] N. Takei and Y. Endo, “Ca2+ ionophore-induced apoptosis on
cultured embryonic rat cortical neurons,” Brain Research,
vol. 652, no. 1, pp. 65–70, 1994. [23] J. A. Pfister, C. Ma, B. E. Morrison, and S. R. D'Mello, “Oppos-
ing effects of sirtuins on neuronal survival: SIRT1-mediated
neuroprotection is independent of its deacetylase activity,”
PLoS One, vol. 3, no. 12, article e4090, 2008. [8] D. W. Choi, “Glutamate neurotoxicity and diseases of the
nervous system,” Neuron, vol. 1, no. 8, pp. 623–634, 1988. [9] X. Zhou, D. Hollern, J. Liao, E. References Escande, and E. N. Chini, “Resveratrol
delays Wallerian degeneration in a NAD+ and DBC1
dependent
manner,”
Experimental
Neurology,
vol. 251,
pp. 91–100, 2014. [17] M. Jiang, J. Wang, J. Fu et al., “Neuroprotective role of Sirt1 in
mammalian models of Huntington’s disease through activa-
tion of multiple Sirt1 targets,” Nature Medicine, vol. 18,
no. 1, pp. 153–158, 2011. [18] V. K. Nimmagadda, C. T. Bever, N. R. Vattikunta et al., “Over-
expression of SIRT1 protein in neurons protects against exper-
imental autoimmune encephalomyelitis through activation of
|
https://openalex.org/W4391275244
|
https://link.springer.com/content/pdf/10.1007/s13369-023-08672-1.pdf
|
English
| null |
A Deep Ensemble Approach for Long-Term Traffic Flow Prediction
|
Arabian journal for science and engineering
| 2,024
|
cc-by
| 10,719
|
Arabian Journal for Science and Engineering (2024) 49:12377–12392
https://doi.org/10.1007/s13369-023-08672-1 Arabian Journal for Science and Engineering (2024) 49:12377–12392
https://doi.org/10.1007/s13369-023-08672-1 RESEARCH ARTICLE-COMPUTER ENGINEERING AND COMPUTER SCIENCE RESEARCH ARTICLE-COMPUTER ENGINEERING AND COMPUTER SCIENCE Abstract In the last 50 years, with the growth of cities and increase in the number of vehicles and mobility, traffic has become
troublesome. As a result, traffic flow prediction started to attract attention as an important research area. However, despite
the extensive literature, traffic flow prediction still remains as an open research problem, specifically for long-term traffic
flow prediction. Compared to the models developed for short-term traffic flow prediction, the number of models developed
for long-term traffic flow prediction is very few. Based on this shortcoming, in this study, we focus on long-term traffic flow
prediction and propose a novel deep ensemble model (DEM). In order to build this ensemble model, first, we developed a
convolutional neural network (CNN), a long short-term memory (LSTM) network and a gated recurrent unit (GRU) network
as deep learning models, which formed the base learners. In the next step, we combine the output of these models according
to their individual forecasting success. We use another deep learning model to determine the success of the individual models. Our proposed model is a flexible ensemble prediction model that can be updated based on traffic data. To evaluate the
performance of the proposed model, we use a publicly available dataset. Experimental results show that the developed DEM
model has a mean square error of 0.06 and a mean absolute error of 0.15 for single-step prediction; it shows that achieves a
mean square error of 0.25 and a mean absolute error of 0.32 for multi-step prediction. We compared our proposed model with
many models in different categories; individual deep learning models (i.e., LSTM, CNN, GRU), selected traditional machine
learning models (i.e., linear regression, decision tree regression, k-nearest-neighbors regression) and other ensemble models
such as random-forest regression. These results also support the claim that ensemble learning models perform better than
individual models. Keywords Deep learning · Traffic flow prediction · Ensemble learning · Long short-term memory · Convolutional neural
networks · Gated recurrent unit Keywords Deep learning · Traffic flow prediction · Ensemble learning · Long short-term memory · Convolutional neural
networks · Gated recurrent unit A Deep Ensemble Approach for Long-Term Traffic Flow Prediction Nevin Cini1,2
· Zafer Aydin2 Nevin Cini1,2
· Zafer Aydin2 Received: 3 May 2023 / Accepted: 21 December 2023 / Published online: 27 January 2024
© The Author(s) 2024 1
Artificial Intelligence Research and Development Center,
¸Sanlıurfa, Turkey 2
Department of Computer Engineering, Abdullah Gul
University, Kayseri, Turkey 1 Introduction pollution and even accidents from time to time. Although
increasing the cost of travel is an economic problem, increas-
ing fuel consumption also leads to environmental problems
such as air pollution [1]. Traffic congestion causes many problems that we can exam-
ine under different headings such as economic, environmen-
tal and social. Among them, the most emphasized is the
increase in cost with the lengthening of the travel time. These
two key issues lead to the emergence of other problems. For example, prolonged travel time causes social and psy-
chological problems, environmental problems such as noise Decision makers who reach information about when and
where traffic congestion may occur with traffic flow fore-
casting can direct drivers to safer roads so that resources can
be used more efficiently. With a more effective planning, it
is possible to use public transportation more efficiently as
well. In this way, the environmental impact caused by traffic
can be reduced. For this reason, traffic flow forecasting is
of key importance in controlling traffic congestion and solv-
ing many problems that may occur, and is an indispensable
component for intelligent transportation systems [2]. That is why, in recent years, many studies in this area have
focused on developing reliable and realistic traffic flow pre-
diction models using the latest technologies [3–5]. However,
most of these studies have presented short-term traffic flow 12 3 Arabian Journal for Science and Engineering (2024) 49:12377–12392 12378 Fig. 1 Different performances of base learners prediction models [6]. Few of the proposed forecasting mod-
els are capable of long-term forecasting. However, long-term
forecasting is as important and useful as short-term forecast-
ing [1, 7, 8]. Furthermore, long-term traffic flow forecasting
is of practical importance for decision makers. An accurate
forecast model will facilitate traffic management even during
the rush hours, and will enable effective measures to be taken
by informing in advance of possible negative events. However, long-term forecasting is a challenging task. This
is due to the stochastic nature of the dynamics that make
up the traffic flow data, which is nonlinear and contains
complex dependencies [9]. It is also not identical in both
temporal and spatial dimensions. Modeling dynamic tem-
poral and spatial dependencies for traffic flow prediction is
very burdensome and arduous. These complex dependencies
increase in number and become more and more complex
in long-term predictions. 1 Introduction As the forecast horizon increases,
even in the best models, the prediction quality decreases and
the average error increases [10]. Fig. 1 Different performances of base learners which we compared the proposed model with widely used
prediction models. There are three main contributions of this study: As a result, reliable long-term prediction becomes a dif-
ficult task, and it is almost impossible to model long-term
dependencies of traffic flow with simple and traditional pre-
diction models [8, 11, 12]. • In this study, we proposed a fully DL-based ensemble
learning framework for long-term traffic flow prediction. • We used three different DL models as base learners. In the
model we developed, we use LSTM and GRU together. We have not come across a model in the literature that
uses these two techniques together. Since these two tech-
niques are versions of recurrent neural networks, it is
not preferred to use them together in a prediction model. However, although these two techniques are similar to
each other, their performances are quite different as seen
in Fig.1. So where one fails, the other can be quite suc-
cessful. For this reason, we preferred to use these two
techniques together. In this study, we propose a deep learning-based ensem-
ble framework for long-term traffic flow prediction. While
deep learning (DL) models can learn dynamic and complex
dependencies of traffic data better than traditional learn-
ing algorithms, ensemble learning (EL) provides flexibility
by increasing the generalization ability of the final model. Because many different predictive models collaborate to
solve the given problem in ensemble learning, it is often
expected that the ensemble model will exceed the predictive
success of a single model. • We use deep learning architectures in our model, both
as base learners and the meta-learner. Thanks to a feed
forward neural network (FFNN), which is the most basic
deep learning technique, we decide the weights of the
base learners. We train a feed forward neural network as a
meta-learner in order to obtain the final prediction result. In this way, we ensured that the base learners dynami-
cally contribute in the final prediction result according to
their prediction success (more successful ones contribute
more, less successful ones contribute less). • We use deep learning architectures in our model, both
as base learners and the meta-learner. 2.1 Traffic Flow Prediction Traffic flow refers to the number of vehicles passing a certain
road section per unit time. This data is collected automati-
cally,usuallywiththehelpofsensors.Sincevehiclescanonly
move on the roads prepared for them, accurate estimation of
the traffic flow in a certain area prevents possible congestion
and ensures more efficient use of the roads [1, 2, 4]. 3 Related Technologies f s
t =
⎡
⎢⎢⎢⎢⎢⎢⎣
f 1
1
f 1
2
f 1
3
. . . f 1
t
f 2
1
f 2
2
f 2
3
. . . f 2
t
. . . . . . . . . . . . . . . . . . . . . f s
1
f s
2
f s
3
. . . f s
t
⎤
⎥⎥⎥⎥⎥⎥⎦
(1) f s
t =
⎡
⎢⎢⎢⎢⎢⎢⎣
f 1
1
f 1
2
f 1
3
. . . f 1
t
f 2
1
f 2
2
f 2
3
. . . f 2
t
. . . . . . . . . . . . . . . . . . . . . f s
1
f s
2
f s
3
. . . f s
t
⎤
⎥⎥⎥⎥⎥⎥⎦
(1) f s
t =
⎡
⎢⎢⎢⎢⎢⎢⎣
f 1
1
f 1
2
f 1
3
. . . f 1
t
f 2
1
f 2
2
f 2
3
. . . f 2
t
. . . . . . . . . . . . . . . . . . . . . f s
1
f s
2
f s
3
. . . f s
t
⎤
⎥⎥⎥⎥⎥⎥⎦
(1) In this section, the related technologies used in the proposed
model are described. (1) 2.2 Problem Formulation Traffic flow forecasting models are often based on a simple
assumption: the future depends on the past. In other words,
data that generated traffic conditions in the past will affect
current and future traffic situation. Therefore, continuity of
data is important. Traffic flow prediction is a time series prob-
lem, and as with all time series problems, past values are
used as target function parameters in the traffic flow esti-
mation problem. In other words, the target/prediction value
at time Tn becomes one of the target function parameters at
time T(n+1). This is for single-step prediction. In multi-step
prediction, more than one value at consecutive time steps par-
ticipates in the process at the same time. To formulate this
problem mathematically, we use the notation f i
t to define
traffic flow from station i at time t. In order to extract spa-
tial and temporal features of traffic flow here, we construct
spatial-temporal feature matrix as follows: 2 Background Here, since we use historical flow data to predict future
flows, the matrix on the left hand side represents historical
flow data and the matrix on the right hand side represents
prediction values. The traffic flow prediction model is repre-
sented by a prediction function, which is represented by θ. f d denotes traffic flow from station d. β is the looked-back
steps, and h is the prediction horizon. In this section, we will define the traffic flow prediction prob-
lem. 1 Introduction Thanks to a feed
forward neural network (FFNN), which is the most basic
deep learning technique, we decide the weights of the
base learners. We train a feed forward neural network as a
meta-learner in order to obtain the final prediction result. In this way, we ensured that the base learners dynami-
cally contribute in the final prediction result according to
their prediction success (more successful ones contribute
more, less successful ones contribute less). The most important feature of the proposed EL model
is that we employ three different DL models (i.e., CNN,
LSTM and GRU) as base learners. This increases model
diversity so that a failure of one model can be compensated
by another model. As shown in Fig.1, the performances of
all three models change as traffic conditions change. From
this figure, it is clear that we cannot achieve the best predic-
tion performance with a single model. Because each model
has strengths and weaknesses, the contribution of the base
models to the final prediction result cannot be equal. In a
successful ensemble model, a base model with high predic-
tive performance is expected to contribute more to the final
result than less successful models. In our ensemble model,
we have developed a meta-learner to provide this. Owing to
this meta-learner, we have dynamically weighted the base
models, that is, we have ensured that each model contribute
in the final prediction result according to its current predic-
tion performance. We leverage this capability of ensemble
learning to improve long-term prediction accuracy. In order
to assess the accuracy, we conducted several experiments, in The paper is organized as follows: In Sect. 2, we provide
a background section. We explain the related technologies in
Sect. 3. A brief overview of current literature on traffic flow
prediction is provided in Sect. 4. In Sect. 5, we introduce
the details of our deep learning-based ensemble framework. Then, we present dataset, preprocessing steps, experimental
results and discussion in Sect. 6. Section 7 includes conclu-
sion and future work. 12 3 12379 Arabian Journal for Science and Engineering (2024) 49:12377–12392 2.3 The Differences Between Short- and Long-Term
Prediction In fact, the difference between the short- and long-term fore-
cast goes far beyond the period we determine with only the
prediction horizon. In the literature, long-term forecasting is
categorized as predicting an hour later or a few steps later
(usually five steps or more), while short-term forecasting is
defined as predicting one step or a few minutes later. Here,
we can say that a categorization based on this definition is
not reliable due to the lack of a consensus in terms of the
prediction horizon. However, according to the assessments
made taking into account the time interval of the data, it is
reasonable in our opinion to consider five steps and beyond
as a long-term forecast. However, we believe that it would
be a more correct approach to classify forecast models that
can make reliable forecasts not only for the specified time
horizon but also beyond, as long-term forecast models, with-
out taking into account the prediction horizon of the model. Therefore, in this study, we test our model with several time
horizons and compare their performance. 3.2 Convolutional Neural Network (CNN) According to this formula, the ensemble learning model
gives weight to each individual model. The most common
approach in the literature, for this purpose, is to give equal
weight to each model. One issue of this approach is that each
model conduces equally to the final prediction, without con-
sidering the prediction performance of single models. When
we give a fixed weight to each single model, we limit the
performance of the ensemble model due to a reduction in its
generalization ability. Therefore to improve the prediction
accuracy, we propose a flexible and robust deep ensemble
model in this study. The proposed model assigns the weights
based on the individual model performance and traffic situa-
tion change. Since CNN was not developed for time series problems, it
was not used for time series prediction for a long time. How-
ever, with the increase in the amount of data, the increase in
computational load and the inability to parallelize the RNN
algorithm efficiently led to new searches. CNN is promising
for time series problems as it can be parallelized and pro-
duces faster results. In recent years, successful CNN-based
time series forecasting models have been developed. Due to
its architecture, CNN is used to reveal the relationships of dif-
ferent time series, especially in problems that need learning
temporal dependencies and spatial dependencies together. A
simple CNN model includes convolution layers, pooling lay-
ers, fully connected layers (FC) and an output layer. Filters
are used in the convolution and pooling layers and the results
are combined in the FC layer. In this way, learning is provided
at each convolution layer. 3.1 Recurrent Neural Network (RNN) ThismodelisoneofthemostimportantDLtechniquespartic-
ularly developed for time series problems. RNN has a simple
feedback loop in order to learn dependencies among the dif-
ferent time intervals. However, the basic RNN architecture
is insufficient to capture complex relationships in long time
intervals, so two different versions have been proposed. The
first of these versions is long short-term memory (LSTM) and
the other is gated recurrent network (GRU). LSTM has three
different gates (input, output, forget) while GRU has two
different gates (update and reset) and by the agency of these
gates, they remove unnecessary information from the model
that comes from the past states, and allowing the model to
focus on only useful information. In this way, the model can
learn long-term dependencies with ease. Figure2 presents
the general structure of the RNN, LSTM and GRU. Here, s denotes the number of stations. We construct this
flow matrix with temporal information horizontally and spa-
tial information vertically. In the next step, we can formulate
the traffic flow prediction problem as follows: ⎡
⎢⎢⎢⎢⎢⎢⎢⎢⎢⎣
f d
t−β
f d
t−(β−1)
f d
t−(β−2)
. . . f d
t−1
⎤
⎥⎥⎥⎥⎥⎥⎥⎥⎥⎦
T
θ−→
⎡
⎢⎢⎢⎢⎢⎢⎢⎢⎢⎣
f d
t
f d
t+1
f d
t+2
. . . f d
t+(h−1)
⎤
⎥⎥⎥⎥⎥⎥⎥⎥⎥⎦
T
(2) (2) 12 3 Arabian Journal for Science and Engineering (2024) 49:12377–12392 12380 Fig. 2 Structure of a RNN (on the left), LSTM (in the middle) and GRU
(on the right) individual model deals with one aspect of the final model, as
a result, the final model provides a more general represen-
tation and achieves a higher predictive accuracy compared
to individual models. To this end, we focus on ensemble
learning approaches in this study and propose a novel deep
ensemble model for traffic flow prediction. The formula for
an ensemble model is as follows: FPM(t) =
K
k=1
Wkαk(t). (3) (3) Fig. 2 Structure of a RNN (on the left), LSTM (in the middle) and GRU
(on the right) where FPM is final prediction model, αk is the kth individual
model, Wk is the weight of the kth individual model, and K
is the number of individual models. 3.3 Deep Ensemble Learning Ensemble learning models combine several base or individ-
ual models with different strategies in order to provide better
generalization and improve final prediction performance [13,
14]. Moreover, today, deep learning models with complex
and layered architecture outperform traditional prediction
models. Deep ensemble learning models, on the other hand,
aim to build a more successful prediction model by com-
bining the peculiar advantages of these two models. There
are many models developed for traffic flow prediction in
the literature, but few of them are ensemble learning-based. However, ensemble learning-based models provide higher
accuracyandgeneralizabilitybecausetheyareconstructedby
combining either individual models developed with different
combinationsofthesamemethodorindividualmodelsdevel-
oped using different methods. Combining multiple models in
thiswayfortrafficflowforecastingcanincreasethefinalfore-
casting accuracy while preventing overfitting. Because each 4 Related Work The importance of traffic flow prediction in transportation
engineering is increasing, and accordingly, we can say that
there is a very large literature in this field. Most studies pro-
pose a model to predict traffic flow. We will examine these
proposed models under two topics by following the tradi-
tion in the literature: parametric models and nonparametric
models [3, 12, 15–19]. We summarize the related literature
in Table 1. 4.1 Parametric Models Models in this class can be explained by traffic flow theories
of transportation engineering, statistics and probability. In a
parametric model, traffic flow is represented as a function
of random variables (e.g., accident, instantaneous decisions
of drivers), time-dependent variables (e.g., time of day, day
of the week or season) and auxiliary variables (e.g., weather,
public holidays, sports or concert events). That is, traffic flow
is defined as the total number of vehicles passing through
a certain road segment at a certain time period under the
influence of many dependent or independent variables, each 12 3 Arabian Journal for Science and Engineering (2024) 49:12377–12392 12381 Arabian Journal for Science and Engineering (2024) 49:12377–12392 Table 1 Summary of the related literature
References
Horizon
Input data
Data size
Method/technologies
used
Evaluation metrics
[9]
24h
Highways agency
network traffic
flow data
15min resolution
from 1 Jul 2018 to
28 Jan 2020
Wavelet
decomposition,
convolutional
neural
network-long and
short- term
memory neural
network
Root mean square
error (RMSE)
mean absolute
error (MAE), R
square
[7]
Up to 24h
Caltrans
performance
measurement
system (PeMS)
dataset
5min interval. 4.1 Parametric Models Data
size not mentioned
LSTM
encoder–decoder
RMSE and
symmetric mean
absolute
percentage error
(SMAPE)
[1]
24h
Dataset obtained
from the
DRIVENET
Between February 1,
2015 to March 31,
2016
Deep neural network
(DNN)
Absolute percentage
error (APE), mean
absolute
percentage error
(MAPE)
[8]
Up to 4h
GPS-data taken from
the GAIA
The data contains
trips from October
to November 2016
Graph CNN-LSTM
neural network
RMSE, MSE, MAE
and MAPE
[12]
Up to 1h
Urban corridor of 30
road segments with
24 intersections
along Victoria
Street (Melbourne)
One-year data of the
year 2016
Convolutional GRU
with attention
network
Weighted mean
absolute
percentage error
(WMAPE),
RMSE, MAE
[16]
Single step
(unspecified)
PeMS
The 5-min traffic
flow data of
District 5 named
Central Coast in
2013
Ensemble learning,
CNN
MAE, RMSE, mean
relative error
(MRE) and the
standard deviations
MAE, MRE and
RMSE
[17]
Up to three-step
Data from the
Portland-
Vancouver
Metropolitan
region
During a 4-month
period from March
4 to June 28, 2019
Ensemble learning,
ensemble
empirical mode
decomposition,
DBN (deep belief
networks)
RMSE, MAPE
[22]
1-h
Princes highway,
Victoria Road,
Canterbury Road,
and M1 in Sydney
Hourly traffic count
from November
2017 to November
2018
Ensemble learning,
ARIMA
RMSE, MAPE
[30]
5h
Data from Hangzhou
Integrated
Transportation
Research Center
and PeMSD10
From 16th October,
2013 to 3rd
October, 2014 and
from 1st January,
2018 to 31st
March, 2018
(15min resolution)
Graph convolutional
network, recurrent
neural network
RMSE and MAPE
[33]
1h (12 steps)
PeMSD4 and
PeMSD8
From January to
February in 2018
and from July to
August in 2016
Encoder–decoder,
attention network,
LSTM
RMSE, MAE,
MAPE, median
absolute error
(MdAE), mean [33] 12 Arabian Journal for Science and Engineering (2024) 49:12377–12392 12382 Table 1 continued
References
Horizon
Input data
Data size
Method/technologies
used
Evaluation metrics
[38]
1h (4 steps)
Arterial sensors in
Arcadia, CA in
2015
15-min interval data
Ensemble learning,
ARMAX, partial
least squares,
support vector
regression, kernel
ridge regression,
Gaussian process
regression
MAE and StdAE
(standard
deviation)
[39]
Up to 30 min
PeMS (“freeway
segment located in
San Diego”)
From September 1,
2019 to September
30, 2019, and the
sampling time
interval is 5 min
Ensemble empirical
mode
decomposition,
wavelet, LSTM
RMSE, MAE,
MAPE
[47]
Up to 30 min
Data of Yuanda
Road, Furong
District, Changsha
City
From September to
October in 2013,
excluding weekend
data for a total of
40 days, time
interval is 5 min
Optimized
variational mode
decomposition
(OVMD) and
combined long
short-term memory
network (LSTM)
RMSE, MAE of which is dynamic in itself. 4.1 Parametric Models Modeling with parametric
approaches is relatively easy, but these models are suitable
for uncomplicated small-sized datasets [20]. defined patterns, future traffic patterns are tried to be pre-
dicted [25]. Another parametric model used in traffic flow predic-
tion is SVM [15]. Although the estimation accuracy can
be increased by using different “kernels,” the computational
load of model training is quite high, especially compared to
KNN and NN. Therefore, it is not practical for large datasets. Indeed, KNN and SVM are not popular models developed for
trafficflowforecast.Themostpopularmodelsinthiscategory
are the NN-based models. And the reasons why NN-based
models are so popular can be listed as follows: (1) they are
suitable for big data, (2) they have fast convergence, and (3)
they can achieve high prediction accuracy. A wide variety of
NN models have been proposed for traffic flow prediction [1,
7, 12, 18]. The most widely used parametric approaches in the liter-
ature are ARIMA, Kalman filtering and linear regression. ARIMA is a time series modeling approach that explores
the temporal relationship between data points of a time series. There are many traffic flow forecasting models developed
using ARIMA and its advanced versions i.e., ARIMAX,
SARIMA, SARIMAX in the literature [21, 22]. Kalman filtering is a widely used traffic flow prediction
method. Its main idea is to predict future traffic flow using
historical traffic flow data with a recursive or iterative process
[23]. Linear regression is a pretty simple parametric approach. This method describes the traffic flow as a linear combination
of the independent variables [24]. 4.2 Nonparametric Models Although deep learning models are also nonparametric mod-
els, we wanted to examine these studies separately since they
have been very popular in this field in recent years and have
a fairly wide literature. The simplest DL models that can be
found in the literature in this field are multilayer-perceptron
(MLP)-based models developed using multiple hidden layers
[26]. However, the most widely used DL technique in solv-
ing the traffic flow prediction problem is recurrent neural
networks (RNN). Especially, GRU and LSTM techniques,
which are variants of RNN, are the most common meth-
ods since they are successful in capturing dependencies at
different times. For example, [27] developed a two-layer Models in this category are more advanced than parametric
models, and their performance varies according to the quality
and size of the dataset. These models can achieve satisfactory
predictionsuccesswithbigdata,butthisrequiresquitealotof
computational capacity. K-nearest neighbor (KNN), support
vector machine (SVM) and neural networks (NN) are the
approaches we can count in this category. KNN can be used for classification or regression. In this
model, common patterns are tried to be extracted from his-
torical traffic flow data. By using the best match with these 123 Arabian Journal for Science and Engineering (2024) 49:12377–12392 12383 LSTM-based model. It used a fully connected layer as the
extraction layer in the first layer, and the LSTM layer as
the prediction layer in the second layer. The proposed other
LSTM-based models are in [6]. A GRU-based model is pro-
posed in [28]. In this study, weather data was used in addition
to traffic data. Apart from RNN, CNN-based models also
have been proposed for short-term traffic flow forecasting
problems [29–31]. CNN-based models are especially pre-
ferred because they can produce results faster than other
neural networks [32]. mize the internal parameters of each base model and select
the best models with the highest accuracy. After selecting
the base models, in the third stage, we decide how much
each base model will contribute to the final model according
to their performance. That is we develop a meta-learner to
dynamically weight each base learner. For this, we first form
new training, validation and test sets using each base model,
then by using these new datasets we build a feed forward
neural network-based (FFNN) model with deep architecture
and the outputs of this final model (i.e., FFNN) or meta-
learner are the weights of each base model. 1 Source: www.transportation.gov/data, and it is available at: https://
cloud.google.com/bigquery/public-data. 4.2.2 Hybrid and Ensemble Models Understanding that it is not possible to model the complex-
ity of traffic data with simple and traditional methods, many
researchers have turned to hybrid models, especially in recent
years. While in early studies we can see the combination of
several parametric models, in recent studies, many of the
hybrid models were built by combining two or more non-
parametric methods [7, 8, 11, 33, 34]. Especially LSTM and
CNN are used together in recently developed hybrid mod-
els [30, 35, 36]. There are also hybrid models developed by
using parametric and nonparametric methods together [11,
37]. As we mentioned in Sect.3.3, an ensemble model can be
built in two different approaches: It can be built by combining
either individual models developed with different combina-
tions of the same method or individual models developed
using different methods. The novelty of our model is that it
combines these two approaches. Moreover, the base learners
and meta-learner we use in our model all have deep architec-
ture, and we do not use a fixed weight for each base learner,
we introduce a meta-learner with the ability of dynamically
weighting the base learners according to their predictive suc-
cess. On the other hand, EL-based models emerge as a new
trend [38–43]. There are only a limited number of EL-based
prediction models in the literature [16, 17, 22, 44–46]. How-
ever, none of these studies focus on long-term forecasting. And this is a research gap that we want to fill in this study. 4.2 Nonparametric Models Thus, the weight
of each base model is determined automatically and dynam-
ically. Meanwhile, in order to capture the traffic condition
changes, we use raw input data as well during the construc-
tion of the final model. Consequently, we separate the base
models weighting step from the base models selection and
tuning step so that the ensemble model can be dynamic and
can change with the traffic conditions. LSTM-based model. It used a fully connected layer as the
extraction layer in the first layer, and the LSTM layer as
the prediction layer in the second layer. The proposed other
LSTM-based models are in [6]. A GRU-based model is pro-
posed in [28]. In this study, weather data was used in addition
to traffic data. Apart from RNN, CNN-based models also
have been proposed for short-term traffic flow forecasting
problems [29–31]. CNN-based models are especially pre-
ferred because they can produce results faster than other
neural networks [32]. 6.1 Dataset and Preprocessing The proposed model is a deep ensemble model which is
capable of properly fusing the prediction results of multiple
deep learning models. Our model learns the strengths and
weaknesses of individual models and weights the predictions
of single models according to their prediction performance. In addition, our model is flexible and performs well under
different traffic conditions since our model receives actual
data as well as prediction results from each model to obtain
the final result. We conducted this study with a publicly available and a real-
world dataset.1 The dataset contains a total of 274 stations. The data was collected from January 1st, 2015 to Decem-
ber 31st, 2015, which contains both weekends and weekdays
and aggregated 1h intervals. Although the dataset contains
274 stations, some stations only have data for 3–4 months,
for instance, station 116820 has data only for the 2nd, 9th,
11th and 12th months. That is, for some stations there are
too many missing values, and this disrupts the continuity of
the dataset. However, this is not desirable for time series and
can significantly reduce the forecasting quality. Therefore,
both because our computational resources are limited and
because we want our model to produce more reliable pre-
dictive results, we have selected 100 stations with as few Figure 3 demonstrates the details of our deep ensemble
traffic flow prediction framework. Our proposed model con-
sists of three stages. The first stage is the preprocessing and
dataset preparation. We will explain this stage in detail in
the next section. The second stage is base model selection. At this stage, we adjust the configurations of the three base
models, namely LSTM, GRU and CNN. For this, we run
models LSTM, GRU and CNN multiple times with different
time lags, numbers of hidden layers and neurons. We opti- 12 3 Arabian Journal for Science and Engineering (2024) 49:12377–12392 12384 Fig 3 General structure of the proposed model Fig. 3 General structure of the proposed model the last two months as the validation set, and the remaining
part as the test set. And we performed Z-Score normalization. missing values as possible and we tested all prediction mod-
els by using these 100 stations. Figure4 shows the first 45
stations we have selected and Fig.5 shows the locations of
the selected stations. 6.1 Dataset and Preprocessing We filled the missing values of the stations used in the
experiments by averaging the data of the previous and the
next hour. Thus, the total number of data samples is 100 ∗
365. We chose this method to fill the missing data because
data that is closer together, whether spatially or temporally,
is more related to each other than data that are far apart. This
idea is based on Tobler’s first law. Tobler’s first law says that
things that are close together are more related to each other
[22]. Inspired by this, we used this method to fill the missing
data. 123 flow matrix as input (X) and output (Y) matrix as follows: Xs1
h =
⎡
⎢⎢⎣
f s1
t
f s1
(t+1)
. . . f s1
t(W−1)
f s1
(t+d) f s1
(t+d+1) . . . f s1
t(W−1)+d
:
:
. . . :
:
:
. . . :
⎤
⎥⎥⎦
(4)
Y s1
h
=
⎡
⎢⎣
f s1
t(W−1)+1 f s1
t(W−1)+2 . . f s1
t(W−1)+h
f s1
t(W−1)+2 f s1
t(W−1)+3 . . f s1
t(W−1)+h+1
:
:
. . :
:
:
. . :
⎤
⎥⎦
(5) (4) (5) In Eqs. 4 and 5, f s1
t
indicates the traffic flow of station
1 at time t. h represents prediction horizon, W denotes time
lag (or time-window size), and d is the stride value. Fig. 5 Road network used for experiments 3 www.keras.io. 4 www.scikit-learn.org. 2 www.tensorflow.org. 6.2 Constructing Traffic Flow Matrix We tried to find the optimum time lag by running each deep
learning model (base learner) many times with different time
lags, i.e., the current traffic flow depends on how many steps
in the past traffic flow. Thus, we have obtained an optimum
time lag for each base learner. If we show the time-lag value
with W; we set W hours as the time lag and added W new
features, each of which indicates hourly traffic volumes in
a W-h period. In this way, prediction models try to predict
the traffic volume in the (W + 1)th hour by using previous
W hours of data. We tested the proposed model for multiple
horizon values: The prediction horizon h is specified as 1
for single-step prediction, and 2, 3, 4, 5, 9, 12, 24 for multi-
step prediction (i.e., long-term prediction). That is, we used
W hours historical data to predict the following h hour(s)
traffic flow value. Accordingly, we constructed the traffic While choosing the stations we will use in our experi-
ments, we also took into account the road type to which the
station belongs, in addition to the amount of data because we
wanted to show how robust and generalizable our model is
for different road types. The road types we use are shown in
Table 2. We separated the dataset into three: We organized 65% of
the dataset (about the first eight months) as the training set, 123 Arabian Journal for Science and Engineering (2024) 49:12377–12392
12385
Fig. 4 Dataset (the first 45 stations)
flow matrix as input (X) and output (Y) matrix as follows:
Xs1
h =
⎡
⎢⎢⎣
f s1
t
f s1
(t+1)
. . . f s1
t(W−1)
f s1
(t+d) f s1
(t+d+1) . . . f s1
t(W−1)+d
:
:
. . . :
:
:
. . . :
⎤
⎥⎥⎦
(4)
Y s1
h
=
⎡
⎢⎣
f s1
t(W−1)+1 f s1
t(W−1)+2 . . f s1
t(W−1)+h
f s1
t(W−1)+2 f s1
t(W−1)+3 . . f s1
t(W−1)+h+1
:
:
. . :
:
:
. . :
⎤
⎥⎦
(5)
In Eqs. 4 and 5, f s1
t
indicates the traffic flow of station
1 at time t. h represents prediction horizon, W denotes time
lag (or time-window size), and d is the stride value. 2 www.tensorflow.org.
3 www.keras.io. 6.2 Constructing Traffic Flow Matrix 6.3 Experiments Settings
TensorFlow2 and Keras,3 which are open-source libraries
of Python, were used to build the proposed deep ensemble
2 www.tensorflow.org. 3 www.keras.io. Fig. 5 Road network used for experiments
model and other deep learning models. We used the scikit-
learn4 as the machine learning library to implement the linear
regression (LR), KNN, DT, RF models. 4 www.scikit-learn.org. 1 Arabian Journal for Science and Engineering (2024) 49:12377–12392 12385 Fi
4 D t
t (th fit 45 t ti
) Fig. 4 Dataset (the first 45 stations) Fig. 4 Dataset (the first 45 stations) Fig. 5 Road network used for experiments Fig. 5 Road network used for experiments flow matrix as input (X) and output (Y) matrix as follows: 6.4 Comparison Metrics Table 2 Road types
Urban: principal arterial—other
Urban: principal arterial—other freeways or expressways
Urban: principal arterial—interstate
Urban: minor arterial
Rural: principal arterial—other
Rural: minor arterial
Rural: principal arterial—interstate
Rural: major collector We use four metrics to measure the performance of the devel-
oped models, mean absolute error (MAE), mean squared
error (MSE), mean squared logarithmic error (MSLE) and
R-squared score which are the most frequently used metrics
for traffic forecasting. MAE, MSE, MSLE, R2 are defined as: MAE, MSE, MSLE, R2 are defined as: MAE, MSE, MSLE, R2 are defined as: MSE = 1
T
T
n=1
(tn −pn)2. (6)
MAE = 1
T
T
n=1
|tn −pn|. (7)
R2 = 1 −
T
n=1(tn −pn)2
T
n=1(tn −v)2 . (8)
MSLE = 1
T
T
n=1
(log(1 + tn) −log(1 + pn))2. (9) (6) We made a lot of trials to determine the best time lag. As
a result of these trials, we found that the best time lag is 24h
for all models. (7) We optimized the hyper-parameters of each model sepa-
rately. For deep learning models, the number of hidden neu-
rons, activation function, dropout rate and learning rate were
optimized by using ‘Bayesian Search’ algorithm. Table 3
shows the hyper-parameter values that we obtained as a result
of optimization for each deep learning model. We used ‘Ran-
dom Search’ algorithm for optimizing hyper-parameters of
LR, KNN, DT, RF models. The Adam algorithm is used to
optimize the loss function of all deep learning models and
the ensemble model. The maximum number of epochs is set
to 100; however due to early stopping, there was no model
that reached 100 epochs. (8) (9) where t, p and T indicate the actual value, prediction value
andthetotalnumberofsamples,respectively.Andv indicates
that mean value of the actual traffic flow data. 6.3 Experiments Settings model and other deep learning models. We used the scikit-
learn4 as the machine learning library to implement the linear
regression (LR), KNN, DT, RF models. model and other deep learning models. We used the scikit-
learn4 as the machine learning library to implement the linear
regression (LR), KNN, DT, RF models. TensorFlow2 and Keras,3 which are open-source libraries
of Python, were used to build the proposed deep ensemble 4 www.scikit-learn.org. 12 123 12386 Arabian Journal for Science and Engineering (2024) 49:12377–12392 6.4 Comparison Metrics Table 3 Optimum
hyper-parameter settings
Model
Parameter
Value
LSTM hyper-parameters (base learner)
Number of layers
4
Number of units
512,512,32,32
Activations
relu, relu, relu,tanh
Dropout rate
0.0
Learning rate
0.0001
GRU hyper-parameters (base learner)
Number of layers
4
Number of units
512, 512, 32, 96
Activations
tanh,relu,relu,relu
Dropout rate
0.5
Learning rate
0.0001
CNN hyper-parameters (base learner)
Number of hidden layers
3
Number of units
512, 96, 128
Filter size
64
Activations
tanh, relu, tanh, relu
Dropout rate
0.0
Learning rate
0.0001
Meta-Learner hyper-parameters
Number of layers
4
Number of units
352, 512, 96, 416
Activations
tanh, tanh, tanh, tanh
Dropout rate
0.2
Learning rate
0.0001 Table 3 Optimum
hyper-parameter settings Table 3 Optimum
hyper-parameter settings 12 123 Arabian Journal for Science and Engineering (2024) 49:12377–12392 12387 Arabian Journal for Science and Engineering (2024) 49:12377–12392
12387
Table 4 Comparison of prediction performances of the proposed ensemble model and other competitive models for eight time horizons
Prediction horizon (h)
Metrics
LR
DT
KNN
RF
LSTM
GRU
CNN
Ensl
Ens2
1
MSE
0.1007
0.1162
0.0882
0.0840
0.0676
0.0687
0.0676
0.0641
0.0613
MAE
0.2079
0.2065
0.1728
0.1740
0.1656
0.1657
0.1675
0.1590
0.1553
2
MSE
0.1501
0.1702
0.1218
0.1245
0.1139
0.1095
0.0989
0.0899
0.0862
MAE
0.2510
0.2484
0.2057
0.2075
0.2252
0.2166
0.2039
0.1889
0.1840
3
MSE
0.1918
0.2196
0.1460
0.1640
0.1499
0.1195
0.1271
0.1141
0.1119
MAE
0.2801
0.2794
0.2226
0.2333
0.2401
0.2168
0.2296
0.2102
0.2065
4
MSE
0.2161
0.2911
0.1705
0.1877
0.2330
0.1506
0.1406
0.1464
0.1313
MAE
0.2994
0.3188
0.2488
0.2566
0.3412
0.2560
0.2449
0.2408
0.2302
5
MSE
0.2395
0.2940
0.1780
0.2174
0.1517
0.1514
0.1881
0.1408
0.1388
MAE
0.3136
0.3117
0.2463
0.2611
0.2455
0.2438
0.2911
0.2329
0.2300
9
MSE
0.2817
0.4037
0.2164
0.3058
0.2090
0.2213
0.2066
0.1893
0.1904
MAE
0.3391
0.3622
0.2767
0.3097
0.3017
0.3071
0.2923
0.2767
0.2752
12
MSE
0.2938
0.5464
0.2418
0.4308
0.2160
0.3016
0.2117
0.2165
0.2079
MAE
0.3472
0.4100
0.2999
0.3530
0.3021
0.3725
0.2938
0.2911
0.2847
24
MSE
0.3256
0.4957
0.3019
0.3874
0.2598
0.2682
0.2754
0.2548
0.2539
MAE
0.3648
0.4569
0.3452
0.3914
0.3224
0.3396
0.3467
0.3186
0.3180
Bold text is used to highlight the result of the most successful model for each prediction horizon
Table 5 Comparison of
prediction performances of the
proposed ensemble models and
base models for eight time
horizons
Prediction horizon (h)
Metrics
LSTM
GRU
CNN
Ensl
Ens2
1
MSLE
0.0140
0.0142
0.0139
0.0134
0.0125
R2
0.9248
0.9213
0.9258
0.9301
0.9366
2
MSLE
0.0229
0.0222
0.0196
0.0183
0.0172
R2
0.8722
0.8715
0.8860
0.9028
0.9050
3
MSLE
0.0250
0.0221
0.0243
0.0213
0.0208
R2
0.8245
0.8605
0.8583
0.8730
0.8737
4
MSLE
0.0400
0.0280
0.0267
0.0266
0.0240
R2
0.6247
0.8321
0.8349
0.8422
0.8543
5
MSLE
0.0277
0.0273
0.0332
0.0253
0.0248
R2
0.8181
0.8281
0.7876
0.8311
0.8367
9
MSLE
0.0383
0.0384
0.0360
0.0332
0.0331
R2
0.7254
0.7469
0.7506
0.7653
0.7562
12
MSLE
0.0376
0.0560
0.0361
0.0369
0.0352
R2
0.7255
0.5598
0.7101
0.7274
0.7338
24
MSLE
0.0440
0.0456
0.0468
0.0435
0.0426
R2
0.6563
0.6297
0.5832
0.6637
0.6724
6.5 Results and Discussions
In order to assess the performance of our ensemble model
(i.e., Ens2 in Tables 4, 5), we compare the proposed model
with base models, i.e., LSTM, GRU and CNN. 6.4 Comparison Metrics Besides these
models, we also compare our model with some selected tra-
ditional machine learning models including LR, KNN, DT,
RF. We use raw input connection in meta-learner so that our
ensemble model can be dynamic and capture the traffic con
the same architecture as the proposed ensemble model except
the raw input connection (i.e., Ens1 in Tables 4, 5). We also
compared the prediction performance of this model with the
model we propose. The 3-day forecasting performance of the base models
and the ensemble models are shown in Fig.6a, b. Figure6a
shows single-step prediction (i.e., prediction horizon = 1) and
Fig.6b shows multi-step predictions (i.e., prediction horizon
24) The forecasting graphs belonging to the same period the same architecture as the proposed ensemble model except
the raw input connection (i.e., Ens1 in Tables 4, 5). We also
compared the prediction performance of this model with the
model we propose. 6.5 Results and Discussions In order to assess the performance of our ensemble model
(i.e., Ens2 in Tables 4, 5), we compare the proposed model
with base models, i.e., LSTM, GRU and CNN. Besides these
models, we also compare our model with some selected tra-
ditional machine learning models including LR, KNN, DT,
RF. We use raw input connection in meta-learner so that our
ensemble model can be dynamic and capture the traffic con-
ditions well. We construct an alternative model, which has The 3-day forecasting performance of the base models
and the ensemble models are shown in Fig.6a, b. Figure6a
shows single-step prediction (i.e., prediction horizon = 1) and
Fig.6b shows multi-step predictions (i.e., prediction horizon
= 24). The forecasting graphs belonging to the same period
(month/day/hour) were selected to compare the performance 12 12388 Arabian Journal for Science and Engineering (2024) 49:12377–12392 Fig. 6 Comparison of the prediction results: a TimeHorizon = 1 (single-step prediction). b TimeHorizon = 24 (multi-step prediction) Fig. 6 Comparison of the prediction results: a TimeHorizon = 1 (single-step prediction). b TimeHorizon = 24 (multi-step prediction) ig. 6 Comparison of the prediction results: a TimeHorizon = 1 (single-step prediction). b TimeHorizon = 24 (multi- of the prediction results: a TimeHorizon = 1 (single-step prediction). b TimeHorizon = 24 (multi-step prediction) • Traditional machine learning models are not sufficient
for long-term prediction. • Traditional machine learning models are not sufficient
for long-term prediction. of the top and bottom forecasting horizons. Based on these
figures, the predictive success of the proposed deep ensem-
ble model has increased considerably for both time horizons
comparedtosinglemodels. Inthesamefigure, thehistograms
in Fig.7a–e show forecasting performances for time horizon
= 1 for different days of the week and different times of the
day. To obtain these histograms, for each model, first, the
difference between the each forecasting point and its ground
truth was taken separately. Then, for each forecasting point,
the model with the smallest of this difference was awarded
a score, and in the end, the models’ scores were added up. These histograms show the total score of each model. The
highest score in all five histograms belongs to the proposed
ensemble model. In fact, these histograms show that the pro-
posed model is decisively ahead of the other models. • Among the traditional machine learning-based (ML-
based) models we have compared, the best performance
belongs to KNN. 6.5 Results and Discussions However, the performance of tree-based
models is quite low. The prediction performance of ran-
dom forest (RF), which is a tree-based ensemble model,
is quite far behind KNN. • Among the deep learning (DL) models, CNN has shown
the best performance in many horizons. This result is
interesting because CNN was not originally developed
for time series problems. But before we can gener-
alize this result, we need to do more experiments. For the dataset we used in this study, CNN performs
quite successfully. In other words, we can say that we
have developed a prediction model compatible with the
dataset. Nevertheless, we cannot make a general conclu-
sion that CNN is the most successful DL technique for
time series problems. However, we can say that CNN is
promising for such problems. Table 4 shows MSE and MAE results we have obtained as
a result of our experiments for eight time horizons. As can be
seen in the table, our model (i.e., Ens2) is the most successful
model in all time horizons except time horizon = 9, in which
Ens1 model was the most successful. These results prove that
the ensemble models perform better, especially in long-term
traffic flow prediction. • According to the results of our experiments, although
CNN is more successful DL model than others, the per-
formance of the other DL models is roughly competitive
with CNN. In addition to these main results, we can list the other
results we achieved when we carefully examine the table as
follows: 123 12 Arabian Journal for Science and Engineering (2024) 49:12377–12392 12389 Fig. 7 Comparison of the prediction results: a prediction results on weekdays. b Prediction results on weekends. c Prediction results on rush hours. d Prediction results on off-peak hours. e Overall performance (TimeHorizon = 1) Fig. 7 Comparison of the prediction results: a prediction results on weekdays. b Prediction results on weekends. c Prediction results on rush hours. d Prediction results on off-peak hours. e Overall performance (TimeHorizon = 1) Fig. 7 Comparison of the prediction results: a prediction results on weekdays. b Prediction results on weekends. c Prediction results on rush hours. d Prediction results on off-peak hours. e Overall performance (TimeHorizon = 1) • These results indicate that DL-based models offer the
opportunity to develop more successful prediction mod-
els because they can better capture long-term dependen-
cies. 6.5 Results and Discussions When these graphs are examined in detail,
the superiority of our proposed model over other models
is clearly seen. • Figure 8 shows the residual plots of our proposed model
and the models we compared. The residual plot shows the
difference between actual values and predicted values. difficult to obtain values close to 1 for difficult problems
such as traffic flow prediction. Therefore, each prediction
problem should be evaluated on its own. • When we examine both metrics values in Table 5, we see
that the model we recommend is the best model for all
prediction horizons except 9h. This shows us the results
in Table 5 are consistent with the results in Table 4. The performance of the model is directly proportional
to the closeness of the points forming the graph to the
starting point. When these graphs are examined in detail,
the superiority of our proposed model over other models
is clearly seen. • The results show that we can achieve a significant per-
formance improvement when we combine ensemble
learning architecture and deep learning techniques. • In general, as prediction horizon increases, the prediction
performance of all the models we compare, including the
model we propose, decreases, which proves that the long-
term prediction is more difficult. 6.5 Results and Discussions • Although the prediction performance of our proposed
model decreases as the prediction horizon increases, this
decrease is small compared to the other models we com-
pared. For example, when the forecast horizon is 4h, the
MSE of our model is 0.1313, and this value increases to
0.1388 when the forecast horizon is 5h. In contrast, when
the forecast horizon is 4h, the MSE of CNN is 0.1406. However, when the forecast horizon increases to 5h, this
• Figure 8 shows the residual plots of our proposed model
and the models we compared. The residual plot shows the
difference between actual values and predicted values. The performance of the model is directly proportional
to the closeness of the points forming the graph to the
starting point. When these graphs are examined in detail,
the superiority of our proposed model over other models
is clearly seen. 7 Conclusion and Future Work
Long-term traffic flow forecasting is vital for traffic man-
agement issues such as congestion control and better route
selection. This importance will become more evident in the
future with the development of related technologies. There-
fore, it is critical to try to improve long-term traffic flow
forecasting performance. That is why, this study proposed
a novel ensemble model for long-term traffic flow predic-
tion. The proposed model is a deep ensemble model built by
properly combining three different deep learning techniques
b
d l
d
i
d
d l h
d
i
ll Arabian Journal for Science and Engineering (2024) 49:12377–12392 12390 Fig. 8 Residual plots: a CNN, b GRU, c LSTM, d Ens1 and e Ens2 Fig. 8 Residual plots: a CNN, b GRU, c LSTM, d Ens1 and e Ens2 • Figure 8 shows the residual plots of our proposed model
and the models we compared. The residual plot shows the
difference between actual values and predicted values. The performance of the model is directly proportional
to the closeness of the points forming the graph to the
starting point. When these graphs are examined in detail,
the superiority of our proposed model over other models
is clearly seen. • Figure 8 shows the residual plots of our proposed model
and the models we compared. The residual plot shows the
difference between actual values and predicted values. The performance of the model is directly proportional
to the closeness of the points forming the graph to the
starting point. 6.5 Results and Discussions good both on weekdays and during peak hours. However,
the performance of CNN among the single models is the
lowest for these two categories. • When we examine the histograms in Fig.7, we see that
the most successful single model is GRU. GRU outper-
forms even our alternative ensemble model (Ens1) for all
categories. • Based on our observations during our experiments, we
can also make a comparison between DL-based models
in terms of computation times. CNN also performs best in
termsofcomputationtimeamongDL-basedmodels.This
is probably due to the fact that CNN can be parallelized
more efficiently than GRU and LSTM. However, LSTM
was the model with the worst performance in terms of
computation time. • In Table 5, we compare ensemble models with base mod-
els using MSLE and R2 metrics. We chose these two
metrics because the first metric measures relative error
rather than actual error. That is, it gives approximately
equal weight to small and large differences between
actual and predicted values. The second metric is used
to compare the quality of models with each other, rather
than to decide the overall quality of a model. This metric
takes a value between 0 and 1, and the model is consid-
ered good as the value gets closer to 1. However, it is very • When Fig.6a, b is compared, it is seen that the pro-
posed model is relatively more successful in sharp ups
and downs. • When the histograms in Fig.7 are examined, it is seen that
thepredictionperformanceoftheproposedmodelisquite 12 3 12390
Arabian Journal for Science and Engineering (2024) 49:12377–12392
Fig. 8 Residual plots: a CNN, b GRU, c LSTM, d Ens1 and e Ens2
difficult to obtain values close to 1 for difficult problems
such as traffic flow prediction. Therefore, each prediction
problem should be evaluated on its own. • When we examine both metrics values in Table 5, we see
that the model we recommend is the best model for all
prediction horizons except 9h. This shows us the results
in Table 5 are consistent with the results in Table 4. • The results show that we can achieve a significant per-
formance improvement when we combine ensemble
learning architecture and deep learning techniques. • In general, as prediction horizon increases, the prediction
performance of all the models we compare, including the
model we propose, decreases, which proves that the long-
term prediction is more difficult. 7 Conclusion and Future Work It would
also be beneficial to try to understand the temporal and spa-
tial components to which our ensemble model gives more
weight. 5. Abbar, S.; Stanojevic, R.; Mustafa, S.; Mokbel, M.: Traffic routing
in the ever-changing city of Doha. Commun. ACM 64(4), 67–68
(2021) 6. Do˘gan, E.: LSTM training set analysis and clustering model devel-
opment for short-term traffic flow prediction. Neural Comput. Appl. 33(17), 11175–11188 (2021) 7. Wang, Z.; Su, X.; Ding, Z.: Long-term traffic prediction based
on lstm encoder–decoder architecture. IEEE Trans. Intell. Transp. Syst. 22(10), 6561–6571 (2020) 8. Bogaerts, T.; Masegosa, A.D.; Angarita-Zapata, J.S.; Onieva, E.;
Hellinckx, P.: A graph CNN-LSTM neural network for short and
long-term traffic forecasting based on trajectory data. Transp. Res. Part C Emerg. Technol. 112, 62–77 (2020) Part C Emerg. Technol. 112, 62–77 (2020) 9. Li, Y.; Chai, S.; Ma, Z.; Wang, G.: A hybrid deep learning
framework for long-term traffic flow prediction. IEEE Access 9,
11264–11271 (2021) 10. Belhadi, A.; Djenouri, Y.; Djenouri, D.; Lin, J.C.-W.: A recurrent
neural network for urban long-term traffic flow forecasting. Appl. Intell. 50, 3252–3265 (2020) 11. Chen, H.; Grant-Muller, S.; Mussone, L.; Montgomery, F.: A study
of hybrid neural network approaches and the effects of missing data
on traffic forecasting. Neural Comput. Appl. 10, 277–286 (2001) 12. Do, L.N.; Vu, H.L.; Vo, B.Q.; Liu, Z.; Phung, D.: An effective
spatial-temporal attention based neural network for traffic flow pre-
diction. Transp. Res. Part C Emerg. Technol. 108, 12–28 (2019) 13. Sagi, O.; Rokach, L.: Ensemble learning: a survey. Wiley Interdis-
cip. Rev. Data Min. Knowl. Discov. 8(4), 1249 (2018) 14. Dong, X.; Yu, Z.; Cao, W.; Shi, Y.; Ma, Q.: A survey on ensemble
learning. Front. Comput. Sci. 14, 241–258 (2020) 15. Hong, W.-C.; Dong, Y.; Zheng, F.; Lai, C.-Y.: Forecasting urban
traffic flow by SVR with continuous ACO. Appl. Math. Model. 35(3), 1282–1291 (2011) Funding Open access funding provided by the Scientific and Techno-
logical Research Council of Türkiye (TÜB˙ITAK). 16. Chen, Y.; Chen, H.; Ye, P.; Lv, Y.; Wang, F.-Y.: Acting as a deci-
sion maker: traffic-condition-aware ensemble learning for traffic
flow prediction. IEEE Trans. Intell. Transp. Syst. 7 Conclusion and Future Work Long-term traffic flow forecasting is vital for traffic man-
agement issues such as congestion control and better route
selection. This importance will become more evident in the
future with the development of related technologies. There-
fore, it is critical to try to improve long-term traffic flow
forecasting performance. That is why, this study proposed
a novel ensemble model for long-term traffic flow predic-
tion. The proposed model is a deep ensemble model built by
properly combining three different deep learning techniques
as base models. We designed our model that can dynamically
produce the weights of the base models based on both each • Although the prediction performance of our proposed
model decreases as the prediction horizon increases, this
decrease is small compared to the other models we com-
pared. For example, when the forecast horizon is 4h, the
MSE of our model is 0.1313, and this value increases to
0.1388 when the forecast horizon is 5h. In contrast, when
the forecast horizon is 4h, the MSE of CNN is 0.1406. However, when the forecast horizon increases to 5h, this
value increases to 0.1881. 12 3 12391 Arabian Journal for Science and Engineering (2024) 49:12377–12392 base model’s performance and traffic condition. Experimen-
tal results show that the proposed approach outperforms all
the models compared. In future research, we plan to investi-
gate the effectiveness of our model with using different base
models and datasets. We will also implement a 1D-CNN fol-
lowed by a recurrent neural network (such as LSTM or GRU)
as base learner and investigate the effect of including this net-
work into our ensemble model. In addition, the fact that the
CNN-based prediction model we developed was quite suc-
cessfulcomparedtootherDLmodelsmotivatedustoconduct
more research in this area. As a future work, we plan to make
more experiments to compare the forecasting performance of
CNN using different time series datasets. More than that, we
will try to understand why CNN is performing better. We
also plan to address the issue of interpretability of DL-based
models. Although deep learning algorithms provide high pre-
diction performance, the interpretability of DL-based models
is very low. This is also true for our model. Therefore, as a
future study, we plan to analyze the outputs of the base learn-
ers of our model separately. Thus, we will try to discover the
critical hours that affect the outcome for each model. 7 Conclusion and Future Work 23(4), 3190–3200
(2020) Open Access This article is licensed under a Creative Commons
Attribution 4.0 International License, which permits use, sharing, adap-
tation, distribution and reproduction in any medium or format, as
long as you give appropriate credit to the original author(s) and the
source, provide a link to the Creative Commons licence, and indi-
cate if changes were made. The images or other third party material
in this article are included in the article’s Creative Commons licence,
unless indicated otherwise in a credit line to the material. If material
is not included in the article’s Creative Commons licence and your
intended use is not permitted by statutory regulation or exceeds the
permitteduse,youwillneedtoobtainpermissiondirectlyfromthecopy-
right holder. To view a copy of this licence, visit http://creativecomm
ons.org/licenses/by/4.0/. 17. Liu, J.; Wu, N.; Qiao, Y.; Li, Z.: Short-term traffic flow forecasting
using ensemble approach based on deep belief networks. IEEE
Trans. Intell. Transp. Syst. 23(1), 404–417 (2020) 18. Gu, Y.; Lu, W.; Xu, X.; Qin, L.; Shao, Z.; Zhang, H.: An improved
Bayesian combination model for short-term traffic prediction with
deep learning. IEEE Trans. Intell. Transp. Syst. 21(3), 1332–1342
(2019) (
)
19. Aljuaydi, F.; Wiwatanapataphee, B.; Wu, Y.H.: Multivariate
machine learning-based prediction models of freeway traffic flow
under non-recurrent events. Alex. Eng. J. 65, 151–162 (2023) 20. Guo, C.; Li, D.; Chen, X.: Unequal interval dynamic traffic flow
prediction with singular point detection. Appl. Sci. 13(15), 8973
(2023) 21. Williams, B.M.: Multivariate vehicular traffic flow prediction: eval-
uation of ARIMAX modeling. Transp. Res. Rec. 1776(1), 194–200
(2001) References Zhang, W.; Yu, Y.; Qi, Y.; Shu, F.; Wang, Y.: Short-term traffic
flow prediction based on spatio-temporal analysis and CNN deep
learning. Transportmetrica A Transp. Sci. 15(2), 1688–1711 (2019) 40. Zheng, G.; Chai, W.K.; Katos, V.; Walton, M.: A joint temporal-
spatial ensemble model for short-term traffic prediction. Neuro-
computing 457, 26–39 (2021) 30. Lv, M.; Hong, Z.; Chen, L.; Chen, T.; Zhu, T.; Ji, S.: Temporal
multi-graph convolutional network for traffic flow prediction. IEEE
Trans. Intell. Transp. Syst. 22(6), 3337–3348 (2020) 41. Liu, Y.; Liu, Z.; Vu, H.L.; Lyu, C.: A spatio-temporal ensemble
method for large-scale traffic state prediction. Comput.-Aided Civ. Infrastruct. Eng. 35(1), 26–44 (2020) 31. Chen, C.; Li, K.; Teo, S.G.; Zou, X.; Li, K.; Zeng, Z.: Citywide
traffic flow prediction based on multiple gated spatio-temporal
convolutional neural networks. ACM Trans. Knowl. Discov. Data
(TKDD) 14(4), 1–23 (2020) 42. Chen, L., Chen, C.P.: Ensemble learning approach for freeway
short-term traffic flow prediction. In: 2007 IEEE International Con-
ference on System of Systems Engineering, pp. 1–6. IEEE (2007) 32. Méndez, M.; Merayo, M.G.; Núñez, M.: Long-term traffic flow
forecasting using a hybrid CNN-BiLSTM model. Eng. Appl. Artif. Intell. 121, 106041 (2023) 43. Yan, H.; Fu, L.; Qi, Y.; Yu, D.-J.; Ye, Q.: Robust ensemble method
for short-term traffic flow prediction. Future Gener. Comput. Syst. 133, 395–410 (2022) 33. Shi, X.; Qi, H.; Shen, Y.; Wu, G.; Yin, B.: A spatial-temporal atten-
tion approach for traffic prediction. IEEE Trans. Intell. Transp. Syst. 22(8), 4909–4918 (2020) 44. Zhang, Y.; Xin, D.: A diverse ensemble deep learning method for
short-term traffic flow prediction based on spatiotemporal correla-
tions. IEEE Trans. Intell. Transp. Syst. 23(9), 16715–16727 (2021) 34. Chen, Z.; Lu, Z.; Chen, Q.; Zhong, H.; Zhang, Y.; Xue, J.; Wu, C.:
Spatial-temporal short-term traffic flow prediction model based on
dynamical-learning graph convolution mechanism. Inf. Sci. 611,
522–539 (2022) 45. Zhao, F.; Zeng, G.-Q.; Lu, K.-D.: EnLSTM-WPEO: short-term
traffic flow prediction by ensemble LSTM, NNCT weight inte-
gration, and population extremal optimization. IEEE Trans. Veh. Technol. 69(1), 101–113 (2019) 35. Wu, Y.; Tan, H.; Qin, L.; Ran, B.; Jiang, Z.: A hybrid deep learning
based traffic flow prediction method and its understanding. Transp. Res. Part C Emerg. Technol. 90, 166–180 (2018) 46. Xiao, J.; Xiao, Z.; Wang, D.; Bai, J.; Havyarimana, V.; Zeng,
F.: Short-term traffic volume prediction by ensemble learning in
concept drifting environments. Knowl.-Based Syst. 164, 213–225
(2019) 36. References 22. Shahriari, S.; Ghasri, M.; Sisson, S.; Rashidi, T.: Ensemble of
ARIMA: combining parametric and bootstrapping technique for
traffic flow prediction. Transportmetrica A Transp. Sci. 16(3),
1552–1573 (2020) 1. Qu, L.; Li, W.; Li, W.; Ma, D.; Wang, Y.: Daily long-term traffic
flow forecasting based on a deep neural network. Expert Syst. Appl. 121, 304–312 (2019) 23. Okutani, I.; Stephanedes, Y.J.: Dynamic prediction of traffic
volume through Kalman filtering theory. Transp. Res. Part B
Methodol. 18(1), 1–11 (1984) 2. Sattar, K.; Chikh Oughali, F.; Assi, K.; Ratrout, N.; Jamal, A.;
Masiur Rahman, S.: Transparent deep machine learning framework
for predicting traffic crash severity. Neural Comput. Appl. 35(2),
1535–1547 (2023) 24. Sun, H.; Liu, H.X.; Xiao, H.; He, R.R.; Ran, B.: Use of local linear
regression model for short-term traffic forecasting. Transp. Res. Rec. 1836(1), 143–150 (2003) 3. Lana, I.; Del Ser, J.; Velez, M.; Vlahogianni, E.I.: Road traffic fore-
casting: recent advances and new challenges. IEEE Intell. Transp. Syst. Mag. 10(2), 93–109 (2018) 25. Cai, P.; Wang, Y.; Lu, G.; Chen, P.; Ding, C.; Sun, J.: A spatiotem-
poral correlative k-nearest neighbor model for short-term traffic
multistep forecasting. Transp. Res. Part C Emerg. Technol. 62, 21–
34 (2016) 4. Do, L.N.; Taherifar, N.; Vu, H.L.: Survey of neural network-based
models for short-term traffic state prediction. Wiley Interdiscip. Rev. Data Min. Knowl. Discov. 9(1), 1285 (2019) 12 3 Arabian Journal for Science and Engineering (2024) 49:12377–12392 12392 26. Polson, N.G.; Sokolov, V.O.: Deep learning for short-term traffic
flow prediction. Transp. Res. Part C Emerg. Technol. 79, 1–17
(2017) 37. Lu, S.; Zhang, Q.; Chen, G.; Seng, D.: A combined method for
short-term traffic flow prediction based on recurrent neural net-
work. Alex. Eng. J. 60(1), 87–94 (2021) 27. Dai, X., Fu, R., Lin, Y., Li, L., Wang, F.-Y.: Deeptrend: a
deep hierarchical neural network for traffic flow prediction. arXiv:1707.03213 (2017) 38. Zhan, H.; Gomes, G.; Li, X.S.; Madduri, K.; Sim, A.; Wu, K.:
Consensus ensemble system for traffic flow prediction. IEEE Trans. Intell. Transp. Syst. 19(12), 3903–3914 (2018) 28. Zhang, D.; Kabuka, M.R.: Combining weather condition data to
predict traffic flow: a GRU-based deep learning approach. IET
Intell. Transp. Syst. 12(7), 578–585 (2018) 39. Chen, X.; Chen, H.; Yang, Y.; Wu, H.; Zhang, W.; Zhao, J.; Xiong,
Y.: Traffic flow prediction by an ensemble framework with data
denoising and deep learning model. Phys. A Stat. Mech. Appl. 565, 125574 (2021) 29. References Zheng, H.; Lin, F.; Feng, X.; Chen, Y.: A hybrid deep learning
model with attention-based conv-LSTM networks for short-term
traffic flow prediction. IEEE Trans. Intell. Transp. Syst. 22(11),
6910–6920 (2020) 47. Dai, G.; Tang, J.; Luo, W.: Short-term traffic flow prediction: an
ensemble machine learning approach. Alex. Eng. J. 74, 467–480
(2023) 123 123 12
|
https://openalex.org/W2336851006
|
http://www.librelloph.com/journalofhumansecurity/article/download/johs-12.1.37/pdf
|
English
| null |
A Piece of Land or Peace on the Land: How Much Is a Peasant's Life Worth in Brazil?
|
Journal of human security
| 2,016
|
cc-by
| 11,845
|
A Piece of Land or Peace on the Land: How Much Is a Peasant’s
Life Worth in Brazil? Artur Zimerman Engineering, Modeling and Applied Social Sciences Center (CECS), Universidade Federal do ABC (UFABC), São Paulo
State, Brazil; E-Mail: artur.zimerman@gmail.com; Tel.: +55 1149968261 Submitted: 20 August 2015 | In revised form: 19 February 2016 | Accepted: 22 March 2016 |
Published: 18 April 2016 Abstract: Land inequality in Brazil is alarming and several poor individuals living in rural areas do not
have enough income to survive decently. The struggle to access land should lead to a paradigm shift with
social movements leading this process since democratization. Their strategies vary, but usually focus on
complementary activities of mass mobilization that culminate in the occupation of unproductive land that is
not fulfilling its social function in order to force expropriation and the creation of new settlements. This study
aims to investigate, through empirical evidence, if such strategies are having the desired effect of allowing
the poor to access land, without increasing the already high numbers, and potentially aggravating the violent
characteristics, of such disputes. During the Cardoso and Lula presidential administrations the relation
between the number of new settlements and the number of deaths caused by land disputes increased. However, there is still a long way to go to improve this policy and achieve positive results. Overall, is this
struggle for the reduction of inequality in the Brazilian countryside being won? Is the sacrifice paying off? And what is the price regarding the relation between land conflict victims and the creation of new rural
settlements? Keywords: Agrarian conflicts; agrarian reform; Brazil; mobilization; land inequality; social moveme Journal of Human Security | 2016 | Volume 12 | Issue 1 | Pages 37–51
DOI: 10.12924/johs2016.12010037
ISSN: 1835–3800
Journal of
Human Security Journal of Human Security | 2016 | Volume 12 | Issue 1 | Pages 37–51
DOI: 10.12924/johs2016.12010037
ISSN: 1835–3800
Journal of
Human Security Journal of
Human Security Journal of
Human Security Journal of Human Security | 2016 | Volume 12 | Issue 1 | Pages 37–51
DOI: 10.12924/johs2016.12010037
ISSN: 1835–3800 Research Article © 2016 by the authors; licensee Librello, Switzerland. This open access article was published
under a Creative Commons Attribution License (http://creativecommons.org/licenses/by/4.0/). 1. Introduction ness and human and animal food crops being replaced by
renewable energy crops, land concentration worldwide is
growing, and tends to grow even more, in countries with a
Brazilian or Latin American profile [3,4]. Brazil has vast dimensions, fertile land and climate condi-
tions that are favorable for agricultural activity. Hence, in
2012, the country ranked third among the world’s largest
exporters of agricultural products ([1] p. 2) . Given these
natural attributes, it is hard to conceive that this abundance
would lead to violence caused by land disputes, since well-
distributed land among the peasant population would con-
tribute to rural inhabitants deriving their livelihoods from
the land. The main problem is that the land is extremely
unequally distributed: 1% of landowners occupy about 50%
of arable lands [2], which is one of the highest rates of land
concentration in the world. With the expansion of agribusi- Land in Brazil is an abundant natural resource though
very unequally distributed and, as such, part of the existing
literature discusses the association between abundant nat-
ural resources and violent conflicts, especially as a cause
for civil wars [5–10] and other episodes of violence. However, there are not many studies relating natural re-
sources in general, and land in particular, to lower-intensity
conflicts that also cause a great number of deaths, as well
as forced displacement, arrests, injuries, threats and the like. In existing studies, the type of political regime is often librello librello They were also aided by registry offices and notaries
in the region, who legitimized the fraud. This is known as
land grabbing (or in Portuguese - grilagem de terra). This study aims to evaluate the cost of human lives that
derives from land dispute conflicts in Brazil, analyzing if so-
cial movements and non-governmental organizations that
mobilize the rural (and urban) population to fight for fairer
land distribution, and consequently reduce poverty, achieve
the desired effect. That is, if the means applied to acquire a
plot of land for the poorer rural population is producing the
practical outcome of a more equitable land division. A great part of the land disputes originated [18] from the
high concentration of owned land occupied by squatters,
and from conflict between plantation owners and small-
holder farmers, since large tracts of land generally required
expansion to meet the agribusiness model, a relatively mod-
ern phenomenon. The democratization process in Brazil began in 1985
with the downfall of the military regime and was consoli-
dated in 1990, with the country’s first free and universal
presidential elections. Concurrently, non-governmental or-
ganizations (NGOs) and the third sector were allowed to
represent the different sectors of the population. Conse-
quently, individuals and groups, particularly in the poorer
segments that had no representation in parliament or in
civil society as a whole, received assistance and were no
longer suppressed. Citizenship was being built. Such or-
ganizations and social movements mobilized peasants in
order to pressure government representatives to undertake
a land reform that would decentralize land tenure. There
was heavy opposition from groups of large landowners who
wanted to preserve the status quo (the Ruralist Democratic
Union, known by its Portuguese acronym as UDN). Such
groups were well organized and exerted significant influ-
ence on congress. In 1987, the president of the UDN admit-
ted his organization had bought 4,000 guns and had 70,000
firearms at the group’s disposal ([19] p. 30) for protection
against occupation of the members’ private property. For this purpose, a new database from DATALUTA will
be examined on the different types of peasant mobilization,
land appropriations and settlements implemented in Brazil
between 1988 and 2011. 1988 marks the year of the new
Federal Constitution. librello Article 186 of this most recent consti-
tution requires proof that the land fulfills its social function
and is not unproductive in order for it not to be destined
to expropriation and distribution to very small owners or
landless peasants. Data on agrarian violence will also be
presented, with the aim of clarifying the number of people
that have died in the struggle for land, the mobilization activ-
ities that third sector entities implement to keep this struggle
alive, and the number of people that have been settled. In
short, is this fight being won? Or furthermore: Is this fight
paying off in terms of the ratio of victims of land conflicts to
settlements created? librello In 1987, the president of the U
ted his organization had bought 4,000 guns and ha
firearms at the group’s disposal ([19] p. 30) for p
against occupation of the members’ private prope
The agrarian issue is embroiled in violence included in models as one of the variables that can lead
to the insurgency of collective violence; violent conflicts
have a greater tendency to happen within anocratic [11]
(or hybrid/intermediate) political regimes. These regimes
are known for being a middle ground between a consoli-
dated democracy and an autocracy, mixing democratic and
autocratic elements in which the population does not have
freedom and the heterogeneity of groups is not respected. On the extreme, civil and political liberties are violated and
protests are strongly suppressed [12–15]. facilitate credit and give vast areas of land to large compa-
nies that had no expertise in the field, since they were from
the automotive industry, financial markets, and other fields
unfamiliar to the agricultural sector [17]. People raced to areas where they believed to be vast
tracts of unowned land. At the time, the motto was “Land
without men for men without land”. However, a great part
of the area that seemed unoccupied had been donated or
financed at very low interest rates to companies who did not
cultivate the lands, especially in the Amazon region, in the
northern part of the country. Rural workers arrived in search
of land and occupied already owned areas. However, the
region did not have the necessary infrastructure to transport
agricultural products, and the newly arrived workers were
vulnerable to several diseases. Since democratization in 1985, due to the country’s
abundance of natural resources, particularly land (the
theme this study aims at addressing), and the very little
time the country has had to strengthen its institutions, Brazil
has gone through agrarian conflicts caused by the histor-
ically unequal distribution of land, which in turn produces
poverty and misery in the countryside. About 1,500 peo-
ple have died since the country’s democratic opening as a
result of the struggle for land; the great majority of deaths
occurring among the peasants or their supporters [16]. Many local elites forged and artificially aged documents
by putting them in drawers full of crickets, which made pa-
pers appear old, as if the properties had been inherited long
ago. librello © 2016 by the authors; licensee Librello, Switzerland. This open access article was published
under a Creative Commons Attribution License (http://creativecommons.org/licenses/by/4.0/). included in models as one of the variables that can lead
to the insurgency of collective violence; violent conflicts
have a greater tendency to happen within anocratic [11]
(or hybrid/intermediate) political regimes. These regimes
are known for being a middle ground between a consoli-
dated democracy and an autocracy, mixing democratic and
autocratic elements in which the population does not have
freedom and the heterogeneity of groups is not respected. On the extreme, civil and political liberties are violated and
protests are strongly suppressed [12–15]. Since democratization in 1985, due to the country’s
abundance of natural resources, particularly land (the
theme this study aims at addressing), and the very little
time the country has had to strengthen its institutions, Brazil
has gone through agrarian conflicts caused by the histor-
ically unequal distribution of land, which in turn produces
poverty and misery in the countryside. About 1,500 peo-
ple have died since the country’s democratic opening as a
result of the struggle for land; the great majority of deaths
occurring among the peasants or their supporters [16]. This study aims to evaluate the cost of human lives that
derives from land dispute conflicts in Brazil, analyzing if so-
cial movements and non-governmental organizations that
mobilize the rural (and urban) population to fight for fairer
land distribution, and consequently reduce poverty, achieve
the desired effect. That is, if the means applied to acquire a
plot of land for the poorer rural population is producing the
practical outcome of a more equitable land division. For this purpose, a new database from DATALUTA will
be examined on the different types of peasant mobilization,
land appropriations and settlements implemented in Brazil
between 1988 and 2011. 1988 marks the year of the new
Federal Constitution. Article 186 of this most recent consti-
tution requires proof that the land fulfills its social function
and is not unproductive in order for it not to be destined
to expropriation and distribution to very small owners or
landless peasants. Data on agrarian violence will also be
presented, with the aim of clarifying the number of people
that have died in the struggle for land, the mobilization activ-
ities that third sector entities implement to keep this struggle
alive, and the number of people that have been settled. librello In
short, is this fight being won? Or furthermore: Is this fight
paying off in terms of the ratio of victims of land conflicts to
settlements created? 2. Background of Violence in the Brazilian
Countryside
One of the Brazilian military government’s goals in 1964
facilitate credit and give vast areas of land to larg
nies that had no expertise in the field, since they w
the automotive industry, financial markets, and ot
unfamiliar to the agricultural sector [17]. People raced to areas where they believed t
tracts of unowned land. At the time, the motto w
without men for men without land”. However, a g
of the area that seemed unoccupied had been d
financed at very low interest rates to companies w
cultivate the lands, especially in the Amazon regi
northern part of the country. Rural workers arrived
of land and occupied already owned areas. How
region did not have the necessary infrastructure to
agricultural products, and the newly arrived work
vulnerable to several diseases. Many local elites forged and artificially aged d
by putting them in drawers full of crickets, which
pers appear old, as if the properties had been inhe
ago. They were also aided by registry offices and
in the region, who legitimized the fraud. This is
land grabbing (or in Portuguese - grilagem de ter
A great part of the land disputes originated [18
high concentration of owned land occupied by s
and from conflict between plantation owners a
holder farmers, since large tracts of land generally
expansion to meet the agribusiness model, a relat
ern phenomenon. The democratization process in Brazil began
with the downfall of the military regime and wa
dated in 1990, with the country’s first free and
presidential elections. Concurrently, non-governm
ganizations (NGOs) and the third sector were a
represent the different sectors of the population
quently, individuals and groups, particularly in th
segments that had no representation in parliam
civil society as a whole, received assistance and
longer suppressed. Citizenship was being built. ganizations and social movements mobilized pe
order to pressure government representatives to u
a land reform that would decentralize land tenu
was heavy opposition from groups of large landow
wanted to preserve the status quo (the Ruralist D
Union, known by its Portuguese acronym as UD
groups were well organized and exerted signific
ence on congress. 2. Background of Violence in the Brazilian
Countryside One of the Brazilian military government’s goals in 1964
was to develop the country’s agriculture. Shortly after the
military coup, during the military government’s first year, the
Land Statute (Estatuto de Terra) was created. This docu-
ment emerged with the purpose to reduce conflicts over
land and to prevent communism from spreading across
the country and, amid the Cold War, counted on support
from the US, especially after it had ‘lost Cuba’ to the Soviet
sphere of influence. One way to accomplish this was to The agrarian issue is embroiled in violence because,
on the one hand, peasants want to take over land that is
supposedly owned by third parties or the State and, on the
other, landowners (or presumed owners) try to protect their
territory from occupants by hiring gunmen or the police. Meanwhile, the State withdraws itself by not defining a clear
policy to appropriate land and redistributes it to the poor
living in rural areas. This is of utter importance considering
that three out of every four of the world’s poor live in rural 38 areas ([20] p. 36). Agrarian reform should be no less than
a social policy to combat poverty where it is mostly con-
centrated: in rural areas. By alleviating rural poverty, the
country would strengthen itself. One of the main problems
in Brazil is that the State lacks the capability to carry out a
substantial agrarian reform. Moreover, pro-market consid-
erations disregard productive land as land to be included
in the reform ([21] p. 221). The agribusiness sector is very
important for balancing accounts given that commodity ex-
ports are a major source of foreign currency for the country. As observed, even workers’ political representatives when
in power, tend to support agribusiness at the expense of
land reform, and Brazil is no exception. nancial loans, shortage of technical assistance, distance
from towns, and susceptibility to diseases. The new settled
owners receive their titles and formally own their plots, and
can even sell them if they so desire. However, in such cases,
they won’t be allowed to participate in a future land reform
and this can generate conflicts within the social movements,
since the fight for land is not aimed at obtaining financial
gains, but survival and a life with dignity. ”Landowners and local authorities in the Brazilian coun-
tryside frequently respond to occupations with violent repres-
sion. 2. Background of Violence in the Brazilian
Countryside Their action reflects the hybrid character of the Brazilian
state, modern and rational in cities and at the federal level
but, in many rural areas, still clientelistic and marked by non-
legitimate violence. . . [On the other hand], the action of the
land occupiers, however, is legitimated by the claim of civil
disobedience while the efforts to repress them cannot lay
claim to legitimacy on that basis” ([22] p. 156). Land occupation and its retaliation is a unique process
that must be understood as a practice based on political
context and performed as collective action ([22] p. 169)
from both sides: those who occupy the land and those who
defend it from occupation. The difference between these
two forces is that those who occupy do not do it to prac-
tice violence against third parties, in the direct sense of
physically victimizing people. However, retaliation of these
occupations is frequently performed with the intent to cause
physical harm to the activists. This duality that represents a contradiction between the
city and rural areas, makes economic and political power
decisive in the countryside, favoring large landholders and
the agribusiness elite by employing State authorities and
the use of violence, to the detriment of the most vulnerable
rural worker and landless people. There is also a legal problem. The occupation of unpro-
ductive land, as mentioned above, can be legally justified
through the Constitution, and the defense of public prop-
erty is in accordance with the Brazilian Civil Code. While
the National Institute of Colonization and Agrarian Reform
(INCRA) defends the peasants, justice courts often favor
the owners. None of these institutions prevail over each
other, thus, land dispute conflicts continue ([23] p. 137). The current model incentivizes both sides of the conflict
to use violence, taking into consideration that one side
has weapons at its disposal, while the other is formed by
an unarmed and poor population willing to revert the long
standing land inequality scenario that persists among these
divergent actors. However, as a counterpoint, the police crackdown on
the protests had the opposite effect than was intended, and
ultimately increased the protests, forcing the State to lend
a hand to farmers by expanding the agrarian reform. State
violence serves the purposes of those interested in getting
a piece of land through agrarian reform. This occurred in
Brazil. 2. Background of Violence in the Brazilian
Countryside The intensification of the protests in 1995 and 1996,
followed by repression, created a situation in which it was
possible for the population to put in practice collective action
mechanisms and to include the issue of police violence and
land reform in the agenda, pressing the Brazilian govern-
ment to expedite and expand the scope of land distribution
[25]. The State’s violence probably favored the creation
of more rural settlements. The question is to what extent
does the cost in lives encourage the State to take action to
implement the necessary policies? Land occupation is the strategy most commonly used by
social movements and rural workers’ unions aiming greater
opportunities for access to land. Landless workers and
smallholder farmers with small areas of land for subsistence
pressure the government through local NGOs to solve land
conflicts and implement rural settlements ([24] p. 73). 3. Empirical Study There is a sense of urgency for the State to intervene
when a peasant group, affiliated to a rural social movement
or union, occupies (or threatens to occupy) an area that
is supposedly unproductive and belongs to another owner,
so that violence does not get out of control. In this case,
the area is evaluated to ensure it is unproductive (usually
the peasant leadership has already done this to map the
possible targets) and begin the expropriation process (usu-
ally through compensation), which can take months or even
years, depending on procedural progress. During this time,
the territory is referred to as a land occupation. Upon ap-
proval and completion of the process, the land becomes a
settlement. Nonetheless, there are many problems entailed,
such as lack of infrastructure, lack of housing, reduced fi- Due to all the issues involved in land acquisition through
redistribution and agrarian reform, those waiting to receive
a portion of land for subsistence end up spending a large
amount of time intensely mobilized by the organizational
structure. The aim is to pressure the various levels of
government through different means, demonstrating the
precarious situation and violence they are subjected to. The democratization process allowed for the creation
of a large number of organizations and civil society move-
ments, whose aim is to mobilize landless workers in order
to occupy unproductive land across Brazil. According to
data from DATALUTA [26,27], to be explained ahead, there
were 108 different associations active between 2000–2011 39 mobilizing people to, among other activities, occupy land
(see Appendix). spreading across the country and building its organizational
structure; and Institutionalization (1990s) when the move-
ment intensified the peasantry’s resistance actions, dealing
with a very repressive State apparatus while occupying land
([29] pp. 11–12). The vast majority of these movements are local (more
than 70 per cent were found acting within just one out of
the 27 different Brazilian states) and do not act nationally. However, there are a few organizations that are engaged in
activities throughout the country and, at times, even acting
internationally, aiding landless peasants from other coun-
tries in their fight for a plot of land [28]. They often form
partnerships in order to maximize their efforts to occupy
land. Only a few social movements act nationwide, there-
fore, this paper will deal mainly with organized movements
that are spread throughout Brazil. 4. MST and CONTAG, Two Nationwide Movements Because of our scope, it is mandatory to detail in brief
how these main social movements were created, how they
operate according to their own rules and the differences
between them. The Rural Landless Workers Movement, known as MST
(from the Portuguese acronym for Movimento dos Trabal-
hadores Rurais Sem Terra), was formed after the Catholic
Church installed the Pastoral Land Commission, CPT (from
the Portuguese acronym for Comissão Pastoral da Terra),
during the military regime in 1975, to counteract a specific
modernization policy which consisted in granting large tracts
of land to the various companies from different economic
sectors so that they could transform what was understood
as a development lag into a very effective agricultural sec-
tor. After acting in conjunction with CPT for some time, the
Movement gained autonomy and started to operate inde-
pendently, although there were some areas in which they
held similar interests and actions. The National Confederation of Agricultural Workers,
CONTAG (Confederação Nacional dos Trabalhadores na
Agricultura), was created in 1963 and is the official rural
workers union [32]. This form of rural unionism defined its
actions, which are aimed at defending labor rights in various
areas. It is a less ideological movement than the MST. Nonetheless, during the military period, the regime’s op-
position, represented by other social movements saw CON-
TAG skeptically and NGOS criticized the Confederation at
that time. The rural department of the Unified Worker’s
Central, CUT (Central Única dos Trabalhadores), criticized
CONTAG for being submissive, legalistic, and for not solv-
ing the problems of small farmers. Since its origin, CUT’s
rural department stated its intention to form an alternative
union, opposed to CONTAG, closer to the grassroots, less
bureaucratic and more combative in defending the interests
of farmers ([33] p. 4). Nevertheless, CUT and CONTAG
joined forces in 1995, but split up in the late ’90s due to
disagreements. The modernization of agriculture, as proposed by the
government, ignored peasants and gave capitalist compa-
nies the main role in modernizing and developing Brazil’s
countryside. CPT’s main task was to organize peasants’
movements that emerged during the military regime ([29] p. 10). The State’s oppression was directed against those so-
cial movements such as the MST, which was first organized
in 1978 after occupying a farm in the state of Rio Grande
do Sul, with 110 families [30]. 4. MST and CONTAG, Two Nationwide Movements In time the MST expanded
its objectives and included the restoration of the democratic
rule in the country and the pursuit for representation in
Congress - certainly an innovative path for dispossessed
people in Brazil, who had never been politically represented. Within a decade, the movement had turned very popular
among peasants and gained growing support from urban
sectors as well, particularly in universities and other NGOs. Land occupations were adopted as the movement’s
main form of action, but additional strategies for striving
for a piece of land were also included in its operation. Ac-
cording to Fernandes [29], MST’s formation consisted of
three different periods: Creation (1978–1985), mainly in the
South and South-East of Brazil, occupying land from lati-
fundia and large corporations; Consolidation (1985–1990) The modernization of agriculture, as proposed by the
government, ignored peasants and gave capitalist compa-
nies the main role in modernizing and developing Brazil’s
countryside. CPT’s main task was to organize peasants’
movements that emerged during the military regime ([29] p. 10). The State’s oppression was directed against those so-
cial movements such as the MST, which was first organized
in 1978 after occupying a farm in the state of Rio Grande
do Sul, with 110 families [30]. In time the MST expanded
its objectives and included the restoration of the democratic
rule in the country and the pursuit for representation in
Congress - certainly an innovative path for dispossessed
people in Brazil, who had never been politically represented. Within a decade, the movement had turned very popular
among peasants and gained growing support from urban
sectors as well, particularly in universities and other NGOs. A large regional newspaper in Brazil published a report
that highlights differences between MST and CONTAG, re-
garding the organization of an act of protest that occurred in
the year 2000 called ‘The Cry of the Land’ (Grito da Terra),
as follows: Unlike the MST, which rents buses with hard fiberglass
seats, CONTAG rented mostly luxury and double-decker buses
with air conditioning that belong to tour companies. CONTAG
rented eight large tents to receive the protesters. A security
company was responsible for the organization of space and
for watching the belongings of those who slept on site. Land occupations were adopted as the movement’s
main form of action, but additional strategies for striving
for a piece of land were also included in its operation. 3. Empirical Study An additional difference between the members of the
MST in relation to other movements of landless workers
is their high quality of life, as demonstrated by a national
spectrum research conducted by government bodies such
as the Ministry of Agrarian Development (MDA) and INCRA. The reason is the MST’s organizational structure, in which
its affiliates need to get involved and are expected to com-
mit to participatory education that is run according to the
Movement’s goals. The results of this national survey indi-
cated that MST landless workers are considered to have a
higher self-perceived social status in relation to other move-
ments, taking into account social justice, racial democracy
and humanitarian values [31]. Despite its ideological edu-
cation for its rural communities, the MST is also perceived
as an organization that prepares and educates children and
adults with the basic elements for being full citizens and
that demands public services from the authorities. Certainly,
these features help the Movement gain more support and
to rank it as one of the largest in Latin America. 5. DATALUTA DATALUTA is a land-conflicts database containing national
information on land occupation, rural settlements, socio-
territorial organizations and land structure as of 1988. Cat-
egories such as socio-territorial movements, land structure
and peasantry demonstrations were recently incorporated. The dataset was created by the Center for Studies, Re-
search and Projects on Agrarian Reform (NERA) of the São
Paulo State University (Universidade Estadual Paulista, UN-
ESP) and it is part of a larger body, the DATALUTA Network,
which seeks to undertake joint efforts by bringing together
nine research groups from universities located in several
states in Brazil, as follows: UNESP (São Paulo State),
Universidade Federal de Uberlândia (Minas Gerais State),
Universidade Estadual do Oeste do Paraná (Paraná State),
Universidade Federal do Rio Grande do Sul (Rio Grande
do Sul State), Universidade Federal de Mato Grosso (Mato
Grosso State), Universidade Federal de Sergipe (Sergipe
State), Universidade Federal do Espírito Santo (Espírito
Santo State), Universidade Federal da Paraíba (Paraíba
State), and Universidade Federal do Mato Grosso do Sul
(Mato Grosso do Sul State). Apart from these meals, CONTAG prepared a bag with
three ham sandwiches for every protester. What is observed in both movements’ demonstrations
is that the MST is opting for bolder actions, such as the
invasion of public buildings, while CONTAG has given pref-
erence to marches and public events. Both movements,
however, have the option for land invasions in common (5
October 2000) [34,35]. Other differences besides those mentioned above, per-
tain to how the movements understand the role of modern-
ization in agriculture. Even the Brazilian government during
the military regime intended to modernize agriculture but
with a completely different approach than that aspired by
certain social movements such as the MST. While the mili-
tary regime focused on the large capitalist conglomerates
as central actors in this process of modernization, the MST
had a more ideological and clear political view in this re-
gard. Its objectives are to combat the commodification of
land reform, fight agribusiness and transnational companies
because these want to control seeds, production and agri-
cultural trade, besides keeping workers in the field under
labor conditions analogous to slavery and other forms of
subordination. MST proposes that agriculture be focused
on the domestic market, that it respect the environment and
encourages the agricultural cooperation and autonomy of
workers [36]. This network helps to obtain more accurate data on the
agrarian issue. 4. MST and CONTAG, Two Nationwide Movements Ac-
cording to Fernandes [29], MST’s formation consisted of
three different periods: Creation (1978–1985), mainly in the
South and South-East of Brazil, occupying land from lati-
fundia and large corporations; Consolidation (1985–1990) The MST, when camping in the capital Brasilia, usually
occupies public spaces with no lighting, toilets or sanitary 40 conditions. As for CONTAG, it bought 24,000 bottles of
mineral water at the opening of a protest and two big loads
of soft drinks, which were distributed by a truck that fol-
lowed the protest on the Promenade. The MST usually
uses water-tank trucks. The MST offers protesters bread
without butter in the morning and a warm lunch with rice,
beans and low-quality meat while CONTAG provided a com-
plete breakfast and barbecues with several types of meat
and vegetables for lunch and dinner. [36]; and (d) both have national penetration and their mem-
bers are scattered around the country, unlike most regional
agrarian movements. 5. DATALUTA Thus, DATALUTA became a national and
international reference for agrarian conflicts and has en-
abled research exchanges with other countries around the
world [27]. The researchers that contribute to the dataset
give consultancy to social movements such as MST and
CPT, among others. Nonetheless, it is yet not widely known
by scholars abroad at this stage. However, there are few
sources of trustful data in this area and they all present
deficiencies. DATALUTA is no exception, but it is improving
constantly and seems to present more advantages than
methodological problems. Figure 1 shows relevant informa-
tion about Dataluta in its various dimensions, from scale-
level data, categories included in this database, and the
sources consulted for its creation. Nevertheless, the two largest groups defending peas-
ants (MST and CONTAG) have common goals such as (a)
the expropriation of all land holdings belonging to foreign
capital and banks; (b) the expropriation of land in which
‘modern slavery’ is used; (c) both strive for the demarcation
of all the lands of indigenous and quilombo communities SCALE
BRAZIL-MACROREGIONS- STATES
CATEGORIES
Settlements
Occupations
Agrarian Structure
Socio-territorial
Movements
Demonstrations
SOURCES
INCRA
CPT, OAN, DATALUTA
SNCR
CPT, OAN, DATALUTA
CPT, DATALUTA
DATALUTA
Figure 1. DATALUTA scales and categories. Source: [27], p.10. DATALUTA BRAZIL-MACROREGIONS- STATES Figure 1. DATALUTA scales and categories. Source: [27], p.10. 41 41 a total of 682,629 families that took part in land occupa-
tions between 2000 and 2011, 64% were mobilized by the
MST followed by CONTAG, with 8.2%. Table 1 shows 6
different entities, the number of occupations and families
involved with them during the analyzed period, as well as
the respective percentages. Data on land occupation, demonstrations and socio-
territorial movements are collected through secondary re-
search in various journals and academic and non-academic
institutions in states where the DATALUTA network’s re-
search groups are located. When needed, field research
is carried out to better understand different realities. Data
is gathered from different sources, confronted and system-
atized through DATALUTA reports. Although land occupations garner great visibility, there
are many other activities organized by these social move-
ments that also draw attention to the issue of lack of ac-
cess to land faced by a great number of peasants. 5. DATALUTA Among
these participatory strategies are: camping in relevant lo-
calities; solidarity actions; petitions in favor of the cause;
hearings with authorities; road blockages; registration of
interested parties; marches; religious celebrations; siege
of construction sites; concentration in public spaces; meet-
ings; strikes (including hunger and thirst strikes); access
bans; manifestos; task forces; occupation of bank branches;
occupation of public and private buildings; leafleting; vehicle
retentions; pilgrimages; looting; thematic demonstrations;
vigils; among others. DATALUTA’s methodology consists of systematizing data
from primary and secondary sources and organizing it on
different scales: municipal, micro-regional, state, macro-
regional and national levels. Reports in DATALUTA work
with land occupation (since 1988), rural settlements (from
1979), socio-territorial movements (since 2000), land struc-
ture (for the years 1998, 2003, 2010, 2011 and 2012) and
demonstrations (since 2000). Data regarding land occupa-
tion, families and socio-territorial movements are organized
from the following sources: CPT, National Agrarian Ombuds-
man (OAN; 2004–2009) and the data collected from national
and regional journals by the network’s research groups. Data from rural settlements are extracted from INCRA
and confronted annually with the Land Institute of the State
of São Paulo (ITESP) and the National Association of State
Land Bodies (Anoter Foundation). As Brazilian universities
and institutions are not well known for collecting data, not
much could be found on the subject. CPT is an excep-
tion that contributes annually with reports [16] regarding
land conflicts in Brazil since the 1970s. Besides CPT and
DATALUTA, no major or trustful datasets about the issue
were released for empirical research; therefore, more efforts
for collecting and processing data are needed. DATALUTA
is certainly a step in the right direction. The amount of participants involved reflects the relative
level of impact and media coverage these activities receive. Consequently, there is a greater possibility for a favorable
public opinion towards land redistribution. Therefore, not
only potential beneficiaries, but urban supporters also can
(and actually do) join the masses in these events. Many
times these activities take place in large urban centers to
form public opinion and mobilize supporters other than those
directly involved in the struggle for land. Scholars, students
and union workers are some of the potential targets to raise
support and serve as a force in forming public opinion and
pressuring decision makers on issues related to agrarian
reform and land redistribution. The media’s approach and
coverage also plays a relevant role in mass mobilizations. 6. Data Analysis As shown in Table 2, in the twelve-year period (2000–
2011) analyzed, over five million people participated in
these mobilization activities as a strategy to pressure de-
cision makers. This number does not include those who
were not registered or counted, which would potentially
increase the total figure. Also, people can participate in
more than one kind of activity. Therefore, the final number
should be interpreted not as the total number of different
people involved in actions, but the total number of over-
all participants. The most popular activities that mobilized
a substantial number of participants were: pilgrimages;
concentration in public spaces; road blockages; thematic
demonstrations; marches; camps; and walks. Of all the land occupations that happened between 2000
and 2011 in Brazil, the MST participated in no less than
50.2% and CONTAG in, at least, 10.4%. Thus, 62.9% of all
land occupations in the country were primarily organized by
these two movements [37], and that is why this paper has
dedicated a separate subsection for the two movements
mentioned above. Therefore, despite the abundance and
proliferation of representative entities that aim to mobilize
peasants and supporters on a variety of activities in order
to decentralize land tenure and achieve land distribution
in the country, the majority of occupations and most land-
less peasants are under the leadership of few groups. Of Table 1. Different manifestations aimed at attaining visibility to the struggle for land, 2000–2011. MST
CONTAG
CPT
FETRAF
INDIANS
MLST
Total
Number of occupations (%)
2,673 (52.5)
530 (10.4)
165 (3.2)
137 (2.7)
152 (3.0)
127 (2.5)
3,784 (74.3)
Number of families (%)
438,819 (64.0)
56,015 (8.2)
13,763 (2.0)
17,617 (2.6)
11,995 (1.7)
15,583 (2.3)
553,792 (80.8)
Source: Based on [27]. Table 1. Different manifestations aimed at attaining visibility to the struggle for land, 2000–2011. 42 Table 2. Different manifestations organized to gain visibility for the struggle for land, 2000–2011. Actions
Events
People
Actions
Events
People
Circling public spaces
2
3,600
Manifestos
29
12,240
Camping
468
265,960
Marches
334
498,487
Solidary Actions
2
37
Task Forces
1
150
Petitions
2
n/d
Occupation of bank branch
224
48,626
Hearings
41
3,894
Occup. of private building
99
48,414
Protests on boats
7
950
Occup. 6. Data Analysis of public building
1472
558,745
Road blockages
1,517
584,985
Occupations of public/private buildings
131
24798
Registration
1
n/d
Leafleting
4
4,500
Walks
233
194,023
Tolls
2
200
Religious Celebrations
47
17,935
Vehicle Retention
4
700
Sieges of construct. sites
35
15,400
Pilgrimages
166
1,112,868
Concent. in public spaces
1,975
932,237
Lootings
33
7,727
Meetings
8
2,060
No information
70
31,213
Strikes
1
n/d
Thematic demonstration
570
604,853
Hunger strikes
9
74
Occupation attempts
7
5,400
Thirst strikes
1
2
Looting attempts
4
170
Interdictions
15
15,804
Tratoraço [37]
1
800
Fasting
21
1,053
Vigils
74
21,859
Total
7610
5,018,777
Source: Based on [27]. Table 2. Different manifestations organized to gain visibility for the struggle for land, 2000–2011. Of the 7,610 events registered in these twelve years, the
MST participated in 3,305 (43%). This movement is excep-
tional at mobilizing rural workers. Supporters of their cause
tend to organize on the national and international levels—
the MST provides training to Latin American peasants and
exchanges knowledge with different organizations in the
field, such as the Bolivian MST and La Via Campesina. The
CPT is one of the most important entities that support the
struggle for land distribution, mobilizes peasants, and also
provides legal and social assistance. The CPT participated
in 935 mobilization activities; the Brazilian Movement of
People Affected by Dams (MAB) participated in 446 mobi-
lizations; the Rural Workers’ Union (STR) took part in 291
activities; the Federation of Agricultural Workers (FETAGRI)
participated in 210 activities; and the Confederation of Agri-
cultural Workers (CONTAG), in 109 mobilizations. Contrary
to the occupations’ figures, CONTAG had a much more
restrained participation. It is important to note that several
of these actions were jointly organized by more than one of
these entities. of demonstrations and participants. This was probably to
the benefit of the PT candidate and did not influence neg-
atively the party’s support given it has the most affinity to
land distribution issues. The year 2010 was the third time
a presidential candidate from the PT was running for office
with real chances to win, and the results of a tiny agrar-
ian reform and dissatisfaction with that policy could impact
demonstrations against the PT government by agrarian re-
form supporters. However, the actual figures on the number
of settlements created during the two terms of President
Cardoso do not differ considerably from those created dur-
ing Lula’s two terms. 6. Data Analysis Both, the Brazilian Social Democratic
Party, PSDB (Partido Social-Democrata Brasileiro) and the
Workers’ Party, PT (Partido dos Trabalhadores) achieved
advances for the rural population with progressive public
policies. However, they both chose to carry out a conser-
vative land reform greatly benefiting agribusiness instead
of reducing inequality in the countryside. The result was
more land concentration and less income for the poor rural
population [29], expanding areas for sugarcane (for the pro-
duction of ethanol) and soybeans intended for agribusiness. There was an increase in the number of victims of land
conflicts both among peasants and agribusiness industrial
farmers [27]. Figure 2 presents data on the number of demonstra-
tions and participants in a compact and overlapping manner,
showing fluctuations and indicating a trend. The changes in both the number of demonstrations and
the number of family participants are cyclical. On presi-
dential election years (2002, 2006 and 2010) in which a
candidate from the Workers Party (PT) had real chances
of winning (and in fact, they won all three elections: Lula
won in 2002 and 2006, and Dilma Rousseff won in 2010)
[38], there was a significant reduction in activities. The
exception was 2010, with a slight increase in the number Alongside mobilization activities aimed at communicat-
ing demands to Brazilian society and decision makers, land
occupations are strategically planned to rouse Government
intervention and accelerate the process of agrarian reform. Therefore, the following figure (Figure 3) shows the disper-
sion of land occupations organized by social, religious and
union movements. 43 Figure 2. Number of demonstrations and participants,
2000–2011. Source: Based on [27]. Figure 3. Number of land occupations and people involved,
1988–2011. Source: Based on [27]. in the governing coalition during Cardoso’s, as well as Lula’s
and Dilma Rousseff’s administrations. This suggests a vot-
ing ambiguity on issues relating to land distribution and
agricultural activity in both houses of the Brazilian Congress. To illustrate this idea, in the 1994 and 1998 elections there
were about 150 congressmen forming the Rural Caucus. Its ideological scope ranged from right to center-left, which
made land redistribution in the country politically unfeasible
([22] p. 164). This can be considered a great obstacle in
achieving fairer land distribution in the country. 6. Data Analysis Both mobilization activities and land occupations orga-
nized by civil society entities encumber participants’ involve-
ment in order to receive a plot of land, which, in the end,
is nothing more than selective incentives ([40] p. 51) to
participate in this risky endeavor. Figure 2. Number of demonstrations and participants,
2000–2011. Source: Based on [27]. Figure 2. Number of demonstrations and participants,
2000–2011. Source: Based on [27]. Figure 3. Number of land occupations and people involved,
1988–2011. Source: Based on [27]. With the risks involved in collective action of this kind,
Figure 4 shows the number of peasants and supporters
killed over issues pertaining land dispute. This figure in-
cludes participation in mobilization activities organized by
social movements. The data confirms that participation in
the struggle for land can result in death, as illustrated by
data from the CPT. An interesting fact observed in the data is that, after 1988
(when the democratic regime came into effect in Brazil) the
number of killings over land disputes was significantly re-
duced. The year coincides with the promulgation of the
1988 Constitution, which transformed agrarian issues and
facilitated expropriation in the case of unproductive land and
non-compliance with the social function as described in the
document. During the first three years of the democratic
regime (1985–1988), the number of killings over land dis-
putes was huge. The new Constitution may have played
an important part in changing this, influenced by additional
factors, such as the emergence and strengthening of social
movements. That is, from that year on, there were laws
that benefited disadvantaged rural groups by favoring land
distribution and expropriations by compensating landowners
in case the land did not comply with the minimum production
standards. Moreover, non-governmental organizations that
were legally supported by the Government advocated for the
peasants’ interests, providing attorneys, land surveyors and
other available assistance. Therefore, the peasants were
no longer alone against large landowners and were increas-
ingly aware of the laws and strategies to legally acquire a
plot of land, although they often gave the impression of civil
disobedience by occupying land that belonged to others. Figure 3. Number of land occupations and people involved,
1988–2011. Source: Based on [27]. 6. Data Analysis Data is available as of the year of the promulgation of
the new Brazilian Constitution, in 1988, when the social
function of the land became an intrinsic part of mobiliza-
tions surrounding the issue of agrarian reform, encouraging
groups of landless rural workers to occupy primarily private
land provided it was unproductive. The same fluctuations can be observed both on the
number of occupations and the number of people involved
in them. Similarly, in election years when a candidate from
the Workers Party (PT) had a real chance of winning (2002,
2006 and 2010) the number of occupations and of partici-
pants dropped. The opposite occurred in previous election
years, 1994 and 1998, when Cardoso was the front-runner
candidate and subsequently won. This indicates that there
is political coherence in the movements participating in
the struggle for agrarian reform, although in absolute num-
bers, as previously mentioned, the number of settlements
achieved does not differ drastically. During the eight years
of the Cardoso administration, 584,655 families were set-
tled, compared to 614,093 settled families during Lula’s two
terms ([39] p. 46). Yet, the available area for these set-
tlements was significantly larger in the PT administrations,
demonstrating the party’s affinity for agrarian reform. How-
ever, in terms of governance, when alliances are not always
ideologically coherent, it is more difficult to implement land
redistribution policies that benefit poorer peasants. Despite the decrease, killings over land dispute continue
to happen, even though the number has been stable since
the promulgation of the 1988 Constitution. Violence ensues
and peasants are being killed and suffering numerous hu-
man rights violations as a result of the struggle for land
redistribution. It has been stated that several mobilization activities of
the poor rural populations are carried out by organizations
that seek agrarian reform. The occupation of unproductive
land, which in part [41] is consistent with the law, is one of
the strategies that achieve the greatest impact. Among members of Congress, the so-called Rural Cau-
cus (Bancada Ruralista) was formed to advocate privileges
for large landowners and oppose the agrarian reform. These
politicians belong to different political parties and many were 44 Figure 4. People killed over land disputes, 1985–2011. Source: Based on [16]. 6. Data Analysis That is, in election years when the Left had realistic
chances of winning the presidential (re)elections, the num-
ber of mobilization activities, along with land occupations
resulting in settlements, was greatly reduced. On other oc-
casions, the number in each of those categories increases,
including the figures on land occupation and settlements in
election years when the chances of a Center party candi-
date winning the Presidency were higher. Overall, the activities and strategies of the social move-
ments to optimize land distribution in democratic Brazil are
coherent and the results are consistent with mobilization
activities. Figure 4. People killed over land disputes, 1985–2011. Source: Based on [16]. Therefore, what does this progress mean in terms of
land redistribution? This cannot be referred to as agrarian
reform, because it is forced by land occupation and other ac-
tivities that pressure authorities. It is not a planned agrarian
reform, executed without pressure from the public opinion
and avoidance of further bloodshed. In 2003, there was an abrupt increase in the number
of people that died as a result of land conflicts. It was in
the beginning of Lula’s first term in Office. This hike in the
number of deaths caused by land conflicts can be inter-
preted as a result of the peasants’ and their supporters’
high expectations generated by the workers’ party assum-
ing the Federal Government for the first time. Consequently,
there is an explicit increase in land conflicts because the
peasants and smallholders understand that the government
is on their side and that it can finally implement the agrarian
reform more vigorously. Concurrently, the large landowners
and agribusiness entrepreneurs are willing to defend their
territory and, thus, more deaths result from these clashes. By analyzing Table 3, it is observed that there was a
quantitative leap in the two terms of the Cardoso and Lula
administrations compared to previous governments with re-
gards to land redistribution in Brazil. In the two terms of
President Cardoso’s administration there were 4,310 settle-
ments created, while 3,602 settlements were created during
the eight years of Lula’s administration. However, the ex-
tension of settlements increased during Lula’s government;
while a smaller number of families were settled, the area size
per family was much larger. This allows both for subsistence
and trade to take place in the establishments [42]. 6. Data Analysis Given these extreme demonstrations and the risk to human
life pertaining to the struggle to access land for peasant sub-
sistence, how are these inputs being translated into outputs? That is, do land occupations and other mobilization activities
result in a higher number of settlements? Figure 5 shows the
number of implemented settlements since the promulgation of
the new Constitution in 1988 to the present day. p
[
]
We have to differentiate two terms used in the interna-
tional literature, which are not applied in the Portuguese
and Spanish languages for lack of the expression “Land
Reform”, but only Agrarian Reform. Thus, there is confu-
sion in the discussion of the agrarian issue in Latin America. Agrarian reform generally embodies land reform as its most
important element. “Agrarian reform includes both redis-
tributing land and assisting new landowners by assuring
them inputs and markets, extending credit and imparting
certain technology that will help them to become agricultural
producers” ([43] p. 12). Of prime importance is the rural
credit and technical knowledge. Peasants and peasants’
cooperatives in possession of newly acquired title lands can
accomplish little, if anything, without access to adequate fi-
nancial resources, enabling them to purchase seeds, stock,
tools, fertilizers, pesticides, etc. and install water supplies
in some cases. Therefore, the Lula administration’s larger
average-sized settlements do not imply more quality than
those of previous administrations. Nonetheless, given data
restrictions, it is not possible to analyze land fertility or in-
frastructure elements such as access to schools, electricity,
roads, medical services, and so on. The lines representing settlements and settlers follow
the same direction as the previous figures, also showing an
important relation to election years, starting on the year of
the new Constitution. By placing Figures 2, 3 and 5 side by side and compara-
tively analyzing them as if they were a temporal combination,
that is, as fitting in pre-determined goals, culminating in an
agrarian reform and in fairer and less unequal land distri-
bution, they all meet the movements’ purposes of social
justice for peasants. Figure 5. Number of settlements and settlers, 1988–2011. Source: Based on [27]. Violence over land disputes has dramatically decreased
when coupled with land redistribution. When there is real
distribution to families who need it most, it enables produc-
tivity, and these families leave the poverty line, reducing the
pressure for survival and for land disputes. Figure 5. 6. Data Analysis In terms of agrarian policy, the ineffectiveness of the
State must be considered, regardless of which government
currently holds office. According to data from NERA, there
is a disparity that stands out between the number of fami-
lies each settlement could hold and the number of families
actually placed in these settlements. That is, in the twelve-
year period observed in this study (2000–2011), 496,368
families could have been placed in the created settlements. However, the actual number of settled families was 366,621,
or a 26 percent underperformance. This figure is not negli-
gible, and could even improve the performance of the State
with regards to civil society’s demands for agrarian reform. Land concentration in the hands of a few former plan-
tation owners, who are currently operating in agribusiness,
allied to a financial sector, which invests in agriculture, re-
quires State action. It is understandable that strong in-
terest groups have put considerable pressure to increase
the amount of land destined to agribusiness in the country. However, democracy should serve all citizens, and allow the
poor to raise their living standards in rural areas by giving
them access to land. The best social program for peasants
is a plot of land so that they can derive their own livelihood
from their work. Certainly, the Bolsa-Família Program [45]
helps the poor, but it has a temporary role to momentarily
alleviate poverty. As long as it is supported by other pub-
lic policy tools, like financing, agricultural capacitation, in
addition to basic education, health and security services
needed by all, access to land is the permanent solution for
overcoming poverty. What can be inferred from the data presented in this study
is that violence in rural Brazil is derived from a lack of political
interest to carry out land reform and achieve a more equitable
distribution by the different levels of Government, which would
correct mistakes of the past and break the vicious cycle the
country has been dwelling in for centuries. Rural Brazil, unlike
its urban counterpart, is lagging, as cronyism prevails and
laws are not respected [22], leading to a struggle between the
strongest and the dispossessed. Extreme inequality among
people creates insurmountable social chasms and requires
state intervention to be repaired. The Brazilian State always
favors large landowners and agribusiness. 6. Data Analysis As evidenced by the increased number of rural settle-
ments created by the State in the past two decades, the
strategies used by social movements to achieve land distri-
bution for low-income peasants have had a positive effect. There is still a lot to be done to repair this extreme inequality,
but the first steps have been taken and are developing. The
violence applied by large landowners and the reaction of
the State towards the peaceful activities organized by social
movements perpetuate the victimization of the peasants. In this sense, public policies to decrease violence caused 6. Data Analysis Number of settlements and settlers, 1988–2011. Source: Based on [27]. 45 Table 3. Results in land distribution relating to agrarian violence in democratic Brazil. Recent Democratic Governments
Land distribution and violence
Sarney
Collor/Franco
Cardoso
Cardoso
Lula
Lula
1985–1990
1991–1994
1995–1998
1999–2002
2003–2006
2007–2010
Settlements Created
537
348
2,329
1,981
2,386
1,216
Families per settlement
115
81
100
72
76
65
Hectares per settled family
72
93
53
63
180
214
Average of hectares per settlement
8,278
7,515
5,319
4,540
13,683
13,941
Killings over land disputes
672
199
150
120
189
115
Average of killings over land dispute/year
112
49.75
37.5
30
47.25
28.75
Ratio between created settlements and people killed
4.8
7
62.1
66
50.5
42.3
Source: Based on [16] and [27]. Table 3. Results in land distribution relating to agrarian violence in democratic Brazil. tation in State institutions. It has also retrieved citizenship
for most of the poor, as well as bolstered the struggle for a
more equal country where everyone can have the minimum
to survive. While during the Sarney administration there was one per-
son killed for every 5 settlements created, in the Collor/Franco
administration there was one person killed for every 7 new
settlements. In the Cardoso administration, out of every 64
settlements, one person was killed over land disputes, and dur-
ing Lula’s government, the number was 46 new settlements
per one killing over land disputes. That is the price paid in
human lives to implement a public policy that is necessary to
the survival of the peasantry, especially of the poorest. As
more settlements are created, the tension between agrarian
sectors over land acquisition decreases, as it allows more
people to support themselves by farming their own land. Brazil is far from enabling everyone to live a decent life
and from being able to support all of its citizens during times
of social and economic hardships. However, as presented,
there have been significant advancements for the poorer
sectors of the population living in rural areas in the past two
decades. Brazil was the 7th world economy in 2012 [44],
but ranked 81st in per capita income [44], which means that
the country is rich, but the majority of the population is poor. Due to the country’s size, there is still much to be done, and
inequality must be reduced. References and Notes nance, and Conflict. Journal of Conflict Resolution. 2005;49:625–633. nance, and Conflict. Journal of Conflict Resolution. 2005;49:625–633. [1] Agricultural trade in 2012: A good story to tell in a
difficult year? European Commission; 2013. Available
from: http://ec.europa.eu/agriculture/trade-analysis/
map/2013-1_en.pdf. [15] Hegre Tanja Ellingsen and Scott Gates and Nils Pet-
ter Gleditsch H. Toward a Democratic Civil Peace? Democracy, Political Change, and Civil War. American
Political Science Review. 2001;95(01):33–48. [16] Conflitos no campo Brasil. Goiânia, Goiás, Brazil:
Comissão Pastoral da Terra; 1985–2011. Available
from: http://cptnacional.org.br/index.php/component/
jdownloads/viewcategory/43-conflitos-no-campo-
brasil-publicacao. [2] Wolford W. This land is our now: Social mobilization
and the meanings of land in Brazil. Durham, NC, USA:
Duke University Press; 2010. [3] Clapp J. Food. Cambridge, UK: Polity Press; 2012. [4] Paarlberg R. Food politics: What everyone needs to
know. New York, NY, USA: Oxford University Press;
2010. [17] Based on interviews with the Land Pastoral Commis-
sion (CPT)s attorney in November 2007 in the city of
Marabá, state of Pará, Brazil. By the way, many details
throughout this paper are based on several interviews
the author gave on the ground to agrarian leaders. That is the reason why the quotes are not overused. [5] Maystadt JF, De Luca G, Sekeris PG, Ulimwengu
J. Mineral resources and conflicts in DRC: A case
of ecological fallacy? Oxford Economic Papers. 2014;66(3):721–749. [6] Rustad AS, Binningsbo HM. A price worth fighting for? Natural resource and conflict recurrence. Journal of
Peace Research. 2012;49(4):531–546. [18] The origin is actually historical. In the XV century, the
Portuguese Crown, concerned with local production
and tax collection, gave land to people for a period
of time (sesmarias) so they would produce and pay
taxes. This kind of production generated Plantations,
which utilized slave labor in the Americas explored by
the Europeans. This lasted until 1822, and originated
the well-known latifúndio in the country. [7] Paivi L. The spoils of nature: Armed civil conflict and
rebel access to natural resources. Journal of Peace
Research. 2010;47(1):15–28. [8] Basedau M, Lay J. Resource Curse or Rentier Peace? The Ambiguous Effects of Oil Wealth and Oil Depen-
dence on Violent Conflict. Journal of Peace Research. 2009;46(6):757–776. [19] Violência rural no Brasil. Human Rights Watch - Amer-
icas Watch Report; 1991. [9] Simmons CS. Territorializing land conflict: Space,
place, and contentious politics in the Brazilian Ama-
zon. GeoJournal. 2005;64(4):307–317. [20] Borras Jr SM, Franco JC, Borras Jr SM, Franco, Bor-
ras SMJ. 7. Final Remarks Democracy has ushered the freedom of individuals to be
affiliated to social movements and allowed for their represen- 46 areas to defend the poor and enforce the law, advancing
democracy for all in Brazil. by land disputes must be created, giving land to those that
need it most. In order to avoid further bloodshed, agrar-
ian reform must be a long-term commitment by the State,
instead of belonging to this or that government at a given
time in history. Acknowledgements I wish to express my sincere gratitude to the Land Project
discussion group, headed by Wendy Wolford at Cornell
University, where I spent twelve months during the aca-
demic year 2013–2014. I also thank FAPESP (São Paulo
Research Foundation in Brazil), which gave me financial
support during my stay at Cornell, and afterwards to present
this paper at the International Studies Association (ISA) an-
nual convention at Atlanta in 2016. The ‘price’ of a peasant’s life in terms of new settlements
is increasing and, today, as more settlements are created
fewer peasants die as a result of land conflicts. However,
the path to further reducing that ratio is long and difficult. Both the lack of State intervention in land disputes and/or
the failure of the police to protect rural workers and landless
people could be seen as reasons for peasants to die in
agrarian conflicts in Brazil. More needs to be done in rural References and Notes Global Land Grabbing and Trajectories of
Agrarian Change: A Preliminary Analysis. Journal of
Agrarian Change. 2012;12(1):34–59. [10] Ross ML. What do you know about natural re-
sources and civil war? Journal of Peace Research. 2004;41(3):337–356. [21] Hall A. Land tenure and land reform in Brazil. In:
Prosterman, R L ; Temple, M N ; Hamstead TM, edi-
tor. Agrarian reform and grassroots development: Ten
case studies. Boulder, CO, USA: Lyenne Rienner;
1990. pp. 205–232. [11] See the complete definition on Polity IV, available at
http://www.systemicpeace.org/inscr/p4manualv2012. pdf. [22] Hammond JL. Land Occupations, Violence, and the
Politics of Agrarian Reform in Brazil. Latin American
Perspectives. 2009;36(4):156–177. [12] Ahmadov AK. Oil,
Democracy,
and Context:
A Meta-Analysis. Comparative Political Studies. 2014;47(9):1238–1267. [23] Alston LJ, Libecap GD, Mueller B. A model of rural
conflict: Violence and land reform policy in Brazil. Envi-
ronment and Development Economics. 1999;4(2):135–
160. [13] Vreeland JR. The Effect of Political Regime on Civil
War Unpacking Anocracy. Journal of Conflict Resolu-
tion. 2008;52(3):401–425. ( )
[14] Collier P, Hoeffler A. Resource Rents, Gover- 47 [24] Feliciano C. O movimento camponês rebelde e a
geografia da reforma agrária. [M.A. Dissertation]. Uni-
versity of São Paulo, Department of Geography, São
Paulo, SP, Brazil; 2003. 1-milhao-na-organizacao-do-grito-da-terra. [35] Despite newspapers calling it land invasions, a more
neutral term used by social movements is land occu-
pations. [25] Ondetti G. Repression, Opportunity, and Protest: Ex-
plaining the Takeoff of Brazil’s Landless Movement. Latin American Politics & Society. 2006;48(2):61–94. [36] Scherer-Warren I. The social movements’ politics for
the rural world. Estudos, Sociedade e Agricultura. 2007;15(1):5–22. [26] Center of Studies, Research and Projects of Agrarian
Reform (NERA) of the State University of São Paulo
(UNESP). [37] This figure can be even higher if partnerships between
movements, which are common in many nooks in the
country, were counted. [27] Dataluta:
Banco de dados da luta pela terra. Presidente Prudente, SP, Brazil:
UNESP; 2013. Available
from:
www2.fct.unesp.br/nera/projetos/
dataluta_brasil_2013.pdf. [38] Actually, Dilma Rousseff won both elections for presi-
dent (2010 and 2014), but this paper does not cover
the 2014 presidential election). [39] Zimerman A. Land kills: The Brazilian experience. Population Review. 2012;51(2):41–58. [28] In this specific case, the MST trains the leaderships
of neighboring countries. Additionally, the La Via
Campesina movement, founded in 1994, is also active
in Brazil. References and Notes The movement is transnational and com-
prises about 150 different peasant movements in ap-
proximately 70 countries around the globe, represent-
ing 200 million smallholder farmers. [40] Olson M. The Logic of Collective Action: Public Goods
and the Theory of Groups. Cambridge, MA, USA:
Harvard University Press; 1965. [41] The “in part” is included because an occupation should
not be performed before a procedural decision. How-
ever, it is through these occupations that the State
is pressured and the movement gains strength and
requests prompt decisions to avoid greater loss and
further deaths. [29] Fernandes BM. Re-peasantization, resistance and
subordination: The struggle for land and agrarian re-
form in Brazil. Agrarian South Journal of Political Econ-
omy. 2013;2(3):1–21. [42] Obviously the size of the area necessary for subsis-
tence depends on various factors, such as cultivated
crop, quality of the soil, irrigation, use of fertilizers and
pesticides, climate, among others. [30] Medeiros LS. História dos movimentos sociais no
campo. Rio de Janeiro, RJ, Brazil: Fase Editora; 1989. [31] Valente RR, Berry BJL. Countering Inequality:
Brazil’s Movimento Sem-Terra. Geographical Review. 2015;105(3):263–282. [43] Thiesenhusen WC. Broken promises: Agrarian reform
and the Latin American campesino. Boulder, CO, USA:
Westview Press; 1995. [32] Watanabe CH. Desenvolvimento rural sustentável
solidário e a formação sindical: A experiência da CON-
TAG. Raízes. 2010;29(1):166–173. [44] National accounts main aggregates database. United
Nations; 2013. Available from: http://unstats.un.org/
unsd/snaama/dnltransfer.asp?fID=2. [33] Picolotto EL. Quando novos atores entram em cena
o que muda? A construção da representação política
dos agricultores de base familiar no sul do Brasil. Re-
vista espaço acadêmico. 2008;91:1–7. [45] The Bolsa-Familia Program is an income-transfer pro-
gram carried out by the Brazilian Federal Government,
and was considered one of the most effective instru-
ments to combat poverty, increasing access to educa-
tion, which in itself reduces poverty in the medium and
long terms. [34] CONTAG
gasta
R$1
milhão
na
organiza-
ção
do
Grito
da
Terra. Diário
do
Grande
ABC. 2000
May
10;
Available
from:
http:
//www.dgabc.com.br/Noticia/265486/contag-gasta-rs- 48 Appendix Table A1. Socio-territorial Movements and Brazilian States in which they operate (2000–2011). Table A1. Socio-territorial Movements and Brazilian States in which they operate (2000–2011). Acronyms
Translation of movements names in Brazil
Brazilian States where actions were taken
ABUST
Brazilian Association of the Social Land Use
SP
ACRQ
Association of Remnants of Quilombo Commu-
nities
MG, PE
ACRQBC
Association of Remaining Quilombo Communi-
ties of the Creoles Heath
MG
ACUTRMU
United Community Association of Rural Work-
ers
MG
ADT
Association of Land Right
GO
AMIGREAL
Association of residents of micro regions of the
state of Alagoas
AL
AMPA
Association of Small Farmers Movement
AP
ARST
Association Renewal of the Landless
SP
ASA
San Antonio Association
MT
ASPARMAB
Association of Small Farmers of Maraba
PA
ASTECA
Technical Association of Agricultural Coopera-
tion
MT
AST
Association of Landless
PA
ASTST
Association of Landless and Homeless
MG
ATUVA
Association of United Workers of Vila Aparecida
PA
CAA
Alternative Agriculture Center
MG
CAR
Center of Roraima Settlers
RR
CCL
Citizenship and Leadership Center
MG
CETA
Coordenação Estadual de Trabalhadores As-
sentados
BA, RS
CLST
Liberation Way of the Landless
MG
CODEVISE
Defence Committee of Victims of Santa Elina
RO
CONAQ
National Joint Coordination of Quilombo Com-
munities
ES
CONLUTAS
National Coordination of Struggles
SP
CONTAG
National Confederation of Agricultural Workers
AC, AL, BA, CE, DF, ES, GO, MA, MG, MS, MT,
PA, PB, PE, PI, PR, RJ, RN, SP, TO
COOTERRA
Cooperative Farmers in the Struggle for Land
BA
CPT
Pastoral Land Commission
AL, BA, CE, MG, MS, MT, PA, PB, PE, PI RN,
RJ
CTV
Land Alive Center
SP
CUT
Workers’ Unitary Central
AC, MS, RS, SP
FATRES
Support foundation to rural workers and farmers
of the Sisal and semi-arid
BA
FERAESP
Federation of Rural Workers of the State of São
Paulo
SP
FETRAF
Workers Federation of Family Farm
DF, GO, MG, MS, PA, PE, PI, SP
FRUTO DA TERRA
Land Fruit
DF, GO, MG, MS, PA, PE, PI, SP
FST
Triangle Social Forum
MG
FUVI
United Families of Valley Vilhena
MS
GERAIZEIROS
North of Minas Gerais Geraizeiros
MG
GRUPO XAMBRE
XAMBRE Group
PR
LCC
Peasant League Corumbiara
RO 49 Table A1: Cont. Table A1: Cont. Appendix Acronyms
Translation of movements names in Brazil
Brazilian States where actions were taken
LCP
League of Poor Peasants
AL, CE, GO, MG, PA, RO, SP
LOC
Workers Peasant League
MG
MAB
Movement of People Affected by Dams
CE, MG, MT, PB, PR, RS , SC
MAST
Movement of Landless Farmers
SP, PR
MATR
Movement in Support of Rural Workers
DF, GO
MBUQT
Brazilian United Movement Searching for Land
SP
MCC
Peasants’ Movement of Corumbiara
RO
MCNT
Movement Conquering Our Land
PA
MCP
Movement of the People’s Councils
CE
MCST
Movement of Landless Needy
SP
MLST
Movement for the Liberation of the Landless
AL, GO, MG, PE, PR, RN, SP
MLSTL
Movement for the Liberation of the Landless
Fight
MG
MLT
Movement of Struggle for Land
AL, BA, MG, SP
MLUPT
Movement United Fight for Land
MG
MMA
Movement of Women Farmers
CE, PR, SC
MMC
Movement of Rural Women
AL
MOVIMENTOS INDÍGENAS
Indigenous Movements
AL, AM, BA, CE, ES, GO, MG, MS, MT, PB, PE,
PR, RO, RR, RS, SC, SP, TO
MNF
Landless Movement New Force
SP
MPA
Small Farmers Movement
DF, ES, PA, RJ, RO, RS
MPRA
Popular Movement for Agrarian Reform
MG
MPST
Popular Movement of the Landless
MG
MPT
Pacific Movement
SP
MRC
Peasant Resistance Movement
BA
MSO
Organized Social Movement
PR
MSONT
Land Dream Movement
PR
MSST
Social Movement of the Landless
AL, PR, RJ
MST
Landless Workers’ Movement
AL, BA, CE, DF, ES, GO, MA, MG, MS, MT, PA,
PB, PE, PI, PR, RJ, RN, RO, RR, RS, SC, SE,
SP, TO
MST DA BASE
Landless Workers’ Movement Base
SP
MSTA
Landless Movement of the Amazon
AM
MSTR
Trade Union Movement of Rural Workers
ES, RO
MT
Workers Movement
AL, PE
MTA
Settlers Movement of Workers
DF, MT, RO
MTAA
Movement of Camped Workers and Settlers of
Mato Grosso
MT
MTB
Brazil Land Movement
PE, PR, SP
MTD
Movement of Unemployed Workers
BA, DF, RJ
MTBST
Landless Brazilian Workers Movement
PE
MTL
Land, Work, and Freedom Movement
AL, BA, GO, MG, PB, PE, RJ
MTL-DI
Movement Land, Labor and Freedom - Demo-
cratic and Independent
GO
MTP
Land, Work and Progress Movement
AL
MTR
Movement of Rural Workers
MG, MS, MT, PR
MTRST
Landless Workers’ Movement
ES
MTRSTB
The Brazilian Landless Workers’ Movement
SP
MTRSTP
Movement of Landless Workers of Paraná
PR
MTRUB
Movement of Rural and Urban Workers
PE
MTS
Movement for a Socialist Tendency
RJ Translation of movements names in Brazil 50 Table A1: Cont. Appendix Table A1: Cont. Table A1: Cont. Acronyms
Translation of movements names in Brazil
Brazilian States where actions were taken
MTST
Movement of Landless Workers
MG, PE, PR, SP
MTSTCB
Movement of Landless Workers and Central of
Brazil
SP
MTV
Land Alive Movement
SP
MUB
Brazilian United Movement
SP
MUST
United Movement of the Landless
SP
MUT
Movement United for Land
PR
OAC
Agrarian Peasant Organization
PR
OITRA
Workers Inclusion Organization for Agrarian Re-
form
SP
OLC
Organization for Action in the Field
BA, PE
OLST
Organization for the Liberation of the Landless
MG
OTC
Organization of Workers in the Field
CE, GO, MG, PA, PR, RO, RS, SP, TO
QUILOMBOLAS
Quilombolas (Maroons)
BA, ES, MG, PE, PR, RJ, SP
RACAA SUL
Service Network of Camped and Settlers of
Southern Bahia
BA
RC
Peasant resistance
PI
SINPRA
Union of Small and Medium Farmers
PA
ST
Landless
SE, SP
STL
Union of Workers in Farming
RN
TERRA LIVRE
Field and City Popular Movement
GO
TUPÃ 3e
Workers of the Municipality of Tupanciretã
RS
UAPE
Union of Pernambuco Farmers
PE
UFT
Union Strength and Land
MS
UNASFP
The Pasture Fund Associations Union
BA
UNIDOS PELA TERRA
United by the Land
SP
UNITERRA
Union of Social Movements for the Land
MG
USST
Union of Landless Santanenses
RS
UST
Workers Trade Union
SP
VIA CAMPESINA
Via Campesina
BA, MG, PB, PI, PR, RS, SP Translation of movements names in Brazil
Brazilian States where actions were taken Translation of movements names in Brazil 51
|
https://openalex.org/W3217729804
|
https://www.nature.com/articles/s41598-021-02058-9.pdf
|
English
| null |
Design and analysis of Maxwell fisheye lens based beamformer
|
Scientific reports
| 2,021
|
cc-by
| 6,242
|
Design and analysis of Maxwell
fisheye lens based beamformer
OPEN Muhammad Ali Babar Abbasi1*, Rafay I. Ansari2, Gabriel G. Machado1 & Vincent F. Fus Antenna arrays and multi-antenna systems are essential in beyond 5G wireless networks for
providing wireless connectivity, especially in the context of Internet-of-Everything. To facilitate this
requirement, beamforming technology is emerging as a key enabling solution for adaptive on-demand
wireless coverage. Despite digital beamforming being the primary choice for adaptive wireless
coverage, a set of applications rely on pure analogue beamforming approaches, e.g., in point-to-
multi point and physical-layer secure communication links. In this work, we present a novel scalable
analogue beamforming hardware architecture that is capable of adaptive 2.5-dimensional beam
steering and beam shaping to fulfil the coverage requirements. Beamformer hardware comprises of
a finite size Maxwell fisheye lens used as a scalable feed network solution for a semi-circular array of
monopole antennas. This unique hardware architecture enables a flexibility of using 2 to 8 antenna
elements. Beamformer development stages are presented while experimental beam steering and
beam shaping results show good agreement with the estimated performance. The vision for Beyond 5G (B5G) networks paves the path towards ultra-reliable low-latency (URLLC) com-
munications. Recent times have seen a sharp rise in the number of connected devices, especially in the context
of internet-of-everything (IoE)1. Wireless communication technologies have found widespread use in different
applications that impact agriculture, transport and healthcare to name a few. Therefore, new and innovative tech-
niques have been proposed for providing seamless coverage and improved link quality to provide high data rates. In the new communication infrastructure, multi antenna systems have been utilized to provide better antenna
gain, enabling enhanced communication link quality2. Additionally, providing seamless connectivity becomes
more challenging for applications in mobile platforms, e.g. unmanned aerial vehicle (UAV), vehicular and satellite
networks where electrically scanned arrays are beneficial. To overcome these challenges, beamforming techniques
have been utilized to enhance the network coverage and overcome the limitations due to interference and path
loss. Similarly, beam steering allows the transmissions to be directed at a particular angle(s), ensuring seamless
connectivity for mobile devices3.h y
Three well-known beamforming techniques are analog, digital and hybrid beamforming. Generally a sin-
gle radio frequency (RF)-chain is used to connect all the antenna elements to a transmitter/receiver module
in analog beamforming architecture, while in the digital beamforming, a dedicated RF-chain is required for
each antenna element. www.nature.com/scientificreports www.nature.com/scientificreports 1Institute of Electronics, Communications and Information Technology (ECIT), Queen’s University Belfast, Belfast,
UK. 2Department of Computer and Information Sciences, Northumbria University, Newcastle upon Tyne NE1 8ST,
UK. *email: m.abbasi@qub.ac.uk nstitute of Electronics, Communications and Information Technology (ECIT), Queen’s University Belfast, Belfast,
UK. 2Department of Computer and Information Sciences, Northumbria University, Newcastle upon Tyne NE1 8ST,
UK. *email: m.abbasi@qub.ac.uk Design and analysis of Maxwell
fisheye lens based beamformer
OPEN (a) Maxwell Fisheye lens based beamformer with transceiver input in first half of the lens, and
monopole circular antenna array in the second half of the lens, (b) Ray tracing diagram depicting source and
image locations in a metallic mirror enclosed Maxwell fisheye lens (c) Simulated absolute E-field inside the lens
structure when a signal at 10 GHz is excited at the source probe of the lens, (d) Comparison between far-field
directivity of an ideal semicircular antenna array and the proposed Maxwell fisheye lens based beamformer
when relative phase shifting given in Table 2 is applied. Figure 1. (a) Maxwell Fisheye lens based beamformer with transceiver input in first half of the lens, and
monopole circular antenna array in the second half of the lens, (b) Ray tracing diagram depicting source and
image locations in a metallic mirror enclosed Maxwell fisheye lens (c) Simulated absolute E-field inside the lens
structure when a signal at 10 GHz is excited at the source probe of the lens, (d) Comparison between far-field
directivity of an ideal semicircular antenna array and the proposed Maxwell fisheye lens based beamformer
when relative phase shifting given in Table 2 is applied. Table 1. Monopole antenna element position relative to the Maxwell fisheye lens centered at x = y = z = 0
mm. Element position
1
2
3
4
5
6
7
8
x (mm)
− 29.4
− 24.9
− 16.7
0
0
16.7
24.9
29.4
z (mm)
0
− 16.7
− 24.9
− 29.4
29.4
24.9
16.7
0 Table 1. Monopole antenna element position relative to the Maxwell fisheye lens centered at x = y = z =
mm. with a dedicated number of ports to excite spatially distributed antenna elements operating as an array. Specifi-
cally, the contributions of this work are as follows: First, we propose a novel beamformer design architecture
using Maxwell fisheye lens whose capability of on demand beam shaping and beam steering is demonstrated
for the first time; Second, we demonstrate that our proposed approach does not require a fixed corporate feed
network and is scalable to handle between 2 to 8 antenna array elements; Third, we validate the beam shaping and
beam steering using proof-of-concept prototype. The remainder of the paper is organized as follows. In “Maxwell
fisheye lens-based beamformer” section, we discuss the proposed beamformer design, present the operation of
the Maxwell fisheye lens and its corresponding theoretical model. Design and analysis of Maxwell
fisheye lens based beamformer
OPEN The hybrid beamforming uses a two stage structure, employing both the analog and
digital beamforming, thereby enhancing the array gain (analogue beamforming) and mitigating interference
(digital beamforming)4. All three beamforming types require multi-antenna systems that helps to improve the
signal-to-noise-ratio (SNR) when used at a receiver side. Interference can also be mitigated through generating
highly directional beams along the desired direction when used at a transmitter side. This work is motivated by
advantages offered by the analogue beamforming, either on its own or as a part a hybrid architecture, and here
we explored the utilization of Maxwell fisheye lens to propose a unique scalable analogue beamformer solution.i pi
y
p
p
q
g
Maxwell fisheye belongs to a class of gradient-index materials, where the refractive index changes with the
geometry of the material5. Other gradient-index materials including Luneberg Lens6, Eaton Lens7 and Fresnel
lens8 utilizes the energy focusing capability of the lens structure at the antenna end, while this work uses energy
focusing capability of a Maxwell Fisheye lens at a feed network level. Maxwell fisheye lens is an inhomogeneous
optical system, where rays emerging from one point within the lens follow a circular arc-shaped path about the
origin (center) of the lens9. Maxwell fisheye lens possesses a characteristic that gives impetus to its use in analogue
beamforming, i.e., for each point on the Maxwell fisheye lens, there is an image, where the image is a conjugate
of the source. The optical length between two conjugate points on the circular arc-shaped trajectory is the same. Therefore, the capability of translating source signal to image signal by the Maxwell fisheye lens motivates its use
for developing a self-scalable multi-antenna system feed network that does not require a corporate feed network 1Institute of Electronics, Communications and Information Technology (ECIT), Queen’s University Belfast, Belfast,
UK. 2Department of Computer and Information Sciences, Northumbria University, Newcastle upon Tyne NE1 8ST,
UK. *email: m.abbasi@qub.ac.uk Scientific Reports | (2021) 11:22739 | https://doi.org/10.1038/s41598-021-02058-9 www.nature.com/scientificreports/ Figure 1. Design and analysis of Maxwell
fisheye lens based beamformer
OPEN Prototype development is explained in “Pro-
totype development method” section, followed by results and discussion in “Results and discussion” section. While “Conclusions and future work” section concludes the paper and presents future directions to this work. www.nature.com/scientificreports/ Our design comprises of ideally conducting parallel plates, where the field propagating parallel to the plates has
a general form of propagation constant given by (4)
km =
k2 −m2
π
d
2 (4) where k = ω√εr/c. and m ∈1, 2, ... . The propagation will take place when km is real. For k < π√εr/d , where
km is imaginary and the propagating waves undergo an exponential decay in the direction of propagation. This
leads to development of parallel plate waveguide structure that can act as a high pass filter with a cut-off fre-
quency given by (5)
ωc =
mπc
d√ε(ω),
when m ∈{1, 2, ...} (5) where m represents the mode number. The wave will propagate at a phase velocity v = ω/km for ω > ωc and
energy will be transported at a group velocity vg = dω/dkm . The group velocity can be controlled by creating a
graded index profile through Eq. (3). The parallel plate waveguide is partially filled with a dielectric material of
permittivity εr with a varying thickness across the lens, which leads to an effective dielectric ´εr given by (6)
´εr = 1 −t(r)
1 + εr
d
. (6) In Eq. (6), r denotes the parallel plate’s radius, while t is dielectric material thickness. The effective permittivity
´εr can be controlled by the method explained in21. Equations (3), (5) and (6) are used to formulate the dielectric
profile between the parallel plates shown in Fig. 1a, where coax probes are use to excite a signal at 10 GHz within
the parallel plate region.h There is a physical limitation regarding the design of the lens which is related to the near zero index area of
the lens profile resulting due to the approach given in Eqs. (3) and (6), which is the requirement of an infinitely
large lens to realize the perfect fisheye operation. We mitigate this by including a cylindrical reflecting metal sheet
to limit the wave propagation within a finite area as postulated in22. Figure 1b demonstrates the function of the
mirror where multiple circular ray trajectories from the source are reflected by the mirror and form an image
point, while this operation occurs within a confined space. In practical terms, the signal excited at the source
point will have an extension at the image as shown in Fig. 1c. www.nature.com/scientificreports/ ation of the ray. As shown previously in14, the rays from a 3D system projects onto a plane without any loss of
generality, and this principle can be used to translate a 3D Maxwell fisheye lens on a finite plane.hi ation of the ray. As shown previously in14, the rays from a 3D system projects onto a plane without any loss of
generality, and this principle can be used to translate a 3D Maxwell fisheye lens on a finite plane.hi The Rexolite lens profile in Fig. 1a is centered at the origin of the Cartesian coordinate system, while the base
of the lens as well as the RF ground plane are at z = 0 . By mapping the ray trajectories from spherical points
on the xz-plane15; the differential of the path length on the sphere corresponds to the differential of the optical
length on the xz-plane. This allowed us to map the geodesic points on multiple circles that produces the sphere
to a Maxwell fisheye lens onto the refractive index profile in xz-plane, which can be written as16 (1)
x =
1 −sinθ
cosθ
2
cos(2φ) and z =
1 −sinθ
cosθ
2
sin(2φ) (1) In this stereographic projection, the line segments dx and dy can be represented using spherical coordinates,
hence the cylindrical lens profile can be represented as In this stereographic projection, the line segments dx and dy can be represented using spherical coordinates,
hence the cylindrical lens profile can be represented as (2)
n2(x, y, 0)[dx2 + dy2] = n2
0[dθ2 + sin2θdφ2] n2(x, y, 0)[dx2 + dy2] = n2
0[dθ2 + sin2θdφ2] (2) where, at the equator of the reference sphere index n = n0 . Using this principle, the Maxwell fisheye lens profile
can be represented by14,17: where, at the equator of the reference sphere index n = n0 . Using this principle, the Maxwell fisheye lens profile
can be represented by14,17: (3)
n = √εr =
2no
1 + r2
R2
,
when r ∈[0, ∞)] (3) where R represents the radius of the reference sphere and 2no is the refractive index at the center of the lens. The
traditional methods used to create a refractive index profile for fisheye focusing include thin plates18 and holy
parallel plates19. In this paper, we use mode theory of parallel-plate waveguide20 for realizing the fisheye principle. Maxwell fisheye lens‑based beamformerhi i
y
Beamformer design. The proposed Maxwell fisheye lens based beamforming semi-circular array design
is shown in Fig. 1a. The figure shows the transceiver coaxial input emanating from the a metallic sheet acting as
a RF ground plane. On the bottom side of this RF ground plane is a planar Maxwell fish-eye lens, where probe
of the transceiver coaxial input is exciting the lens. The RF ground plane is bounded by a copper mirror, which
is necessary to confine the microwave signal within the lens structure. The circular lens structure can be char-
acterised into two halves, where half of the circle represents the input-side and the remaining half represents
the monopole side. The dielectric profile is surrounded by the input and output side probes. At the monopole
side half of the Maxwell fisheye lens, coax probes (similar to that of the input-side) are connected to monopole
antennas, which are vertically placed at the top side of the RF ground plane. Monopole antennas connected to
the output probes of the Maxwell fisheye lens creates a semi-circular monopole antenna array, which is backed
by a metallic reflector placed at a distance of /2 . The lens structure and monopoles are designed to operate at
10 GHz. Relative locations of the monopole antenna elements are provided in Table 1. The source and drain
locations used in this work coincide with the location identification via time reversed scattering transformation
approach shown in10,11, while general form of geometrical solution of Maxwell fisheye lens can be found in12. Maxwell fisheye lens operation. To understand the operation of Maxwell fisheye lens and how it is used
to develop a beamformer, let us consider a three dimensional Cartesian coordinate system shown in Fig. 1a
when the trajectory of the rays within a 3 dimensional (3D) Maxwell fisheye lens is projected onto xz- plane13. Mathematically, ray trajectory can be represented as r(t) = x(t), y(t), z(t) , where t signifies the parametric vari- https://doi.org/10.1038/s41598-021-02058-9 Scientific Reports | (2021) 11:22739 | www.nature.com/scientificreports/ www.nature.com/scientificreports/ It shows that the wave leaving a Maxwell fisheye
lens follows a circular path and converges on an image point. This illustrates the focusing capability of the lens,
here 85% of the inserted field at the source port can be extracted from the drain probe at operation frequency of
10 GHz. In this case, the electromagnetic field energy loss is related to the imperfect imaging at the drain located
at the image point of the lens, which is discussed in11,23. Beamformer operation. When a 10 GHz signal is simultaneously excited at the transceiver coax inputs,
monopole antennas are excited, hence forming a beam along the −z-direction. Figure 1d presents the beam-
forming results validated via full-wave electromagnetic simulation tool CST Microwave Studio, and compared
them with that of a mathematical semi-circular antenna array factor. To achieve beam shaping with maximum
directivity, phase alignment at the array excitation is required, thus phase shifting depicted in Table 2 is applied
at each of the 8 input ports of the beamformer. The results demonstrate that directivity of the Maxwell fisheye
lens based antenna array and an ideal circular array with monopoles are comparable, validating the utility of https://doi.org/10.1038/s41598-021-02058-9 Scientific Reports | (2021) 11:22739 | https://doi.org/10.1038/s41598-021-02058-9 www.nature.com/scientificreports/ Table 2. Phase correction applied to each antenna element. Element
1
2
3
4
5
6
7
8
Phase ( ◦)
− 242.84
− 153
− 53.76
0
0
− 53.76
− 153
− 242.84 Table 2. Phase correction applied to each antenna element. Element
1
2
3
4
5
6
7
8
Phase ( ◦)
− 242.84
− 153
− 53.76
0
0
− 53.76
− 153
− 242.84 Figure 2. (a) Design configuration of the Maxwell fisheye lens beamformer with transmission line network
developed for prototyping and performance evaluation, (b) Prototype hardware, and (c) Simulated E-field
mapped on multiple 2D surfaces on the beamformer prototype. Figure 2. (a) Design configuration of the Maxwell fisheye lens beamformer with transmission line network
developed for prototyping and performance evaluation, (b) Prototype hardware, and (c) Simulated E-field
mapped on multiple 2D surfaces on the beamformer prototype. the fish-eye lens structure. Moreover, we also observe that the directivity patterns have significantly high back
lobes, prompting a need of metallic reflector shown in Fig. 1a, which is added to redirect the radiated energy
towards −z-direction. www.nature.com/scientificreports/ When beam steering is performed in an ideal circular array, the maximum directivity
level remains the same, as shown in the Fig. 1d. Although, there is a slight degradation ( < 1dB ) in the maxi-
mum directivity when the steering angle of 25◦ is applied to the Maxwell fisheye lens based beamformer, that
doesn’t effect the beam steering capability. Additionally, it can be observed that the half power beamwidth for
an ideal semicircular array stays consistent when beam steering is performed. Finally, for a Maxwell fisheye lens
based antenna array we can observe a side lobe at steering angle of 50◦ , however with 6.8dB difference from the
maximum directivity the overall beamfomring response is comparable with the ideal semicircular antenna array. Results and discussion From the simulations, we observed an impedance mismatch when the reflective mirror is placed close to the
probe in the Maxwell fisheye lens. This impacts the device operation and reduces its efficiency. The reason being
the classical substrate-filled parallel plate waveguide excitation impedance principle24 does not directly apply to
parallel plate waveguide with growing/decaying substrate profile, and an additional step of impedance match-
ing optimization is needed. When the probe location and depth is optimized for maximum device efficiency in
full-wave electromagnetic simulator, the source excitation forms an image at the drain position. This is done by
optimizing close-to-perfect phase alignment of the propagating waves emerging from source side, and terminat-
ing at the drain point. In particular, where the Rexolite material is not present, the outer edge of the lens acts
purely as a parallel plate waveguide, in which a portion of the signal propagates at uniform velocity.hi p
y
p
p
g
p
g
p
p g
y
The solution of wave equation is characterized by m in Eq. (5), while the TEm and TMm modes are defined on
discrete wavelengths, having specific cut-off frequencies. It is pertinent to note that the wave equation solution
with no magnetic-fields along the direction of propagation, i.e., TMm has a special case of having mode TM0 with
no cutoff frequency. Hence, the lens structure supports the TM0 mode along the direction of propagation, where
this mode doesn’t possess an electric or magnetic field, leading to a quasi-TEM mode. The mode of propaga-
tion in the lens structure close to metallic mirror is different from that of within the Rexolite filled parallel plate
waveguide (evident from Fig. 1c). g
g
Moreover, the transitions between sections of the lens and other parts of the beamformer prototype contrib-
ute differently to the signal propagation, which are: (i) the part of the lens where air transition to the Rexolite
substrate; (ii) transition between dielectric filled plates to the coax probe; (iii) transition between microstrip
transmission line and source probe, and (iv) the transition between drain probe and monopole antenna array. All of these transitions have different wave propagation characteristics. During prototype development stage,
these transitions were separately optimized for maximum power transfer and low return loss operation before
combining together in the form of Maxwell fisheye lens-based beamformer shown in Fig. Prototype development method yp
p
Figure 2a shows the dimensions of the Maxwell fisheye lens developed for the beamforming performance evalu-
ation, where the design configuration of the prototype is depicted. The schematic shows the phase aligned
transmission lines which are connecting the external signal input through SMA connectors to the inputs of the
Maxwell fisheye lens. The monopole antenna array elements are at the output of the Maxwell fisheye lens backed
by a metallic reflector when looking at the top side, whereas the Rexolite lens profile is shown when looking at the
bottom side of the beamformer in Fig. 2a. Length of the monopole antenna and gap from the metallic reflector
is also provided. The Maxwell fisheye lens has two perfect electrically conducting (PEC) plates spaced d = 5 mm
apart. Note that the bottom plate from the Maxwell fisheye lens is removed for the purposes of showing the
Rexolite lens and probes assembly. The substrate between the parallel plates is Rexolite with εr = 2.53 , disper-
sion factor=0.00066 and the coefficient of linear thermal expansion = 3.8 × 10−51/oF inch. These properties
make Rexolite a suitable candidate for lens development. A copper mirror surrounds the Rexolite lens, where
the diameter of the circular band is 100 mm. The lens is excited through input coaxial probes with a height (h)
of 4.5mm to provide impedance matching to the 50 line. Accordingly, the output probes have the same height
as the input probes, and the dimension of the input and output coaxial probes are also provided. The lens profile
is made from a cylindrical Rexolite stock using Triumph Duplex milling machine. To ensure the best Maxwell
fisheye lens performance, we made sure that the surface roughness across the lens profile is < /8 at 10 GHz,
while this was realised through polishing the Rexolite lens surface with a Tech-Gen precision finisher. The https://doi.org/10.1038/s41598-021-02058-9 Scientific Reports | (2021) 11:22739 | www.nature.com/scientificreports/ Figure 3. (a) Beam shaping achieved when multiple input transceiver ports are excited with same magnitude
and phase, (b) Beam steering when ideal phase ramp is applied to a shaped beam. Figure 3. (a) Beam shaping achieved when multiple input transceiver ports are excited with same magnitude
and phase, (b) Beam steering when ideal phase ramp is applied to a shaped beam. prototype lens structure has overall diameter of 100mm, whereas Rexolite lens has 60mm diameter. Prototype development method To conduct
experiments, we have developed phased aligned module in which all the transmission lines incur the same phase
shift. It is important to mention that these phase align transmission lines are used to connect the lens structure
to the edge of the board and are not a part of the proposed beamformer. Any length of transmission should work
the same way provided that the phase incurred by the propagating wave at 10 GHz is the same.i y
y
g
g
As in Fig. 1a, in the prototype shown in Fig. 2a, the monopole side of the Maxwell fisheye lens beamformer is
where the metal backed monopole antennas are connected to the output of the Maxwell fisheye lens beamformer. Figure 2b show the fabricated prototype used to measure the beamformer performance. Figure 2c shows that
the signals travel in the phased align transmission lines independently and the mutual coupling between signals
in parallel transmission lines is minimized deliberately. It is important to note that the overall transmission line
length is kept as small as possible to avoid any additional signal insertion loss from the edge of the substrate to
the input of the Maxwell fisheye lens. Figure 2c also presents a side-view of the transition between the microstrip
transmission line, coaxial input, Rexolite lens and coaxial output, where signal propagating through the transmis-
sion line is delivered to the input of the monopole antenna through the Maxwell fisheye lens. Results and discussion 2a,b.f g
gi
y
g
Figure 3 shows the beam shaping and the directivity of the propagating waves when different number of
ports of the prototype are excited. The 2 port excitation signifies signal input at two central ports, and so on. It
can be observed that the directivity increases as we increase the number of ports from 2 to 8, i.e., a directivity of
approximated 6 dBi is observed when 2 ports are excited. The directivity increases to approximately 11 dBi when
8 ports are excited. Moreover, if 2 ports are excited then a similar level of directivity is achieved between ±35◦ . When 4 ports are excited, a beam with higher directivity is achievable with a smooth slope between ±50◦ . For 6
port excitation, a similar sharp beam shaping is achievable, while the slope can be observed at around ±35◦ . It is
important to note that the sidelobe level for 4, 6 and 8 port excitation is higher as compared to sidelobe level at Scientific Reports | (2021) 11:22739 | https://doi.org/10.1038/s41598-021-02058-9 www.nature.com/scientificreports/ Figure 4. Beam shaping in response to the simultaneous excitation of 2, 4, 6 and 8 ports. Figure 4. Beam shaping in response to the simultaneous excitation of 2, 4, 6 and 8 ports. Figure 5. Beam steering as a response to the phase ramp. Figure 5. Beam steering as a response to the phase ramp. 2 port excitation. However, the difference between main lobe and 1st side lobe is around 8 dB. Figure 3a repre-
sents the beam steering of Maxwell fisheye lens beamformer,when an ideal phase ramp is applied to all the ports. Although the beam shapes are non-ideal, these are useful for this feasibility study about the behavior of beams
steering using this disruptive concept as shown in the Fig. 3b. A phase ramp of 15◦ , 30◦ , 45◦ and 60◦ is applied
separately to form the beam 2, 3, 4, and 5, respectively. It can be observed that when beam 1 is steered towards
beam 2 direction, the maximum directivity is slightly reduced from 10.2 dB to 10.1 dB, while the maximum
directivity becomes 10.7 dB for beam 3. Furthermore, we observe that beam shifting impacts the beam shape. However, this difference is not significant for this study and the overall device operation is not compromised. Results and discussion f
gi
y
p
p
Figure 4 depicts the three dimensional directivity view of the beam shaping in response to the excitation of
2,4,6 and 8 ports. The beam becomes narrower as the number of ports excited increases from 2 to 6. While the
wide equalized directivity beam can be used to realize broadcast mode, the sharp beam with 6-port excitation
can be used for applications demanding high directivity. When all 8 ports of the beamformer were excited, we
did not observe any additional benefit for boresight transmission as the beamshaping function is similar to
that of the 6 port excitation. Hence it is better to use 6 port excitation to achieve similar operation, while the
remaining two ports are interchangeably used for beam steering. The patterns in Figs. 4 and 5 shows contours
drawn after every 2dB power difference. The half power beamwidth (HPBW) for the 4 port excitation is narrower
when compared to 6 ports excitation, while the maximum area of equal power is wider in 2 port excitation as
previously shown in Fig. 3a. In a situation where a narrow beam is required, we can use the 4 port excitation
while for a wider beam 2 port excitation can be utilized. The 6 port excitation provides a beam shaping that is
intermediary between 2 and 4 port excitation. In all the cases, the sidelobe level is below 8 dB from the main
lobe. One important point to note here is that 2, 4 and 6 port excitation can be used for beam steering without
using an external phase shifter, whilst the latter is required for beam steering using 8 port excitation and the
results are depicted in Fig. 5 which signifies the beam steering in response to the phase ramp applied in a similar
fashion as described in Fig. 3a. Beam 1 is generated through an 8 port excitation. The rest of the beams 2,3 and 4
are generated as a result of applying the phase ramp also with 8 ports. It can be observed that the beam steering
is achieved but the mean maximum radiation area is disturbed. However, the performance of beam steering is
reliable even at higher steering angles, when observed at the 3 dB contour. For example, if we observe beam 5
with higher beam steering, the mean maximum radiation area is still in the desired angular direction i.e. 50◦ . Results and discussion This aspect reveals the benefits and flexibility of utilizing the circular array structure.h hil
y
g
y
The prototype shown in Fig. 2b is used to evaluate beam shaping and beam steering capabilities of the
Maxwell fisheye lens beamformer. The prototype was placed in a 10 m × 5 m × 5 m far-field anechoic chamber
facility at Queen’s University Belfast, where each port was separately excited by a 10 GHz signal. Co-polarized
component in far-field was recorded along the xz-plane (same as in Fig. 2a), for all the 8 input SMA ports of
the prototype. The far-field measurement data for each port excitation was combined in a post processing step
and the normalised results are compared with that of full-wave electromagnetic simulations, shown in Fig. 6. The beam shaping depicted in Fig. 6a,c is close to the simulated predictions. Beam steering depicted in Fig. 6b
shows a slight difference in simulated and measured 3 dB beamwidth, while the beam shape and side lobe level is
maintained as predicted. The beam shaping for two port excitation resulting in equalized radiations fields along Scientific Reports | (2021) 11:22739 | https://doi.org/10.1038/s41598-021-02058-9 www.nature.com/scientificreports/ Figure 6. Comparison between simulation and measurement normalized far-field radiation patterns along
yz-plane. Figure 6. Comparison between simulation and measurement normalized far-field radiation patterns along
yz-plane. Figure 6. Comparison between simulation and measurement normalized far-field radiation patterns along
yz-plane. azimuth direction is verified with the help of measurements results shown in Fig. 6c. The measurement res
clearly validate the performance of the Maxwell fisheye lens based proposed beamformer hardware. Conclusions and future work
h
k
l
l b In this work, we present a novel scalable analogue beamforming hardware architecture that helps in realizing
adaptive beam steering and beam shaping to fulfil the coverage requirements. The proposed beamforming hard-
ware comprises of a finite size Maxwell fisheye lens used as a feed network for a semi-circular array of monopole
antennas. The hardware architecture enables on-demand beamforming with a flexibility of using 2 to 8 antenna
elements. The unique aspect of the proposed approach is that it does not require dedication of the number of
ports in a feed network and is scalable to handle between 2 to 8 antenna array elements. The simulation and
measured results are presented to validate the performance of the beamformer in terms of beam shaping and
beam steering. As a future direction to this work, we plan to investigate means to achieve better impedance
matching between transitions. Moreover, the proposed design can be expanded to act as a high performance
wireless repeater due to the simplicity of the feed network. Received: 26 August 2021; Accepted: 9 November 2021 Received: 26 August 2021; Accepted: 9 November 2021 References Miao, Y. et al. Fair and dynamic data sharing framework in cloud-assisted internet of everything. IEEE Internet of Things J. 6,
201–7212. https://doi.org/10.1109/JIOT.2019.2915123 (2019). 1. Miao, Y. et al. Fair and dynamic data sharing framework in cloud-assisted internet of everything. IEEE Internet of Things
7201–7212. https://doi.org/10.1109/JIOT.2019.2915123 (2019). p
g
2. Zhang, J. et al. Prospective multiple antenna technologies for beyond 5g. IEEE J. Sel. Areas Commun. 38, 1637–1660. https://doi.
org/10.1109/JSAC.2020.3000826 (2020). p
g
2. Zhang, J. et al. Prospective multiple antenna technologies for beyond 5g. IEEE J. Sel. Areas Commun. 38, 1637–1660. https://doi.
org/10.1109/JSAC.2020.3000826 (2020). g
3. Ho, K. C. & Tsai, S. H. A novel multiuser beamforming system with reduced complexity and beam optimizations. IEEE Trans
Wirel. Commun. 18, 4544–4557. https://doi.org/10.1109/TWC.2019.2926718 (2019). p
g
4. Busari, S. A. et al. Generalized hybrid beamforming for vehicular connectivity using thz massive mimo. IEEE Trans. Veh. Technol
68, 8372–8383. https://doi.org/10.1109/TVT.2019.2921563 (2019).h 68, 8372–8383. https://doi.org/10.1109/TVT.2019.2921563 (2019) ,
p
g
(
)
5. Ma, H. F. et al. Three-dimensional gradient-index materials and their applications in microwave lens antennas. IEEE Trans. Anten-
nas Propag. 61, 2561–2569. https://doi.org/10.1109/TAP.2012.2237534 (2013). p
g
(
)
5. Ma, H. F. et al. Three-dimensional gradient-index materials and their applications in microwave lens antennas. IEEE Trans. Anten-
nas Propag. 61, 2561–2569. https://doi.org/10.1109/TAP.2012.2237534 (2013). p g
p
g
(
)
6. Zhang, Z., Yang, S., Chen, Y., Qu, S. & Hu, J. Fast analysis of parallel-plate cylindrical luneberg lens antennas through dyadic greens
functions. IEEE Trans. Microw. Theory Techn. 66, 4327–4337. https://doi.org/10.1109/TMTT.2018.2856239 (2018).h h
y
,
p
g
(
)
7. Li, J. Y. & Ng Mou Kehn, M. The 90◦ rotating eaton lens synthesized by metasurfaces. IEEE Antennas Wirel. Propag. Lett. 17,
1247–1251. https://doi.org/10.1109/LAWP.2018.2841187 (2018).fi h
y
p
g
(
)
7. Li, J. Y. & Ng Mou Kehn, M. The 90◦ rotating eaton lens synthesized by metasurfaces. IEEE Antennas Wirel. Propag. Lett. 17,
1247–1251. https://doi.org/10.1109/LAWP.2018.2841187 (2018).fi p
g
(
)
8. Jouade, A., Himdi, M. & Lafond, O. Fresnel lens at millimeter-wave: Enhancement of efficiency and radiation frequency bandwidth. IEEE Trans. Antennas Propag. 65, 5776–5786. https://doi.org/10.1109/TAP.2017.2755120 (2017).i p
g
(
)
8. Jouade, A., Himdi, M. & Lafond, O. Fresnel lens at millimeter-wave: Enhancement of efficiency and radiation frequency bandwidth. IEEE Trans. Antennas Propag. 65, 5776–5786. https://doi.org/10.1109/TAP.2017.2755120 (2017).i p g
g
9. Frank, A., Leyvraz, F. & Wolf, K. B. Hidden symmetry and potential group of the maxwell fish-eye. J. References Math. Phys. 31, 2757–2768
(1990). 10. Ma, Y., Wu, T. & Ong, C. Subwavelength drains designed for mirror-closed dielectric lenses. J. Opt. 15, 125705 (2013). 10. Ma, Y., Wu, T. & Ong, C. Subwavelength drains designed for mirror-closed dielectric lenses. J. Opt. 15, 125705 (2013). 11. Kinsler, P. & Favaro, A. Comment on reply to comment on perfect imaging without negative refraction. New J. Phys. 13, 02
(2011).hi 12. Quevedo-Teruel, O., Mitchell-Thomas, R. & Hao, Y. Frequency dependence and passive drains in fish-eye lenses. Phys. Rev. A 86,
053817 (2012).i (
)
13. Eskandari, H., Majedi, M. S., Attari, A. R. & Quevedo-Teruel, O. Elliptical generalized maxwell fish-eye lens using conformal
mapping. New J. Phys. 21, 063010 (2019).i pp
g
y
14. Abbasi, M. A. B. & Fusco, V. F. Maxwell fisheye lens based retrodirective array. Sci. Rep. 9, 1–8 (2019).h i
15. Luneburg, R. K. Mathematical Theory of Optics (University of California press, 1964). gh
y f
p
y
p
16. Pozar, D. M. Microwave Engineering (Wiley, 2011).i h
16. Pozar, D. M. Microwave Engineering (Wiley, 2011i g
g
y
17. Tai, C. T. Maxwell fish-eye treated by maxwell equations. Nature 182, 1600–1601 (1958).i i
18. Lefebvre, G. et al. Experiments on maxwells fish-eye dynamics in elastic plates. Appl. Phys. Lett. 106, 024101 (2015). l
T
á
h
h
d
l b
db
d
l
d
f
fl
l
h i
18. Lefebvre, G. et al. Experiments on maxwells fish-eye dynamics in elastic plates. Appl. Phys. L 19. Climente, A., Torrent, D. & Sánchez-Dehesa, J. Omnidirectional broadband insulating device for flexural waves in thin plates. J. Appl. Phys. 114, 214903 (2013). pp
y
20. Leal-Sevillano, C. A., Ruiz-Cruz, J. A., Montejo-Garai, J. R. & Rebollar, J. M. Rigorous analysis of the parallel plate waveguide:
From the transverse electromagnetic mode to the surface plasmon polariton. Radio Sci. 47, 1–8 (2012). g
p
p
1. Sato, K. & Ujiie, H. A plate luneberg lens with the permittivity distribution controlled by hole density. Electron. Commun. Jpn
(Part I: Commun.) 85, 1–12 (2002). https://doi.org/10.1038/s41598-021-02058-9 Scientific Reports | (2021) 11:22739 | https://doi.org/10.1038/s41598-021-02058-9 www.nature.com/scientificreports/ iii
23. Blaikie, R. Comment on perfect imaging without negative refraction. New J. Phys. 12, 058001 (2010). 24. Marcuvitz, N. Waveguide Handbook Vol. 21 (Iet, 1951). Acknowledgements g
Authors would like to thank W. McKeown and C. Colligan for their assistance in the lens machining. This work
was partially supported by the Engineering and Physical Sciences Research Council (EPSRC) under Grant EP/
P000673/1 and Grant EP/S007954/1. Competing interests h p
g
The authors declare no competing interests. Author contributions M.A.B.A. and V.F. conceived the idea. M.A.B.A. designed and simulated the proposed beamformer and per-
formed the measurements. R.I.A., G.G.M. and M.A.B.A. interpreted the results. R.I.A., M.A.B.B. and G.G.M. wrote the paper. V.F. supervised the research. All authors reviewed the manuscript. © The Author(s) 2021 Additional information Correspondence and requests for materials should be addressed to M.A.B.A. Correspondence and requests for materials should be addressed to M.A.B.A. Reprints and permissions information is available at www.nature.com/reprints. Reprints and permissions information is available at www.nature.com/reprints. Publisher’s note Springer Nature remains neutral with regard to jurisdictional claims in published maps and
institutional affiliations. Publisher’s note Springer Nature remains neutral with regard to jurisdictional claims in published maps and
institutional affiliations. Open Access This article is licensed under a Creative Commons Attribution 4.0 International
License, which permits use, sharing, adaptation, distribution and reproduction in any medium or
format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the
Creative Commons licence, and indicate if changes were made. The images or other third party material in this
article are included in the article’s Creative Commons licence, unless indicated otherwise in a credit line to the
material. If material is not included in the article’s Creative Commons licence and your intended use is not
permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from
the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by/4.0/. © The Author(s) 2021 https://doi.org/10.1038/s41598-021-02058-9 Scientific Reports | (2021) 11:22739 |
|
https://openalex.org/W2145133052
|
https://europepmc.org/articles/pmc2825248?pdf=render
|
English
| null |
Prosaposin down-modulation decreases metastatic prostate cancer cell adhesion, migration, and invasion
|
Molecular cancer
| 2,010
|
cc-by
| 16,019
|
© 2010 Hu et al; licensee BioMed Central Ltd. This is an Open Access article distributed under the terms of the Creative Commons
Attribution License (http://creativecommons.org/licenses/by/2.0), which permits unrestricted use, distribution, and reproduction in
any medium, provided the original work is properly cited. Prosaposin down-modulation decreases
metastatic prostate cancer cell adhesion,
migration, and invasion Siyi Hu1, Nathalie Delorme1, Zhenzhen Liu1, Tao Liu1, Cruz Velasco-Gonzalez2, Jone Garai1, Ashok Pullikuth3,
Shahriar Koochekpour1,4,5,6* Abstract Background: Factors responsible for invasive and metastatic progression of prostate cancer (PCa) remain largely
unknown. Previously, we reported cloning of prosaposin (PSAP) and its genomic amplification and/or
overexpression in several androgen-independent metastatic PCa cell lines and lymph node metastases. PSAP is the
lysosomal precursor of saposins, which serve as activators for lysosomal hydrolases involved in the degradation of
ceramide (Cer) and other sphingolipids. Results: Our current data show that, in metastatic PCa cells, stable down-modulation of PSAP by RNA-interference
via a lysosomal proteolysis-dependent pathway decreased b1A-integrin expression, its cell-surface clustering, and
adhesion to basement membrane proteins; led to disassembly of focal adhesion complex; and decreased
phosphorylative activity of focal adhesion kinase and its downstream adaptor molecule, paxillin. Cathepsin D
(CathD) expression and proteolytic activity, migration, and invasion were also significantly decreased in PSAP
knock-down cells. Transient-transfection studies with b1A integrin- or CathD-siRNA oligos confirmed the cause and
effect relationship between PSAP and CathD or PSAP and Cer-b1A integrin, regulating PCa cell migration and
invasion. Conclusion: Our findings suggest that by a coordinated regulation of Cer levels, CathD and b1A-integrin
expression, and attenuation of “inside-out” integrin-signaling pathway, PSAP is involved in PCa invasion and
therefore might be used as a molecular target for PCa therapy. Hu et al. Molecular Cancer 2010, 9:30
http://www.molecular-cancer.com/content/9/1/30 Hu et al. Molecular Cancer 2010, 9:30
http://www.molecular-cancer.com/content/9/1/30 Hu et al. Molecular Cancer 2010, 9:30 http://www.molecular-cancer.com/content/9/1/30 * Correspondence: skooch@lsuhsc.edu
1Stanley S. Scott Cancer Center, School of Medicine, Louisiana State
University Health Sciences Center, New Orleans, LA 70112, USA Background fluid [2]. Secreted PSAP is a well-known potent neuro-
trophic factor [7,8]. Total PSAP deficiency is lethal in
both man and mice [2]. However, deficiency of indivi-
dual saposin proteins is responsible for a number of
lipid storage diseases [9-11]. Prosaposin (PSAP) is a dual-function highly conserved
glycoprotein that exists as the lysosomal precursor of
four small sphingolipid activator proteins, known as
saposins A, B, C and D [1-3]. Saposins are generated by
proteolytic cleavage of another lysosomal protease,
cathepsin D (CathD) [4-6]. In lysosomes, mature sapo-
sins are intensively involved in metabolism of sphingoli-
pids and ceramide (Cer), functioning either as essential
co-factors for sphingolipid hydrolases and/or destabiliz-
ing the complex of lipids and membranes [3]. PSAP also
exists as a secreted protein, which has been found in
various body fluids such as milk, serum, and seminal Homozygous inactivation of PSAP gene in mice led to
shrinkage and atrophic changes in the male reproductive
organs, with gross pathological features including a
reduction in size and weight of the testes, seminal vesi-
cle, and prostate gland [12]. Histological examination of
the involuted prostate tissue revealed the presence of
undifferentiated epithelial cells. Collectively, these data
support a developmental role for PSAP in prostate
gland. During our search for a prostate tumor marker,
we cloned PSAP as a secreted protein from the highly
invasive and metastatic PCa cell line PC-3 [13]. In addi-
tion, we discovered its overexpression and/or genomic * Correspondence: skooch@lsuhsc.edu
1Stanley S. Scott Cancer Center, School of Medicine, Louisiana State
University Health Sciences Center, New Orleans, LA 70112, USA Hu et al. Molecular Cancer 2010, 9:30
http://www.molecular-cancer.com/content/9/1/30 Page 2 of 18 proliferation assay, we did not find a significant reduc-
tion (≤5%) in the PSAP-KD cells growth (data not
shown). In addition, we also tested the effect of PSAP
down-modulation in soft agar colony formation assay. This study also did not show any significant changes in
the size or number of colonies in PSAP-KD clones com-
pared to control transfectants. To evaluate the biological
activities of PSAP, we generated a mammalian expres-
sion vector expressing the biologically active full-length
rhPSAP (see additional files 2, 3, 4, 5). Treatment of
both the control and PSAP-KD clones with rhPSAP at
0.1 to 10 nM did not stimulate their growth rate (data
not shown). Background Overall, these results suggest that altera-
tions in the intracellular or extracellular PSAP levels do
not affect the anchorage-dependent and -independent
growth of PC-3 and DU-145 cells. amplification in several androgen-independent (AI) and/
or metastatic PCa cell lines and in punch biopsy sam-
ples of LuCaP PCa xenograft and lymph node metas-
tases. Interestingly, PSAP expression in C4-2B, an AI-
bone metastatic PCa cell line was significantly higher
than in its parental isogenic and marginally tumorigenic
cell line, LNCaP [13]. Recently, we demonstrated that
saposin C and TX14A-synthetic peptide, two well-
known bioactive derivatives of PSAP, act as cell survival
and anti-apoptotic factors, stimulate migration and inva-
sion, and activate PI3K/Akt- and MAPK-signaling path-
ways in PCa cell lines [14-16]. However, the underlying
mechanisms of PSAP regulation of PCa cell migration
and invasion have not been investigated. In this study, we evaluated the contribution of PSAP
in multistep process of invasion by using an RNA-inter-
ference strategy and transient or stable transfectants of
metastatic PCa cell lines. Down-modulation of PSAP
expression did not alter PCa cell growth. However, by
increasing cellular Cer levels and decreasing b1A-integrin
and CathD expression, PSAP significantly decreased the
cell adhesion, migration, and invasion abilities of AI-
PCa cells. Taken together, our data support a role for
PSAP in invasive and metastatic progression of PCa. g
PSAP down-modulation decreases PCa cells adhesion,
migration, and invasion During routine cell culture and trypsinization, we
noticed that in both cell lines, the PSAP-KD clones
were detached more easily than their control clones or
parental cell types. Therefore, we investigated cell adhe-
sion to the major components of the basement mem-
brane such as laminin (LN) and fibronectin (FN). We
found that PSAP-KD clones showed a significant reduc-
tion of cell adhesion on FN- or LN-coated plates as
compared with the control clones (Fig. 1C). In the PC-3
cell line, compared to control transfectants, the adhesion
of the PSAP-KD clones on FN and LN was reduced by
78% and 71%, respectively. Likewise, the adhesion of the
PSAP-KD clones in the DU-145 cells was decreased by
49% on FN and 69% on LN. We obtained a comparable
decrease in cell adhesion for the other extracellular
matrix (ECM) proteins such as collagen I or IV in the
PSAP-KD clones (data not shown). It is noteworthy that
the decreased ability of cell adhesion to ECM proteins
was associated with clear morphological changes in
PSAP-KD clones as compared with their control coun-
terparts. Control transfectants demonstrated morpholo-
gical indications of adhesion phenotype such as
spreading, membrane protrusion and ruffles, and polar-
ity on all ECM proteins examined. In contrast, PSAP-
KD cells appear lower in number and condensed with
smaller and either delayed or multi-polar membrane
protrusion (Fig. 1D). PSAP is overexpressed in metastatic PCa cells PSAP is overexpressed in metastatic PCa cells
As shown in Fig. 1A, PSAP and saposin C are expressed
at higher levels in metastatic PCa cell lines than in the
normal prostate epithelial cells (Pr.Ep). In addition,
using other PCa progression models of isogenic cell
lines, we observed consistent data for increased PSAP
expression level from normal, poorly tumorigenic, or
non-tumorigenic cells to androgen-independent and/or
highly invasive and metastatic cell lines such as LNCaP/
C4-2B, PC-3/PC-3M, and p69-M12-M2182 (see addi-
tional file 1). The biological significances of PSAP as an
intracellular or extracellular soluble protein in PCa cells
are largely unknown. Our attempts to increase the
expression of PSAP in PC-3 and DU-145 cells beyond
their endogenous level failed. Therefore, we decided to
use RNA interference strategy to specifically down-mod-
ulate PSAP expression. After establishing several control
or PSAP-KD clones, we randomly selected two clones
for each category for further analysis. As shown in Fig. 1B, stable transfection of the two metastatic PCa cell
lines with a PSAP-shRNA vector decreased the PSAP
mRNA level. In addition, both the intracellular and
extracellular PSAP and saposin C protein levels in the
two PSAP-KD clones (P5 and P16 in PC-3 and P15 and
P32 in DU-145 cell line) were significantly reduced by
≥70% as compared with two control clones (C1 and C3
in PC-3 and C9 and C13 in DU-145 cell line). Using
direct cell counting with haemocytometer and MTS cell Defective adhesion might reflect itself in migration
and invasion as the two important malignancy-asso-
ciated phenotypes. Our previous studies revealed that
active molecular derivatives of PSAP (i.e. saposin C or
TX14A peptide) stimulate PCa cell motility and invasion
[14,16]. Next, we examined the effect of PSAP down-
modulation on these phenotypes by using the conven-
tional Boyden Chamber (transwell filter) assays. We
found that the PSAP-KD clones showed a significant Hu et al. Molecular Cancer 2010, 9:30
http://www.molecular-cancer.com/content/9/1/30 Page 3 of 18 Figure 1 PSAP gene silencing decreases metastatic PCa cell adhesion to basement membrane proteins. (A) PSAP over-expression in
metastatic prostate cancer cell lines. Equal amount of cell lysates or supernatants from PC-3, DU-145, and normal prostate epithelial (Pr.Ep) cells
were subjected to western blotting with anti-PSAP and saposin C antibodies. (B) Stable PSAP down-modulation was accomplished by short-
hairpin RNA targeted at PSAP gene. PSAP is overexpressed in metastatic PCa cells PC-3 and DU-145 metastatic PCa cell lines were stably transfected with a G418-resistant vector containing a
shRNA sequence specific for human PSAP or a scrambled control sequence. Total RNA was extracted for RT-PCR (top). GAPDH transcript was
used as an internal control for RNA loading. Cell lysates and culture supernatants were subjected to immunoblotting with PSAP antibody. GAPDH antibody was used for protein loading. (C and D) PC-3 and DU-145 cell adhesion to ECM proteins was examined by seeding 1.5 × 104
cells per well in 96-well plates pre-coated with 10 μg/ml fibronectin (FN) or laminin (LN). After 2 h incubation, adhered cells were fixed and
stained with toluidine blue. Cells were photographed and counted from ten random fields at 100 × magnification. Columns, mean of three
independent samples run together; bars, ± SEM, p < 0.0001, ANOVA was used to compare PSAP-KD and control clones. Each experiment was
repeated three times independently. C1 and C3 in PC-3 and C9 and C13 in DU-145 were control clones (shRNA-scrambled vector) and P5 and
P16 in PC-3 and P15 and P32 in DU-145 were PSAP-KD clones (shRNA-PSAP). Figure 1 PSAP gene silencing decreases metastatic PCa cell adhesion to basement membrane proteins. (A) PSAP over-expression in
metastatic prostate cancer cell lines. Equal amount of cell lysates or supernatants from PC-3, DU-145, and normal prostate epithelial (Pr.Ep) cells
were subjected to western blotting with anti-PSAP and saposin C antibodies. (B) Stable PSAP down-modulation was accomplished by short-
hairpin RNA targeted at PSAP gene. PC-3 and DU-145 metastatic PCa cell lines were stably transfected with a G418-resistant vector containing a
shRNA sequence specific for human PSAP or a scrambled control sequence. Total RNA was extracted for RT-PCR (top). GAPDH transcript was
used as an internal control for RNA loading. Cell lysates and culture supernatants were subjected to immunoblotting with PSAP antibody. GAPDH antibody was used for protein loading. (C and D) PC-3 and DU-145 cell adhesion to ECM proteins was examined by seeding 1.5 × 104
cells per well in 96-well plates pre-coated with 10 μg/ml fibronectin (FN) or laminin (LN). After 2 h incubation, adhered cells were fixed and
stained with toluidine blue. Cells were photographed and counted from ten random fields at 100 × magnification. Columns, mean of three
independent samples run together; bars, ± SEM, p < 0.0001, ANOVA was used to compare PSAP-KD and control clones. PSAP is overexpressed in metastatic PCa cells Each experiment was
repeated three times independently. C1 and C3 in PC-3 and C9 and C13 in DU-145 were control clones (shRNA-scrambled vector) and P5 and
P16 in PC-3 and P15 and P32 in DU-145 were PSAP-KD clones (shRNA-PSAP). Hu et al. Molecular Cancer 2010, 9:30
http://www.molecular-cancer.com/content/9/1/30 Hu et al. Molecular Cancer 2010, 9:30
http://www.molecular-cancer.com/content/9/1/30 Page 4 of 18 Hu et al. Molecular Cancer 2010, 9:30
http://www.molecular-cancer.com/content/9/1/30 To examine whether changes in protein stability could
be responsible for the reduced b1A expression in PSAP-
KD clones, we investigated the half-life of the b1A pro-
tein by treating a representative clone from both control
and PSAP-KD cells with protein synthesis inhibitor,
cycloheximide (CHX). In agreement with previously
reported data [20], we found that the b1A protein half-
life was approximately 20 h in the control clones, while
it decreased to ~14 h in the PSAP-KD clones in both
cell lines (Fig. 3D &3E). The differences between PSAP-
KD and control clones could be due to the enhanced
degradation rate of the b1A protein in PSAP-KD which
allows its earlier disappearance while synthesis of new
proteins are inhibited by CHX. decrease of migration by 70% in PC-3 and 79% in DU-
145 compared to the control clones (Fig. 2A &2B). In
addition, PSAP down-modulation further reduced the
ability of cell invasion through the Matrigel-coated
membrane by 78% in PC-3 cells and by 85% in DU-145
cells. We also found that treatment of both control and
PSAP-KD cells with rhPSAP in a dose-dependent man-
ner increased their migratory and invasive behavior (Fig. 2C). However, the overall ability of PSAP-KD cells to
migrate and invade through Matrigel were significantly
less than the control cells indicating a major role for
intracellular PSAP expression in the regulation of cell
migration and invasion. PSAP down-modulation reduces b1A-integrin expression
Reduction of cell-substrate adhesion in PSAP-KD cells
could be the result of changes in the expression and/or
usage of adhesion receptors such as the intregrin super-
family which exist as a- and b-subunits. As heterodi-
mers, these subunits could recognize different ECM
proteins. Using RT-PCR and immunoblotting, we
screened control and PSAP-KD clones of PC-3 and DU-
145 cells for a/b-subunit expression. PSAP is overexpressed in metastatic PCa cells Consistent with
previous reports, using specific primers and antibodies
against integrin-subunits, we were able to detect moder-
ate to high level of expression for a1, a2, a3, a5, a6, aV,
b3, and b4 integrin subunits [17,18]; no differences
between PSAP-KD and control clones were noted. The
b1-integrin is the most abundant subunit expressed in
PCa cells and tissues; it is capable of forming heterodi-
mers that can bind to FN, LN, and collagen IV [19]. Previous studies showed that PCa cells expressed three
different b1 isoforms: b1A, b1B and b1C, with b1A as the
most abundant isoform [18]. We found that in PSAP-
KD clones only the b1A isoform expression at the pro-
tein level was reduced while b1B or b1C did not change. Compared to the control clones, the expression level of
both the pre-mature-b1A (105 kDa) and the mature b1A-
isoform (125 kDa) were significantly decreased in PSAP-
KD clones (Fig. 3A). As it was expected, the changes in
the b1A expression pattern were very similar to the total
b1-integrin. Furthermore, to confirm the role of the b1A-
integrin expression in PCa cell adhesion on ECM pro-
teins, we repeated the adhesion assays and used control
clones that were transiently-transfected with a specific
human integrin b1-siRNA oligos. The protein levels of
total integrin b1 as well as b1A isoform were reduced by
80-90% in control clones in both cell lines (Fig. 3B). We
found that, down-modulation of the b1-integrin expres-
sion decreased cell adhesion by 83% for FN and 66% for
LN in PC-3 and by 52% for FN and 69% for LN in DU-
145 (Fig. 3C). These results suggested that reduced
expression of b1A-integrin expression contributed to the
decreased ability of PSAP-KD clones to adhere to base-
ment membrane proteins. p
y
To understand the posttranslational mechanisms
responsible for the reduced b1A half-life in PSAP-KD
cells, we investigated the involvement of the lysosomal-,
the calpain- and the ubiquitin-mediated proteolysis
pathways. PSAP-KD and control clones were incubated
for different time periods (6 to 24 h) with a non-toxic
dosage of leupeptin or NH4Cl (lysosomal protease inhi-
bitors), ALLN (calpain inhibitor), or MG132 (protea-
some inhibitor). Treatment of both the control and
PSAP-KD clones, with leupeptin or NH4Cl, increased
b1A expression in a time-dependent manner beginning
as early as 6 hours (Fig. 3F). PSAP is overexpressed in metastatic PCa cells The increase in the b1A-
integrin expression was more evident in PSAP-KD
clones than in the control clones. However, the b1A pro-
tein expression level was not affected by inhibitors of
proteasome or calpain (data not shown). These data
show that down-modulation of PSAP via a lysosomal
proteolysis-dependent pathway increases b1A-integrin
degradation rate. Under our experimental conditions,
cell viability at the end of the treatment period with
CHX or other pharmacological agents was ≥95%, as
exhibited by a trypan blue dye exclusion assay. PSAP down-modulation prevents focal adhesion kinase
activation and focal adhesion complex formation PSAP down-modulation prevents focal adhesion kinase
activation and focal adhesion complex formation PSAP-KD cells appeared small and condensed and did
not show morphological evidence of adhesion phenotype
such as spreading, directional membrane protrusion, and
ruffles. These data promoted us to investigate the activ-
ity, expression, or subcellular localization of certain
structural molecules (actin and vinculin), focal adhesion
kinase (FAK) as the most important integrin-regulated
signaling molecule, and adaptor protein (i.e., paxillin)
which are collectively involved in the assembly of focal
adhesion complex. Using whole cell lysates prepared
from subconfluent cells and following their adhesion to
FN or LN, we examined the phosphorylation of FAK at
different tyrosine residues and paxillin by immunopreci-
pitation (IP) of FAK and western blotting with phospho-
specific antibodies. As shown in Fig. 4A, total FAK and
paxillin protein levels were not affected by PSAP down- Hu et al. Molecular Cancer 2010, 9:30
http://www.molecular-cancer.com/conte Page 5 of 18 Hu et al. Molecular Cancer 2010, 9:30 Hu et al. Molecular Cancer 2010, 9:30
http://www.molecular-cancer.com/content/9/1/30 Figure 2 The PSAP expression correlated with migratory and invasive potential of prostate cancer cell lines. (A) Transwell filter migration
and invasion assays. Stable PSAP-KD clones of PC-3 and DU-145 were seeded on transwell filters and incubated for 24 h. Basal medium
containing 5% FBS was used as chemo-attractant. For the invasion assay, the membrane was pre-coated with 50 μg Matrigel. (B) Cells migrated
or invaded were counted from ten random fields at 100× magnification using a phase-contrast microscope. (C) Effects of rhPSAP on cell
migration and invasion. A representative control and PSAP-KD clone from each cell line was seeded on transwell filters and incubated for 24 h
(migration) or 48 h (invasion). As a chemo-attractant, basal medium containing 0.5% FBS and various amount of rhPSAP (0, 1, 10, 50 nM) were
included in the lower compartment of the transwell filters. Each bar represented the mean ± SEM of three independent experiments, each in
quadruplicates. ANOVA was used to examine the significance of the data (p < 0.0001) comparing the PSAP-KD clones relative to control clones
in each cell line or among different treatment concentrations for rhPSAP and control. http://www.molecular-cancer.com/content/9/1/30 http://www.molecular-cancer.com/content/9/1/30 Figure 2 The PSAP expression correlated with migratory and invasive potential of prostate cancer cell lines. (A) Transwell filter migration
and invasion assays. Stable PSAP-KD clones of PC-3 and DU-145 were seeded on transwell filters and incubated for 24 h. PSAP down-modulation prevents focal adhesion kinase
activation and focal adhesion complex formation Cell
lysates were analyzed for b1A protein expression by immunoblotting. The b1A-integrin degradation curve was calculated as described above. Columns, mean of three independent samples run together; bars, ± SEM, p < 0.0001, Two-sample t-tests with Satterthwaite corrections were
Hu et al. Molecular Cancer 2010, 9:30
http://www.molecular-cancer.com/content/9/1/30
Page 6 of 18 http://www.molecular-cancer.com/content/9/1/30 re 3 Effect of PSAP down-modulation on b1A-integrin expression and PCa cell adhesion to FN and LN. (A) PSAP d
b
T
l RNA
d f
RT PCR
h
f
f
b
b
d b
d GA Figure 3 Effect of PSAP down-modulation on b1A-integrin expression and PCa cell adhesion to FN and LN. (A) PSAP down-modulation
reduces b1-integrin expression. Total RNA was extracted for RT-PCR with primers specific for integrin-b1A, -b1B, and -b1C, and GAPDH. Cell lysates
were analyzed by western blotting with antibodies against b1-integrin and its isoforms b1A, b1B, and b1C and GAPDH. (B) Transient down
modulation of b1-integrin expression. Previously established control stable clones of PC-3 and DU-145 cells were transiently transfected with b1
integrin- or scrambled-siRNA oligos. After 48 h, cell lysates were analyzed for integrin b1 and b1A expression by western blotting. (C) Inhibition of
PCa cell adhesion by transient transfection with integrin b1-siRNA. After siRNA transfection, cells were subjected to adhesion assay on FN- or LN-
coated 96-well plates as described in “Materials and Methods”. (D) b1A-integrin stability. The steady-state b1A protein levels were evaluated by
treating cells with cycloheximide (12.5 μg/ml) for up to 72 h and immunoblotting with anti-b1A antibody. (E) The half-live of b1A-integrin was
evaluated by densitometric analysis of immunoblotting bands using the Quantity One software and the b1A protein levels were calculated as
percentage of non-treatment values after normalization using GAPDH for loading control. (F) Effect of inhibition of lysosomal proteolysis on b1A-
integrin expression. Cells were incubated in the presence or absence of the lysosomal proteolysis inhibitor, NH4Cl (50 mM) for up to 24 h. Cell
lysates were analyzed for b1A protein expression by immunoblotting. The b1A-integrin degradation curve was calculated as described above. Columns, mean of three independent samples run together; bars, ± SEM, p < 0.0001, Two-sample t-tests with Satterthwaite corrections were
used to compare b1-siRNA versus scrambled siRNA oligos transfected cells following adhesion to FN or LN. ANOVA was used to examine the
significance of the data (p < 0.05) among different cycloheximide treatment periods in PSAP-KD versus control clones. PSAP down-modulation prevents focal adhesion kinase
activation and focal adhesion complex formation Basal medium
containing 5% FBS was used as chemo-attractant. For the invasion assay, the membrane was pre-coated with 50 μg Matrigel. (B) Cells migrated
or invaded were counted from ten random fields at 100× magnification using a phase-contrast microscope. (C) Effects of rhPSAP on cell
migration and invasion. A representative control and PSAP-KD clone from each cell line was seeded on transwell filters and incubated for 24 h
(migration) or 48 h (invasion). As a chemo-attractant, basal medium containing 0.5% FBS and various amount of rhPSAP (0, 1, 10, 50 nM) were
included in the lower compartment of the transwell filters. Each bar represented the mean ± SEM of three independent experiments, each in
quadruplicates. ANOVA was used to examine the significance of the data (p < 0.0001) comparing the PSAP-KD clones relative to control clones
in each cell line or among different treatment concentrations for rhPSAP and control. Hu et al. Molecular Cancer 2010, 9:30
http://www.molecular-cancer.com/content/9/1/30 Page 6 of 18 Figure 3 Effect of PSAP down-modulation on b1A-integrin expression and PCa cell adhesion to FN and LN. (A) PSAP down-modulation
reduces b1-integrin expression. Total RNA was extracted for RT-PCR with primers specific for integrin-b1A, -b1B, and -b1C, and GAPDH. Cell lysates
were analyzed by western blotting with antibodies against b1-integrin and its isoforms b1A, b1B, and b1C and GAPDH. (B) Transient down
modulation of b1-integrin expression. Previously established control stable clones of PC-3 and DU-145 cells were transiently transfected with b1
integrin- or scrambled-siRNA oligos. After 48 h, cell lysates were analyzed for integrin b1 and b1A expression by western blotting. (C) Inhibition of
PCa cell adhesion by transient transfection with integrin b1-siRNA. After siRNA transfection, cells were subjected to adhesion assay on FN- or LN-
coated 96-well plates as described in “Materials and Methods”. (D) b1A-integrin stability. The steady-state b1A protein levels were evaluated by
treating cells with cycloheximide (12.5 μg/ml) for up to 72 h and immunoblotting with anti-b1A antibody. (E) The half-live of b1A-integrin was
evaluated by densitometric analysis of immunoblotting bands using the Quantity One software and the b1A protein levels were calculated as
percentage of non-treatment values after normalization using GAPDH for loading control. (F) Effect of inhibition of lysosomal proteolysis on b1A-
integrin expression. Cells were incubated in the presence or absence of the lysosomal proteolysis inhibitor, NH4Cl (50 mM) for up to 24 h. PSAP down-modulation prevents focal adhesion kinase
activation and focal adhesion complex formation However, the PSAP-KD
cells showed a small and round morphology and a weak
b1-integrin staining which remained non-clustered and
largely in the cytoplasmic region. Furthermore, the con-
trol cells formed several focal contacts as visualized by
phospho-specific antibodies against FAK (Tyr-397) and
paxillin (Tyr-118). The control cells also exhibited a
greater extent of co-localization of FAK and paxillin
proteins. However, the PSAP-KD cells showed clearly
attenuated activation of focal adhesions characterized by
a smaller size and lesse number of focal contacts as well
as irregular localization of FAK and paxillin (Fig. 4B). By
using the antibody against vinculin, another cytoskeletal
protein, similar attenuation in the formation of focal
adhesions was also observed in the PSAP-KD clones
(data not shown). To assess the involvement of CathD in PCa cell inva-
sion, the PC-3 and DU-145 control clones were transi-
ently transfected with human CathD siRNA-oligos. As
shown in Fig. 5C, CathD-siRNA reduced CathD expres-
sion by 90% in both cell lines. Interestingly, knock-down
of CathD expression also decreased the intracellular
expression of PSAP and saposin C (Fig. 5C). In addition,
in vitro migration and invasion assays revealed that
CathD-siRNA decreased cell migration by 76% in PC-3
and by 71% in DU-145, as well as cell invasion by 82%
in PC-3 and 77% in DU-145 (Fig. 5D). Cell viability
assays showed that the decrease of cell migration and
invasion was independent of cell proliferation. Overall,
these data strongly support a close metabolic and func-
tional relationship between PSAP and ProCathD in the
process of migration and invasion in PCa cells. PSAP down-modulation increases ceramide levels in PCa
cells PSAP is the precursor of saposins which serve as the
essential co-factors of lysosomal sphingolipid hydrolases. Therefore, the presence and relative abundance of PSAP
greatly influence the balance between sphingolipid
synthesis and degradation. In the lysosomes, saposins,
derived from PSAP, degrade Cer which physiologically
serves as one of the most important second messenger
systems in the regulation of fundamental biological pro-
cesses such as growth, differentiation, cell adhesion, and
migration as well as the activation of several important
signaling pathways. By using MALDI-mass spectrometry
analysis, we found that a number of cellular Cer with
various hydro-carbon lengths, including Cer-C16, -C18,
and -C24 are accumulated in the PSAP-KD clones. Total Cer levels in PSAP-KD clones were increased by
46% and 56% in PC-3 and DU-145 cells, respectively
(Fig. 6A). PSAP down-modulation prevents focal adhesion kinase
activation and focal adhesion complex formation However, we did not observe a significant
change in the levels of sphingosine, a metabolic product
of Cer, as well as glycosphingolipids with short oligosac-
charides, such as lactosyl-Cer and glucosyl-Cer (data not
shown). Next, we examined if Cer can affect b1A-integ-
rin expression in stable control clones of PC-3 and DU-
145 cells (Fig. 6B). C6-D-e-Cer treatment decreased
both the expression and maturation of b1A-integrin. Inactive Cer failed to induce any changes in b1A-integrin
expression (data not shown). To determine whether
intracellular Cer accumulation might be responsible for In addition, stress fibers (F-actin) were also arranged
as long fibers co-localized with vinculin and in parallel
with membrane protrusions in control transfectants. In
contrast, such topological evidence of adhesion pheno-
type was absent in PSAP-KD cells. Overall, these data
suggest that the reduction of b1A-integrin expression
secondary to PSAP-down modulation via the interrup-
tion of the “inside-out” signaling mechanism signifi-
cantly inhibits FAK activity and the proper assembly of
focal adhesion complex and contributes to impaired cell
adhesion and migratory phenotype in PSAP-KD cells. PSAP down-modulation prevents focal adhesion kinase
activation and focal adhesion complex formation To visualize the impairment of cell adhesion in
relation to the changes in b1A-integrin and the assembly
of focal adhesion plaque, we used immunofluoresence
(IF) staining of a representative clone of the control and
PSAP-KD cells. As shown in Fig. 4B, the control cells
spread out on the ECM-coated slides and showed a
strong b1-integrin staining that was mainly localized at
or near the cell membrane region, suggesting a func-
tionally activated b1-integrin. However, the PSAP-KD
cells showed a small and round morphology and a weak
b1-integrin staining which remained non-clustered and
largely in the cytoplasmic region. Furthermore, the con-
trol cells formed several focal contacts as visualized by
phospho-specific antibodies against FAK (Tyr-397) and
paxillin (Tyr-118). The control cells also exhibited a
greater extent of co-localization of FAK and paxillin
proteins. However, the PSAP-KD cells showed clearly
attenuated activation of focal adhesions characterized by
a smaller size and lesse number of focal contacts as well
as irregular localization of FAK and paxillin (Fig. 4B). By
using the antibody against vinculin, another cytoskeletal
protein, similar attenuation in the formation of focal
adhesions was also observed in the PSAP-KD clones
(data not shown). observed a significant reduction in the expression levels
of inactive proCathD (~53 KDa), active intermediate
(~48 KDa), and mature (~31 KDa) forms in all PSAP-
KD clones compared to their control counterparts. The
secreted levels of proCathD were also reduced by PSAP
down-modulation. In addition, our analysis showed that
CathD proteolytic activity in the PSAP-KD clones
decreased by 22% in PC-3 cells and by 48% in DU-145
cells (Fig. 5B). modulation. FAK was constitutively phosphorylated on
tyrosine residues in control transfectants to the levels
similar to PSAP-KD clones. Following 45 or 90 min
adhesion to FN or LN, FAK phosphorylation at Tyr-397,
Tyr-576, Tyr-861, and Tyr-925 and the level of paxillin
phosphorylation at Tyr-118 increased at higher amounts
in the control clones than the PSAP-KD clones (Fig. 4A). To visualize the impairment of cell adhesion in
relation to the changes in b1A-integrin and the assembly
of focal adhesion plaque, we used immunofluoresence
(IF) staining of a representative clone of the control and
PSAP-KD cells. As shown in Fig. 4B, the control cells
spread out on the ECM-coated slides and showed a
strong b1-integrin staining that was mainly localized at
or near the cell membrane region, suggesting a func-
tionally activated b1-integrin. PSAP down-modulation prevents focal adhesion kinase
activation and focal adhesion complex formation Similar results were
obtained from three independent experiments. Figure 3 Effect of PSAP down-modulation on b1A-integrin expression and PCa cell adhesion to FN and LN. (A) PSAP down-modulation
reduces b1-integrin expression. Total RNA was extracted for RT-PCR with primers specific for integrin-b1A, -b1B, and -b1C, and GAPDH. Cell lysates
were analyzed by western blotting with antibodies against b1-integrin and its isoforms b1A, b1B, and b1C and GAPDH. (B) Transient down
modulation of b1-integrin expression. Previously established control stable clones of PC-3 and DU-145 cells were transiently transfected with b1
integrin- or scrambled-siRNA oligos. After 48 h, cell lysates were analyzed for integrin b1 and b1A expression by western blotting. (C) Inhibition of
PCa cell adhesion by transient transfection with integrin b1-siRNA. After siRNA transfection, cells were subjected to adhesion assay on FN- or LN-
coated 96-well plates as described in “Materials and Methods”. (D) b1A-integrin stability. The steady-state b1A protein levels were evaluated by
treating cells with cycloheximide (12.5 μg/ml) for up to 72 h and immunoblotting with anti-b1A antibody. (E) The half-live of b1A-integrin was
evaluated by densitometric analysis of immunoblotting bands using the Quantity One software and the b1A protein levels were calculated as
percentage of non-treatment values after normalization using GAPDH for loading control. (F) Effect of inhibition of lysosomal proteolysis on b1A-
integrin expression. Cells were incubated in the presence or absence of the lysosomal proteolysis inhibitor, NH4Cl (50 mM) for up to 24 h. Cell
lysates were analyzed for b1A protein expression by immunoblotting. The b1A-integrin degradation curve was calculated as described above. Columns, mean of three independent samples run together; bars, ± SEM, p < 0.0001, Two-sample t-tests with Satterthwaite corrections were
used to compare b1-siRNA versus scrambled siRNA oligos transfected cells following adhesion to FN or LN. ANOVA was used to examine the
significance of the data (p < 0.05) among different cycloheximide treatment periods in PSAP-KD versus control clones. Similar results were
obtained from three independent experiments. Hu et al. Molecular Cancer 2010, 9:30
http://www.molecular-cancer.com/content/9/1/30 Page 7 of 18 modulation. FAK was constitutively phosphorylated on
tyrosine residues in control transfectants to the levels
similar to PSAP-KD clones. Following 45 or 90 min
adhesion to FN or LN, FAK phosphorylation at Tyr-397,
Tyr-576, Tyr-861, and Tyr-925 and the level of paxillin
phosphorylation at Tyr-118 increased at higher amounts
in the control clones than the PSAP-KD clones (Fig. 4A). PSAP down-modulation decreases cathepsin D expression
and proteolytic activity in PCa cells The multi-step process of invasion phenotype requires
the involvement of matrix-degrading proteolytic
enzymes. Among different classes of proteolytic
enzymes, several lines of evidence demonstrated a
dynamic active physical and functional interaction
between CathD and PSAP [4,6,21]. Therefore, we exam-
ined if down-modulation of PSAP affects CathD expres-
sion and activity. As shown in Fig. 5A, CathD mRNA
expression was not affected by PSAP down-modulation
in any of the cell lines investigated. However, we Hu et al. Molecular Cancer 2010, 9:30
http://www.molecular-cancer.com/content/9/1/30 Hu et al. Molecular Cancer 2010, 9:30 Page 8 of 18 http://www.molecular-cancer.com/content/9/1/30 Figure 4 PSAP down-modulation decreased FAK activity and prevented b1A-integrin clustering and proper assembly of focal adhesion
complex. (A) PSAP down-modulation reduced phosphorylation of FAK and paxillin. Cells were incubated in suspension with gentle rotation for
45 min and then plated onto FN- or LN-coated dishes for 45 or 90 min. Whole cell lysates were extracted and equal amount of proteins were
used for immunoprecipitation with anti-FAK or-paxillin antibody and immunoblotting with phospho-specific antibodies against Tyr-397, -576,
-861, -925 of FAK or Tyr-118 of paxillin. (B) Effect of PSAP down-modulation on b1A-integrin clustering and focal adhesion complex assembly. Cells were plated onto FN- or LN-coated slides for 2 h, fixed and permeabilized. Immunofluorescence staining was performed with primary Figure 4 PSAP down-modulation decreased FAK activity and prevented b1A-integrin clustering and proper assembly of focal adhes
complex. (A) PSAP down-modulation reduced phosphorylation of FAK and paxillin. Cells were incubated in suspension with gentle rotation f
45 min and then plated onto FN- or LN-coated dishes for 45 or 90 min. Whole cell lysates were extracted and equal amount of proteins wer
used for immunoprecipitation with anti-FAK or-paxillin antibody and immunoblotting with phospho-specific antibodies against Tyr-397, -576,
-861, -925 of FAK or Tyr-118 of paxillin. (B) Effect of PSAP down-modulation on b1A-integrin clustering and focal adhesion complex assembly. Cells were plated onto FN- or LN-coated slides for 2 h, fixed and permeabilized. Immunofluorescence staining was performed with primary
antibodies against integrin b1A, FAK pY397 and paxillin pY118 followed by Cy3 (red) or FITC (green)-conjugated secondary antibodies. F-actin
was stained by Oregon Green 488-phalloidin (green). All images were taken by a Leica DM RA2 fluorescence microscope. Consistent data we
obtained from three independent experiments. Figure 4 PSAP down-modulation decreased FAK activity and prevented b1A-integrin clustering and proper assembly of focal adhesion
complex. PSAP down-modulation decreases cathepsin D expression
and proteolytic activity in PCa cells Figure 5 Down-regulation of cathepsin D expression and activity decreased migration and invasion in PSAP-KD cells. (A) PSAP down-
modulation reduced CathD expression and activity. Total RNA was subjected to RT-PCR using specific primers for CathD. Cell lysates and culture
supernatants were prepared from parallel dishes and analyzed by immunoblotting with an anti-CathD monoclonal antibody which recognizes
proCathD (P), intermediate CathD (I), and mature CathD (M). (B) Whole cell extracts were also assayed for CathD enzymatic activity using a kit
with a fluorimetric substrate. The enzymatic activity of CathD was calculated as units/mg total protein. Columns, mean of three independent
samples run together; bars, ± SEM. ANOVA was used to examine the significance of the data (p < 0.0001) comparing PSAP-KD clones versus
control clones. (C) Transient down-modulation of CathD and its effect on PSAP expression. Control clones of PC-3 and DU-145 cell lines were
transiently transfected with specific CathD- or control-siRNA oligos. After 48 h, cell lysates were analyzed for CathD, PSAP, and saposin C
expression by immunoblotting. (D) Migration and invasion assays were performed on parallel-transfected tissue culture plates as described in the
legend to Fig. 2. Columns, mean of three independent samples run together; bars, ± SEM, p < 0.0001, Two-sample t-tests with Satterthwaite
corrections were used to compare CathD-siRNA versus scrambled-siRNA oligos transfected cells. Similar results were obtained from three
independent experiments. re 5 Down-regulation of cathepsin D expression and activity decreased migration and invasion in PSAP-KD cells the impairment of cell adhesion to ECM proteins, we
investigated the possible effects of Cer on cell adhesion,
migration and invasion by treating the cells with C6-D-
e-Cer. C6-D-e-Cer in a dose-dependent manner
decreased cell adhesion to FN and LN by ≥50% in con-
trol clones from both cell lines (Fig. 6C &6D). It is note-
worthy that the inhibition of cell adhesion on FN and
LN was associated with clear morphological features
such as decreased spreading of the cells and lack of
polarity and membrane protrusions (data not shown) similar to our observation in Fig. 1D and 4B. In addi-
tion,
C6-D-e-Cer,
in
a
dose-dependent
manner
decreased cell migration and invasion by 56% and 64%
in control clones of PC-3 cells and by 59% and 66% in
control clones of DU-145 cells, respectively (Fig. 6E
&6F). PSAP down-modulation decreases cathepsin D expression
and proteolytic activity in PCa cells Figure 5 Down-regulation of cathepsin D expression and activity decreased migration and invasion in PSAP-KD cells. (A) PSAP down-
modulation reduced CathD expression and activity. Total RNA was subjected to RT-PCR using specific primers for CathD. Cell lysates and culture
supernatants were prepared from parallel dishes and analyzed by immunoblotting with an anti-CathD monoclonal antibody which recognizes
proCathD (P), intermediate CathD (I), and mature CathD (M). (B) Whole cell extracts were also assayed for CathD enzymatic activity using a kit
with a fluorimetric substrate. The enzymatic activity of CathD was calculated as units/mg total protein. Columns, mean of three independent
samples run together; bars, ± SEM. ANOVA was used to examine the significance of the data (p < 0.0001) comparing PSAP-KD clones versus
control clones. (C) Transient down-modulation of CathD and its effect on PSAP expression. Control clones of PC-3 and DU-145 cell lines were
transiently transfected with specific CathD- or control-siRNA oligos. After 48 h, cell lysates were analyzed for CathD, PSAP, and saposin C Figure 5 Down-regulation of cathepsin D expression and activity decreased migration and invasion in PSAP-KD cells. (A) PSAP down-
modulation reduced CathD expression and activity. Total RNA was subjected to RT-PCR using specific primers for CathD. Cell lysates and culture
supernatants were prepared from parallel dishes and analyzed by immunoblotting with an anti-CathD monoclonal antibody which recognizes
proCathD (P), intermediate CathD (I), and mature CathD (M). (B) Whole cell extracts were also assayed for CathD enzymatic activity using a kit
with a fluorimetric substrate. The enzymatic activity of CathD was calculated as units/mg total protein. Columns, mean of three independent
samples run together; bars, ± SEM. ANOVA was used to examine the significance of the data (p < 0.0001) comparing PSAP-KD clones versus
control clones. (C) Transient down-modulation of CathD and its effect on PSAP expression. Control clones of PC-3 and DU-145 cell lines were
transiently transfected with specific CathD- or control-siRNA oligos. After 48 h, cell lysates were analyzed for CathD, PSAP, and saposin C
expression by immunoblotting. (D) Migration and invasion assays were performed on parallel-transfected tissue culture plates as described in the
legend to Fig. 2. Columns, mean of three independent samples run together; bars, ± SEM, p < 0.0001, Two-sample t-tests with Satterthwaite
corrections were used to compare CathD-siRNA versus scrambled-siRNA oligos transfected cells. Similar results were obtained from three
independent experiments. PSAP down-modulation decreases cathepsin D expression
and proteolytic activity in PCa cells (A) PSAP down-modulation reduced phosphorylation of FAK and paxillin. Cells were incubated in suspension with gentle rotation for
45 min and then plated onto FN- or LN-coated dishes for 45 or 90 min. Whole cell lysates were extracted and equal amount of proteins were
used for immunoprecipitation with anti-FAK or-paxillin antibody and immunoblotting with phospho-specific antibodies against Tyr-397, -576,
-861, -925 of FAK or Tyr-118 of paxillin. (B) Effect of PSAP down-modulation on b1A-integrin clustering and focal adhesion complex assembly. Cells were plated onto FN- or LN-coated slides for 2 h, fixed and permeabilized. Immunofluorescence staining was performed with primary
antibodies against integrin b1A, FAK pY397 and paxillin pY118 followed by Cy3 (red) or FITC (green)-conjugated secondary antibodies. F-actin
was stained by Oregon Green 488-phalloidin (green). All images were taken by a Leica DM RA2 fluorescence microscope. Consistent data were
obtained from three independent experiments. Hu et al. Molecular Cancer 2010, 9:30 Hu et al. Molecular Cancer 2010, 9:30
http://www.molecular-cancer.com/content/9/1/30 Page 9 of 18 http://www.molecular-cancer.com/content/9/1/30 Figure 5 Down-regulation of cathepsin D expression and activity decreased migration and invasion in PSAP-KD cells. (A) PSAP down-
modulation reduced CathD expression and activity. Total RNA was subjected to RT-PCR using specific primers for CathD. Cell lysates and culture
supernatants were prepared from parallel dishes and analyzed by immunoblotting with an anti-CathD monoclonal antibody which recognizes
proCathD (P), intermediate CathD (I), and mature CathD (M). (B) Whole cell extracts were also assayed for CathD enzymatic activity using a kit
with a fluorimetric substrate. The enzymatic activity of CathD was calculated as units/mg total protein. Columns, mean of three independent
samples run together; bars, ± SEM. ANOVA was used to examine the significance of the data (p < 0.0001) comparing PSAP-KD clones versus
control clones. (C) Transient down-modulation of CathD and its effect on PSAP expression. Control clones of PC-3 and DU-145 cell lines were
transiently transfected with specific CathD- or control-siRNA oligos. After 48 h, cell lysates were analyzed for CathD, PSAP, and saposin C
expression by immunoblotting. (D) Migration and invasion assays were performed on parallel-transfected tissue culture plates as described in the
legend to Fig. 2. Columns, mean of three independent samples run together; bars, ± SEM, p < 0.0001, Two-sample t-tests with Satterthwaite
corrections were used to compare CathD-siRNA versus scrambled-siRNA oligos transfected cells. Similar results were obtained from three
independent experiments. Discussion We employed an RNA-interference strategy to investi-
gate the molecular events underlying PSAP regulation of
metastatic PCa cell invasion. Stable down-modulation of
PSAP decreased CathD expression and proteolytic activ-
ity, migration, and invasion of the highly invasive and
metastatic PCa cells. In vitro and in vivo studies have
implicated a role for CathD in PCa growth, invasion,
and metastasis [22-25]. Data from transient transfection
studies presented here further support a relationship
between ProCathD and PSAP. The two precursors
PSAP and proCathD are glycoproteins that originate in
the endoplasmic reticulum (ER) which travel together as
a PSAP-proCathD complex and reach the lysosomes by
intracellular trafficking [4,21,26,27]. In endosomal and/
or lysosomal compartments, they undergo mutual pro-
teolytic processing to become the final mature products,
saposins and CathD. PSAP accelerates the activation of
proCathD and stimulates its autocatalytic activity, gener-
ating the enzymatically active intermediate and mature
CathD. In turn, CathD catalyzes the production of sapo-
sins from PSAP [4,27]. In our study, most of the
observed decrease in intracellular CathD was in the
enzymatically active forms (intermediate and mature)
rather than in proCathD levels. This data suggests that
PSAP down-modulation not only affect proCathD con-
version to CathD, but might also influence CathD stabi-
lity and/or synthesis. Taken together our data suggest a
cooperative interaction between PSAP and CathD in
PCa cell migration and invasion. p
yp
The key findings of this report are the significant
reduction of Src binding to FAK and the lack of proper
assembly of focal adhesion complex in PSAP knock-
down cells. Together, they highlight the importance of
PSAP and saposin C in regulating “inside-out” integrin-
mediated signal transduction pathway(s) leading to
decreased PCa cell migration and invasion. Based on
our data, it appears the observed structural and func-
tional outcomes occur primarily due to reduced b1A-
integrin expression following PSAP down modulation. In addition, reduction of Src binding to FAK was paral-
leled with decreased Src activity (phosphorylation) in
PSAP-KD cells (data not shown) and did not affect the
activity level of its upstream targets MAPK and PI3K/
Akt (data not shown). As natural cell membrane and
intracellular proteins, PSAP and its active molecular
derivatives, saposin C and its neuro-active domain (pro-
saptide), might also interact with Src alone or in asso-
ciation with
focal
adhesion complex
and other
interactive adaptor proteins (e.g., paxillin) to stabilize
the dynamic state of focal adhesion plaques. PSAP down-modulation decreases cathepsin D expression
and proteolytic activity in PCa cells However, neither the morphological alterations
nor inhibition of cell adhesion, migration, and invasion
were induced by inactive Cer analog C6-L-e-Cer or the
C6-D-t-Cer, the unnatural stereoisomeres of C6-D-e-
Cer, which suggested that these phenotypes are highly Hu et al. Molecular Cancer 2010, 9:30
http://www.molecular-cancer.com/content/9/1/30 Page 10 of 18 Page 10 of 18 of Src-family kinases and establishes FAK-Src signaling
complex. This association leads to the Src transpho-
sphorylation of FAK within its kinase domain of the
activation loop (Tyr-576) and its C-terminal domain
(Tyr-861 and -925) and to the activation of downstream
adaptor molecules such as paxillin, by phosphorylation
at Tyr-118 [31]. Consistent with the role of integrins in
FAK-Src signaling regulation and downstream activation
of adaptor molecules, we found that decreasing b1A-
integrin expression disrupted these processes in several
ways: a) loss of directional membrane protrusion and
ruffles and clustering of b1-integrin and FAK, b) inability
to form focal adhesion complex, c) decreased Src-bind-
ing to FAK (see additional file 6), d) significant reduc-
tion of phosphorylative activity of FAK at Tyr-397, -576,
-861, and -925, and e) decreased phosphorylation of
paxillin at Tyr-118 in PSAP-KD cells. These data pro-
vide a classical example whereby interruption of integ-
rin-regulated FAK-Src signaling secondary to down-
modulation of PSAP (as a lysosomal protein) leads to a
less adhesive and motile phenotype in PCa cells. dependent on native Cer configuration. Treatment of
cells with 8 to 32 μM of C6-D-e-Cer or its inactive
enantiomers C6-L-e-Cer and C6-D-t-Cer for 24 or 36 h
followed by a trypan blue exclusion assay showed ≥95%
cell viability. Discussion In PCa cells, b1-integrin is the most abundant and
ubiquitously expressed subunit [19]. Experimental modi-
fications of b1-integrin have been demonstratred to
affect development, cell proliferation, migration, and
activation of downstream FAK-Src signaling [28,29]. We
discovered that, b1A-integrin isoform is not only the
most abundant isoform, but also that down-modulating
cellular PSAP levels significantly reduced its expression
in the highly invasive and metastatic PCa cell lines, PC-
3, DU-145, and C4-2B (data not shown). Accumulated Cer levels secondary to PSAP down-
modulation which lead inevitably to reduction of sapo-
sins may be responsible for decreased b1A expression. In
support of this assertion, we found that exogenous Cer
not only decreased PCa cell adhesion, migration, and
invasion, but also reduced b1A-integrin expression in
control clones of PC-3 and DU-145 cell lines. It has
been reported that Cer could inhibit integrin b1-glycosy-
lation and -trafficking to cell surface by disrupting the
function of Golgi complexes [32]. We observed that
PSAP down-modulation induced the accumulation of
cellular
Cer
without
affecting
the
levels
of Several studies have demonstrated that, upon engage-
ment with ECM components, integrins reorganize to
form focal adhesion complexes, activate FAK autopho-
sphorylation at Tyr-397, and establish a mechanical link-
age with cytoskeletal molecules such as actin and
vinculin, which control cell shape and motility [30]. FAK phosphorylation at Tyr-397 also creates a high-
binding affinity site for Src-homology 2 (SH2) domain Hu et al. Molecular Cancer 2010, 9:30
http://www.molecular-cancer.com/content/9/1/30 Hu et al. Molecular Cancer 2010, 9:30 Page 11 of 18 http://www.molecular-cancer.com/content/9/1/30 Figure 6 Effect of ceramide on PCa cell adhesion, migration, and invasion. (A) PSAP down-modulation increased Cer levels in PSAP-KD
cells. PSAP-KD and control clones of PC-3 and DU-145 cells were subjected to matrix-assisted laser desorption mass spectrometric analysis as
described in “Materials and Methods”. The assay was performed in duplicate and repeated twice independently. Cer content was quantified and
calibrated to the intracellular phosphate (Pi) level and depicted as Cer (pM)/Pi (nM). (B) Effect of Cer on b1A-integrin expression. Control clones of
PC-3 and DU-145 cells were treated with active or inactive Cer analogs or vehicle (DMSO) at the indicated concentrations for 36 h. Cell lysates
were subjected to immunoblotting with specific antibodies against total b1-integrin or b1A-isoform. Inactive Cer analog and DMSO did not affect
integrins expression level. GAPDH antibody was used as control loading. (C and D) Effect of Cer on PCa cell adhesion. Discussion Parental PC-3 and DU-145
cells were treated with cell permeable natural Cer analog (C6-D-e-Cer; D-Cer), inactive Cer (C6-L-e-Cer; L-Cer), or vehicle (DMSO) at 1 or 2 μM for
three to five days and then, subjected to cell adhesion on FN- or LN-coated plates as described in the legend to Fig. 1. (E and F) Effect of Cer
on PCa cell migration and invasion. The effect of Cer on migration and invasion of control PC-3 and DU-145 transfectants was examined by
treating cells with active or inactive Cer analogs (described above) followed by migration and invasion assay as explained in the legend to Fig. 2. Each bar represents the mean ± SEM of three independent experiments. ANOVA models with Dunnett and Tukey corrections were used to
http://www.molecular-cancer.com/content/9/1/30 Figure 6 Effect of ceramide on PCa cell adhesion, migration, and invasion. (A) PSAP down-modulation increased Cer levels in PSAP-KD
cells. PSAP-KD and control clones of PC-3 and DU-145 cells were subjected to matrix-assisted laser desorption mass spectrometric analysis as
described in “Materials and Methods”. The assay was performed in duplicate and repeated twice independently. Cer content was quantified and
calibrated to the intracellular phosphate (Pi) level and depicted as Cer (pM)/Pi (nM). (B) Effect of Cer on b1A-integrin expression. Control clones of
PC-3 and DU-145 cells were treated with active or inactive Cer analogs or vehicle (DMSO) at the indicated concentrations for 36 h. Cell lysates
were subjected to immunoblotting with specific antibodies against total b1-integrin or b1A-isoform. Inactive Cer analog and DMSO did not affect
integrins expression level. GAPDH antibody was used as control loading. (C and D) Effect of Cer on PCa cell adhesion. Parental PC-3 and DU-145
cells were treated with cell permeable natural Cer analog (C6-D-e-Cer; D-Cer), inactive Cer (C6-L-e-Cer; L-Cer), or vehicle (DMSO) at 1 or 2 μM for
three to five days and then, subjected to cell adhesion on FN- or LN-coated plates as described in the legend to Fig. 1. (E and F) Effect of Cer
on PCa cell migration and invasion. The effect of Cer on migration and invasion of control PC-3 and DU-145 transfectants was examined by
treating cells with active or inactive Cer analogs (described above) followed by migration and invasion assay as explained in the legend to Fig. 2. Each bar represents the mean ± SEM of three independent experiments. Discussion While these reports suggest a role for
PSAP in invasive and metastatic progression of prostate,
breast, and pancreatic tumors, a recent report has sug-
gested that PSAP may inhibit breast and PCa metastasis
by paracrine and endocrine stimulation of thrombos-
pondin-1 expression in a p53-dependent manner in
fibroblasts of primary tumors and distant metastases
[40]. With respect to PCa, the study was based on PC-
3M, a metastatic subline of PC-3 cell line. In addition,
the authors used PC-3M-LN4, a lymph node-metastatic
subline of PC-3M that had been subjected to four cycles
of injection to prostate and harvesting from the lymph
d
f th
i
d
i
[41] D
t
t
i
l
l glycosphingolipids. This result is somewhat different
from those other studies of total PASP deficiency in
patients and in experimental mouse models, in which
significant accumulation of Cer as well as lactosyl-Cer
and glucosyl-Cer has been observed [9,10,33]. We spec-
ulate that the balance of Cer metabolism is more sensi-
tive to the relative changes in PSAP expression than is
the metabolism of glycosphingolipids, which essentially
dependes on the presence of a low (threshold) PSAP
level, similar to the residual amount of PSAP in the
PSAP-KD clones, which is comparable to normal pros-
tate epithelial cells (data not shown). It is noteworthy
that the endogenous Cer levels are coordinately regu-
lated by several specialized enzymes and hydrolases
which produce Cer or use Cer as substrate [34]. Ele-
vated PSAP expression may shift the balance of Cer by
activating certain hydrolases or even by directly regulat-
ing their expression through functional saposins. For
example, saposin D can stimulate the activity of acid
ceramidase, which mediates the conversion of Cer into
sphingosine [35]. This hypothesis is supported by our
finding that ceramidase expression is reduced in PSAP-
KD cells (unpublished data). The Cer level is frequently
decreased in cancer cells and correlates inversely with
the degree of malignant progression [36]. Therefore, it is
conceivable that PSAP overexpression may greatly con-
tributes to Cer-level reduction in invasive and metastatic
cancer cells. Considering the complexity of Cer as a
bioactive sphinogolipid, the underlying mechanisms by
which Cer inhibits PCa cell motility and invasiveness
require further detailed investigation. Our data indicate a role for soluble PSAP as a para-
crine regulatory factor in migration and invasion. Discussion ANOVA models with Dunnett and Tukey corrections were used to
compare cells treated with DMSO, L-Cer, or D-Cer. Statistically significant differences were set at p < 0.05. Consistent data were obtained from
three independent experiments. Hu et al. Molecular Cancer 2010, 9:30
http://www.molecular-cancer.com/content/9/1/30 Page 12 of 18 PSAP has been demonstrated to be overexpressed in
conditioned media of estrogen receptor (ER)-positive
MCF-7 and ER-negative MDA-MB-231 breast cancer
cell lines as well as in a human SV40-transformed breast
epithelial cells, HBL100 [37]. In MCF-7 conditioned
media, the PSAP expression pattern closely resembled
that of proCathD. Interestingly, the same authors
demonstrated that estrogen (17b-estradiol) increased
secretion of both proteins in a dose-dependent manner. These observations together with our data support the
hypothesis that the close functional association between
proCathD and PSAP may eliminate tissue barriers by
facilitating proteolytic degradation of basement mem-
brane glycoproteins. PSAP was also identified as a gene
with causative role during functional screening for
tamoxifen-resistance in breast cancer cell line, ZR-75-1
[28]. Further investigation of clinical samples using
qRT-PCR analysis of mRNA levels in 223 ER-positive
primary breast cancers from patients who had recurrent
metastatic disease and were treated with tamoxifen as a
first-line therapy, revealed a high PSAP expression level
for 182 out of 223 patients. In addition, Cox univariate
and multivariate analyses for progression-free survival
correlated the high PSAP expression levels in these
patients with shorter progression-free survival [29]. Independently, using Mass Spectrometry based-proteo-
mic analyses and qRT-PCR for comparative analysis of
non-metastatic primary breast cancer and lymph node
metastases, PSAP was found to be significantly increased
(2 fold) in lymph node metastasis [38]. Similarly, in our
previous studies using punch biopsy samples of meta-
static PCa, genomic amplification of PSAP was detected
in 2 out of 5 lymph node metastases [13]. In addition to
breast cancer and PCa, in a comparative analysis of the
secretomes of an immortalized pancreatic duct normal
epithelial cell (HPDE) and a pancreatic ductal adenocar-
cinoma cell line (Panc1), PSAP expression in Panc1 was
found to 11-fold higher than in the HPDE cell line [39]. Interestingly, PSAP upregulation in Panc1 was asso-
ciated with CathD (6.7-fold) and b1-integrin (3.4-fold)
overexpression. Transient transfection assays Cells were seeded in 6-well plates overnight and trans-
fected with 50 pmol of human CathD-, integrin-b1, or
control-siRNA oligos (Santa Cruz Biotechnology, Santa
Cruz, CA) and 5 μl Lipofectamine RNAiMAX for 8 h. The transfected cells were cultured in complete medium
for 16 h and then, in basal medium for additional 24 h
before performing functional assays or harvesting cell
lysates and/or supernatants for protein expression
analyses. Establishment of stable transfectants of PSAP knock-
down cell lines
5 Cells were seeded at 2 × 105 per well in 6-well plates
overnight and transfected with 2 μg short hairpin (sh)-
RNA plasmid containing a siRNA sequence targeted
against human PSAP or a scrambled control sequence
(SuperArray Biosciences, Germantown, MD) and 5 μl
Lipofectamine 2000 according to the manufacturer’s
instructions (Invitrogen, Carlsbad, CA). After 8 hours of
incubation at 37°C, the transfection medium was
removed and cells were cultured in complete medium
for 48 h. Cells were trypsinized and cultured in the pre-
sence of 1 mg/ml G418 for the selection of antibiotic-
resistant colonies over a period of 2 to 3 weeks. Several
PSAP-knockdown (KD) and -control clones were iso-
lated and analyzed for PSAP expression by western blot-
ting and RT-PCR. We randomly selected two PSAP-KD
clones (P4 and P7 for PC-3 cell line and P15 and P32
for DU-145) and two control clones (C4 and C8 for PC-
3 cell line and C9 and C13 for DU-145 cell line) for
functional studies. The stable cell lines were routinely
examined for PSAP expression and maintained in the
complete medium containing 300 μg/ml of G418. In summary, we provide mechanistic evidence that
PSAP
down-modulation
upregulates
Cer
levels,
decreases b1A-integrin and CathD expression, attenuates
the “inside-out” integrin-signaling pathway, and signifi-
cantly decreases PCa cell adhesion, migration, and inva-
sion. The fact that PSAP is frequently overexpressed in
human malignant cells warrants further investigation of
its role in carcinogenesis and in invasive and metastatic
progression of cancer cells. Materials and methods
Cell culture Cell lines used in this study were essentially maintained
as described before [13,14]. Cycloheximide (CHX), leu-
peptin, MG-132, and ALLN were obtained from Sigma
(Saint Louis, MO). Expression and purification of recombinant human PSAP
in CHO-K1 cells Discussion Based
on our study, this paracrine regulatory effect is not suffi-
cient to bypass the intracellular regulatory mechanisms
responsible for significant suppression of migratory and
invasive phenotypes secondary to PSAP down-modula-
tion. It is likely that the receptor-mediated signaling
mechanisms and post-receptor downstream effectors
responsible for the paracrine effect of PSAP may be dif-
ferent from the intracrine regulatory pathways. Our previous studies also showed that exogenous
saposin C and prosaptide treatment could stimulate PCa
cell growth, involving activating several signaling path-
ways [14]. However, our current data show that under
our experimental conditions, the growth properties of
PCa cells was not affected by either intracellular down-
modulation of PSAP or treatment with rhPSAP. Furthermore, neither PSAP down-modulation nor
rhPSAP treatment affected the MAPK and PI3K activity
level (unpublished data). Therefore, the observed effect
of exogenous saposin C does not necessarily reflect the
physiological function of extracellularly secreted PSAP
or an intracellular pool of this protein. Page 13 of 18 Hu et al. Molecular Cancer 2010, 9:30
http://www.molecular-cancer.com/content/9/1/30 Hu et al. Molecular Cancer 2010, 9:30
http://www.molecular-cancer.com/content/9/1/30 Hu et al. Molecular Cancer 2010, 9:30
http://www.molecular-cancer.com/content/9/1/30 between the influence of clonal selection and a cause-
and-effect relationship for the PSAP contribution as an
inhibitor of PCa metastasis. Our analysis of three inde-
pendent PCa progression models based on isogenic cell
lines has revealed a steady-state increase in PSAP
expression levels in invasive and metastatic cells as com-
pared to their parental cells (see additional file 1). These
data show that PSAP expression in PC-3M is at least 3-
fold higher than in its isogenic parental cell line, PC-3,
indicating that, upon metastatic progression, PSAP
expression increases. By analyzing gene microarray
expression data from different sources, the authors also
reported that the relative PSAP mRNA expression in
metastatic PCa was 30% lower than in localized primary
tumors [40]. This analysis is based solely on bioinfor-
matics evaluation which does not necessarily represent
the mRNA and protein expression levels of tumor cells. As such, a cause-and-effect relationship between PSAP
and the complex multistep process of metastatic pheno-
type in PCa can not be concluded from the study. Clari-
fication of PSAP’s role in invasive and metastatic
progression of PCa and other malignancies requires
additional detailed investigations. Superflow Resins (Qiagen, Santa Clara, CA). Discussion The mole-
cular size of recombinant PSAP expressed in CHO-K1
cells was similar to that of native PSAP secreted by PC-
3 cells. The size and purity of the purified proteins were
determined by using 4-20% Tris-Glycine gel electro-
phoresis, coomassie blue staining, silver staining, and
western blotting with previously characterized anti-
PSAP antibodies (see additional files 2, 3, 4, 5) [14]. RNA extraction, cDNA synthesis, and semi-quantitative
RT-PCR The full-length cDNA of PSAP gene (GenBank Acces-
sion No. NP002769) was synthesized, tagged at the C-
terminal with hexa-histidine (His6), and subcloned into
the mammalian expression vector pSectag2A (Invitro-
gen, Carlsbad, CA). The pSectag2A vector contained the
Ig leader sequence which allows the secretion of
recombinant proteins. After bacterial transformation,
the sequence accuracy was verified by automated
sequencing in both directions. Stable CHO-K1 clones
expressing high levels of the secreted recombinant
human (rh)-PSAP was obtained using Zeocin as a selec-
tion antibiotic. Recombinant PSAP protein was purified
from culture supernatant using imidazole and Ni-NTA RNA was isolated by using the RNeasy Kit according to
the manufacturer’s instructions (Qiagen, Santa Clara, CA). For cDNAsynthesis, the template (5 μg RNA per sample)
was reverse-transcribed using AffinityScript cDNA Synth-
esis Kit (Stratagene, La Jolla, CA). Semi-quantitative PCR
was performed in total 20 μl volume containing 1 μl
cDNA, 0.2 μM dNTPs, 0.4 μM primers, and 0.4 μl Taq
DNA polymerase (Promega, Madison, MI). Primers were
synthesized by Integrated DNA Technologies Inc (San
Jose, CA). The oligonucleotides used (according to the
human
gene
sequences
with
accession
numbers: Page 14 of 18 Page 14 of 18 Hu et al. Molecular Cancer 2010, 9:30
http://www.molecular-cancer.com/content/9/1/30 Page 14 of 18 Hu et al. Molecular Cancer 2010, 9:30
http://www.molecular-cancer.com/content/9/1/30 NM_002211 for b1A, NM_033666 for b1B, NM_033667 for
b1C, NM_002778 for PSAP, NM_001909 for CathD, and
NM_002046 for GAPDH, as deposited at the NCBI/gen-
ome data bank) were as follows: b1A, 5’-AGAATCCA-
GAGTGTCCCACTGG-3’
(sense)
and
5’-
TTTCCCTCATACTTC GGATTG-3’ (antisense); b1B, 5’-
AAGACTTATGTATTAGCTGTCAG-3’ (sense) and 5’-
CCATTGAATAGCTTG CTACAC-3’ (antisense); b1C, 5’-
TCTGTCGCCCAGCCTGGAGTG-3’ (sense) and 5’-
TTTCCCTCATACTTCGGATTG-3’ (antisense); PSAP,
5’- CCA GAG CTG GAC ATG ACT GA-3’ (sense) and
5’-CAGTTCCCAACAAGGGCTTA-3’ (antisense); CathD,
5’-CTGCACAAGTTCACGTCCAT-3’ (sense) and 5’-
TTCTGCTGCATCAGGTTGTC-3’ (antisense); and
GAPDH, 5’-GGTCGGAGTCAACGGATTTGGTCG-3’
(sense) and 5’-CCTCCGACGCCTGCTTCACCAC-3’
(antisense). After cDNA synthesis, PCR was completed
using a T-gradient model (Biometra, Horsham, PA) under
the following conditions: a denaturation cycle at 95°C for
2 min, 95°C for 45 s, annealing at 58°C for 45 s and elon-
gation at 72°C for 40 s, and a final extension at 72°C for 5
min. The sizes of amplified cDNA fragments and the
number of PCR cycles were: 238 bp/20 cycles for b1A, 278
bp/32 cycles for b1B, 172 bp/32 cycles for b1C, 1000 bp/20
cycles for PSAP, 590 bp/20 cycles for CathD, and 780 bp/
20 cycles for GAPDH. RNA extraction, cDNA synthesis, and semi-quantitative
RT-PCR The PCR product was confirmed as
a single band using 1.5% agarose gel electrophoresis. A
non-template control was included in each PCR experi-
ment. Each experiment was repeated three times
independently. AB1952P at 1:1000), rabbit anti-integrin b1B and b1C
antibodies provided by Dr. L.R. Languino, University of
Massachusetts (at 1:100), monoclonal anti-paxillin
(Millipore, clone 5H11, 1:1000), and rabbit anti-paxillin-
pY118 (Cell signaling, 1:1000). Anti-GAPDH (Santa
Cruz; 1:4000) was used as control loading. AB1952P at 1:1000), rabbit anti-integrin b1B and b1C
antibodies provided by Dr. L.R. Languino, University of
Massachusetts (at 1:100), monoclonal anti-paxillin
(Millipore, clone 5H11, 1:1000), and rabbit anti-paxillin-
pY118 (Cell signaling, 1:1000). Anti-GAPDH (Santa
Cruz; 1:4000) was used as control loading. For immunoprecipitation of cell adhesion molecules,
tissue culture plates were coated with 10 μg/ml fibro-
nectin (FN, Sigma, Saint Louis, MO) or laminin-1 (LN,
R&D Systems, Minnneapolis, MN) overnight at 4°C and
blocked with 1% BSA/PBS for 1 h at room temperature. Subconfluent cultured cells were washed with PBS, incu-
bated with Versene (Invitrogen) for 30 min at 37°C and
a final incubation with 0.0025% trypsin for 10 min. The
detached cells were collected by centrifugation and 6 ×
106 cells in basal medium and were incubated for 45
min at 37°C with gentle rotation. Equal volume of cell
suspension were either lysed immediately or added to
FN- or LN-coated plates for 45 or 90 min. At the end
of incubation periods, plates were washed with cold PBS
on ice and cells were collected by centrifugation at 300
× g for 3 min. One mg protein lysates were incubated
with 1 μg anti-FAK monoclonal antibody (clone 4.47,
Millipore, Billerica, MA) overnight followed by 2 h incu-
bation with rotation at 4°C in the presence of 20 μl Pro-
tein A/G agarose beads (Santa Cruz). After three washes
with the lysis buffer, the beads were re-suspended in
sample-loading buffer and aliquots of cleared superna-
tant were subjected to SDS-PAGE and immunoblotting
with a mouse anti-FAK (clone 4.47; Millipore) at 1:1000
dilution, anti-FAK-pY397 (MAB1144; Millipore) at 1:250
dilution, rabbit anti-FAK-pY576 (Invitrogen) at 1:1000
dilution, anti-FAK-pY861 (Invitrogen) at 1:1000 dilution,
or rabbit anti-FAK-pY925 (Cell Signaling) at 1:1000
dilution. HRP-conjugated secondary antibodies (Santa
Cruz) were used at 1:1000 dilutions. RNA extraction, cDNA synthesis, and semi-quantitative
RT-PCR Where indicated,
subconfluent culture plates were incubated in the basal
medium for 24 h and then, incubated with the protein
synthesis inhibitor, CHX (12.5 μg/ml; Sigma), the lyso-
somal protease inhibitors, leupeptin (100 μM; Sigma) or
NH4Cl (50 mM), the calpain inhibitor ALLN (10 μM;
Sigma), the proteasome inhibitor MG132 (10 μM;
Sigma), or the vehicle alone (DMSO; Sigma) before har-
vesting protein lysates. Immunoblotting bands were
quantified by densitometric analysis using the Quantity
One software (Bio-Rad) and the protein levels were cal-
culated as percentage of non-treated cells with GAPDH
normalization. C ll
lif
i Protein extraction, immunoblotting, and
immunoprecipitation Cell-free culture supernatants were collected and con-
centrated up to 10 times by using a centrifuge concen-
trator with a 3.0 kDa molecular weight cut-off
membrane (Vivascience, Stonehouse, UK). Protein sam-
ples (2 μg for CathD and 5 μg for PSAP) were subjected
to SDS-PAGE and immunoblotting as previously
described [24]. Normalization of culture supernatants
was based on the total cell number and/or protein con-
tent. Whole cell lysates were also prepared from the
same tissue culture plates and used for immunoblotting
(2 μg for CathD and 15 μg for PSAP). Membranes were
probed with mouse anti-human PSAP (1 μg/ml; Abnova,
Taibei, Taiwan) or mouse anti-human CathD (clone
CTD-19 at 1:2000; Santa Cruz) antibodies and signals
were detected by ECL detection system (GE Healthcare,
London, UK). Direct immunoblotting for cell adhesion
molecules was performed on protein lysates: 1 μg for
integrin b1,2 μg for integrin b1A, 20 μg for integrin b1B,
30 μg for integrin b1C, and 5 μg for paxillin. We used
the following antibodies for immunoblotting of cell
adhesion molecules: Mouse anti-integrin b1 (Santa Cruz,
clone JB1B at 1:200), rabbit anti-integrin b1A (Millipore, Cell migration and invasion assays The effect of PSAP down-modulation on cell migration
and invasion was performed using 8-μm transwell fil-
ters (Costar, Corning, NY) with modification as
described previously [14]. For the invasion assay, the
upper compartment was coated with 50 μg Matrigel
(BD Biosciences, San Jose, CA) to form a matrix bar-
rier. A suspension of cells (5 × 104 for PC-3 or 2 ×
104 for DU-145) in basal medium containing 0.1% BSA
was added to the upper compartment. The lower com-
partment was filled with 400 μl basal medium contain-
ing 5% FBS as chemoattractant. After 48 h for PC-3 or
24 h for DU-145, the non-migratory cells on the upper
surface were removed by a cotton swab and the cells
on the lower surface were fixed and stained with the
Diff-Quick solution (Dade Behring, Deerfield, Illinois). To test the effect of rhPSAP on cell migration and
invasion in stable transfectants, 2 × 104 PC-3 or 1 ×
104 DU-145 cells were added to each well and incu-
bated 24 h for migration or 48 h for invasion. Basal
medium containing 0.5% FBS in the absence or pre-
sence of rhPSAP at 0.1, 1, 10, or 50 nM was used as
chemoattractant in the lower transwell compartment. Cell adhesion assays To determine the effect of PSAP down-modulation on
adhesion, subconfluent cultured cells were harvested
by versene treatment as described in the immunopreci-
pitation assays for cell adhesion molecules and seeded
at 1.5 × 104 cells/well in basal medium on FN- or LN-
coated 96-well plates as described above. After 2 h of
incubation at 37°C, cells were washed twice with PBS,
fixed with 10% formaldehyde, and stained with 0.25%
tolouidine blue each for 15 min at room temperature. Images were taken at 100× magnification by a video
camera (Nikon DU-S1) fitted to a microscope. The
adhered cells were counted from ten randomly chosen
fields in at least six independent wells. The experiment
was repeated three times independently. Cell proliferation assays To evaluate the effect of PSAP down-modulation on
cell proliferation, PSAP-KD and control clones were
seeded at 2 × 105 in 10 cm tissue culture dishes in
complete medium and incubated for 2, 4, or 6 days. At
the end of incubation period, cells were harvested by Hu et al. Molecular Cancer 2010, 9:30
http://www.molecular-cancer.com/content/9/1/30 Page 15 of 18 Hu et al. Molecular Cancer 2010, 9:30
http://www.molecular-cancer.com/content/9/1/30 Migrated or invaded cells in each transwell filter were
counted. Imaging and cell counting were performed as
described for cell adhesion assays. The experiment was
performed in quadruplicates and repeated at least
three times independently. trypsinization and viable cell number was determined
by trypan blue exclusion assay using a hematocyt-
ometer. To determine the effect of rhPSAP on cell
growth, 2 × 103 cells per well were seeded in 96-well
plates in complete medium for 2 days and, after wash-
ing the plates with PBS, cells were incubated in the
presence or absence of rhPSAP at 0.1, 1, 10 nM or
0.5% FBS in basal medium containing 0.1% BSA. After
2 days, the cell number was measured by MTS assay
using CellTiter 96 AQueous One Solution Cell Prolif-
eration/Cytotoxicity Assay Kit according to manufac-
turer’s instructions (Promega). Briefly, 20 μl MTS
solution was added to each well for 2 h incubation and
the absorbance at 490 nm was determined. We used
twelve replicates for each treatment condition. trypsinization and viable cell number was determined
by trypan blue exclusion assay using a hematocyt-
ometer. To determine the effect of rhPSAP on cell
growth, 2 × 103 cells per well were seeded in 96-well
plates in complete medium for 2 days and, after wash-
ing the plates with PBS, cells were incubated in the
presence or absence of rhPSAP at 0.1, 1, 10 nM or
0.5% FBS in basal medium containing 0.1% BSA. After
2 days, the cell number was measured by MTS assay
using CellTiter 96 AQueous One Solution Cell Prolif-
eration/Cytotoxicity Assay Kit according to manufac-
turer’s instructions (Promega). Briefly, 20 μl MTS
solution was added to each well for 2 h incubation and
the absorbance at 490 nm was determined. We used
twelve replicates for each treatment condition. Cathepsin D activity assays Cells were grown up to 70% confluency in their mainte-
nance medium and serum starved for 24 h. Cells were
collected in extraction buffer containing 10 mM HEPES
pH 7.0, 10 mM KCl, 1.5 mM MgCl2, 0.5% CHAPS and
subjected to three freeze-thaw cycles. After centrifuga-
tion at 300 × g for 10 min, the clarified crude cell
extracts were subjected to CathD Assay Kit (Sigma)
based on the hydrolysis by the enzyme of an internally
quenched fluorimetric substrate (i.e., MCA) according
to manufacturer’s instruction. Immunofluorescence staining and microscopic
analysis y
To visualize the effect of PSAP down-modulation on cell
adhesion molecules, subconfluent culture plates were
detached by versene treatment as described for the
immunoprecipitation assays of cell adhesion molecules. Cell suspensions were incubated in a basal medium for
45 min at 37°C with gentle rotation. Cells were seeded
at 5 × 104 per well on FN- or LN-coated slides (Lab
Tek) and incubated for 2 h at 37°C. Immunofluores-
cence staining was preformed as described previously
[42]. Briefly, cells were fixed in 3.7% paraformaldehyde
for 30 min and then, permeablized with 0.3% Triton X-
100 for 15 min. The slides were blocked with 1% BSA
for 30 min, incubated with primary antibodies against
integrin b1 (clone JB1B at 1:50; Santa Cruz), FAK-pY397
(MAB1144 at 1:50 dilution; Millipore), and paxillin-
pY118 (at 1:200, Cell Signaling) overnight at 4°C, and
then with FITC- or Cy3-conjugated secondary antibo-
dies (at 1:200; Jackson ImmunoResearch Laboratories,
West Grove, PA) for 1 h at room temperature. In some
cases, the slides were further stained with Oregon Green
488-phalloidin (at 1:40; Invitrogen) for 30 min. After
optimization of the immunofluoresence staining, each
test was performed in triplicates and repeated three
times independently. Mass-spectrophotometric analysis of
sphingolipids Subconfluent culture plates were washed twice with
PBS, and incubated in their basal medium for 24 h. After washing the plates twice with ice-cold PBS, cells
were scraped, centrifuged, and cellular Cer levels was
measured by matrix-assisted laser desorption mass spec-
trometry (MALDI-MS) which included a panel of C14 Hu et al. Molecular Cancer 2010, 9:30
http://www.molecular-cancer.com/content/9/1/30 Page 16 of 18 to C26 Cer species: sphingomyeline, sphingosine, sphin-
gosine-1-phosphate (S-1-P), and the dihydro analogues
of sphingosine and S-1-P. The assay was performed in
duplicate and repeated two times independently. Cer
content was quantitated and calibrated to the intracellu-
lar phosphate (Pi) level and depicted as Cer (pM)/Pi
(nM). Additional file 2: Expression and purification of rhPSAP in CHO-K1
cells. The cDNA of the human PSAP gene (NM_002778) was tagged with
c-terminal 6×histidine by PCR amplification with the following primers 5’-
AAA GCG GCC CAG CCG GCC GGC CCG GTC CTT GGA CTG-3’ (forward)
and 5’- CCG CTC GAG CTA GTG ATG GTG ATG GTG ATG GTT CCA CAC
ATG GCG TTT GC-3’ (backward). The PCR product was digested with the
restriction enzymes SfiI and XhoI (New England Biolabs) and sub-cloned
into a mammalian expression vector pSectag2A (Invitrogen). The positive
clones were selected and confirmed by DNA sequencing. CHO-K1 (ATCC)
cells were cultured in F12-K medium supplemented with 10% FBS and
1% penicillin-streptomycin. Cells (3 × 105) were seeded in 60 mm dishes
and cultured overnight to 20-30% confluency. Cells were transfected
with 4 μg DNA of the pSectag2A/rhPSAP-His6 vector and 20 μl Lipofectin
(Invitrogen) for 16 h. Several stable clones were isolated after selection
with 500 μg/ml of Zeocin for two weeks. One stable clone with the
highest rhPSAP expression was used for large-scale, rhPSAP purification. The stable cells were cultured in several T500 Triple layer flasks (Nunc) up
to 90% confluency and replaced with 60 ml of OptiMEM medium
(Invitrogen). After 48 h, the culture medium was pooled for
centrifugation at 500 × g for 10 min at 4°C and the cleared supernatant
was filtered with a 0.22 μm membrane. One liter of the supernatant was
mixed with 5 ml Ni-NTA Superflow Resins (Qiagen) and incubated for 4
h at room temperature or overnight at 4°C by slow gyroscopic spin at
100 rpm. Additional file 4: Silver staining and western analysis of purified
rhPSAP. Recombinant PSAP proteins were subjected to SDS-PAGE using Additional file 4: Silver staining and western analysis of purified
rhPSAP. Recombinant PSAP proteins were subjected to SDS-PAGE using
a 10% Tris-Glycine gel. The gels were subjected to silver staining and
western blotting with an anti-PSAP antibody as indicated in the Materials
and Methods section. Both silver staining and immunoblotting showed
the presence of the same molecule with the expected molecular weight
for PSAP. Additional file 4: Silver staining and western analysis of purified
rhPSAP. Recombinant PSAP proteins were subjected to SDS-PAGE using
a 10% Tris-Glycine gel. The gels were subjected to silver staining and
western blotting with an anti-PSAP antibody as indicated in the Materials
and Methods section. Both silver staining and immunoblotting showed
the presence of the same molecule with the expected molecular weight
for PSAP. Cli k h
f
fil a 10% Tris-Glycine gel. The gels were subjected to silver staining and
western blotting with an anti-PSAP antibody as indicated in the Materials a 10% Tris-Glycine gel. The gels were subjected to silver staining and
western blotting with an anti-PSAP antibody as indicated in the Materials
and Methods section. Both silver staining and immunoblotting showed
the presence of the same molecule with the expected molecular weight
for PSAP. and Methods section. Both silver staining and immunoblotting showed
the presence of the same molecule with the expected molecular weight
for PSAP. Additional file 1: PSAP expression in three different prostate cancer
progression models. Cells were cultured in their complete media up to
75% confluency, washed with PBS, and incubated in their respective
basal media for 24 h. Cell extracts and concentrated culture supernatants
were prepared as described in the Materials and methods section. Equal
amount of protein samples (15 μg cell extracts or concentrated culture
supernatants) were resolved by SDS-PAGE under reducing conditions
and subjected to Western analysis using a mouse monoclonal antibody
against human saposin C. The GAPDH antibody was used for protein
loading. Pr. Ep, normal human prostate epithelial cells; C4-2B, a bone
metastatic AI-subline of androgen-sensitive LNCaP; p69, a human normal
prostate epithelial cells immortalized with SV40 T-antigen; M12, a
metastatic subline of p69 cell line; M2182, a highly invasive and
metastatic subline of M12 cell line; PC-3M, a highly metastatic subline of
PC-3 cells. Click here for file Click here for file [ http://www.biomedcentral.com/content/supplementary/1476-4598-9-30-
S4.PDF ] Additional file 5: Biological activity of the purified rhPSAP. Mass-spectrophotometric analysis of
sphingolipids The resins were washed with the gradient imidazole at 10, 20,
50, and 100 mM in a binding buffer (50 mM Na2HPO4, pH 8.0, 300 mM
NaCl). The appropriate fractions which contained the purified rhPSAP
protein were pooled, and the buffer was exchanged to PBS by
ultrafiltration using Vivaspin concentrators with a 10 KDa cut-off
membrane. The purified rhPSAP were quantified by measuring OD280/
260 nm; then they were filter-sterilized, and stored at -80°C for future
use. Statistical analyses Data were analyzed using SAS v9.1 (SAS Institute, Cary,
NC). Various ANOVA models were used. Nesting of
assayed biological specimens (clones) in treatments were
accounted for, and included as random effects. Post-hoc
comparisons were performed with the Dunnett or
Tukey procedures. Two-sample t-tests with Sat-
terthwaite correction, when needed, were used to com-
pare experimental settings with two groups. Cell counts
were logarithm transformed as necessary before
ANOVA was completed. An experiment-wise signifi-
cance level of 0.05 was used; similarly, simultaneous
confidence intervals of means were obtained. Additional file 3: Coomassie blue staining of the purified rhPSAP
protein Recombinant proteins expressed in 1 liter of culture Additional file 3: Coomassie blue staining of the purified rhPSAP
protein. Recombinant proteins expressed in 1 liter of culture
supernatants were purified by using Ni-NTA Superflow resin and
polypropylene purification column. Forty μl of culture supernatant, 20 μl
of imidazole washing, or elution fraction of CHO-K1 stable transfectants
(clone # 2-2) were mixed with a non-reducing sample-loading buffer and
separated in 4-20% Tris-Glycine gel. Solid arrows indicate imidazole-
eluted rhPSAP (~68-72 kDa). [ http://www.biomedcentral.com/content/supplementary/1476-4598-9-30-
S3.PDF ] Ceramide treatment Cell permeable bioactive N-Hexanoyl-D-erythro-sphin-
gosine (C6-D-e-Cer), inactive N-Hexanoly-L-erythro-
sphingosine (C6-L-e-Cer), and N-Hexanoly-D-threo-
sphingosine (C6-D-t-Cer) were purchased from Matreya,
LLC (Pleasant Gap, PA). To determine the effect of Cer
on b1A-integrin expression, cells were treated with active
or inactive Cer analog at 8 to 32 μM for 36 h in com-
plete medium and then, for 24 h in basal medium
before immunoblotting. The effect of Cer on cell adhe-
sion, migration, and invasion was determined by treating
cells with 1 or 2 μM of active or inactive Cer for 5 days
followed by 24 h incubation in basal medium before the
functional assays. The effect of Cer on cell growth was
measured by MTS assay as described in cell prolifera-
tion assay. Cytotoxicity of Cer was determined in paral-
lel experiments using trypan blue exclusion assay. Click here for file Click here for file [ http://www.biomedcentral.com/content/supplementary/1476-4598-9-30-
S2.PDF ] [ http://www.biomedcentral.com/content/supplementary/1476-4598-9-30-
S2.PDF ] [ http://www.biomedcentral.com/content/supplementary/1476-4598-9-30-
S1.PDF ] Authors’ contributions SK as principle investigator conceived the ideas, trained his research team
for various techniques including cell adhesion, migration, and invasion
assays, performed the initial studies described in the manuscript,
coordinated the experiments, and supervised the entire project in his
laboratory where all experiments were carried out. SK and SH were involved
in the design and execution of the experiments, analyzing data and
interpretation, and writing the manuscript. SH performed most of the
experiments described in various sections of the manuscript including
establishment and characterization of the stable PSAP-KD cells, transient
transfection studies, protein and gene expression assays, cell adhesion,
migration, and invasion assays, experiments on PSAP and Cer, and
determining cathepsin D expression and proteolytic activity. ND performed
the pilot studies to differentiate PSAP-KD cells from control transfectants. ZL
was responsible for routine cell culture and assisted SH on some of the
experiment on cell adhesion and immunoblotting. TL was partially involved
in protein expression analysis for integrins. JG was involved in cell culture
maintenance and performed part of the experiments on
immunoprecipitation. AP provided training and expertise on
immunofluorescence staining, microscopic data collection, and analysis. CV-G
was responsible for all statistic analysis All authors read and approved the 13. Koochekpour S, Zhang Y, Beroukhim R, Hsieh CL, Hofer MD, Zhau HE,
Hiraiwa M, Pattan DY, Ware JL, Luftig RB, Sandhoff K, Sawyers CL, Pienta KJ,
Rubin MA, Vessella RL, Sellers WR, Sartor O: Amplification and
overexpression of prosaposin in prostate cancer. Genes Chrom Cancer
2005, 44:351-364. 14. Koochekpour S, Sartor O, Lee TJ, Zieske A, Patten DY, Hiraiwa M,
Sandhoff K, Remmel N, Minokadeh A: Prosaptide TX14A stimulates
growth, migration, and invasion and activates the Raf-MEK-ERK-RSK-Elk-
1 signaling pathway in prostate cancer cells. Prostate 2004, 61:114-123. 15. Koochekpour S, Sartor O, Hiraiwa M, Lee TJ, Rayford W, Remmel N,
Sandhoff K, Minokadeh A, Patten DY: Saposin C stimulates growth and
invasion, activates p42/44 and SAPK/JNK signaling pathways of MAPK
and up regulatesuPA/uPAR expression in prostate cancer and stromal
cells. Asian J Androl 2005, 7:147-158. 16. Lee TJ, Sartor O, Luftig RB, Koochekpour S: Saposin C promotes survival
and prevents apoptosis via PI3K/Akt-dependent pathway in prostate
cancer cells. Mol Cancer 2004, 3:31-44. 17. Muant SJ, Handley J, Stower M, Reid N, Cussenotn O, Maitland NJ:
Coordinated changes in expression of cell adhesion molecules in
prostate cancer. Eur J Cancer 1997, 33:263-271. 18. References [ http://www.biomedcentral.com/content/supplementary/1476-4598-9-30-
S5.PDF ]
Additional file 6: PSAP down modulation decreased FAK binding to
Src and Paxillin phosphorylation in metastatic PCa cells. PSAP-KD
cells were incubated in their basal medium for 24 h and whole cell
lysates were extracted and subjected to immunoprecipitation for Src and
immunoblotting with anti-FAK or-paxillin antibody as described in the
Materials and Methods section. Similar data were obtained from three
independent experiments. Click here for file
[ http://www.biomedcentral.com/content/supplementary/1476-4598-9-30-
S6.PDF ] [ http://www.biomedcentral.com/content/supplementary/1476-4598-9-30-
S5.PDF ] 1. Kishimoto Y, Hiraiwa M, O’Brien JS: Saposins: Structure, function,
distribution and molecular genetics. J Lipid Res 1992, 33:1255-1267. Additional file 6: PSAP down modulation decreased FAK binding to
Src and Paxillin phosphorylation in metastatic PCa cells. PSAP-KD
cells were incubated in their basal medium for 24 h and whole cell
lysates were extracted and subjected to immunoprecipitation for Src and
immunoblotting with anti-FAK or-paxillin antibody as described in the
Materials and Methods section. Similar data were obtained from three
independent experiments. Click here for file 2. Koochekpour S: PSAP (Prosaposin (variant Gaucher disease and variant
metachromatic leukodystrophy)). Atlas Genet Cytogenet Oncol Haematol
2006http://AtlasGeneticsOncology.org/Genes/PSAPID42980ch10q22.html. 3. Schuette CG, Pierstorff B, Huettler S, Sandhoff K: Sphingolipid activator
proteins: proteins with complex functions in lipid degradation and skin
biogenesis. Glycobiol 2001, 11:81R-90R. 4. Gopalakrishnan MM, Grosch HW, Locatelli-Hoops S, Werth N, Smolenov E,
Nettersheim M, Sandhoff K, Hasilik A: Purified recombinant human
prosaposin forms oligomers that bind procathepsin D and affect its
autoactivation. Biochem J 2004, 383:507-515. [ http://www.biomedcentral.com/content/supplementary/1476-4598-9-30-
S6.PDF ] 5. Hiraiwa M, Martin BM, Kishimoto Y, Conner GE, Tsuj S, O’Brien JS:
Lysosomal proteolysis of prosaposin, the precursor of saposins
(aphingolipid activator proteins): its
mechanism and inhibition by ganglioside. Arch Biochem Biophys 1997,
341:17-24. Authors’ contributions Fornaro M, Tallini G, Bofetiado C, Bosari S, Languino LR: Down-regulation
of beta 1c integrin, an inhibitor of cell proliferation, in prostate
carcinoma. Am J Path 1996, 149:765-773. Achnowledgements We gratefully acknowledge Dr. Lucia R. Languino from University of
Massachusetts for providing antibodies against b1B and b1C integrin
isoforms. We thank Dr. J. L. ware from the Department of Pathology, Medical
College of Virginia for providing the p69, M12, and M2182 cell lines. We
thank Ms. Jonna Ellis and Ms. Heidi L. Davis at Stanley S. Scott Cancer
Center, LSUHSC for editing the manuscript. S.K. was funded by Stanley S. Scott Cancer Center (S.K), Louisiana Cancer Research Consortium (S.K),
Louisiana State University Health Sciences Center and Tulane University
Clinical and Translational Research Education and Commercialization Project
(CTRECP-070; to SK), NIH-NCRR; 1P20 RR021970 (A. Ochoa/S.K), and NCI 1R21
CA120625 (to S.K.). 6. Laurent-Matha V, Lucas A, Huttler S, Sandhoff K, Garcia M, Rochefort H:
Procathepsin D interacts with prosaposin in cancer cells but its
internalization is not mediated by LDL
receptor-related protein. Exp Cell Res 2002, 277:210-219. 7. O’Brien JS, Carson GS, Seo HC, Hiraiwa M, Kishimoto Y: Identification of
prosaposin as a neurotrophic factor. Proc Natl Acad Sci USA 1994,
91:9593-9596. 7. O’Brien JS, Carson GS, Seo HC, Hiraiwa M, Kishimoto Y: Identification of
prosaposin as a neurotrophic factor. Proc Natl Acad Sci USA 1994,
91:9593-9596. 8. O’Brien JS, Carson GS, Seo HC, Hiraiwa M, Weiler S, Tomich JM,
Barranger JA, Kahn M, Azuma N, Kishimoto Y: Identification of the
neurotryphic factor sequence of prosaposin. FASEB J 1995, 9:681-685. y
9. Elleder M, Jerabkova M, Befekadu A, Hřebíček M, Berná L, Ledvinová J,
Hůlková H, Rosewich H, Schymik N, Paton BC, Harzer K: Prosaposin
deficiency, a rale diagnosed, rapidly progressing, neonatal
neuroviscearal lipid storage disease. Report of a further patient. Neuropediatrics 2005, 36:171-180. Author details
1 1Stanley S. Scott Cancer Center, School of Medicine, Louisiana State
University Health Sciences Center, New Orleans, LA 70112, USA. 2School of
Public Health, School of Medicine, Louisiana State University Health Sciences
Center, New Orleans, LA 70112, USA. 3Departments of Pharmacology, School
of Medicine, Louisiana State University Health Sciences Center, New Orleans,
LA 70112, USA. 4Department of Biochemistry and Molecular Biology, School
of Medicine, Louisiana State University Health Sciences Center, New Orleans,
LA
0
2 USA 5D
f Mi
bi l
d I
l
S h
l
f 1Stanley S. Scott Cancer Center, School of Medicine, Louisiana State
University Health Sciences Center, New Orleans, LA 70112, USA. 2School of
Public Health, School of Medicine, Louisiana State University Health Sciences
Center, New Orleans, LA 70112, USA. 3Departments of Pharmacology, School
of Medicine, Louisiana State University Health Sciences Center, New Orleans,
LA 70112, USA. 4Department of Biochemistry and Molecular Biology, School
of Medicine, Louisiana State University Health Sciences Center, New Orleans,
LA 70112, USA. 5Department of Microbiology and Immunology, School of
Medicine, Louisiana State University Health Sciences Center, New Orleans, LA
70112, USA. 6Department of Urology, School of Medicine, Louisiana State
University Health Sciences Center, New Orleans, LA 70112, USA. 10. Hulková M, Cervenková M, Ledvinová J, Tochácková M, Hrebícek M,
Poupetová H, Befekadu A, Berná L, Paton BC, Harzer K, Böör A, Smíd F,
Elleder M: A novel mutation in the coding region of the prosaposin gene
leads to a complete deficiency of prosaposin and saposins, and is
associated with a complex sphingolipidosis dominated lactosylceramide
accumulation. Hum Mol Genet 2001, 9:927-940. LA 70112, USA. Department of Microbiology and Immunology, School of
Medicine, Louisiana State University Health Sciences Center, New Orleans, LA
70112, USA. 6Department of Urology, School of Medicine, Louisiana State
University Health Sciences Center, New Orleans, LA 70112, USA. 11. Tylki-Szymańska A, Czartoryska B, Vanier MT, Poorthuis BJ, Groener JA,
Ługowska A, Millat G, Vaccaro AM, Jurkiewicz E: Non-neuronopathic
Gaucher disease due to saposin C deficiency. Clin Genet 2007, 72:538-542. 12. Morales CR, Zhao Q, El-Alfy M, Suzuki K: Targeted disruption of the mouse
prosaposin gene affects the development of the prostate gland and
other male reproductive organs. J Androl 2000, 21:765-775. Additional file 4: Silver staining and western analysis of purified
rhPSAP. Recombinant PSAP proteins were subjected to SDS-PAGE using The
effect of the rhPSAP protein on prostate stromal cells was determined in
an in vitro migration assay. Briefly, cells were seeded in the upper
compartment of transwell filters in a basal medium supplemented with
0.1% BSA. The lower compartment was filled with 400 μl of either 5%
FBS (as a positive control) or basal media supplemented with 0.1% BSA
with or without rhPSAP at the indicated concentrations. After 24 h of
incubation, the cells were fixed and stained with Diff-Quick. Non-
migrated cells were removed by a cotton swap and the total cell
number per filter was counted. Each sample was assayed in
quadruplicates. Data represented the average of three independent
experiments ± SEM. Statistical significance (p < 0.05) between the control
and treatment groups was evaluated by one-way ANOVA test with
Bonferroni adjustment. [ http://www.biomedcentral.com/content/supplementary/1476-4598-9-30-
S1.PDF ] Click here for file Page 17 of 18 Page 17 of 18 Hu et al. Molecular Cancer 2010, 9:30
http://www.molecular-cancer.com/content/9/1/30 Hu et al. Molecular Cancer 2010, 9:30
http://www.molecular-cancer.com/content/9/1/30 Zhu Y, Conner GE: Intermolecular association of lysosomal protein
precursors during biosynthesis. J Biol Chem 1994, 269:3846-3851. 28. Meijer D, van Agthoven T, Bosma PT, Nooter K, Dorssers LC: Functional
screen for genes responsible for tamoxifen resistance in human breast
cancer cells. Mol Cancer Res 2006, 4:379-386. 29. Meijer D, Jansen MP, Look MP, Ruigrok-Ritstier K, van Staveren IL,
Sieuwerts AM, van Agthoven T, Foekens JA, Dorssers LC, Berns EM:
TSC22D1 and PSAP predict clinical outcome of tamoxifen treatment in
patients with recurrent breast cancer. Breast Cancer Res Treat 2009,
113:253-260. 30. McLean GW, Carragher NO, Avizienyte E, Evans J, Brunton VG, Frame MC:
The role of focal-adhesion kinase in cancer–a new therapeutic
opportunity. Nature Rev Cancer 2005, 5:505-515. 31. Mitra SK, Schlaepfer DD: Integrin-regulated FAK-Src signaling in normal
and cancer cells Curr Opin Cell Biol 2006 18:516-523 31. Mitra SK, Schlaepfer DD: Integrin-regulated FAK-Src s and cancer cells. Curr Opin Cell Biol 2006, 18:516-523 32. Hu W, Xu R, Zhang G, Jin J, Szulc ZM, Bielawski J, Hannun YA, Obeid LM,
Mao C: Golgi fragmentation is associated with ceramide-induced cellular
effects. Mol Biol Cell 2005, 16:1555-1567. 33. Fujita N, Suzuki K, Vanier MT, Popko B, Maeda N, Klein A, Henseler M,
Sandhoff K, Nakayasu H, Suzuki K: Targeted disruption of the mouse
sphingolipid activator protein gene: a complex phenotype, including
severe leukodystrophy and wide-spread storage of multiple
sphingolipids. Hum Mol Genet 1996, 5:711-725. p
g
p
,
34. Hannun YA, Obeid LM: Principles of bioactive lipid signalling: lessons
from sphingolipids. Nat Rev Mol Cell Biol 2008, 9:139-150. 35. Azuma N, Obrien JS, Moser HW, Kishimoto Y: Stimulation of acid
ceramidase activity by saposin D. Arch Biochem Biophys 1994, 322:354-357
36. Ogretmen B, Hannun YA: Biologically active sphingolipids in cancer
pathogenesis and treatment. Nat Rev Cancer 2004, 4:604-616. 34. Hannun YA, Obeid LM: Principles of bioactive lipid signalling: lessons
from sphingolipids. Nat Rev Mol Cell Biol 2008, 9:139-150. 35. Azuma N, Obrien JS, Moser HW, Kishimoto Y: Stimulation of acid
ceramidase activity by saposin D. Arch Biochem Biophys 1994, 322:354-357. 36. Ogretmen B, Hannun YA: Biologically active sphingolipids in cancer
pathogenesis and treatment. Nat Rev Cancer 2004, 4:604-616. 37. Campana WM, O’Brien JS, Hiraiwa M, Patton S: Secretion of prosaposin, a
multifunctional protein by breast cancer cells. Biochim Biophys Acta 1999,
1427:392-400. 38. Hu et al. Molecular Cancer 2010, 9:30
http://www.molecular-cancer.com/content/9/1/30 Hu et al. Molecular Cancer 2010, 9:30
http://www.molecular-cancer.com/content/9/1/30 Hu et al. Molecular Cancer 2010, 9:30
http://www.molecular-cancer.com/content/9/1/30 Page 18 of 18 carcinomas using orthotopic implantation in nude mice. Clin Cancer Res
1996, 2:1627-1636. 42. Pullikuth A, McKinnon E, Schaeffer HJ, Catling AD: The MEK1 scaffolding
protein MP1 regulates cell spreading by integrating PAK1 and Rho
Signals. Mol Cel Biol 2005, 25:5119-5133. doi:10.1186/1476-4598-9-30
Cite this article as: Hu et al.: Prosaposin down-modulation decreases
metastatic prostate cancer cell adhesion, migration, and invasion. Molecular Cancer 2010 9:30. carcinomas using orthotopic implantation in nude mice. Clin Cancer Res
1996, 2:1627-1636. 42. Pullikuth A, McKinnon E, Schaeffer HJ, Catling AD: The MEK1 scaffolding
protein MP1 regulates cell spreading by integrating PAK1 and Rho
Signals. Mol Cel Biol 2005, 25:5119-5133. doi:10.1186/1476-4598-9-30
Cite this article as: Hu et al.: Prosaposin down-modulation decreases
metastatic prostate cancer cell adhesion, migration, and invasion. Molecular Cancer 2010 9:30. 20. Moro L, Perlino E, Marra E, Languino LR, Greco M: Regulation of b1C and
b1A integrin expression in prostate carcinoma cells. J Biol Chem 2004,
279:1692-1702. carcinomas using orthotopic implantation in nude mice. Clin Cancer Res
1996, 2:1627-1636. 21. Grässel S, Hasilik A: Human cathepsin D precursor is associated with a 60
kDa
glycosylated polypeptide. Biochem Biophys Res Commun 1992,
182:276-282. doi:10.1186/1476-4598-9-30
Cite this article as: Hu et al.: Prosaposin down-modulation decreases
metastatic prostate cancer cell adhesion, migration, and invasion. Molecular Cancer 2010 9:30. 22. Konno S, Cherry JP, Mordente JA, Chapman JR, Choudhury MS, Mallouh C,
Tazaki H: Role of cathepsin D in prostatic cancer cell growth and its
regulation by brefeldin A. World J Urol 2001, 19:234-239. 23. Nomura T, Katunuma N: Involvement of cathepsins in the invasion,
metastasis and
proliferation of cancer cells J Med Invest 2005 52:1 9 proliferation of cancer cells. J Med Invest 2005, 52:1-9. 24. Sourla A, Doillon C, Koutsilieris M: Three-dimensional type I collagen gel
system containing MG-63 osteoblasts-like cells as a model for studying
local bone reaction caused by metastatic cancer cells. Anticancer Res
1996, 16:2773-2780. 25. Miyake D, Hara I, Eto H: Prediction of the extent of prostate cancer by
the combined use of systematic biopsy and serum level of cathepsin D. Int J Urol 2003, 10:196-200. 26. Godbold GD, Kyujeong A, Yeyeodu S, Lee LF, Ting JPY, Erickson AH:
Biosynthesis and intracellular targeting of the lysosomal aspartic
proteinase cathepsin D. Adv Exp Med Biol 1998, 436:153-162. 27. Competing interests The authors declare that they have no competing interests. 19. Haywood-Reid PL, Zipf DR, Springer WR, Haywood-Reid PL: Quantification
of integrin subunits on human prostatic cell lines-comparison of
nontumorigenic and tumorigenic lines. Prostate 1997, 31:1-8. 19. Haywood-Reid PL, Zipf DR, Springer WR, Haywood-Reid PL: Quantification
of integrin subunits on human prostatic cell lines-comparison of
nontumorigenic and tumorigenic lines. Prostate 1997, 31:1-8. Received: 24 November 2009
Accepted: 4 February 2010 Published: 4 February 2010 Received: 24 November 2009
Accepted: 4 February 2010 Published: 4 February 2010 Received: 24 November 2009
Accepted: 4 February 2010 Published: 4 February 2010 Hu et al. Molecular Cancer 2010, 9:30
http://www.molecular-cancer.com/content/9/1/30 Bouchal P, Roumeliotis T, Hrstka R, Nenutil R, Vojtesek B, Garbis SD:
Biomarker discovery in low-grade breast cancer using isobaric stable
isotope tags and two-dimensional liquid chromatography-tandem mass
spectrometry (iTRAQ-2DLC-MS/MS) based quantitative proteomic
analysis. J Proteome Res 2009, 8:362-373. Submit your next manuscript to BioMed Central
and take full advantage of:
• Convenient online submission
• Thorough peer review
• No space constraints or color figure charges
• Immediate publication on acceptance
• Inclusion in PubMed, CAS, Scopus and Google Scholar
• Research which is freely available for redistribution
Submit your manuscript at
www.biomedcentral.com/submit Submit your next manuscript to BioMed Central
and take full advantage of:
• Convenient online submission
• Thorough peer review
• No space constraints or color figure charges
• Immediate publication on acceptance
• Inclusion in PubMed, CAS, Scopus and Google Scholar
• Research which is freely available for redistribution
Submit your manuscript at
www.biomedcentral.com/submit Submit your next manuscript to BioMed Central
and take full advantage of: 39. Grønborg M, Kristiansen TZ, Iwahori A, Chang R, Reddy R, Sato N, Molina H,
Jensen ON, Hruban RH, Goggins MG, Maitra A, Pandey A: Biomarker
discovery from pancreatic cancer secretome using a differential
proteomic approach. Mol Cell Proteomics 2006, 5:157-171. 40. Kang SY, Halvorsen OJ, Gravdal K, Bhattacharya N, Lee JM, Liu NW,
Johnston BT, Johnston AB, Haukaas SA, Aamodt K, Yoo S, Akslen LA,
Watnick RS: Prosaposin inhibits tumor metastasis via paracrine and
endocrine stimulation of stromal p53 and Tsp-1. Proc Natl Acad Sci USA
2009, 106:12115-12120. 41. Pettaway CA, Pathak S, Greene G, Ramirez E, Wilson MR, Killion JJ, Fidler IJ:
Selection of highly metastatic variants of different human prostatic
|
https://openalex.org/W2169547031
|
https://dash.harvard.edu/bitstream/1/10419415/1/3360009.pdf
|
English
| null |
Human Papillomavirus Type 18 E6 and E7 Genes Integrate into Human Hepatoma Derived Cell Line Hep G2
|
PloS one
| 2,012
|
cc-by
| 7,335
|
Terms of Use This article was downloaded from Harvard University’s DASH repository, and is made available
under the terms and conditions applicable to Other Posted Material, as set forth at http://
nrs.harvard.edu/urn-3:HUL.InstRepos:dash.current.terms-of-use#LAA Permanent link http://nrs.harvard.edu/urn-3:HUL.InstRepos:10419415 Published Version
doi:10.1371/journal.pone.0037964 Published Version
doi:10.1371/journal.pone.0037964 Citation Ma, Tianzhong, Zhongjing Su, Ling Chen, Shuyan Liu, Ningxia Zhu, Lifeng Wen, Yan Yuan, et al. 2012. Human papillomavirus type 18 E6 and E7 genes integrate into human hepatoma derived
cell line Hep G2. PLoS ONE 7(5): e37964. Received February 10, 2012; Accepted April 26, 2012; Published May 24, 2012 Copyright: 2012 Ma et al. This is an open-access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted
use, distribution, and reproduction in any medium, provided the original author and source are credited. Funding: This study was supported by Natural Science Foundation of Guangdong Province, China (Grant No. 10151503102000034) and Research Foundation of
Shantou University Medical College (Grant No. LC0401). The funders had no role in study design, data collection and analysis, decision to publish, or preparation
of the manuscript. Competing Interests: The authors have declared that no competing interests exist. * E-mail: chenhb@stu.edu.cn Hep G2 cell line contained the integrated DNA of HPV 18. All the
HPV viruses live exclusively in the superficial tissues that cover our
body parts: the skin, the lining of the genital organs, urethra,
bladder, rectum, vocal cords, and esophagus. It remains unclear
whether there is an association between HPV infection and
hepatocellular carcinoma. Our interest in such a putative associ-
ation was the impetus that led us to investigate the expression of
E6 and E7 oncogenes in the Hep G2 cell line, and furthermore, to
see if such expression is required for the maintenance of the
proliferative and malignant phenotypes of Hep G2 cell line. Human Papillomavirus Type 18 E6 and E7 Genes
Integrate into Human Hepatoma Derived Cell Line Hep
G2 Tianzhong Ma1, Zhongjing Su1, Ling Chen2, Shuyan Liu1, Ningxia Zhu1, Lifeng Wen1, Yan Yuan1, Leili Lv1,
Xiancai Chen1, Jianmin Huang3, Haibin Chen1* 1 Department of Histology and Embryology, Shantou University Medical College, Guangdong Province, China, 2 Department of Biochemistry and Molecular Biology,
Shantou University Medical College, Guangdong Province, China, 3 Pediatric Endocrine Unit, Massachusetts General Hospital for Children and Harvard Medical School,
Boston, Massachusetts, United States of America Share Your Story The Harvard community has made this article openly available. Please share how this access benefits you. Submit a story . Accessibility PLoS ONE | www.plosone.org Abstract Background and Objectives: Human papillomaviruses have been linked causally to some human cancers such as cervical
carcinoma, but there is very little research addressing the effect of HPV infection on human liver cells. We chose the human
hepatoma derived cell line Hep G2 to investigate whether HPV gene integration took place in liver cells as well. Methods: We applied PCR to detect the possible integration of HPV genes in Hep G2 cells. We also investigated the
expression of the integrated E6 and E7 genes by using RT-PCR and Western blotting. Then, we silenced E6 and E7
expression and checked the cell proliferation and apoptosis in Hep G2 cells. Furthermore, we analyzed the potential genes
involved in cell cycle and apoptosis regulatory pathways. Finally, we used in situ hybridization to detect HPV 16/18 in
hepatocellular carcinoma samples. Results: Hep G2 cell line contains integrated HPV 18 DNA, leading to the expression of the E6 and E7 oncogenic proteins. Knockdown of the E7 and E6 genes expression reduced cell proliferation, caused the cell cycle arrest at the S phase, and
increased apoptosis. The human cell cycle and apoptosis real-time PCR arrays analysis demonstrated E6 and E7-mediated
regulation of some genes such as Cyclin H, UBA1, E2F4, p53, p107, FASLG, NOL3 and CASP14. HPV16/18 was found in only
9% (9/100) of patients with hepatocellular carcinoma. Conclusion: Our investigations showed that HPV 18 E6 and E7 genes can be integrated into the Hep G2, and we observed
a low prevalence of HPV 16/18 in hepatocellular carcinoma samples. However, the precise risk of HPV as causative agent of
hepatocellular carcinoma needs further study. Citation: Ma T, Su Z, Chen L, Liu S, Zhu N, et al. (2012) Human Papillomavirus Type 18 E6 and E7 Genes Integrate into Human Hepatoma Derived Cell Line Hep
G2. PLoS ONE 7(5): e37964. doi:10.1371/journal.pone.0037964
Editor: Syed A. Aziz, Health Canada, Canada Citation: Ma T, Su Z, Chen L, Liu S, Zhu N, et al. (2012) Human Papillomavirus Type 18 E6 and E7 Genes Integrate into Human Hepatoma Derived Cell Line Hep
G2. PLoS ONE 7(5): e37964. doi:10.1371/journal.pone.0037964 Editor: Syed A. Aziz, Health Canada, Canada Received February 10, 2012; Accepted April 26, 2012; Published May 24, 2012 Received February 10, 2012; Accepted April 26, 2012; Published May 24, 2012 Knockdown of HPV 18 E7 and E6 genes inhibited cell
growth Knockdown of HPV 18 E7 and E6 genes inhibited cell
growth Inhibition of both E6 and E7 genes expression by HPV 18
E7 siRNA We designed three siRNAs targeting E7 gene and screened for
more effective siRNAs using RT-PCR assay. We found that two of
three siRNAs (siRNA E7-63 and siRNA E7-112) were more
effective (Fig. 3A). The level of E7 mRNA in Hep G2 cells
transfected
with
siRNA
E7-63
and
E7-112
decreased
to
14.51%62.06% and 35.89%61.00% respectively compared with
that of the NC E7 (negative control) transfected cells. The level of
E6 mRNA in Hep G2 cells transfected with siRNA E7-63 and E7-
112 decreased to 36.50%610.12% and 24.50%67.32% re-
spectively compared with that of the NC E7 (negative control)
transfected cells (p,0.01, Fig. 3B). To overcome limitations of
RNAi experiments, such as off-target effects, we mixed siRNA E7-
63 and E7-112 to obtain our siRNA pools. The expression of E7
and E6 at mRNA and protein levels was assessed at different times
after transfection. There were significant decreases in E6 and E7
mRNA levels at 24 and 48 hours after siRNA treatment when Real-time PCR array analysis of shRNA E7 induced
apoptosis and cell cycle gene transcriptional profiles Results compared with that of control siRNA treated cells (Fig. 3C). At
24 h and 48 h after transfections, the level of E7 mRNA in Hep
G2
cells
transfected
with
siRNA
E7
pools
decreased
to
57.93%60.31% and 8.79%60.42% respectively compared with
that of the NC E7 transfected cells. At 24 h and 48 h after
transfections, the level of E6 mRNA in Hep G2 cells transfected
with
siRNA
E7
pools
decreased
to
68.14%60.53%
and
0.63%60.02% respectively compared with that of the NC E7
transfected cells (p,0.01, Fig. 3D). Introduction Epidemiological studies have shown that HPV infection is the
main etiological factor for cervical cancer [1] and high-risk HPV
type viral DNAs are frequently integrated into the host cell
genome in HPV-related cervical carcinomas [2]. This integration
has been associated with dysregulation of E6 and E7 viral genes
expression, which accounts for the major oncogenic activity of the
HPV DNA [3]. Expression of these genes can lead to immortal-
ization of keratinocytes, the natural host cells of HPV [4]. However, little information is available about the integration of
HPV into human liver cells. The human hepatoma derived cell line Hep G2 was derived
from biopsies taken during extended lobectomy of a 15-year-old
Caucasian male from Argentina [5]. This cell line has been used in
many laboratories around the world and we discovered that the May 2012 | Volume 7 | Issue 5 | e37964 1 PLoS ONE | www.plosone.org HPV 18 Integrated into Hep G2 Cells Immunohistochemistry revealed that the Hep G2 cell line
was typical of liver cells In order to characterize the Hep G2 cell line more explicitly, we
used the anti-human hepatocyte antibody to verify the character-
istics of the hepatoma cells. Anti-human hepatocyte immunohis-
tochemical evaluation of Hep G2 with the hepatocyte-specific
marker confirmed that the Hep G2 cell line was positive in liver
cell antigens (Fig. 1A), but the HeLa cells were negative (Fig. 1B). Hep G2 cells with integrated HPV 18 DNA expressed E6
and E7 mRNAs and proteins Hep G2 cells with integrated HPV 18 DNA expressed E6
and E7 mRNAs and proteins To determine whether the inhibition of HPV 18 E7 and E6
genes was sufficient to suppress the proliferation of Hep G2 cells,
cells were assessed using EdU assays for analyzing the S-phase
fraction (SPF). After transfections of Hep G2 cells with E7-siRNA,
a time-dependent reduction of cell S-phase fraction was observed
at 48 h, 60 h and 72 h (Fig. 4A). At 48 h, 60 h and 72 h following
E7-siRNA transfections, the percents of S-phase cells were
9.39%63.55%, 17.29%65.85% and 30.87%64.26% compared
with 18.74%66.66%, 24.03%65.35% and 41.97%68.73% in
NC-siRNA transfected cells (P,0.05, Fig. 4B). An amplified fragment of 847 bp was present in both the Hep
G2 and EC109 cells (HPV 18 positive) while it was absent in K562
cells (HPV 18 negative), as was expected (Fig. 2A). Transcription of the HPV 18 E6 and E7 genes in the Hep G2
and EC109 cell lines were evaluated by RT-PCR. The results
showed that the expected fragments of E6 (196 bp) and E7
(332 bp) were present in both Hep G2 and EC109 cells, but not in
K562 cells (Fig. 2B). Western blot analysis of cell extracts was also
carried
out
to
determine
whether
mRNA
expression
was
correlated with translation of the gene products. Again, a specific
protein band (18 kD) of E6 and E7 was observed in both Hep G2
and EC109 cells, but not in K562 cells (Fig. 2C). b-actin was used
as an internal control. Transfection with E7-siRNA induced apoptosis To assess the ratio of cell death, the Annexin V apoptosis assay
was performed at 24 h, 48 h and 72 h after siRNA transfections
and the stained cells were analyzed by flow cytometry in the
experimental and negative control groups. As shown in Fig. 5A, B,
C and D, 7.26%60.29%, 22.03%60.23% and 19.20%60.78%
in siRNA E7-transfected Hep G2 cells after 24 h, 48 h and 72 h
underwent total apoptosis compared with only 5.25%60.76% in
NC-E7 transfected cells. The fourth and first quadrant cells were
counted and expressed as the percentages of apoptotic cells. Therefore, knockdown of E7 resulted in about a 4-fold increase in
apoptosis (P,0.01, Fig. 5E). Inhibition of both E6 and E7 genes expression by HPV 18
E7 siRNA Real-time PCR array analysis of shRNA E7 induced
apoptosis and cell cycle gene transcriptional profiles In this study, we compared the gene expression patterns of cell
cycle and apoptosis signaling pathways in Hep G2 cells transfected
with shRNA E7 or negative control shRNA at 48 h after
transfections. A total of 40 out of 168 genes in PCR Array
showed differential expressions when cells transfected with shRNA
E7 and NC-shRNA were compared. Data analysis indicated that
among the cell cycle PCR array, 7 genes were up-regulated and 12 Figure 1. Immunohistochemistry demonstrated that the Hep G2 cell line was of characteristics typical of liver cells. Immunohistochemical evaluation of Hep G2 cells with anti-human hepatocyte antibody indicated that the Hep G2 cell cytoplasm (A), but not
HeLa cells (B), exhibited typical hepatocyte antigens. doi:10.1371/journal.pone.0037964.g001 Figure 1. Immunohistochemistry demonstrated that the Hep G2 cell line was of characteristics typical of liver cells. Immunohistochemical evaluation of Hep G2 cells with anti-human hepatocyte antibody indicated that the Hep G2 cell cytoplasm (A), but not
HeLa cells (B), exhibited typical hepatocyte antigens. doi:10.1371/journal.pone.0037964.g001 May 2012 | Volume 7 | Issue 5 | e37964 PLoS ONE | www.plosone.org 2 HPV 18 Integrated into Hep G2 Cells Figure 2. Hep G2 cells with integrated HPV 18 DNA expressed E6 and E7 mRNAs and proteins. (A) PCR amplification of HPV 18 E6E7 gene
was assessed in samples from Hep G2, EC109, and K562 cells. Then an amplified fragment of 847 bp was present in both Hep G2 and EC109 cells (HPV
18 positive), but absent in K562 cells (HPV 18 negative). (B) The expression of HPV 18 E6 and E7 mRNA was detected by RT-PCR. The expected
fragments of E6 (196 bp) and E7 (332 bp) were present in both Hep G2 and EC109 cells, but not in K562 cells. (C) Western blotting showed the
expression of the HPV 18 E6 and E7 proteins. EC109 and K562 were used as controls. b-actin was used as an internal control. doi:10.1371/journal.pone.0037964.g002 Figure 2. Hep G2 cells with integrated HPV 18 DNA expressed E6 and E7 mRNAs and proteins. (A) PCR amplification of HPV 18 E6E7 gene
was assessed in samples from Hep G2, EC109, and K562 cells. Then an amplified fragment of 847 bp was present in both Hep G2 and EC109 cells (HPV
18 positive), but absent in K562 cells (HPV 18 negative). (B) The expression of HPV 18 E6 and E7 mRNA was detected by RT-PCR. Real-time PCR array analysis of shRNA E7 induced
apoptosis and cell cycle gene transcriptional profiles The expected
fragments of E6 (196 bp) and E7 (332 bp) were present in both Hep G2 and EC109 cells, but not in K562 cells. (C) Western blotting showed the
expression of the HPV 18 E6 and E7 proteins. EC109 and K562 were used as controls. b-actin was used as an internal control. doi:10.1371/journal.pone.0037964.g002 genes were down-regulated after E7 gene was silenced (Table 1). Cyclin H, CDK2, E2F4, p53, Cullin3, MKI67, UBA1, p107 and
others were affected. After the knockdown of E7, in Apoptosis
PCR Array, 16 genes were up-regulated and 5 genes were down-
regulated. Affected genes were BAK1, BCL2L10, CARD8,
CASP2, CASP8, CASP14, FASLG, NOL3, CD27 and other
genes (Table 2). In situ hybridization on tissue microarrays for HPV 16/18
The HPV signals observed in the punctate nuclear staining of
Hep G2 cells (Fig. 6A) and HeLa cells (positive control)
demonstrated an integrated form of HPV DNA (Fig. 6B). HPV
signals in the HPV-ISH-positive case were of diffuse staining in
tumor cells (Fig. 6C). No signal was found in hepatoma carcinoma In situ hybridization on tissue microarrays for HPV 16/18
The HPV signals observed in the punctate nuclear staining of
Hep G2 cells (Fig. 6A) and HeLa cells (positive control)
demonstrated an integrated form of HPV DNA (Fig. 6B). HPV
signals in the HPV-ISH-positive case were of diffuse staining in
tumor cells (Fig. 6C). No signal was found in hepatoma carcinoma Figure 3. HPV18 E6 and E7 mRNA and relative protein expressions in Hep G2 transfected with E7-siRNA. (A) Effective siRNAs were
screened using RT-PCR assays from three siRNAs targeting the E7 gene and two of three siRNAs (siRNA E7-63 and siRNA E7-63 112) were more
effective. (B) The level of E7 mRNA in Hep G2 cells transfected with siRNA E7-63 and siRNA E7-112 decreased to 14.51%62.06% and 35.89%61.00%
respectively compared with that of the NC siRNA E7 (Negative control) transfected cells (p,0.01). (C) The expression of E7 and E6 mRNA was assessed
through RT-PCR at different time after transfection. (D) At 24 h and 48 h after transfection, the level of E7 mRNA in Hep G2 cells transfected with
siRNA E7 decreased to 57.93%60.31% and 8.79%60.42% respectively compared with that of the NC siRNA E7 transfected cells (p,0.01). doi:10.1371/journal.pone.0037964.g003 Figure 3. HPV18 E6 and E7 mRNA and relative protein expressions in Hep G2 transfected with E7-siRNA. genes were down-regulated after E7 gene was silenced (Table 1).
Cyclin H, CDK2, E2F4, p53, Cullin3, MKI67, UBA1, p107 and
others were affected. After the knockdown of E7, in Apoptosis
PCR Array, 16 genes were up-regulated and 5 genes were down-
regulated. Affected genes were BAK1, BCL2L10, CARD8,
CASP2, CASP8, CASP14, FASLG, NOL3, CD27 and other
genes (Table 2). Discussion Chronic viral and bacterial infections and nematode infestations
have been estimated to be associated with approximately one out
of five human cancers worldwide [6]. The HPV genome has been
detected in esophagus, lung, colon, ovary, breast, and prostate. However, there is little research addressing the ability of HPV to
infect hepatic cells and the effect it would have upon these cells. Scinicariello F. et al detected human papillomavirus 16 and 18 in
primary hepatocellular carcinoma and found that tumors in 1 of
16 samples showed HPV 16 positive and 2 of 16 samples showed
HPV 18 positive [7].HPV is a simple DNA virus that has evolved
to escape immune attack by a combination of stealth and
interference. HPV life cycle has rare by-products to trigger
immune response and E6E7 oncoproteins also target cytokine
expression to modulate cell proliferation and interferon responses,
contributing to immune evasion [8]. When the immune system is
weakened, HPV is likely carried to the liver of the patients from
a distant site via blood circulation. This may explain why HPV
viremia occurs more frequently than previously believed. Our research have shown that the hepatoma derived cell line
Hep G2 was positive for HPV 18 and this infection led to the
integration of viral DNA into the host genome and expression of
the E6 and E7 oncogenes. For a long time, there has been
controversies about the precise origin of Hep G2 cell line. The
HeLa cells do contaminate a number of established cell lines. A
major issue here was to determine if this line was a cervical
adenocarcinoma, a carcinoma of the cervix metastatic to the liver,
a primary hepatoma cell line, or a contaminated pailloma positive
HeLa cell line. Anti-hepatocyte immunohistochemistry convinc-
ingly demonstrated that Hep G2 cell line was of typical
characteristics of liver cells. This evidence suggested a possible
association between HPV 18 infection and hepatocellular carci-
noma (HCC). As a basic biological characteristic, HPV virus
causes proliferation of infected host cells. This is often manifested
clinically in skin or mucous membrane warts. It is unknown
whether HPV plays a role in HCC carcinogenesis, or whether In recent years, most researchers have reported that down-
regulation of HPV18 E6 and E7 expression by RNAi led to
retarded growth of HPV18-positive cervical cancer cell lines [15]
and proposed that siRNA-induced E6 and E7 simultaneously
silencing had a major therapeutic potential. Real-time PCR array analysis of shRNA E7 induced
apoptosis and cell cycle gene transcriptional profiles (A) Effective siRNAs were
screened using RT-PCR assays from three siRNAs targeting the E7 gene and two of three siRNAs (siRNA E7-63 and siRNA E7-63 112) were more
effective. (B) The level of E7 mRNA in Hep G2 cells transfected with siRNA E7-63 and siRNA E7-112 decreased to 14.51%62.06% and 35.89%61.00%
respectively compared with that of the NC siRNA E7 (Negative control) transfected cells (p,0.01). (C) The expression of E7 and E6 mRNA was assessed
through RT-PCR at different time after transfection. (D) At 24 h and 48 h after transfection, the level of E7 mRNA in Hep G2 cells transfected with
siRNA E7 decreased to 57.93%60.31% and 8.79%60.42% respectively compared with that of the NC siRNA E7 transfected cells (p,0.01). doi:10.1371/journal.pone.0037964.g003 PLoS ONE | www.plosone.org May 2012 | Volume 7 | Issue 5 | e37964 3 3 HPV 18 Integrated into Hep G2 Cells Figure 4. Inhibition of HPV 18 E7 gene inhibited cell growth in Hep G2. (A) Hep G2 cells transfected with E7-siRNA or control siRNA were
evaluated in EdU assays at 0 h, 12 h, 24 h, 36 h, 48 h, 60 h and 72 h after transfection. After transfection of Hep G2 cells with E7-siRNA, a time-
dependent reduction of cell proliferation was observed at 48 h, 60 h and 72 h. (B) At 48 h, 60 h and 72 h following E7-siRNA transfection, the
percents of S-phase cells were 9.39%63.55%, 17.29%65.85% and 30.87%64.26% compared with 18.74%66.66%, 24.03%65.35% and
41.97%68.73% in NC-siRNA transfected cells (p,0.05). doi:10.1371/journal.pone.0037964.g004 Figure 4. Inhibition of HPV 18 E7 gene inhibited cell growth in Hep G2. (A) Hep G2 cells transfected with E7-siRNA or control siRNA were
evaluated in EdU assays at 0 h, 12 h, 24 h, 36 h, 48 h, 60 h and 72 h after transfection. After transfection of Hep G2 cells with E7-siRNA, a time-
dependent reduction of cell proliferation was observed at 48 h, 60 h and 72 h. (B) At 48 h, 60 h and 72 h following E7-siRNA transfection, the
percents of S-phase cells were 9.39%63.55%, 17.29%65.85% and 30.87%64.26% compared with 18.74%66.66%, 24.03%65.35% and
41.97%68.73% in NC-siRNA transfected cells (p,0.05). doi:10.1371/journal.pone.0037964.g004 HPV and HBV co-infection may have a synergistic effect. Real-time PCR array analysis of shRNA E7 induced
apoptosis and cell cycle gene transcriptional profiles Lee et al
found that HPV 16 E6 oncoprotein and HBV X protein could
synergistically induce the activation of the AP1 site of E element in
the enhancer I (EnI) to increase the transcription activity of the
HBV and other oncogenes containing an AP1 site in the promoter
in human liver cell [9]. Both HBV and HPV are DNA viruses, and
share, along with oncogenic retroviruses, a replication strategy
involving reverse transcriptase and a characteristic life cycle that
includes integration of viral DNA into the host genome [10]. Viral
integration into or adjacent to the human telomerase reverse
transcriptase (hTERT) gene, for example, may lead to hTERT
telomerase activation, cellular immortalization, and a predisposi-
tion to carcinogenesis in liver and cervical cancers [11]. The HPV
oncoproteins E6 and E7 are the primary viral factors responsible
for initiation and progression of cancer. A primary target of E7 is
the retinoblastoma (Rb) family of proteins that control the activity
of E2F transcription factors, which are key regulators of S phase
genes. Inactivation of Rb is important for the differentiation-
dependent productive viral lifecycle and for tumor progression.Th-
ese high-risk E6 and E7 oncoproteins then lead to cell proliferation
through their association with PDZ domain proteins and, through
p53 degradation, prevent the normal repair of chance mutations
in the cellular genome [12]. E7 oncoprotein also alters cell cycle
control through interactions with histone deacetylases, cyclins and
cyclin-dependent kinase inhibitors. These factors are crucial
regulators of immune evasion, cell cycle progression, telomere
maintenance, apoptosis and chromosomal stability. HPV onco-
proteins modify these activities to play an important role in the
basic mechanisms of oncogenesis [13,14] cells of this HPV- negative specimen. (Fig. 6D). With regard to
tumor grade, HPV16/18 was positive in 2 of 28 patients (7.1%)
with grade I carcinoma, in 6 of 65 patients (8.8%) with grade II
carcinoma, and in 1 of 7 patients (14.2%) with grade III carcinoma
(Table 3). There was no association between carcinoma grade and
HPV infection (P = 0.719). PLoS ONE | www.plosone.org Discussion When we silenced the
E7 and E6 by RNAi, Hep G2 cells in the S-phase reduced and
apoptosis increased. We found that silencing HPV E7 and E6
could affect the expression of genes involved in cell-cycle
regulation, such as Cyclin H, CDK2, E2F4, p53, Cullin3,
MKI67, UBA1, p107 and so on. In our study, the knockdown
of HPV E7 could affect the genes of cell apoptosis including
BAK1, BCL2L10, CARD8, CASP2, CASP8, CASP14, FASLG,
NOL3 and CD27. Therefore, integration of the HPV E7 and E6
genes into Hep G2 cells may play an important role in the
immortalization and carcinogenesis of this hepatoma cell line. May 2012 | Volume 7 | Issue 5 | e37964 PLoS ONE | www.plosone.org 4 HPV 18 Integrated into Hep G2 Cells Figure 5. Transfection with E7-siRNA induced apoptosis. (A, B, C and D) The percent of apoptotic Hep G2 cells was measured using the
Annexin V assay at 24 h, 48 h and 72 h after transfection, and then the stained cells were analyzed through flow cytometry. (E) The results showed
that 7.26%60.29%, 22.03%60.23% and 19.20%60.78% in siRNA E7 transfected Hep G2 cells after 24 h, 48 h and 72 h underwent total apoptosis
compared with only 5.25%60.76% in NC siRNA E7 transfected cells (p,0.05). doi:10.1371/journal.pone.0037964.g005 Figure 5. Transfection with E7-siRNA induced apoptosis. (A, B, C and D) The percent of apoptotic Hep G2 cells was meas
Annexin V assay at 24 h, 48 h and 72 h after transfection, and then the stained cells were analyzed through flow cytometry. (E) The
that 7.26%60.29%, 22.03%60.23% and 19.20%60.78% in siRNA E7 transfected Hep G2 cells after 24 h, 48 h and 72 h underwent t
compared with only 5.25%60.76% in NC siRNA E7 transfected cells (p,0.05). doi:10.1371/journal.pone.0037964.g005 Figure 5. Transfection with E7-siRNA induced apoptosis. (A, B, C and D) The percent of apoptotic Hep G2 cells was measured using the
Annexin V assay at 24 h, 48 h and 72 h after transfection, and then the stained cells were analyzed through flow cytometry. (E) The results showed
that 7.26%60.29%, 22.03%60.23% and 19.20%60.78% in siRNA E7 transfected Hep G2 cells after 24 h, 48 h and 72 h underwent total apoptosis
compared with only 5.25%60.76% in NC siRNA E7 transfected cells (p,0.05). doi:10.1371/journal.pone.0037964.g005 leading to the expression of the E6 and E7 oncogenic proteins. PLoS ONE | www.plosone.org Discussion Furthermore our results demonstrated that RNAi inhibited the
expression of E7 as well as E6, and down-regulation of these viral
oncogenes initiated the process of growth arrest and induced
apoptosis in Hep G2 cells. Although 100 tumor samples were
analyzed and only 9% of cases tested positive for HPV 16/18 in
hepatocellular carcinoma, further research on the carcinogenic
potential of HPV infection in hepatic cells is warranted. These
findings
should
encourage
clinicians
and
medical
professionals to consider the effect that HPV infection has on
the development, investigation, and treatment of HCC. It is
standard practice to inoculate for HBV to prevent the de-
velopment
of
chronic
HBV
infection
and
HCC. Perhaps
consideration should also be given to the administration of HPV
vaccine in adult immunization, or to include HPV testing as part
of routine outpatient care. Patients with HCC who are HBV
positive undergo liver transplantation require antiviral therapy. There is a benefit from viral suppression in the form of reducing
HBV recurrence after transplantation [16]. HPV status, on the
other hand, has never played a role in the consideration of liver
resection versus transplantation in patients with HCC. Perhaps
this stance should be investigated further. May 2012 | Volume 7 | Issue 5 | e37964 Cell lines and cell culture The human hepatoma derived cell line Hep G2 [5] and human
esophageal carcinoma cell line EC109 cell lines were obtained
from the Committee of Type Culture Collection of the Chinese
Academy of Sciences (Shanghai, China). The human erythroleu- In conclusion, our research showed that HPV 18 E6 and E7
genes can be integrated into the hepatoma cell line Hep G2, May 2012 | Volume 7 | Issue 5 | e37964 May 2012 | Volume 7 | Issue 5 | e37964 5 HPV 18 Integrated into Hep G2 Cells Table 1. mRNA expression of E7-related genes in Hep G2 cells after the silencing of HPV 18 E7 through RT2 ProfilerTM Human Cell
Cycle PCR Array. Gene names
Genbank accession no. Description
Folds up- or down-regulation
ABL1
NM_005157
C-abl oncogene 1, receptor tyrosine kinase
22.20
ANAPC4
NM_013367
Anaphase promoting complex subunit 4
22.13
ATR
NM_001184
Ataxia telangiectasia and Rad3 related
22.31
CCNH
NM_001239
Cyclin H
22.80
CDK2
NM_001798
Cyclin-dependent kinase 2
2.12
CDKN2B
NM_004936
Cyclin-dependent kinase inhibitor 2B (p15, inhibits CDK4)
2.01
CKS1B
NM_001826
CDC28 protein kinase regulatory subunit 1B
22.09
CUL3
NM_003590
Cullin 3
22.83
E2F4
NM_001950
E2F transcription factor 4, p107/p130-binding
3.27
MKI67
NM_002417
Antigen identified by monoclonal antibody Ki-67
22.48
MNAT1
NM_002431
Menage a trois homolog 1, cyclin H assembly factor
(Xenopus laevis)
2.04
NBN
NM_002485
Nibrin
22.04
PCNA
NM_182649
Proliferating cell nuclear antigen
22.08
RAD1
NM_002853
RAD1 homolog (S. pombe)
22.29
RBL1
NM_002895
Retinoblastoma-like 1 (p107)
2.83
RPA3
NM_002947
Replication protein A3, 14 kDa
22.08
SERTAD1
NM_013376
SERTA domain containing 1
22.20
TP53
NM_000546
Tumor protein p53
3.20
UBA1
NM_003334
Ubiquitin-like modifier activating enzyme 1
2.44
In 84 E7-related genes, 19 genes mRNA transcript were altered markedly compared with the negative control. Standards of eligibility: folds up- or down-regulation .2.0
and p,0.05; n = 3. doi:10.1371/journal.pone.0037964.t001 Table 1. mRNA expression of E7-related genes in Hep G2 cells after the silencing of HPV 18 E7 through RT2 ProfilerTM Human Cell
Cycle PCR Array. In 84 E7-related genes, 19 genes mRNA transcript were altered markedly compared with the negative control. Standards of eligibility: folds up- or down-regulation .2.0
and p,0.05; n = 3. doi:10.1371/journal.pone.0037964.t001 TAGTTTACAACCCGTGCCCTCC39). An initial denaturation
of 5 min at 94uC was followed by 30 cycles of amplification (30 s
at 94uC, 30 s at 62uC and 90 s at 72uC), with a final extension of
10 min at 72uC. ohistochemistry for characterizing Hep G2 cells Immunohistochemistry for characterizing Hep G2 cells
Cells grown on slides were fixed in paraformaldehyde for
15 min and penetrated in Phosphate buffered saline (PBS)
containing 0.3% Triton X-100 for 30 min, then followed by
incubation with 3% hydrogen peroxide for 25 min. Washed in
PBS, slides were incubated with 1.5% normal horse serum in PBS
for 30 min with the purpose of blocking nonspecific binding. After
being rinsed in PBS for three times, slides were incubated
overnight at 4uC with mouse anti-human hepatocyte antibody
(1:100 dilution) (Invitrogen, USA). Slides were then incubated for
30 min with biotinylated anti-mouse antibody (1:200 dilution)
(VECTOR Lab, USA) followed by 30 min in Avidin: Biotinylated
Enzyme
Complex
at
room
temperature. Subsequently
the
antibody-antigen binding sites were identified by applying sub-
strate diaminobenzidine. Finally, slides were dehydrated, and
mounted. HeLa cells were served as controls. The total RNA of cells was extracted using Trizol reagent
according to the manufacturer’s protocol (TaKaRa, Japan). Complementary DNA (cDNA) of cells was synthesized from
1 mg of total RNA using the ImProm-IITM Reverse Transcription
System (Promega, USA). PCR was performed using Pfu DNA
polymerase and the following primers for HPV 18 E6 (forward:
59AAGATTTATTTGTGGTGT 39, reverse: 59GCTGGATT-
CAACGGTTTC 39) and for HPV 18 E7 (forward: 59TATG-
CATGGACCTAAGGC 39, reverse: 59CAGCCATTGTTGCT-
TACT 39). An initial denaturation of 5 min at 94uC was followed
by 30 cycles of amplification (30 s at 94uC, 30 s at 57uC and 30 s
at 72uC), with a final extension of 10 min at 72uC. Western blot analysis Protein samples from the cells were extracted with the
ProteoExtract Subcellular Proteome Extraction Kit (Calbiochem,
USA) according to the manufacturer’s instructions. The protein
samples were boiled for 5 min in 16SDS sample buffer, and 50 mg
of the proteins were loaded in wells of a 15% sodium dodecyl
sulfate-polyacrylamide electrophoresis (SDS-PAGE) gel. Following
separation, the proteins were transferred to a nitrocellulose Cell lines and cell culture The products of this PCR assay were sequenced
using bi-directional sequencing on an Applied Biosystems 3700
automated DNA sequencing machine in the GeneCore Bio-
technologies. kemia line K562 [17] and human cervical carcinoma cell line
HeLa [18] were obtained from the Center for Molecular Biology
of Shantou University Medical College (Shantou, China). The cell
lines were maintained in Dulbecco’s modified Eagle’s medium
(DMEM) supplemented with 10% fetal bovine serum (FBS)
(Invitrogen Gibco, USA) and grown at 37uC in a humidified
atmosphere with 5% CO2. PLoS ONE | www.plosone.org Polymerase chain reactions (PCR) PCR amplification of HPV 18 E6 and E7 genes was performed
using Pfu DNA polymerase (Generay, Shanghai, China) and the
following HPV 18 E6 and E7 primers (forward 59GACACTAG-
TACTATGGCGCGCTTTGA39
and
reverse
59AGTAC- PLoS ONE | www.plosone.org May 2012 | Volume 7 | Issue 5 | e37964 May 2012 | Volume 7 | Issue 5 | e37964 PLoS ONE | www.plosone.org 6 HPV 18 Integrated into Hep G2 Cells Table 2. mRNA expression of E7-related genes in Hep G2 cell after the silencing of HPV 18 E7 in RT2 ProfilerTM Human Apoptosis
PCR Array. Gene names
Genbank accession no. Description
Folds up- or down-regulation
BAK1
NM_001188
BCL2-antagonist/killer 1
2.48
BAX
NM_004324
BCL2-associated X protein
2.07
BCL2
NM_000633
B-cell CLL/lymphoma 2
22.30
BCL2L10
NM_020396
BCL2-like 10 (apoptosis facilitator)
23.05
BIK
NM_001197
BCL2-interacting killer (apoptosis-inducing)
2.05
BIRC3
NM_001165
Baculoviral IAP repeat-containing 3
22.14
CARD6
NM_032587
Caspase recruitment domain family, member 6
2.05
CARD8
NM_014959
Caspase recruitment domain family, member 8
2.13
CASP2
NM_032982
Caspase 2, apoptosis-related cysteine peptidase
2.01
CASP8
NM_001228
Caspase 8, apoptosis-related cysteine peptidase
2.03
CASP14
NM_012114
Caspase 14, apoptosis-related cysteine peptidase
3.71
CIDEA
NM_001279
Cell death-inducing DFFA-like effector a
2.23
CIDEB
NM_014430
Cell death-inducing DFFA-like effector b
2.12
FASLG
NM_000639
Fas ligand (TNF superfamily, member 6)
4.00
LTBR
NM_002342
Lymphotoxin beta receptor (TNFR superfamily,
member 3)
2.04
NOL3
NM_003946
Nucleolar protein 3 (apoptosis repressor with
CARD domain)
23.92
TNFRSF10A
NM_003844
Tumor necrosis factor receptor
superfamily, member 10a
22.06
CD27
NM_001242
CD27 molecule
2.29
TRAF2
NM_021138
TNF receptor-associated factor 2
2.09
TRAF3
NM_003300
TNF receptor-associated factor 3
2.03
TRAF4
NM_004295
TNF receptor-associated factor 4
2.25
In 84 E7-related genes, 21 genes mRNA transcript were altered markedly compared with the negative control. Standards of eligibility: folds up- or down-regulation .2.0
and p,0.05; n = 3. doi:10.1371/journal.pone.0037964.t002 ression of E7-related genes in Hep G2 cell after the silencing of HPV 18 E7 in RT2 ProfilerTM Human Apop Table 2. mRNA expression of E7-related genes in Hep G2 cell after the silencing of HPV 18 E7 in RT2 ProfilerTM Human Apoptosis
PCR Array cycle. At 0 h, 12 h, 24 h, 36 h, 48 h, 60 h, and 72 h after
transfection, cells transfected with siRNA-E7 or NC-E7 were
incubated with 50 mM EdU for two hours at 37uC using the Click-
iTH EdU Alexa FluorH 594 Imaging Kit (Invitrogen, USA)
according to the manufacturer’s instructions. Polymerase chain reactions (PCR) After fixation, cells
were treated with the reagent containing Alexa 594 azide, followed
by the counterstaining with Hoechst 33342 for detection and
imaging using fluorescence microscopy. membrane, after blocking with 5% nonfat dry milk for 1 h at room
temperature, the membrane was incubated overnight at 4uC with
primary antibodies specific for HPV 18 E6, HPV 18 E7 (Santa
Cruz, USA) and b-actin (Cell Signaling Technology, USA). The
immunoblot analysis was performed via enhanced chemilumines-
cence. Annexin V-FITC apoptosis assays Apoptosis of Hep G2 cells were determined using the Annexin
V-FITC Apoptosis Detection Kit (Calbiochem, USA) according to
the manufacturer’s protocol. In brief, at 24 h, 48 h and 72 h after
being transfected with siRNA-E7 or NC-E7, cells were washed
twice in cold PBS and then resuspended in 16 binding buffer at
a concentration of 16106 cells/mL. An aliquot of 500 ml solution
(56105 cells) was then transferred to another tube containing
1.25 mL Annexin V-FITC and the cells were gently vortexed and
incubated for 15 min at room temperature in the dark. Then an
aliquot of 500 mL cold 16binding buffer and 10 mL PI were added
and the cells were analyzed with a FACScalibur flow cytometer
within one hour. RNA interference Synthetic target HPV 18 E7 siRNAs sequences (Table 4) were
obtained from GenePharma (Shanghai, China). The siRNA was
separately cloned into pGPH1 vector and this was verified through
DNA sequencing analysis performed at Shanghai GenePharma
Company. In brief, 56105 Hep G2 cells were cultured in 6-well
plates using 2 mL of DMEM supplemented with 10% FBS. After
incubation for 24 h when cells were about 50% confluent, they
were transfected with 2.16 mg of the siRNA-E7 pools (HPV 18 E7-
63, HPV 18 E7-112 = 1:1) or NC-E7 (Negative control), using
LipofectamineTM 2000 (Invitrogen, USA) transfection reagent at
a siRNA: LipofectamineTM 2000 ratio of 1.08 mg: 2 mL according
to the manufacturer’s instructions. Transfection was terminated
4,6 h later by replacing the medium with 2.0 mL fresh medium
containing 10% FBS. EdU staining for cell proliferation EdU is a novel alternative for BrdU assays to directly measure
active DNA synthesis or the S-phase DNA synthesis of the cell PLoS ONE | www.plosone.org May 2012 | Volume 7 | Issue 5 | e37964 7 HPV 18 Integrated into Hep G2 Cells Figure 6. These photomicrographs show in situ hybridization results for human papillomavirus positive hepatocellular carcinoma. (A) Hep G2 cells were with the punctate signal pattern of HPV DNA. (B) HeLa cells were served as positive control. (C) Hepatocellular carcinoma was
with diffuse signal pattern of HPV staining. (D) No signal was found in hepatoma carcinoma cells of this HPV- negative specimen. doi:10.1371/journal.pone.0037964.g006 Figure 6. These photomicrographs show in situ hybridization results for human papillomavirus positive hepatocellular carcinoma. (A) Hep G2 cells were with the punctate signal pattern of HPV DNA. (B) HeLa cells were served as positive control. (C) Hepatocellular carcinoma was
with diffuse signal pattern of HPV staining. (D) No signal was found in hepatoma carcinoma cells of this HPV- negative specimen. doi:10.1371/journal.pone.0037964.g006 In situ hybridization on tissue microarrays In situ hybridization on tissue microarrays Expression profiling of E7-related genes in Hep G2 cells
using real-time PCR human cell cycle and apoptosis
arrays Tissue samples and microarrays were obtained from National
Engineering Center for BioChip at Shanghai following written
informed consent according to an established protocol approved
by the Ethic Committee of Second Military Medical University. Datas do not contain any information that may lead to the
identification of the patients. All studies were approved by the
Institutional Ethical Review Board at the Shantou University
Medical College. Tissue microarrays were constructed of 2-mm
cores of patient tissues taken from representative areas of
Hepatocellular carcinoma. Analysis was performed on the 100
independent cases including 28 of grade I, 65 of grade II, 7 of
grade III. We performed in situ hybridization (ISH) using
a Subtype-specific HPV probe on all cases. The ISH screening
was performed using a digoxigenin-labeled HPV probe cocktail
detecting DNA-HPV types 16 and 18 (Triplex International The real-time PCR microarrays of cell cycle and apoptosis
signaling pathways were purchased from Qiagen (CAT: PAHS-
020 and PAHS-01) and were used according to the manufacturer’s
instructions. The siRNA E7-63 was separately cloned into pGPH1
vector and this was verified through DNA sequencing analysis
performed at Shanghai GenePharma Company. In brief, the real-
time PCR reactions (40 cycles) were performed with sequential
incubations of 10 min at 95uC, 15 seconds at 95uC, and 1 min at
60uC. The fold- change for each gene from experimental group
and control group was expressed as 22DDCT. If the fold change was
greater than 2, then the result was reported as a fold up-regulation. If the fold change was less than 2, then the negative inverse of the
result was reported as a fold down-regulation. Table 3. Characteristics of the Hepatocellular Carcinoma Cases With HPV Infection. Hepatocellular carcinoma case
Subtype-specific ISH HPV 16/18
Age
Gender
Grade
1
+
61
Male
I
2
+
42
Male
II
3
+
62
Male
II
4
+
76
Male
II
5
+
45
Male
II
6
+
58
Male
II
7
+
68
Male
II
8
+
61
Male
I
9
+
60
Male
III
doi:10.1371/journal.pone.0037964.t003
PLoS ONE | www.plosone.org
8
May 2012 | Volume 7 | Issue 5 | e37964 Table 3. Characteristics of the Hepatocellular Carcinoma Cases With HPV Infection. References 1. Munoz N, Bosch FX, de Sanjose S, Herrero R, Castellsague X, et al. (2003)
Epidemiologic classification of human papillomavirus types associated with
cervical cancer. N Engl J Med 348: 518–527. 1. Munoz N, Bosch FX, de Sanjose S, Herrero R, Castellsague X, et al. (2003)
Epidemiologic classification of human papillomavirus types associated with
cervical cancer. N Engl J Med 348: 518–527. the AP1 site of E element in the enhancer I (EnI) in human liver cell. Biochem
Biophys Res Commun 265: 62–66. 10. Pang RW, Joh JW, Johnson PJ, Monden M, Pawlik TM, et al. (2008) Biology of
hepatocellular carcinoma. Ann Surg Oncol 15: 962–971. g J
2. Schwarz E, Freese UK, Gissmann L, Mayer W, Roggenbuck B, et al. (1985)
Structure and transcription of human papillomavirus sequences in cervical
carcinoma cells. Nature 314: 111–114. 2. Schwarz E, Freese UK, Gissmann L, Mayer W, Roggenbuck B, et al. (1985)
Structure and transcription of human papillomavirus sequences in cervical
carcinoma cells. Nature 314: 111–114. hepatocellular carcinoma. Ann Surg Oncol 15: 962–971. 11. Ferber MJ, Montoya DP, Yu C, Aderca I, McGee A, et al. (2003) Integrations of
the hepatitis B virus (HBV) and human papillomavirus (HPV) into the human
telomerase reverse transcriptase (hTERT) gene in liver and cervical cancers. Oncogene 22: 3813–3820. 3. Jeon S, Allen-Hoffmann BL, Lambert PF (1995) Integration of human
papillomavirus type 16 into the human genome correlates with a selective
growth advantage of cells. J Virol 69: 2989–2997. g
12. Doorbar J (2006) Molecular biology of human papillomavirus infection and
cervical cancer. Clin Sci (Lond) 110: 525–541. g
g
J
4. Barbosa MS, Schlegel R (1989) The E6 and E7 genes of HPV-18 are sufficient
for inducing two-stage in vitro transformation of human keratinocytes. Oncogene 4: 1529–1532. (
)
13. Moody CA, Laimins LA (2010) Human papillomavirus oncoproteins: pathways
to transformation. Nat Rev Cancer 10: 550–560. to transformation. Nat Rev Cancer 10: 550–560. 5. Aden DP, Fogel A, Plotkin S, Damjanov I, Knowles BB (1979) Controlled
synthesis of HBsAg in a differentiated human liver carcinoma-derived cell line. Nature 282: 615–616. 14. Korzeniewski N, Spardy N, Duensing A, Duensing S (2011) Genomic instability
and cancer: lessons learned from human papillomaviruses. Cancer Lett 305:
113–122. 15. Qi Z, Xu X, Zhang B, Li Y, Liu J, et al. (2010) Effect of simultaneous silencing of
HPV-18 E6 and E7 on inducing apoptosis in HeLa cells. Acknowledgments We thank Drs. Junhui Bian and Jennifer Mutch for their assistance in
preparing the manuscript and Mr. Shaohui Liu for preparation of reagents
and equipments. In situ hybridization on tissue microarrays Hepatocellular carcinoma case
Subtype-specific ISH HPV 16/18
Age
Gender
Grade
1
+
61
Male
I
2
+
42
Male
II
3
+
62
Male
II
4
+
76
Male
II
5
+
45
Male
II
6
+
58
Male
II
7
+
68
Male
II
8
+
61
Male
I
9
+
60
Male
III
doi:10.1371/journal.pone.0037964.t003
PLoS ONE | www.plosone.org
8
May 2012 | Volume 7 | Issue 5 | e37964 Table 3. Characteristics of the Hepatocellular Carcinoma Cases With HPV Infection. Table 3. Characteristics of the Hepatocellular Carcinoma Cases With HPV Infection. PLoS ONE | www.plosone.org May 2012 | Volume 7 | Issue 5 | e37964 HPV 18 Integrated into Hep G2 Cells Table 4. siRNA targeting HPV 18 E7 sites and sequences. hematoxylin, dehydrated, and mounted. Positive controls (HeLa
cells) were served with ISH for HPV. Negative controls were
prepared by substituting PBS for the primary antibody. Cases with
HPV signals in the punctate nuclear or diffuse staining of tumor
cells were determined to be positive. Table 4. siRNA targeting HPV 18 E7 sites and sequences. Name
Sequence
HPV 18 E7-63
Sense
59- CGGUUGACCUUCUAUGUCATT-39
Anti -sense
59-UGACAUAGAAGGUCAACCGGA-39
HPV 18 E7-74
Sense
59-CUAUGUCACGAGCAAUUAATT -39
Anti -sense
59-UUAAUUGCUCGUGACAUAGAA -39
HPV 18 E7-112
Sense
59- CGAUGAAAUAGAUGGAGUUTT-39
Anti -sense
59-AACUCCAUCUAUUUCAUCGTT -3
Negative Control
FAM-siRNA
Sense
59-UUCUCCGAACGUGUCACGUTT-39
Anti -sense
59-ACGUGACACGUUCGGAGAATT-
Negative Control siRNA
Sense
59- UUCUCCGAACGUGUCACGUTT -39
Anti -sense
59-ACGUGACACGUUCGGAGAATT
-3
doi:10.1371/journal.pone.0037964.t004 Biosciences, Fujian, China). Sections were counterstained with Biosciences, Fujian, China). Sections were counterstained with Author Contributions Conceived and designed the experiments: HC TM JH. Performed the
experiments: TM ZS NZ LW YY. Analyzed the data: TM ZS NZ. Contributed reagents/materials/analysis tools: LC SL LL XC. Wrote the
paper: TM HC. References Biochem Cell Biol 88:
697–704. 6. De Flora S, Bonanni P (2011) The prevention of infection-associated cancers. Carcinogenesis 32: 787–795. 7. Scinicariello F, Sato T, Lee CS, Hsu HC, Chan TS, et al. (1992) Detection of
human papillomavirus in primary hepatocellular carcinoma. Anticancer Res 12:
763–766. 16. Fink SA, Jacobson IM (2011) Managing patients with hepatitisB-related or
hepatitisC-related decompensated cirrhosis. Nat Rev Gastroenterol Hepatol 8:
285–295. 8. Kanodia S, Fahey LM, Kast WM (2007) Mechanisms used by human
papillomaviruses to escape the host immune response. Curr Cancer Drug
Targets 7: 79–89. 17. Liu XS, Jiang J (2007) Induction of apoptosis and regulation of the MAPK
pathway by ursolic acid in human leukemia K562 cells. Planta Med 73:
1192–1194. 9. Lee DH, Choi BH, Rho HM (1999) The synergistic transactivation of the
hepatitis B viral (HBV) pregenomic promoter by the E6 protein of human
papillomavirus type 16 (HPV-16 E6) with HBV X protein was mediated through 18. Su Z, Li R, Song X, Liu G, Li Y, et al. (2012) Identification of a novel isoform of
DHRS4 protein with a nuclear localization signal. Gene 494: 161–167. PLoS ONE | www.plosone.org May 2012 | Volume 7 | Issue 5 | e37964
|
https://openalex.org/W3159621728
|
https://journals.plos.org/plosone/article/file?id=10.1371/journal.pone.0250717&type=printable
|
English
| null |
In Silico screening of circulating tumor DNA, circulating microRNAs, and long non-coding RNAs as diagnostic molecular biomarkers in ovarian cancer: A comprehensive meta-analysis
|
PloS one
| 2,021
|
cc-by
| 11,152
|
PLOS ONE PLOS ONE RESEARCH ARTICLE Background Ovarian cancer (OC) is a leading cause of death in gynecological malignancies worldwide. Multitudinous studies have suggested the potential of circulating tumor DNA (ctDNA), circu-
lating microRNAs (miRNAs), and long non-coding RNAs (lncRNAs) as novel diagnostic
molecular biomarkers for OC. Here, we include three updated meta-analysis methods using
different molecular biomarkers to evaluate their discriminative value in OC diagnosis. Editor: Shannon M. Hawkins, Indiana University
School of Medicine, UNITED STATES
Received: September 21, 2020
Accepted: April 9, 2021
Published: April 26, 2021
Copyright: © 2021 Zhang et al. This is an open
access article distributed under the terms of the
Creative Commons Attribution License, which
permits unrestricted use, distribution, and
reproduction in any medium, provided the original
author and source are credited. Editor: Shannon M. Hawkins, Indiana University
School of Medicine, UNITED STATES
Received: September 21, 2020
Accepted: April 9, 2021
Published: April 26, 2021
Copyright: © 2021 Zhang et al This is an open In Silico screening of circulating tumor DNA,
circulating microRNAs, and long non-coding
RNAs as diagnostic molecular biomarkers in
ovarian cancer: A comprehensive meta-
analysis Linlin Zhang1☯, Chenyan Hu2☯, Zhongping Huang1, Zhijia Li2, Qin Zhang1*, Yang HeID2*
1 Department of Gynecology, People’s Hospital of Mianzhu City, Deyang, Sichuan, China, 2 College of
Medical Technology, State Key Laboratory of Southwestern Chinese Medicine Resources, Chengdu
University of Traditional Chinese Medicine, Chengdu, Sichuan, China Linlin Zhang1☯, Chenyan Hu2☯, Zhongping Huang1, Zhijia Li2, Qin Zhang1*, Yang HeID2*
1 Department of Gynecology, People’s Hospital of Mianzhu City, Deyang, Sichuan, China, 2 College of
Medical Technology, State Key Laboratory of Southwestern Chinese Medicine Resources, Chengdu
University of Traditional Chinese Medicine, Chengdu, Sichuan, China Linlin Zhang1☯, Chenyan Hu2☯, Zhongping Huang1, Zhijia Li2, Qin Zhang1*, Yang HeID2*
1 Department of Gynecology, People’s Hospital of Mianzhu City, Deyang, Sichuan, China, 2 College of
Medical Technology, State Key Laboratory of Southwestern Chinese Medicine Resources, Chengdu
University of Traditional Chinese Medicine, Chengdu, Sichuan, China a1111111111
a1111111111
a1111111111
a1111111111
a1111111111 ☯These authors contributed equally to this work. * qinzhang0838@sina.com (QZ); heyang@cdutcm.edu.cn (YH) OPEN ACCESS Citation: Zhang L, Hu C, Huang Z, Li Z, Zhang Q,
He Y (2021) In Silico screening of circulating tumor
DNA, circulating microRNAs, and long non-coding
RNAs as diagnostic molecular biomarkers in
ovarian cancer: A comprehensive meta-analysis. PLoS ONE 16(4): e0250717. https://doi.org/
10.1371/journal.pone.0250717 Methods We conducted three meta-analyses after searching different databases, and 23 eligible arti-
cles, including 8 concerning ctDNA, 11 concerning miRNAs, and 4 concerning lncRNAs,
were found. Further, we pooled data concerning the sensitivity, specificity, and other indica-
tors of accuracy for ctDNA/miRNAs/lncRNAs in the diagnosis of OC. The heterogeneity was
further explored by meta-regressions and subgroup analyses, and Deeks’ funnel plots were
used to measure the publication bias of these three meta-analyses. Copyright: © 2021 Zhang et al. This is an open
access article distributed under the terms of the
Creative Commons Attribution License, which
permits unrestricted use, distribution, and
reproduction in any medium, provided the original
author and source are credited. Results Data Availability Statement: All relevant data are
within the paper and its Supporting information
files. Linlin Zhang1☯, Chenyan Hu2☯, Zhongping Huang1, Zhijia Li2, Qin Zhang1*, Yang HeID2*
1 Department of Gynecology, People’s Hospital of Mianzhu City, Deyang, Sichuan, China, 2 College of
Medical Technology, State Key Laboratory of Southwestern Chinese Medicine Resources, Chengdu
University of Traditional Chinese Medicine, Chengdu, Sichuan, China ☯These authors contributed equally to this work.
* qinzhang0838@sina.com (QZ); heyang@cdutcm.edu.cn (YH) Introduction Ovarian cancer (OC), one of the three major gynecological malignancies, is a leading cause of
death among gynecological malignancies worldwide and accounts for 4% of all cancers in
females [1, 2]. Because the ovary is deep within the pelvis, the disease is insidious and asymp-
tomatic at early stages (Stage I or II). Approximately 70% of cases of OC are found to be
advanced (Stage III or IV), and the 5-year survival rate for patients is less than 30% [3, 4]; how-
ever, this percentage can rise sharply to 92.7% at early stages of OC [5]. Thus, the earlier the
precise diagnosis, the more effectively the disease can be controlled at early stages. Therefore,
sensitive and specific diagnostic methods or molecular signatures for the early detection of OC
are urgently needed to improve overall patient survival. Traditionally, the gold standard for OC diagnosis with accurate test results is histopatholog-
ical examination. However, histopathological analysis is not suitable for the early diagnosis of
OC because of it is invasive in nature and time consuming [6]. Currently, serum carbohydrate
antigen 125 (CA125), the most common serum marker in OC, is being used to assist in the
diagnosis of advanced OC and to monitor OC recurrence. However, only 50% of early-stage
patients have elevated serum CA125 levels; furthermore, 1% of healthy females, 3% of females
with benign ovarian tumors, and 6% females with non-ovarian benign diseases may also report
elevated serum CA125 levels. Thus, the sensitivity and specificity of serum CA125 in the early
diagnosis of OC are limited and the false-positive rate of serum CA125 is high [7], which limits
its application in the early diagnosis of OC. Accordingly, sensitive and specific noninvasive
diagnostic molecular biomarkers OC diagnosis are urgently required to improve the prognosis
of patients with OC. Circulating tumor DNA (ctDNA), a part of cell-free DNA (cfDNA), originates from DNA
fragments produced by apoptosis, necrosis, or secretion of tumor cells [8]. ctDNA contains the
same genetic defects as the tumor DNA it originates from, e.g., point mutations, rearrange-
ments, and amplifications [9], and can reflect the dynamic changes of tumors in real time with
a short half-life in blood [10]. In addition, as a method for liquid biopsy, ctDNA detection can
overcome the defects resulting from tumor heterogeneity in tissue biopsy and ensure more
comprehensive detection [11]. Conclusion ctDNA/miRNAs/lncRNAs may be promising molecular biomarkers for OC diagnosis. Further
large-scale studies are needed to verify the potential applicability of ctDNA/miRNAs/
lncRNAs molecular signatures alone or in combination as diagnostic molecular biomarkers
for OC. Data Availability Statement: All relevant data are
within the paper and its Supporting information
files. In all, this meta-analysis included 1732 OC patients and 3958 controls. The sensitivity of
ctDNA for OC diagnosis was superior to that of lncRNA and miRNA (84% vs. 81% vs. 78%). Moreover, the specificity and area under the receiver-operating characteristic (ROC) curve
(AUC) of ctDNA were 91% and 94%, which were significantly higher than those of miRNA
and lncRNAs (78% and 85%; 78% and 86%, respectively). No significant difference was
observed among the two meta-analyses of ctDNA and lncRNA (P > 0.05) with regard to pub-
lication bias, while the meta-analysis of miRNA observed a significantly small publication
bias (P < 0.05). Funding: This study was financially supported by
the Sichuan science and technology project (Grant
No. 2019YFS0402) and by the scientific research
fund of Sichuan health commission (Grant No. 20PJ164). Competing interests: The authors have declared
that no competing interests exist. 1 / 19 PLOS ONE | https://doi.org/10.1371/journal.pone.0250717
April 26, 2021 PLOS ONE A meta-analysis of ovarian cancer using ctDNA, miRNAs, and lncRNAs PLOS ONE | https://doi.org/10.1371/journal.pone.0250717
April 26, 2021 Inclusion and exclusion criteria The inclusion criteria were as follows: (1) the diagnosis of OC patients was based on the histo-
pathological analysis as the gold standard; (2) diagnostic tests used ctDNA/miRNAs/lncRNAs
as molecular biomarkers for diagnosing OC; (3) the study was designed as a case-control
study, with healthy individuals and patients with benign diseases included in the control
group; (4) the study contained enough data to construct 2 × 2 diagnostic tables; (5) the sample
size was more than 5 OC patients to reduce selection bias; and (6) the full text was published in
English. The following studies were excluded: (1) fraud studies; (2) descriptive studies wherein only
the diagnostic value of ctDNA/miRNAs/lncRNAs in the diagnosis of OC was described with-
out control groups; (3) studies with incorrect calculations and incomplete data; (4) confer-
ences, reviews, abstracts, editorials, and case reports; and (5) duplicate data or duplicate
publications. In addition, when several studies used the same patient cohort, only the latest,
largest, or best quality studies was included. Introduction Several researchers have reported that lncRNAs
participate in carcinogenesis by regulating cell proliferation, division, differentiation, and
metastasis [19, 20]. Furthermore, many studies have reported the potential of ctDNA, miR-
NAs, and lncRNAs as novel diagnostic molecular biomarkers for OC [10, 21, 22]. However,
these published studies are inconsistent, and to the best of our knowledge, no preceding meta-
analysis exists in the literature evaluating these three molecular biomarkers simultaneously. We included three meta-analysis methods using different molecular signatures in this
study. In these three meta-analyses, we aim to evaluate the diagnostic values of these different
molecular signatures in OC, and particularly to analyze the discriminative value of ctDNA,
miRNAs, and lncRNAs between OC patients and healthy controls. Literature research strategy These three meta-analyses were conducted in accordance with the PRISMA guidelines (S1
Checklist) [23]. We searched the PubMed, EMBASE, Cochrane Library, and Web of Science
databases for all related articles published from January 1, 2015 until March 20, 2020. A large
number of meta-analyses have been published before 2015 to evaluate the diagnostic value of
ctDNA/miRNAs/lncRNAs [24, 25], but there is a lack of meta-analysis to evaluate these bio-
markers simultaneously during 2015–2020. Thus, we restricted the start date of publications to
January 1, 2015 for an updated and comprehensive meta-analysis. The search keywords were
as follows: “Circulating Tumor DNA/MicroRNAs/RNA, Long Noncoding,” “Ovarian Neo-
plasms,” “early.” There were no language restrictions, but only English articles were included. If the title or abstract met the inclusion criteria, the full text was then evaluated for further
verification. Introduction Thus, the ctDNA detection can be used for early diagnosis and
staging of cancer, tumor efficacy evaluation, tumor recurrence monitoring, and prognosis
evaluation [12, 13]. Circulating microRNAs (miRNAs) are a class of endogenous non-coding
small-molecule single-stranded RNAs sized approximately 18–24 nucleotides in length; these
miRNAs regulate the expression of approximately 30% of human proteins and participate in
the regulation of cell differentiation, growth, apoptosis, and metabolism. Recent studies have
shown that miRNAs have specific expression profiles in various tumor tissues and that they
are associated with various stages of tumorigenesis, tumor development, and metastasis [14]. At each stage, the corresponding miRNAs change along with the genes they regulate [15]. Long non-coding RNAs (lncRNAs) are a new type of non-coding RNA with a length of more
than 200 nucleotides. lncRNAs are characterized by no or limited protein coding potential; 2 / 19 PLOS ONE | https://doi.org/10.1371/journal.pone.0250717
April 26, 2021 PLOS ONE A meta-analysis of ovarian cancer using ctDNA, miRNAs, and lncRNAs further, they regulate gene expression at various levels in the form of RNA, for example via epi-
genetic regulation, transcriptional regulation, and post-transcriptional regulation [16]. Studies
have confirmed that lncRNAs are involved in almost all types of biological processes, such as
innate immunity, development, and tumorigenesis [17, 18]. In particular, the function of
lncRNAs in cancer has widely been explored. Several researchers have reported that lncRNAs
participate in carcinogenesis by regulating cell proliferation, division, differentiation, and
metastasis [19, 20]. Furthermore, many studies have reported the potential of ctDNA, miR-
NAs, and lncRNAs as novel diagnostic molecular biomarkers for OC [10, 21, 22]. However,
these published studies are inconsistent, and to the best of our knowledge, no preceding meta-
analysis exists in the literature evaluating these three molecular biomarkers simultaneously. We included three meta-analysis methods using different molecular signatures in this
study. In these three meta-analyses, we aim to evaluate the diagnostic values of these different
molecular signatures in OC, and particularly to analyze the discriminative value of ctDNA,
miRNAs, and lncRNAs between OC patients and healthy controls. further, they regulate gene expression at various levels in the form of RNA, for example via epi-
genetic regulation, transcriptional regulation, and post-transcriptional regulation [16]. Studies
have confirmed that lncRNAs are involved in almost all types of biological processes, such as
innate immunity, development, and tumorigenesis [17, 18]. In particular, the function of
lncRNAs in cancer has widely been explored. Data extraction and quality assessment Two reviewers independently assessed the eligibility of the retrieved articles. Differences
between reviewers were resolved by consulting with a third reviewer. The data characteristics 3 / 19 PLOS ONE | https://doi.org/10.1371/journal.pone.0250717
April 26, 2021 PLOS ONE A meta-analysis of ovarian cancer using ctDNA, miRNAs, and lncRNAs of each article were as follows: last name of the first author (if the last name and the publication
year of two articles were the same, the articles were arranged according to the last name of the
first corresponding author), publication year, country of origin, specimen types, detection
methods, pathology type, case and control numbers, biomarker types (ctDNA/miRNAs/
lncRNAs), AUC and data of 2 × 2 diagnostic tables (including sensitivity and specificity). The
methodological quality assessment of the articles included in this study was conducted as per
the quality assessment of the diagnostic accuracy study-2 (QUADAS-2) [26]. QUADAS-2 con-
tains four key domains: patient selection, index test, reference standard, and flow and timing,
with “low,” “high,” or “unknown” as the results for the risk of bias. The QUADAS-2 quality
evaluation chart was generated using Review Manager 5.3.3 (Cochrane collaboration, Barce-
lona, Spain). Statistical analysis All analyses were performed using the statistical analysis software Meta Disc 1.4 (Cochrane
Colloquium, Barcelona, Spain) and Stata 15.1 (Stata Corporation, College Station, USA). The
sensitivity, specificity, diagnostic odds ratio (DOR), positive likelihood ratio (PLR), and nega-
tive likelihood ratio (NLR) of ctDNA/miRNAs/lncRNAs in OC diagnosis were analyzed using
the bivariate meta-analysis model (BRM). The bivariate summary ROC (SROC) curve was
generated by graphing the sensitivity and specificity of each included study. AUC values of
0.5–0.7, 0.7–0.9, and 0.9–1.0 indicate that the diagnostic accuracy is low, medium, and high,
respectively. Additionally, Fagan nomograms and likelihood ratio plots were used to detect the
clinical value of ctDNA/miRNAs/lncRNAs in OC diagnosis; in the former, a pre-test probabil-
ity of 20% was assumed and then the post-test probability was calculated using Bayes’ theorem
[27]. The spearman correlation coefficient, calculated using the logarithm of sensitivity and
the logarithm of (1-specificity), was used to detect the threshold effect of the included studies. If threshold effect exists, a positive correlation appears. A spearman correlation coefficient of
greater than 0.6 and a P value of less than 0.05 indicated that the threshold effect was statisti-
cally significant [28]. Heterogeneity between studies was evaluated using Cochran’s Q statistic
and I2 statistic, and a P value of less than 0.1 or I2 value higher than 50% indicated significant
heterogeneity [29, 30]. Random effect models were applied for calculating the pooled effect
when obvious heterogeneity was observed. Otherwise, fixed effect models were chosen for cal-
culating effect values [31]. Heterogeneity was further validated by subgroup and sensitivity
analyses, and the source of heterogeneity was analyzed via single factor meta-regression. In
our meta-analysis, because of heterogeneity, all statistical data were calculated using a random
effects model. Further, Deeks’ funnel plots were used to measure the publication bias in the
three meta-analyses, and a p value of less than 0.05 indicated a statistical publication bias [32]. Search results The literature screening process of ctDNA/miRNAs/lncRNAs is shown in Fig 1a, 1c and 1e,
respectively. In all, 793 records (ctDNA: 130; miRNAs: 536; lncRNAs: 127) were initially iden-
tified on computer literature search, of which 392 duplicates (ctDNA: 49; miRNA: 278;
lncRNA: 65) were excluded. After carefully evaluating the title, abstract, and keywords, 326
articles (ctDNA: 61; miRNA: 227; lncRNA: 38) were excluded because they were either
reviews, conference articles, non-clinical trials, or non-human studies (ctDNA: 43; miRNA:
129; lncRNA: 21) or were not related to human OC and ctDNA/miRNAs/lncRNAs (ctDNA:
18; miRNA: 95; lncRNA: 15). Subsequently, full text of 75 articles (ctDNA: 20; miRNAs: 31;
lncRNAs: 24) was obtained for further detailed review, and 52 of these (ctDNA: 12; miRNAs: PLOS ONE | https://doi.org/10.1371/journal.pone.0250717
April 26, 2021 4 / 19 PLOS ONE A meta-analysis of ovarian cancer using ctDNA, miRNAs, and lncRNAs Fig 1. Flowcharts of literature search and identification and graphs of risk of bias and applicability concerns. (a) A flowchart
of ctDNA. (b) A graph of bias risk and applicability concerns of ctDNA. (c) A flowchart of miRNAs. (d) A graph of bias risk and
applicability concerns of miRNAs. (e) A flowchart of lncRNAs. (f) A graph of bias risk and applicability concerns of lncRNAs. Fig 1. Flowcharts of literature search and identification and graphs of risk of bias and applicability concerns. (a) A flowchart
of ctDNA. (b) A graph of bias risk and applicability concerns of ctDNA. (c) A flowchart of miRNAs. (d) A graph of bias risk and
applicability concerns of miRNAs. (e) A flowchart of lncRNAs. (f) A graph of bias risk and applicability concerns of lncRNAs. https://doi.org/10.1371/journal.pone.0250717.g001 Fig 1. Flowcharts of literature search and identification and graphs of risk of bias and applicability concerns. (a) A flowchart
of ctDNA. (b) A graph of bias risk and applicability concerns of ctDNA. (c) A flowchart of miRNAs. (d) A graph of bias risk and
applicability concerns of miRNAs. (e) A flowchart of lncRNAs. (f) A graph of bias risk and applicability concerns of lncRNAs. Search results https://doi.org/10.1371/journal.pone.0250717.g001 5 / 19 PLOS ONE | https://doi.org/10.1371/journal.pone.0250717
April 26, 2021 PLOS ONE A meta-analysis of ovarian cancer using ctDNA, miRNAs, and lncRNAs 20; lncRNAs: 20) were excluded because they were fraud studies (miRNA: 1; lncRNA: 1) or
irrelevant reports that were not related to diagnosis (ctDNA: 5; miRNA: 5; lncRNA: 9) or
descriptive studies had no quantitative analysis of the diagnostic value of ctDNA/miRNA/
lncRNA for OC (ctDNA: 1; miRNA: 12; lncRNA: 10) or reported insufficient data for con-
structing 2 × 2 diagnostic tables (ctDNA: 6; miRNA: 2). Finally, 23 articles (ctDNA: 8; miR-
NAs: 11; lncRNAs: 4) were included in this diagnostic meta-analysis [33–55]. Study characteristics and quality assessment The characteristics of ctDNA/miRNAs/lncRNAs included in the study are shown in Table 1. The 23 articles (ctDNA: 8; miRNAs: 11; lncRNAs: 4) included 1732 OC patients and 3958 con-
trols (ctDNA: OC patients, 361, controls, 1773; miRNAs: OC patients, 1253, controls, 2002;
lncRNAs: OC patients, 118, controls, 183), and all OC patients were diagnosed with the gold
standard of histopathological analysis. With regard to the origins of these studies, most of the
studies were concentrated in China (ctDNA: 3; miRNAs: 4; lncRNAs: 4), and other studies in
the United States of America (ctDNA: 4), Slovakia (ctDNA: 1), Japan (miRNAs: 3), Hungary
(miRNAs: 1), Germany (miRNAs: 1), Italy (miRNAs: 1), and India (miRNAs: 1). Subsequently,
while analyzing the accuracy of ctDNA/miRNAs/lncRNAs in diagnosing OC, most studies
tended to use plasma as a specimen (ctDNA: 8; miRNAs: 1; lncRNAs: 2), followed by serum
(miRNAs: 9), tissue (lncRNAs: 2), and urine (miRNAs: 1). With regard to tumor lymphade-
nopathy (TNM) classification, 2 studies focused on stage I-II (lncRNAs: 2), 1 study concen-
trated on stage III-IV (miRNA: 1), and 19 studies focused on stage I-IV (ctDNA: 7; miRNAs:
10; lncRNAs: 2). All eligible studies were published between 2015 and 2020. The QUADAS-2 summary dia-
gram for ctDNA/miRNAs/lncRNAs is shown in Fig 1b, 1d and 1f. The results demonstrate
that the overall quality of all the included studies was stable. However, two important issues
emerged. One was that due to the design of case-control studies, patient selection in eighteen
studies may increase the risk of bias and applicability concerns. The other was that only two
studies had set thresholds in advance, which may lead to unknown risks of bias in related
articles. PLOS ONE | https://doi.org/10.1371/journal.pone.0250717
April 26, 2021 PLOS ONE PLOS ONE Table 1. Characteristics of included studies, divided by ctDNA, miRNAs and lncRNAs. PLOS ONE Author
Year
Country
Sample
Method
Pathology
type
Case
(Stage)
Case
(n)
Control
(Type)
Control
(n)
Biomarkers
AUC
Sensitivity
Specificity
ctDNA
Cohen et al
2018
USA
Plasma
Multiplex-PCR
And Sequencing
EOC
I-IV
54
HC
812
BRAF, CDKN2A,
CTNNB1, KRAS,
PIK3CA, TP53
-
0.9815
0.9914
Cristiano
et al
2019
USA
Plasma
WGS
OC
I-IV
28
HC
215
ALK, APC, AR,
CTNNB1, EGFR,
ERBB4, FGFR3,
HNF1A, KIT,
PDGFRA, SKT11,
TP53
0.99
0.8929
0.9535
Douville
et al
2019
USA
Plasma
RealSeqS
EOC
I-III
48
HC
378
-
0.989
0.979
0.905
Dvorska´
et al
2019
Slovakia
Plasma
PCR And
Pyrosequencing
OC
I-IV
33
HC
9
RASSF1, PTEN,
CDH1, PAX1
0.822
0.91
0.56
Wang et al
2017
China
Plasma
MSP
EOC
I-IV
71
HC
123
OPCML, RUNX3,
TFPI2
-
0.9014
0.9187
Wang et al
2018
USA
Plasma
Multiplex-PCR
And Sequencing
OC
I-IV
83
HC
192
PIK3CA, PIK3R1,
PTEN, TP53
-
0.4337
1
Yu et al
2019
China
Plasma
qRT-PCR
(SYBR-Green)
OC
I-IV
20
HC
20
ALU
0.861
0.8
0.6
Zhang et al
2018
China
Plasma
qRT-PCR
(SYBR-Green)
OC
-
24
HC
+BOD
24
ALU-219
0.73
0.667
0.792
miRNAs
Chen et al
2020
China
Serum
qRT-PCR
(TaqMan)
EOC
I-IV
152
HC
+BOD
+BOT
107
miR-125b
0.73
0.76
0.416
Kobayashi
et al
2018
Japan
Serum
qRT-PCR
(TaqMan)
EOC
I-IV
70
HC
13
miR-1290
0.48
0.51
0.57
Liang et al
2018
China
Serum
qRT-PCR (SYBR
Green)
OC
I-IV
101
HC
+BOD
100
miR-183
0.77
0.752
0.8
Ma´rton
et al
2019
Hungary
plasma
qRT-PCR (SYBR
Green)
EOC
I+III
+IV
28
HC+NM
60
miR-200c
0.861
0.7143
0.8667
Meng et al
2015
Germany
Serum
qRT-PCR
(TaqMan)
EOC
I-IV
180
HC
66
miR-429
0.845
0.594
0.955
Todeschini
et al
2017
Italy
Serum
qRT-PCR (SYBR
Green)
EOC
(HGSOC)
III-IV
168
HC
65
miR-1246
0.893
0.87
0.77
Yokoi et al
2018
Japan
Serum
Microarrays
OC
I-IV
160
CF
1379
miR-4532
0.974
0.956
0.928
Yokoi et al
2017
Japan
Serum
qRT-PCR
(TaqMan)
EOC
I-IV
155
HC
63
miR-142
0.847
0.706
0.884
Zhou et al
2015
China
Urine
qRT-PCR
(TaqMan)
EOC
(SOC)
I-IV
34
HC
25
miR-6076
0.693
0.925
0.576
Zhu et al
2017
China
Serum
qRT-PCR
(TaqMan)
EOC
I-IV
135
BOD
54
miR-125b
0.737
0.756
0.685
Zuberi et al
2015
India
Serum
qRT-PCR (SYBR
Green)
EOC
I-IV
70
CF
70
miR-200a
0.81
0.806
0.735
lncRNAs
Cui et al
2020
China
Plasma
qRT-PCR (SYBR
Green)
OC
I-II
17
HC
58
CASC11
0.88
0.765
0.81
Gong et al
2019
China
Plasma
qRT-PCR (SYBR
Green)
OC
I-II
28
HC
54
MIR4435-2HG
0.786
0.786
0.889
Yang et al
2019
China
Tissues
qRT-PCR (SYBR
Green)
OC
I-IV
32
CF
31
FLJ33360
0.793
0.844
0.71
(Continued) Table 1. Diagnostic accuracy of ctDNA/miRNAs/lncRNAs in OC In the three meta-analyses, spearman correlation coefficients for ctDNA, miRNAs, and
lncRNAs were -0.024 (P > 0.05), 0.118 (P > 0.05), and 0.600 (P > 0.05), respectively, which
confirmed that there was no threshold effect and the heterogeneity was caused by other rea-
sons in this study. Forest plots for sensitivity and specificity of ctDNA/miRNAs/lncRNAs in
OC diagnosis are shown in Figs 2a, 3a and 4a. The heterogeneity caused by non-threshold
effects was evaluated by Cochran’s Q statistic and I2 statistic, and I2 values for sensitivity and
specificity of ctDNA/miRNAs/lncRNAs indicated heterogeneity caused by non-threshold
effects (ctDNA: sensitivity, I2 = 89.70%, specificity, I2 = 68.09%; miRNAs: sensitivity, I2 =
88.30%, specificity, I2 = 95.50%; lncRNAs: sensitivity, I2 = 0.00%, specificity, I2 = 54.09%). The
pooled sensitivity, specificity, PLR, NLR, AUC and DOR of ctDNA/miRNAs/lncRNAs for OC
diagnosis is shown in Table 2. According to the SROC curves of the included studies, the AUC
values indicating a high accuracy of ctDNA/miRNAs/lncRNAs in diagnosing OC (Figs 2b, 3b
and 4b). The goodness-of-fit and bivariate normality analysis indicated that the random effects
bivariate model was very suitable for the calculation of summary estimates (Figs 2d and 3d). Sensitivity analysis, influence analysis, and outlier detection analysis of ctDNA/miRNAs/
lncRNAs revealed that the data reported in the record no. 6 of ctDNA, record no. 7 of miR-
NAs, and record no. 2 of lncRNAs are far from the rest of the data, suggesting that they could PLOS ONE | https://doi.org/10.1371/journal.pone.0250717
April 26, 2021 6 / 19 A meta-analysis of ovarian cancer using ctDNA, miRNAs, and lncRNAs PLOS ONE compared with the
remaining three markers (lncRNA CASC11, lncRNA FLJ33360, and lncRNA HAGLROS). be one of the reasons for the heterogeneity (Figs 2c and 2d, 3c and 3d and 4c). After excluding
record no. 6 of ctDNA, no. 7 of miRNAs, and no. 2 of lncRNAs, the random effect model was
used to combine the effects. For ctDNA and miRNAs, the remaining data were still heteroge-
neous (ctDNA: I2 = 89.00%, P < 0.05; miRNAs: I2 = 82.20%, P < 0.05), and the I2 values for
sensitivity and specificity were reduced by 16.88% and 0.75% for ctDNA and 5.27% and 6.64%
for miRNAs, respectively (S1a and S1b Fig). The results for lncRNAs showed that the remain-
ing 3 sets of data were not heterogeneous (I2 = 0.00, P > 0.1). This may be because of the supe-
rior diagnostic specificity of lncRNA MIR4435-2HG studied by Gong et al. compared with the
remaining three markers (lncRNA CASC11, lncRNA FLJ33360, and lncRNA HAGLROS). Meta-regression and subgroup analysis for heterogeneity To further explore the sources of heterogeneity, meta-regression analyses (Fig 5a and 5b) and
subgroup analyses (Tables 3 and 4) were performed for ctDNA and miRNAs. For ctDNA,
results of the meta-regression analysis depicted that methods (p = 0.00) and control size
(p = 0.01) showed a statistically significant impact on specificity. Subgroup analysis indicated
that ctDNA detection accuracy in the Caucasian population (DOR, 154.12; AUC, 0.97) showed
a better diagnostic performance than that in the Asian population (DOR, 14.30; AUC,0.86). Subsequently, we noticed that ctDNA could detect all epithelial ovarian cancer (EOC) cases
(DOR, 326.84; AUC, 0.98) more accurately than OC cases (DOR, 20.80; AUC, 0.86) on com-
paring pathology types. In addition, sample size of > 40 (DOR, 126.89; AUC, 0.97) showed
superior diagnostic performance than that of 40 (DOR, 18.17; AUC, 0.91). Nevertheless, the
subgroup based on control size suggested that a control size of > 30 (DOR, 219.93; AUC, 0.98)
showed a better diagnostic performance than a control size of 30 (DOR, 6.34; AUC, 0.77). With regard to miRNAs, the results of the meta-regression analysis showed that methods of
qRT-PCR (TaqMan) (p = 0.00) and case type (EOC cases or OC cases; p = 0.01) would affect
the sensitivity. Further, ethnic differences (p = 0.03) significantly affected the specificity. Sub-
group analysis revealed that the accuracy of detection using miRNAs in Asians (DOR, 10.98;
AUC, 0.84) was inferior to that in Caucasians (DOR, 21.99; AUC, 0.89). In addition, it was
remarkable that qRT-PCR (SYBR Green) (DOR, 14.87; AUC, 0.86), in contrast to qRT-PCR
(TaqMan) (DOR, 7.04; AUC, 0.79), showed a better performance in predicting OC. PLOS ONE Characteristics of included studies, divided by ctDNA, miRNAs and lncRNAs. PLOS ONE | https://doi.org/10.1371/journal.pone.0250717
April 26, 2021 7 / 19 A meta-analysis of ovarian cancer using ctDNA, miRNAs, and lncRNAs PLOS ONE Table 1. (Continued)
Author
Year
Country
Sample
Method
Pathology
type
Case
(Stage)
Case
(n)
Control
(Type)
Control
(n)
Biomarkers
AUC
Sensitivity
Specificity
Wang et al
2019
China
Tissues
qRT-PCR (SYBR
Green)
OC
I-IV
41
CF
40
HAGLROS
0.751
0.83
0.7
ctDNA, circulating tumor DNA; miRNAs, circulating microRNAs; lncRNAs, long non-coding RNAs; USA, the United States of America; AUC, area under curve; PCR,
polymerase chain reaction; WGS, whole-genome sequencing; RealSeqS, repetitive element aneuploidy sequencing system; MSP, methylation-specific polymerase chain
reaction; qRT-PCR, real-time fluorescence quantitative polymerase chain reaction; OC, ovarian cancer; EOC, epithelial ovarian cancer; HGSOC, high grade serous
ovarian carcinoma; SOC, serous ovarian carcinoma; HC, healthy control; BOD, benign ovarian diseases; BOT, borderline ovarian tumors; NM, non-malignant masses;
CF, cancer free; CA125, serum carbohydrate antigen 125; HE4, human epididymis protein 4. Author
Year
Country
Sample
Method
Pathology
type
Case
(Stage)
Case
(n)
Control
(Type)
Control
(n)
Biomarkers
AUC
Sensitivity
Specificity
Wang et al
2019
China
Tissues
qRT-PCR (SYBR
Green)
OC
I-IV
41
CF
40
HAGLROS
0.751
0.83
0.7 ctDNA, circulating tumor DNA; miRNAs, circulating microRNAs; lncRNAs, long non-coding RNAs; USA, the United States of America; AUC, area under curve; PCR,
polymerase chain reaction; WGS, whole-genome sequencing; RealSeqS, repetitive element aneuploidy sequencing system; MSP, methylation-specific polymerase chain
reaction; qRT-PCR, real-time fluorescence quantitative polymerase chain reaction; OC, ovarian cancer; EOC, epithelial ovarian cancer; HGSOC, high grade serous
ovarian carcinoma; SOC, serous ovarian carcinoma; HC, healthy control; BOD, benign ovarian diseases; BOT, borderline ovarian tumors; NM, non-malignant masses;
CF, cancer free; CA125, serum carbohydrate antigen 125; HE4, human epididymis protein 4. be one of the reasons for the heterogeneity (Figs 2c and 2d, 3c and 3d and 4c). After excluding
record no. 6 of ctDNA, no. 7 of miRNAs, and no. 2 of lncRNAs, the random effect model was
used to combine the effects. For ctDNA and miRNAs, the remaining data were still heteroge-
neous (ctDNA: I2 = 89.00%, P < 0.05; miRNAs: I2 = 82.20%, P < 0.05), and the I2 values for
sensitivity and specificity were reduced by 16.88% and 0.75% for ctDNA and 5.27% and 6.64%
for miRNAs, respectively (S1a and S1b Fig). The results for lncRNAs showed that the remain-
ing 3 sets of data were not heterogeneous (I2 = 0.00, P > 0.1). This may be because of the supe-
rior diagnostic specificity of lncRNA MIR4435-2HG studied by Gong et al. PLOS ONE | https://doi.org/10.1371/journal.pone.0250717
April 26, 2021 Fagan nomogram analysis, likelihood ratio analysis and publication bias Fagan’s nomogram plots were used to verify the probability of OC being detected by ctDNA/
miRNAs/lncRNAs in an otherwise healthy person (Fig 5c–5e). For an individual with a 20%
possibility of developing OC before the test, if the ctDNA, miRNAs, or lncRNAs are positive,
the probability of being diagnosed with OC would reach 72%, 47%, or 49%, respectively. PLOS ONE | https://doi.org/10.1371/journal.pone.0250717
April 26, 2021 8 / 19 A meta-analysis of ovarian cancer using ctDNA, miRNAs, and lncRNAs PLOS ONE However, a negative ctDNA, miRNAs, or lncRNAs result implied that the probability after the
test would reduce to 4%, 6%, or 6%, respectively. Accordingly, a PLR > 10 and NLR < 0.1
demonstrated high diagnostic accuracy [56]. In this study, the pooled PLR > 10 and
NLR > 0 1 indicated that ctDNA showed a significantly high detection rate for OC but
Fig 2. Forest plots of sensitivity and specificity, SROC curve, sensitivity analysis plot, goodness-of-fit and bivariate normality analysis plots, likelihood
ratio plot and Deeks’ funnel plot of ctDNA. (a) Forest plots of sensitivity and specificity to evaluate the diagnostic performance of ctDNA. (b) SROC curve
to describe the diagnostic value of ctDNA. (c) Sensitivity analysis to estimate each study’s value of ctDNA. (d) Goodness-of-fit and bivariate normality
analysis plots to explore the sources of heterogeneity of ctDNA. (e) Likelihood ratio plot to appraise the diagnostic and elimination capabilities of ctDNA. (f)
Deeks’ funnel plot to assess publication bias of ctDNA. https://doi.org/10.1371/journal.pone.0250717.g002 Fig 2. Forest plots of sensitivity and specificity, SROC curve, sensitivity analysis plot, goodness-of-fit and bivariate normality analysis plots, likelihood
ratio plot and Deeks’ funnel plot of ctDNA. (a) Forest plots of sensitivity and specificity to evaluate the diagnostic performance of ctDNA. (b) SROC curve
to describe the diagnostic value of ctDNA. (c) Sensitivity analysis to estimate each study’s value of ctDNA. (d) Goodness-of-fit and bivariate normality
analysis plots to explore the sources of heterogeneity of ctDNA. (e) Likelihood ratio plot to appraise the diagnostic and elimination capabilities of ctDNA. (f)
Deeks’ funnel plot to assess publication bias of ctDNA. https://doi.org/10.1371/journal.pone.0250717.g002 However, a negative ctDNA, miRNAs, or lncRNAs result implied that the probability after the
test would reduce to 4%, 6%, or 6%, respectively. Accordingly, a PLR > 10 and NLR < 0.1
demonstrated high diagnostic accuracy [56]. Fagan nomogram analysis, likelihood ratio analysis and publication bias In this study, the pooled PLR > 10 and
NLR > 0.1 indicated that ctDNA showed a significantly high detection rate for OC, but However, a negative ctDNA, miRNAs, or lncRNAs result implied that the probability after the
test would reduce to 4%, 6%, or 6%, respectively. Accordingly, a PLR > 10 and NLR < 0.1
demonstrated high diagnostic accuracy [56]. In this study, the pooled PLR > 10 and
NLR > 0.1 indicated that ctDNA showed a significantly high detection rate for OC, but PLOS ONE | https://doi.org/10.1371/journal.pone.0250717
April 26, 2021 9 / 19 A meta-analysis of ovarian cancer using ctDNA, miRNAs, and lncRNAs PLOS ONE Fig 3. Forest plots of sensitivity and specificity, SROC curve, sensitivity analysis plot, goodness-of-fit and bivariate normality analysis plots, likelihood
ratio plot and Deeks’ funnel plot of miRNAs. (a) Forest plots of sensitivity and specificity to evaluate the diagnostic performance of miRNAs. (b) SROC
curve to describe the diagnostic value of miRNAs. (c) Sensitivity analysis to estimate each study’s value of miRNAs. (d) Goodness-of-fit and bivariate
normality analysis plots to explore the sources of heterogeneity of miRNAs. (e) Likelihood ratio plot to appraise the diagnostic and elimination capabilities of
miRNAs. (f) Deeks’ funnel plot to assess publication bias of miRNAs. https://doi.org/10.1371/journal.pone.0250717.g003 Fig 3. Forest plots of sensitivity and specificity, SROC curve, sensitivity analysis plot, goodness-of-fit and bivariate normality analysis plots, likelihood
ratio plot and Deeks’ funnel plot of miRNAs. (a) Forest plots of sensitivity and specificity to evaluate the diagnostic performance of miRNAs. (b) SROC
curve to describe the diagnostic value of miRNAs. (c) Sensitivity analysis to estimate each study’s value of miRNAs. (d) Goodness-of-fit and bivariate
normality analysis plots to explore the sources of heterogeneity of miRNAs. (e) Likelihood ratio plot to appraise the diagnostic and elimination capabilities of
miRNAs. (f) Deeks’ funnel plot to assess publication bias of miRNAs. https://doi.org/10.1371/journal.pone.0250717.g003 https://doi.org/10.1371/journal.pone.0250717.g003 exhibited a very low capacity for exclusion (Fig 2e). In other words, the use of ctDNA to diag-
nose OC may not be suitable as an exclusion test; however, ctDNA can be used as a confirma-
tory test. Fagan nomogram analysis, likelihood ratio analysis and publication bias However, the pooled PLR < 10 and NLR > 0.1 of miRNAs indicated that based on
the current research, the clinical application value of miRNAs in the diagnosis of OC is still PLOS ONE | https://doi.org/10.1371/journal.pone.0250717
April 26, 2021 10 / 19 A meta-analysis of ovarian cancer using ctDNA, miRNAs, and lncRNAs PLOS ONE Fig 4. Forest plots of sensitivity and specificity, SROC curve, sensitivity analysis plot and Deeks’ funnel plot of lncRNAs. (a) Forest plots of sensitivity
and specificity to evaluate the diagnostic performance of lncRNAs. (b) SROC curve to describe the diagnostic value of lncRNAs. (c) Sensitivity analysis to
estimate each study’s value of lncRNAs. (d) Deeks’ funnel plot to assess publication bias of lncRNAs. https://doi.org/10.1371/journal.pone.0250717.g004 Fig 4. Forest plots of sensitivity and specificity, SROC curve, sensitivity analysis plot and Deeks’ funnel plot of lncRNAs. (a) Forest plots of sensitivity
and specificity to evaluate the diagnostic performance of lncRNAs. (b) SROC curve to describe the diagnostic value of lncRNAs. (c) Sensitivity analysis to
estimate each study’s value of lncRNAs. (d) Deeks’ funnel plot to assess publication bias of lncRNAs. https://doi.org/10.1371/journal.pone.0250717.g004 https://doi.org/10.1371/journal.pone.0250717.g004 limited and further studies are needed (Fig 3e). The potential publication bias of the included
studies was evaluated using Deeks’ funnel plots, and its slope corresponded to p = 0.05 for
ctDNA and p = 0.03 for miRNAs, suggesting that there is no publication bias for ctDNA in
studies and a significant small publication bias for miRNA in studies (Figs 2f and 3f). Never-
theless, as only 4 articles were included for LncRNAs, although the Deeks’ funnel plot showed
a slope corresponding to p = 0.89 (Fig 4d), there was still publication bias because of the small
number of included articles. Discussion OC is the main cause of gynecological malignant cancer-associated deaths, accounting for 4%
of female malignant tumors worldwide [2, 5]. However, it has few obvious symptoms in the Table 2. Diagnostic efficacy of ctDNA, miRNAs and lncRNAs in OC. n
Sensitivity (95% CI)
Specificity (95% CI)
PLR (95% CI)
NLR (95% CI)
DOR (95% CI)
AUC (95% CI)
ctDNA
8
0.84 (0.67–0.93)
0.91 (0.81–0.96)
10.30 (3.60–29.80)
0.16 (0.08–0.33)
64.00 (16.00–247.00)
0.94 (0.92–0.96)
miRNAs
11
0.78 (0.69–0.84)
0.78 (0.63–0.88)
3.60 (2.20–5.80)
0.28 (0.19–0.42)
13.00 (6.00–29.00)
0.85 (0.82–0.88)
lncRNAs
4
0.81 (0.73–0.87)
0.78 (0.68–0.86)
3.62 (2.50–5.24)
0.25 (0.17–0.36)
15.30 (8.42–27.82)
0.86 (0.82–0.88)
ctDNA circulating tumor DNA; miRNAs circulating microRNAs; lncRNAs long non-coding RNAs; CI confidence interval; PLR positive likelihood ratio; NLR Table 2. Diagnostic efficacy of ctDNA, miRNAs and lncRNAs in OC. Table 2. Diagnostic efficacy of ctDNA, miRNAs and lncRNAs in OC. ctDNA, circulating tumor DNA; miRNAs, circulating microRNAs; lncRNAs, long non-coding RNAs; CI, confidence interval; PLR, positive likelihood ratio; NLR,
negative likelihood ratio; AUC, area under curve; DOR, diagnostic odds ratio. PLOS ONE | https://doi.org/10.1371/journal.pone.0250717
April 26, 2021 11 / 19 PLOS ONE A meta-analysis of ovarian cancer using ctDNA, miRNAs, and lncRNAs Fig 5. Meta-regression and Fagan’s Nomogram of ctDNA/miRNAs. (a) Meta-regression to explore heterogeneity between studies of ctDNA. (b) Meta-
regression to explore heterogeneity between studies of miRNAs. (c) Fagan’s Nomogram to reveal the clinical application of ctDNA in the identification of
OC patients and control individuals. (d) Fagan’s Nomogram to reveal the clinical application of miRNAs in the identification of OC patients and control
individuals. (e) Fagan’s Nomogram to reveal the clinical application of lncRNAs in the identification of OC patients and control individuals. https://doi.org/10.1371/journal.pone.0250717.g005 Fig 5. Meta-regression and Fagan’s Nomogram of ctDNA/miRNAs. (a) Meta-regression to explore heterogeneity between studies of ctDNA. (b) Meta-
regression to explore heterogeneity between studies of miRNAs. (c) Fagan’s Nomogram to reveal the clinical application of ctDNA in the identification of
OC patients and control individuals. (d) Fagan’s Nomogram to reveal the clinical application of miRNAs in the identification of OC patients and control
individuals. (e) Fagan’s Nomogram to reveal the clinical application of lncRNAs in the identification of OC patients and control individuals. https://doi.org/10.1371/journal.pone.0250717.g005 Fig 5. Meta-regression and Fagan’s Nomogram of ctDNA/miRNAs. (a) Meta-regression to explore heterogeneity between studies of ctDNA. (b) Meta-
regression to explore heterogeneity between studies of miRNAs. PLOS ONE | https://doi.org/10.1371/journal.pone.0250717
April 26, 2021 PLOS ONE PLOS ONE PLOS ONE Table 3. Subgroup analysis of the diagnostic efficacy of ctDNA in OC. n
Sensitivity (95%
CI)
Specificity (95%
CI)
PLR (95% CI)
NLR (95% CI)
DOR (95% CI)
AUC
Heterogeneity (I2; P
values)
Race
Asian
3
0.82 (0.74–0.88)
0.85 (0.79–0.90)
3.80 (1.22–
11.80)
0.27 (0.10–
0.72)
14.30 (2.10–97.21)
0.86
(88%; 0.000)
Caucasian
5
0.77 (0.71–0.82)
0.95 (0.94–0.96)
16.43 (5.23–
51.61)
0.13 (0.02–
0.82)
154.12 (29.18–
814.04)
0.97
(82%; 0.000)
Method
Multiplex-PCR and
sequencing
2
0.65 (NA)
0.99 (NA)
74.48 (NA)
0.36 (NA)
208.21 (NA)
NA
NA
Others
6
0.86 (0.81–0.91)
0.90 (0.88–0.92)
5.20 (2.48–
10.89)
0.19 (0.09–
0.40)
29.92 (8.06–111.09)
0.92
(85%; 0.000)
Pathology type
EOC
3
0.93 (0.89–0.96)
0.94 (0.92–0.96)
19.02 (6.44–
56.16)
0.07 (0.03–
0.17)
326.84 (49.62–
2152.93)
0.98
(83%; 0.003)
OC (Various types of OC)
5
0.65 (0.58–0.71)
0.93 (0.91–0.95)
5.65 (1.78–
17.88)
0.35 (0.20–
0.61)
20.80 (4.79–90.30)
0.86
(81%; 0.000)
Sample size
40
3
0.85 (0.76–0.92)
0.90 (0.86–0.93)
3.98 (0.94–
16.85)
0.23 (0.12–
0.45)
18.17 (2.30–143.80)
0.91
(86%; 0.001)
> 40
5
0.76 (0.71–0.81)
0.95 (0.93–0.96)
15.37 (5.78–
40.85)
0.15 (0.04–
0.62)
126.89 (20.28–
793.83)
0.97
(88%; 0.000)
Control size
30
3
0.78 (0.68–0.87)
0.66 (0.53–0.78)
2.10 (1.45–3.04)
0.38 (0.25–
0.59)
6.34 (2.92–13.79)
0.77
(0%; 0.830)
> 30
5
0.78 (0.73–0.83)
0.95 (0.94–0.96)
21.25 (9.59–
47.07)
0.11 (0.02–
0.65)
219.93 (73.95–
654.03)
0.98
(67%; 0.017)
CI, confidence interval; PLR, positive likelihood ratio; NLR, negative likelihood ratio; AUC, area under curve; DOR, diagnostic odds ratio; PCR, polymerase chain
reaction; NA, not available; EOC, epithelial ovarian cancer; OC, ovarian cancer. Table 3. Subgroup analysis of the diagnostic efficacy of ctDNA in OC. Subgroup analysis of the diagnostic efficacy of ctDNA in OC CI, confidence interval; PLR, positive likelihood ratio; NLR, negative likelihood ratio; AUC, area under curve; DOR, diagnostic odds ratio; PCR, polymerase chain
reaction; NA, not available; EOC, epithelial ovarian cancer; OC, ovarian cancer. CI, confidence interval; PLR, positive likelihood ratio; NLR, negative likelihood ratio; AUC, area under curve; DOR, diagno
reaction; NA, not available; EOC, epithelial ovarian cancer; OC, ovarian cancer. combination were 0.99 and 0.45, respectively, implying that serum CA199 and CA125 may not
be suitable to distinguish between OC patients and normal individuals. Therefore, traditional
diagnostic biomarkers for OC are not ideal for improving the diagnostic accuracy. ctDNA originates from tumor cells and can reflect the dynamic changes of tumors in real
time. PLOS ONE | https://doi.org/10.1371/journal.pone.0250717
April 26, 2021 Discussion (c) Fagan’s Nomogram to reveal the clinical application of ctDNA in the identification of
OC patients and control individuals. (d) Fagan’s Nomogram to reveal the clinical application of miRNAs in the identification of OC patients and control
individuals. (e) Fagan’s Nomogram to reveal the clinical application of lncRNAs in the identification of OC patients and control individuals. htt
//d i
/10 1371/j
l
0250717 005 Fig 5. Meta-regression and Fagan’s Nomogram of ctDNA/miRNAs. (a) Meta-regression to explore heterogeneity between studies of ctDNA. (b) Meta-
regression to explore heterogeneity between studies of miRNAs. (c) Fagan’s Nomogram to reveal the clinical application of ctDNA in the identification of
OC patients and control individuals. (d) Fagan’s Nomogram to reveal the clinical application of miRNAs in the identification of OC patients and control
individuals. (e) Fagan’s Nomogram to reveal the clinical application of lncRNAs in the identification of OC patients and control individuals. https://doi.org/10.1371/journal.pone.0250717.g005 https://doi.org/10.1371/journal.pone.0250717.g005 early stages, and most of patients are diagnosed in an advanced tumor stage, leading to high
mortality rates in recent years [24]. Furthermore, there is a large discrepancy between the
5-year survival rates of early and late-stage patients, indicating the urgent need for new diag-
nostic biomarkers for OC. Currently, the commonly used biomarkers for OC have low sensitivity or specificity. Zhen
et al. studied the diagnostic value of serum CA125 and human epididymis protein 4 (HE4) in
OC by constructing ROC curves (serum CA125: sensitivity, 0.74, specificity, 0.83; HE4: sensi-
tivity, 0.74, specificity, 0.90), and reported low sensitivity of these two biomarkers [57]. The
report by Guo et al. demonstrated that serum carbohydrate antigen 199 (CA199) had an
appropriate diagnostic sensitivity (0.73) but low specificity (0.43) [58]. In addition, they also
combined serum CA125 and serum CA199 for analysis, and sensitivity and specificity of this PLOS ONE | https://doi.org/10.1371/journal.pone.0250717
April 26, 2021 12 / 19 A meta-analysis of ovarian cancer using ctDNA, miRNAs, and lncRNAs PLOS ONE PLOS ONE Table 4. Subgroup analysis of the diagnostic efficacy of miRNAs in OC. n Sensitivity (95% CI) Specificity (95% CI)
PLR (95% CI)
NLR (95% CI)
DOR (95% CI)
AUC
Heterogeneity (I2; P
values)
Race
Asian
8
0.78 (0.75–0.80)
0.87 (0.85–0.88)
3.03 (1.36–6.73)
0.29 (0.17–
0.50)
10.98 (3.51–34.36)
0.84
(94%; 0.000)
Caucasian
3
0.73 (0.68–0.77)
0.86 (0.81–0.91)
5.50 (2.93–
10.34)
0.29 (0.15–
0.57)
21.99 (12.77–
37.86)
0.89
(0%; 0.729)
Specimen
Serum
9
0.76 (0.73–0.78)
0.87 (0.85–0.89)
3.72 (1.75–7.89)
0.30 (0.20–
0.47)
13.00 (4.72–35.76)
0.85
(94%; 0.000)
Plasma
1
0.71 (NA)
0.87 (NA)
5.36 (NA)
0.33 (NA)
16.26 (NA)
NA
NA
Urine
1
0.93 (NA)
0.58 (NA)
2.18 (NA)
0.13 (NA)
16.77 (NA)
NA
NA
Method
qRT–PCR (TaqMan)
6
0.69 (0.66–0.72)
0.68 (0.62–0.73)
2.64 (1.38–5.05)
0.43 (0.33–
0.56)
7.04 (2.71–18.26)
0.79
(86%; 0.000)
qRT–PCR (SYBR Green)
4
0.81 (0.77–0.85)
0.79 (0.74–0.83)
3.64 (2.90–4.56)
0.25 (0.19–
0.35)
14.87 (10.04–
22.01)
0.86
(0%; 0.606)
Microarrays
1
0.96 (NA)
0.93 (NA)
13.28 (NA)
0.05 (NA)
280.04 (NA)
NA
NA
Pathology type
EOC
9
0.73 (0.70–0.76)
0.72 (0.68–0.76)
2.99 (1.83–4.88)
0.36 (0.27–
0.47)
9.34 (4.55–19.15)
0.82
(84%; 0.000)
OC (Various types of
OC)
2
0.88 (NA)
0.92 (NA)
10.87 (NA)
0.13 (NA)
81.32 (NA)
NA
NA
Sample size
50
2
0.83 (NA)
0.78 (NA)
3.79 (NA)
0.22 (NA)
17.41 (NA)
NA
NA
> 50
9
0.76 (0.73–0.78)
0.87 (0.85–0.89)
3.72 (1.75–7.89)
0.30 (0.20–
0.47)
13.00 (4.72–35.76)
0.85
(94%; 0.000)
Control size
40
2
0.65 (NA)
0.57 (NA)
1.52 (NA)
0.62 (NA)
2.45 (NA)
NA
NA
> 40
9
0.77 (0.75–0.80)
0.87 (0.86–0.89)
4.41 (2.10–9.23)
0.28 (0.18–
0.41)
16.93 (6.55–43.71)
0.87
(93%; 0.000)
CI, confidence interval; PLR, positive likelihood ratio; NLR, negative likelihood ratio; AUC, area under curve; DOR, diagnostic odds ratio; NA, not available; qRT-PCR,
real-time fluorescence quantitative polymerase chain reaction; EOC, epithelial ovarian cancer; OC, ovarian cancer. https://doi org/10 1371/journal pone 0250717 t004 CI, confidence interval; PLR, positive likelihood ratio; NLR, negative likelihood ratio; AUC, area under curve; DOR, diagnostic odds ratio; NA, not available; qRT-PCR,
real-time fluorescence quantitative polymerase chain reaction; EOC, epithelial ovarian cancer; OC, ovarian cancer. CI, confidence interval; PLR, positive likelihood ratio; NLR, negative likelihood ratio; AUC, area under curve; DOR, dia
real-time fluorescence quantitative polymerase chain reaction; EOC, epithelial ovarian cancer; OC, ovarian cancer. (95% CI, 0.81–0.96). PLOS ONE | https://doi.org/10.1371/journal.pone.0250717
April 26, 2021 PLOS ONE miRNAs promote tumorigenesis by participating in the regulation of cell differentiation,
growth, apoptosis, and metabolism. Further, lncRNAs participate in carcinogenesis by regulat-
ing cell proliferation, division, differentiation, and metastasis. Thus, they may have potential
applications as biomarkers for the early detection of OC. However, the current studies focus-
ing on the diagnostic value of ctDNA/miRNAs/lncRNAs in OC are inconsistent. Differences
in specimen types (such as plasma or tissue), different inclusion criteria, and different detec-
tion techniques may explain the differences between different studies, but the lack of system-
atic evaluation complicates the conclusion. In addition, the understanding of the diversity
value of ctDNA/miRNAs/lncRNAs is vague. Hence, we conducted this comprehensive and
up-to-date study in a clinical context to further analyze the diagnostic value of ctDNA/miR-
NAs/lncRNAs. In this comprehensive meta-analysis, we included 23 diagnostic studies to investigate
whether ctDNA/miRNAs/lncRNAs are useful diagnostic molecular signatures for OC. We
noted that ctDNA had high accuracy for OC diagnosis, with a pooled AUC of 0.94 (95% CI,
0.92–0.96), a pooled sensitivity of 0.84 (95% CI, 0.67–0.93), and a pooled specificity of 0.91 PLOS ONE | https://doi.org/10.1371/journal.pone.0250717
April 26, 2021 13 / 19 A meta-analysis of ovarian cancer using ctDNA, miRNAs, and lncRNAs PLOS ONE The likelihood ratios (LRs) reflect the authenticity of sensitivity and speci-
ficity: the pooled PLR and NLR were 10.30 (95% CI, 3.60–29.80) and 0.16 (95% CI, 0.08–0.33),
respectively, indicating that OC patients have an approximately 10 times greater chance of
showing positive results in the ctDNA assay than healthy controls. When a true negative was
detected in a ctDNA assay-negative test, the error rate was approximately 16%. The DOR,
combining the advantages of sensitivity and specificity, is a significant indicator of diagnostic
accuracy. The pooled DOR of ctDNA was 64.00 (95% CI, 16.00–247.00) in our study, implying
that the diagnostic accuracy of ctDNA is high. Further, Fagan nomogram analysis and likeli-
hood ratio analysis verified the higher diagnostic performance of ctDNA in diagnosing OC. We also found that lncRNAs were more accurate than miRNAs in diagnosing OC. The pooled
AUC for lncRNAs was higher than that for miRNAs (0.86 vs. 0.85). In particular, the pooled
sensitivity of lncRNAs was higher than that of miRNAs (0.81 vs. 0.78), and the specificities of
these two biomarkers were quite similar (0.78 vs. 0.78). The likelihood ratio calculations con-
firmed that lncRNAs were similar to miRNAs in identifying OC (PLR, 3.62 vs. 3.60), whereas (95% CI, 0.81–0.96). The likelihood ratios (LRs) reflect the authenticity of sensitivity and speci-
ficity: the pooled PLR and NLR were 10.30 (95% CI, 3.60–29.80) and 0.16 (95% CI, 0.08–0.33),
respectively, indicating that OC patients have an approximately 10 times greater chance of
showing positive results in the ctDNA assay than healthy controls. When a true negative was
detected in a ctDNA assay-negative test, the error rate was approximately 16%. The DOR,
combining the advantages of sensitivity and specificity, is a significant indicator of diagnostic
accuracy. The pooled DOR of ctDNA was 64.00 (95% CI, 16.00–247.00) in our study, implying
that the diagnostic accuracy of ctDNA is high. Further, Fagan nomogram analysis and likeli-
hood ratio analysis verified the higher diagnostic performance of ctDNA in diagnosing OC. We also found that lncRNAs were more accurate than miRNAs in diagnosing OC. The pooled
AUC for lncRNAs was higher than that for miRNAs (0.86 vs. 0.85). In particular, the pooled
sensitivity of lncRNAs was higher than that of miRNAs (0.81 vs. 0.78), and the specificities of
these two biomarkers were quite similar (0.78 vs. 0.78). PLOS ONE The likelihood ratio calculations con-
firmed that lncRNAs were similar to miRNAs in identifying OC (PLR, 3.62 vs. 3.60), whereas PLOS ONE | https://doi.org/10.1371/journal.pone.0250717
April 26, 2021 14 / 19 PLOS ONE A meta-analysis of ovarian cancer using ctDNA, miRNAs, and lncRNAs lncRNAs were superior to miRNAs in ruling out OC (NLR, 0.25 vs. 0.28). lncRNAs exhibited a
higher pooled DOR compared with miRNAs (15.30 vs. 13.00). These results support the
hypothesis that ctDNA/miRNAs/lncRNAs may be effective biomarkers for the diagnosis of
OC. Additionally, combining these three biomarkers may help avoid the shortcomings of
using a single diagnostic biomarker with insufficient sensitivity or specificity. These results
also suggest the potential of combining ctDNA, miRNAs, and lncRNAs as diagnostic biomark-
ers for OC. Heterogeneity is a significant research issue in meta-analyses. In this meta-analysis, Q-test
and I2 statistical analysis revealed significant heterogeneity of the included studies. The thresh-
old effect is the primary factor affecting the heterogeneity of diagnostic meta-analyses. In this
study, the spearman correction coefficients of ctDNA/miRNAs/lncRNAs (ctDNA: -0.024,
p > 0.05; miRNAs: 0.118, p > 0.05; lncRNAs: 0.600, p > 0.05) indicated that there was no
threshold effect and the heterogeneity was caused by other reasons. To analyze the potential
sources of heterogeneity, we used meta-regression and subgroup analysis for investigating the
characteristics of the included studies, such as ethnic differences, sample types, case types, and
detection methods among others. We noted that for ctDNA, ctDNA could be detected more
accurately in the Caucasian population than in the Asian population. Subsequently, all EOC
cases had a higher diagnostic accuracy of ctDNA for prediction, as compared to OC cases. In
addition, a larger sample size showed a better diagnostic performance than a smaller sample
size. As far as miRNAs are concerned, the accuracy of detection using miRNAs in Asians was
inferior to that in Caucasians. Furthermore, the detection method of qRT-PCR (SYBR-Green)
was verified to have a significantly better performance with regard to OC prediction compared
with qRT-PCR (TaqMan). Publication bias was not significant for ctDNA, indicating that our
meta-analysis results are reliable. S1 Checklist. PRISMA 2009 checklist.
(DOC) S1 Checklist. PRISMA 2009 checklist. (DOC)
S1 Fig. Forest plots for studies on ctDNA/miRNAs used in the diagnosis of OC among rec-
ords (outlier records were excluded). (a) Forest plots of sensitivity and specificity of ctDNA
after excluding outlier record. (b) Forest plots of sensitivity and specificity of miRNAs after
excluding outlier record. (TIF)
S1 File. The search strategy of ctDNA, miRNAs and lncRNAs. (DOCX) S1 Fig. Forest plots for studies on ctDNA/miRNAs used in the diagnosis of OC among rec-
ords (outlier records were excluded). (a) Forest plots of sensitivity and specificity of ctDNA
after excluding outlier record. (b) Forest plots of sensitivity and specificity of miRNAs after
excluding outlier record. (TIF) Author Contributions Conceptualization: Linlin Zhang, Yang He. Conceptualization: Linlin Zhang, Yang He. Data curation: Linlin Zhang, Chenyan Hu, Zhongping Huang, Zhijia Li. Formal analysis: Zhijia Li. Project administration: Qin Zhang, Yang He. Supervision: Qin Zhang, Yang He. Writing – original draft: Linlin Zhang, Chenyan Hu, Yang He. Writing – review & editing: Linlin Zhang, Chenyan Hu, Qin Zhang, Yang He. Data curation: Linlin Zhang, Chenyan Hu, Zhongping Huang, Zhijia Li. Formal analysis: Zhijia Li. Project administration: Qin Zhang, Yang He. Supervision: Qin Zhang, Yang He. Writing
original draft: Linlin Zhang Chenyan Hu Yang He Data curation: Linlin Zhang, Chenyan Hu, Zhongping Huang, Zhijia Li. Formal analysis: Zhijia Li. Writing – original draft: Linlin Zhang, Chenyan Hu, Yang He. Writing – review & editing: Linlin Zhang, Chenyan Hu, Qin Zhang, Yang He. PLOS ONE However, the possible reasons for publication bias of miR-
NAs were as follows: (1) only English studies were included, indicating that language bias may
be the source of publication bias; (2) the evaluation index of the included studies was the con-
sistency of miRNA expression results and histopathological biopsy results, and there was a
possibility that the authors preferred to publish positive results; (3) the small number of
included cases may affect the accuracy of the statistical results. These results also indicate that
the current evidence cannot determine the best sources of heterogeneity (ctDNA/miRNAs) for
the reliable detection of OC. To confirm the findings of the current study, further large-scale
studies with different ethnic groups, sample types, case types, detection methods, and sample
sizes are required. Our current research does have some limitations. First, ctDNA/miRNAs/lncRNAs are
recently discovered tumor biomarkers; accordingly, the number of studies that could be
included in the study was relatively small, resulting in poor stability of some pooled analysis
results. This issue can be improved when more research data on these markers are available. In
addition, the included studies lack CA125 levels, so it cannot be ruled out whether CA125 test-
ing can complement the performance of accuracy of ctDNA/miRNAs/lncRNAs alone or in
combination. Besides, 11 of the 23 eligible studies were conducted in China, which may result
in a selection bias for specific study populations. Although we conducted subgroup analyses to
find the source of heterogeneity, we could not fully explain the heterogeneity. Furthermore,
some data were calculated based on the data extracted from ROC curves, which may not be as
powerful as the data obtained directly from articles. Despite these limitations, our meta-analysis has several important advantages. First, this
was a relatively comprehensive systematic study, wherein the diagnostic value of ctDNA/miR-
NAs/lncRNAs in OC was independently estimated and verified. Furthermore, our method
was strict and followed the guidelines for conducting and reporting systematic reviews. 15 / 19 PLOS ONE | https://doi.org/10.1371/journal.pone.0250717
April 26, 2021 PLOS ONE A meta-analysis of ovarian cancer using ctDNA, miRNAs, and lncRNAs Conclusion In summary, the current evidence indicates that ctDNA/miRNAs/lncRNAs are highly corre-
lated with OC, and may be potential and promising biomarkers for distinguishing OC patients
from healthy controls. Further large-scale studies are needed to verify the potential applicabil-
ity of ctDNA/miRNAs/lncRNAs alone or in combination as OC diagnostic molecular bio-
markers and to explore the potential factors that may affect the accuracy of ctDNA/miRNAs/
lncRNAs in OC diagnosis. PLOS ONE | https://doi.org/10.1371/journal.pone.0250717
April 26, 2021 References https://doi.org/10.1038/
511524a PMID: 25079538 9. Crowley E, Nicolantonio FD, Loupakis F, Bardelli A. Liquid biopsy: Monitoring cancer-genetics in the
blood. Nat Rev Clin Oncol. 2013; 10(8): 472. https://doi.org/10.1038/nrclinonc.2013.110 PMID:
23836314 10. Diehl F, Schmidt K, Choti MA, Romans K, Goodman S, Li M, et al. Circulating mutant DNA to assess
tumor dynamics. Nat Med. 2008; 14(9): 985–990. https://doi.org/10.1038/nm.1789 PMID: 18670422 11. Gerlinger M, Rowan AJ, Horswell S, Math M, Larkin J, Endesfelder D, et al. Intratumor heterogeneity
and branched evolution revealed by multiregion sequencing. N Engl J Med. 2012; 366(10): 883–892. https://doi.org/10.1056/NEJMoa1113205 PMID: 22397650 12. Heitzer E, Ulz P, Geigl JB. Circulating tumor DNA as a liquid biopsy for cancer. Clin Chem. 2015; 61(1):
112–123. https://doi.org/10.1373/clinchem.2014.222679 PMID: 25388429 13. Francis G, Stein S. Circulating cell-free tumour DNA in the management of cancer. Int J Mol Sci. 2015;
16(6): 14122–14142. https://doi.org/10.3390/ijms160614122 PMID: 26101870 14. Esquela KA, Slack J. Oncomirs miRNAs with a role in cancer. Nat Rev Cancer. 2006; 6: 259–569. https://doi.org/10.1038/nrc1840 PMID: 16557279 15. Sotiopoulou G, Pampalakis G, Lianidou E, Mourelatos Z. Emerging of microRNAs as molecular
switches in the integrated circuit of the cancer cell. RNA. 2009; 15(8): 1443–1461. https://doi.org/10. 1261/rna.1534709 PMID: 19561119 16. Du W, Feng Z, Sun Q. LncRNA LINC00319 accelerates ovarian cancer progression through miR-423-
5p/NACC1 pathway. Biochem Bioph Res Co. 2018; 507(1–4): 198–202. https://doi.org/10.1016/j.bbrc. 2018.11.006 PMID: 30442370 17. Liu B, Ye B, Yang L, Zhu X, Huang G, Zhu P, et al. Long noncoding RNA lncKdm2b is required for ILC3
maintenance by initiation of Zfp292 expression. Nat Immunol. 2017; 18(5): 499e508. https://doi.org/10. 1038/ni.3712 PMID: 28319097 18. Zhu P, Wang Y, Huang G, Ye B, Liu B, Wu J, et al. lnc-β-Catm elicits EZH2-dependent beta-catenin sta-
bilization and sustains liver CSC self-renewal. Nat Struct Mol Biol. 2016; 23(7): 631e639. https://doi.org/
10.1038/nsmb.3235 PMID: 27239797 19. Spizzo R, Almeida MI, Colombatti A, Calin GA. Long non-coding RNAs and cancer: A new frontier of
translational research? Oncogene. 2012; 31(43): 4577e4587. https://doi.org/10.1038/onc.2011.621
PMID: 22266873 20. Yu H, Xu Y, Zhang D, Liu G. Long noncoding RNA LUCAT1 promotes malignancy of ovarian cancer
through regulation of miR-612/HOXA13 pathway. Biochem Biophys Res Commun. 2018; 503(3):
2095–2100. https://doi.org/10.1016/j.bbrc.2018.07.165 PMID: 30107913 21. Mitchell PS, Parkin RK, Kroh EM, Fritz BR, Wyman SK, Pogosova–Agadjanyan EL, et al. Circulating
microRNAs as stable blood-based markers for cancer detection. Proc Natl Acad Sci USA. 2008; 105:
10513–10518. https://doi.org/10.1073/pnas.0804549105 PMID: 18663219 22. References 1. Meryet-Figuière M, Lambert B, Gauduchon P, Vigneron N, Brotin E, Poulain L, et al. An overview of
long non–coding RNAs in ovarian cancers. Oncotarget. 2016; 7(28): 44719–44734. https://doi.org/10
18632/oncotarget.8089 PMID: 26992233
2. Moyle P, Sala E. Malignant neoplasms, staging of ovarian cancer. Abdom Imaging. 2013; 107(2):
2029–2040. 3. Collinson F, Qian W, Fossati R, Lissoni A, Williams C, Parmar M, et al. ICON1 collaborators. Optimal
treatment of early-stage ovarian cancer. Ann Oncol. 2014; 25(6): 1165–1171. https://doi.org/10.1093/
annonc/mdu116 PMID: 24631948
4. Tew WP. Ovarian cancer in the older woman. J Geriatr Oncol. 2016; 7(5): 354–361. https://doi.org/10
1016/j.jgo.2016.07.008 PMID: 27499341
5. Wang X, Kong D, Wang C, Ding X, Zhang L, Zhao M, et al. Circulating microRNAs as novel potential
diagnostic biomarkers for ovarian cancer: A systematic review and updated meta-analysis. J Ovarian
Res. 2019; 12(1): 24. https://doi.org/10.1186/s13048-019-0482-8 PMID: 30898156
6. Li B, Pu K, Ge L, Wu X. Diagnostic significance assessment of the circulating cell-free DNA in ovarian
cancer: An updated meta-analysis. Gene. 2019; 714: 143993. https://doi.org/10.1016/j.gene.2019. 143993 PMID: 31330238 1. Meryet-Figuière M, Lambert B, Gauduchon P, Vigneron N, Brotin E, Poulain L, et al. An overview of
long non–coding RNAs in ovarian cancers. Oncotarget. 2016; 7(28): 44719–44734. https://doi.org/10. 18632/oncotarget.8089 PMID: 26992233 2. Moyle P, Sala E. Malignant neoplasms, staging of ovarian cancer. Abdom Imaging. 2013; 107(2):
2029–2040. 3. Collinson F, Qian W, Fossati R, Lissoni A, Williams C, Parmar M, et al. ICON1 collaborators. Optimal
treatment of early-stage ovarian cancer. Ann Oncol. 2014; 25(6): 1165–1171. https://doi.org/10.1093/
annonc/mdu116 PMID: 24631948 4. Tew WP. Ovarian cancer in the older woman. J Geriatr Oncol. 2016; 7(5): 354–361. https://doi.org/10. 1016/j.jgo.2016.07.008 PMID: 27499341 5. Wang X, Kong D, Wang C, Ding X, Zhang L, Zhao M, et al. Circulating microRNAs as novel potential
diagnostic biomarkers for ovarian cancer: A systematic review and updated meta-analysis. J Ovarian
Res. 2019; 12(1): 24. https://doi.org/10.1186/s13048-019-0482-8 PMID: 30898156 6. Li B, Pu K, Ge L, Wu X. Diagnostic significance assessment of the circulating cell-free DNA in ovarian
cancer: An updated meta-analysis. Gene. 2019; 714: 143993. https://doi.org/10.1016/j.gene.2019. 143993 PMID: 31330238 16 / 19 PLOS ONE | https://doi.org/10.1371/journal.pone.0250717
April 26, 2021 PLOS ONE A meta-analysis of ovarian cancer using ctDNA, miRNAs, and lncRNAs 7. Moss EL, Hollingworth J, Reynolds TM. The role of CAl25 in clinical practice. J Clin Pathol. 2005; 58:
308–312. https://doi.org/10.1136/jcp.2004.018077 PMID: 15735166 8. Yong E. Cancer biomarkers: Written in blood. Nature. 2014; 511(7511): 524. References Cristiano S, Leal A, Phallen J, Fiksel J, Adleff V, Bruhm DC, et al. Genome-wide cell-free DNA fragmen-
tation in patients with cancer. Nature. 2019; 570(7761): 385–389. https://doi.org/10.1038/s41586-019-
1272-6 PMID: 31142840 35. Douville C, Cohen JD, Ptak J, Popoli M, Schaefer J, Silliman N, et al. Assessing aneuploidy with repeti-
tive element sequencing. Proc Natl Acad Sci USA. 2020; 117(9): 4858–4863. https://doi.org/10.1073/
pnas.1910041117 PMID: 32075918 36. Dvorska´ D, Brany´ D, Nagy B, Grenda´r M, Poka R, Solte´sz B, et al. Aberrant methylation status of
tumour suppressor genes in ovarian cancer tissue and paired plasma samples. Int J Mol Sci. 2019; 20
(17). https://doi.org/10.3390/ijms20174119 PMID: 31450846 37. Wang B, Yu L, Luo X, Huang L, Li QS, Shao XS, et al. Detection of OPCML methylation, a possible epi-
genetic marker, from free serum circulating DNA to improve the diagnosis of early-stage ovarian epithe-
lial cancer. Oncol Lett. 2017; 14(1): 217–223. https://doi.org/10.3892/ol.2017.6111 PMID: 28693156 38. Wang Y, Li L, Douville C, Cohen JD, Yen TT, Kinde I, et al. Evaluation of liquid from the Papanicolaou
test and other liquid biopsies for the detection of endometrial and ovarian cancers. Sci Transl Med. 2018; 10(433). https://doi.org/10.1126/scitranslmed.aap8793 PMID: 29563323 39. Yu Z, Qin S, Wang H. Alter circulating cell-free DNA variables in plasma of ovarian cancer patients. J
Obstet Gynaecol Res. 2019; 45(11): 2237–2242. https://doi.org/10.1111/jog.14102 PMID: 31502389 40. Zhang R, Pu WY, Zhang SY, Chen L, Zhu WP, Xiao L, et al. Clinical value of ALU concentration and
integrity index for the early diagnosis of ovarian cancer: A retrospective cohort trial. PLoS One. 2018; 13
(2): e0191756. https://doi.org/10.1371/journal.pone.0191756 PMID: 29401471 41. Chen ZH, Guo XL, Sun SK, Lu CX, Wang LM. Serum miR-125b levels associated with epithelial ovarian
cancer (EOC) development and treatment responses. Bioengineered. 2020; 11(1): 311–317. https://
doi.org/10.1080/21655979.2020.1736755 PMID: 32107956 42. Kobayashi M, Sawada K, Nakamura K, Yoshimura A, Miyamoto M, Shimizu A, et al. Exosomal miR-
1290 is a potential biomarker of high-grade serous ovarian carcinoma and can discriminate patients
from those with malignancies of other histological types. J Ovarian Res. 2018; 11(1): 81. https://doi.org/
10.1186/s13048-018-0458-0 PMID: 30219071 43. Liang J, Yang X, Liu L, Qiao L, Peng P, Zhou J. Combined measurement of miRNA-183, HE4, and CA-
125 increases diagnostic efficiency for ovarian cancer. Eur J Gynaecol Oncol. 2020; 41(1): 30–35. 44. Ma´rton E´ , Luka´cs J, Penyige A, Janka E, Hegedu¨s L, Solte´sz B, et al. References Yang JP, Yang XJ, Xiao L, Wang Y. Long noncoding RNA PVT1 as a novel serum biomarker for detec-
tion of cervical cancer. Eur Rev Med Pharmaco. 2016; 20(19): 3980. PMID: 27775803 23. Moher D, Liberati A, Tetzlaff J, Altman DG, Group P. Preferred reporting items for systematic reviews
and meta-analyses: the PRISMA statement. Ann Intern Med. 2009; 151(4): 264–269. https://doi.org/10. 7326/0003-4819-151-4-200908180-00135 PMID: 19622511 24. Zhou Q, Li W, Leng B, Zheng W, He Z, Zuo M, et al. Circulating Cell Free DNA as the Diagnostic Marker
for Ovarian Cancer: A Systematic Review and Meta-Analysis. PLoS One. 2016; 11(6): e0155495. https://doi.org/10.1371/journal.pone.0155495 PMID: 27253331 25. Liang HC, Jiang ZP, Xie GE, Lu Y. Serum microRNA-145 as a novel biomarker in human ovarian can-
cer. Tumour Biol. 2015; 36(7): 5305–5313. https://doi.org/10.1007/s13277-015-3191-y PMID:
25722112 26. Whiting PF, Rutjes AW, Westwood ME, Mallett S, Deeks JJ, Reitsma JB, et al. Group Q-: QUADAS-2:
a revised tool for the quality assessment of diagnostic accuracy studies. Ann Intern Med. 2011; 155(8):
529–36. https://doi.org/10.7326/0003-4819-155-8-201110180-00009 PMID: 22007046 27. Hellmich M, Lehmacher W. A ruler for interpreting diagnostic test results. Methods Inf. Med. 2005; 44:
124–126. PMID: 15778803 28. Walusimbi S, Bwanga F, De Costa A, Haile M, Joloba M, Hoffner S. Meta-analysis to compare the accu-
racy of GeneXpert, MODS and the WHO 2007 algorithm for diagnosis of smear-negative pulmonary
tuberculosis. BMC Infect Dis. 2013; 13: 507. https://doi.org/10.1186/1471-2334-13-507 PMID:
24172543 17 / 19 PLOS ONE | https://doi.org/10.1371/journal.pone.0250717
April 26, 2021 PLOS ONE A meta-analysis of ovarian cancer using ctDNA, miRNAs, and lncRNAs 29. Higgins JP, Thompson SG, Deeks JJ, Altman DG. Measuring inconsistency in meta-analyses. BMJ. 2003; 327(7414): 557–560. https://doi.org/10.1136/bmj.327.7414.557 PMID: 12958120 30. Zamora J, Abraira V, Muriel A, Khan K, Coomarasamy A. Meta-DiSc: a software for meta-analysis of
test accuracy data. BMC Med Res Methodol. 2006; 6: 31. https://doi.org/10.1186/1471-2288-6-31
PMID: 16836745 31. DerSimonian R. Meta-analysis in the design and monitoring of clinical trials. Stat Med. 1996; 15: 1237–
1248. https://doi.org/10.1002/(SICI)1097-0258(19960630)15:12<1237::AID-SIM301>3.0.CO;2-N
PMID: 8817798 32. Deeks JJ, Macaskill P, Irwig L. The performance of tests of publication bias and other sample size
effects in systematic reviews of diagnostic test accuracy was assessed. J Clin Epidemiol. 2005; 58(9):
882–893. https://doi.org/10.1016/j.jclinepi.2005.01.016 PMID: 16085191 33. Cohen JD, Li L, Wang Y, Thoburn C, Afsari B, Danilova L, et al. Detection and localization of surgically
respectable cancers with a multi-analyte blood test. Science. 2018; 359(6378): 926–930. https://doi. org/10.1126/science.aar3247 PMID: 29348365 34. References Circulating epithelial–mesenchy-
mal transition-associated miRNAs are promising biomarkers in ovarian cancer. J Biotechnol. 2019; 297:
58–65. https://doi.org/10.1016/j.jbiotec.2019.04.003 PMID: 30953675 45. Meng XD, Joosse SA, Muller V, Trillsch F, Milde-Langosch K, Mahner S, et al. Diagnostic and prognostic
potential of serum miR-7, miR-16, miR-25, miR-93, miR-182, miR-376a and miR-429 in ovarian cancer
patients. Br J Cancer. 2015; 113(9): 1358–1366. https://doi.org/10.1038/bjc.2015.340 PMID: 26393886 46. Todeschini P, Salviato E, Paracchini L, Ferracin M, Petrillo M, Zanotti L, et al. Circulating miRNA land-
scape identifies miR-1246 as promising diagnostic biomarker in high-grade serous ovarian carcinoma:
A validation across two independent cohorts. Cancer Lett. 2017; 388: 320–327. https://doi.org/10.1016/
j.canlet.2016.12.017 PMID: 28017893 47. Yokoi A, Matsuzaki J, Yamamoto Y, Yoneoka Y, Takahashi K, Shimizu H, et al. Integrated extracellular
microRNA profiling for ovarian cancer screening. Nat Commun. 2018; 9(1): 4319. https://doi.org/10. 1038/s41467-018-06434-4 PMID: 30333487 48. Yokoi A, Yoshioka Y, Hirakawa A, Yamamoto Y, Ishikawa M, Ikeda S, et al. A combination of circulating
miRNAs for the early detection of ovarian cancer. Oncotarget. 2017; 8(52): 89811–89823. https://doi. org/10.18632/oncotarget.20688 PMID: 29163790 18 / 19 PLOS ONE | https://doi.org/10.1371/journal.pone.0250717
April 26, 2021 PLOS ONE A meta-analysis of ovarian cancer using ctDNA, miRNAs, and lncRNAs 49. Zhou J, Gong G, Tan H, Dai F, Zhu X, Chen Y, et al. Urinary microRNA-30a-5p is a potential biomarker
for ovarian serous adenocarcinoma. Oncol Rep. 2015; 33(6): 2915–2923. https://doi.org/10.3892/or. 2015.3937 PMID: 25962395 50. Zhu T, Gao W, Chen X, Zhang Y, Wu M, Zhang P, et al. A Pilot Study of Circulating MicroRNA-125b as
a Diagnostic and Prognostic Biomarker for Epithelial Ovarian Cancer. Int J Gynecol Cancer. 2017; 27
(1): 3–10. https://doi.org/10.1097/IGC.0000000000000846 PMID: 27636713 51. Zuberi M, Mir R, Das J, Ahmad I, Javid J, Yadav P, et al. Expression of serum miR-200a, miR-200b,
and miR-200c as candidate biomarkers in epithelial ovarian cancer and their association with clinico-
pathological features. Clin Transl Oncol. 2015; 17(10):779–787. https://doi.org/10.1007/s12094-015-
1303-1 PMID: 26063644 52. Cui Y, Shen G, Zhou D, Wu F. CASC11 Overexpression Predicts Poor Prognosis and Regulates Cell
Proliferation and Apoptosis in Ovarian Carcinoma. Cancer Manag Res. 2020; 12: 523–529. https://doi. org/10.2147/CMAR.S226801 PMID: 32158258 53. Gong J, Xu X, Zhang X, Zhou Y. LncRNA MIR4435-2HG is a potential early diagnostic marker for ovar-
ian carcinoma. Acta Biochim Biophys Sin (Shanghai). 2019; 51(9): 953–959. https://doi.org/10.1093/
abbs/gmz085 PMID: 31435668 54. Yang M, Zhai Z, Guo S, Li X, Zhu Y, Wang Y. PLOS ONE | https://doi.org/10.1371/journal.pone.0250717
April 26, 2021 References Long non-coding RNA FLJ33360 participates in ovarian
cancer progression by sponging miR-30b-3p. Onco Targets Ther. 2019; 12: 4469–4480. https://doi.org/
10.2147/OTT.S205622 PMID: 31239715 55. Yang M, Zhai Z, Zhang Y, Wang Y. Clinical significance and oncogene function of long noncoding RNA
HAGLROS overexpression in ovarian cancer. Arch Gynecol Obstet. 2019; 300(3): 703–710. https://doi. org/10.1007/s00404-019-05218-5 PMID: 31197441 56. Zhou L, Chen Y, Gao J, Shankar S, Zhang G. Diagnostic value of circulating microRNAs for endometri-
osis: A meta-analysis. Reprod Sci. 2020; 27(3): 793–805. https://doi.org/10.1007/s43032-019-00024-5
PMID: 32096023 57. Zhen S, Bian LH, Chang LL, Gao X. Comparison of serum human epididymis protein 4 and carbohy-
drate antigen 125 as markers in ovarian cancer: A meta-analysis. Mol Clin Oncol. 2014; 2(4): 559–566. https://doi.org/10.3892/mco.2014.279 PMID: 24940495 58. Guo B, Lian W, Liu S, Cao Y, Liu J. Comparison of diagnostic values between CA125 combined with
CA199 and ultrasound combined with CT in ovarian cancer. Oncol Lett. 2019; 17(6): 5523–5528. https://doi.org/10.3892/ol.2019.10264 PMID: 31186773 19 / 19
|
https://openalex.org/W2736388094
|
https://univ-angers.hal.science/hal-02872796/document
|
English
| null |
Mn / Ca intra-test variability in the benthic foraminifer &lt;i&gt;Ammonia tepida&lt;/i&gt;
| null | 2,017
|
cc-by
| 15,440
|
To cite this version: Jassin Petersen, Christine Barras, Antoine Bézos, Carole La, Lennart Jan de Nooijer, et al.. Mn / Ca
intra-test variability in the benthic foraminifer Ammonia tepida. Biogeosciences Discussions, 2018,
15, pp.331-348. 10.5194/bg-2017-273. hal-02872796 Mn / Ca intra-test variability in the benthic foraminifer
Ammonia tepida
Jassin Petersen, Christine Barras, Antoine Bézos, Carole La, Lennart Jan de
Nooijer, Filip Meysman, Aurélia Mouret, Caroline Slomp, Frans Jorissen Distributed under a Creative Commons Attribution 4.0 International License Correspondence: Jassin Petersen (jassin.petersen@univ-nantes.fr) Correspondence: Jassin Petersen (jassin.petersen@univ-nantes.fr) Received: 28 June 2017 – Discussion started: 21 July 2017 Received: 28 June 2017 – Discussion started: 21 July 2017 Revised: 21 November 2017 – Accepted: 29 November 2017 – Published: 16 January 2018 Revised: 21 November 2017 – Accepted: 29 November 2017 – Published: 16 January 2018 Abstract. The adaptation of some benthic foraminiferal
species to low-oxygen conditions provides the prospect of
using the chemical composition of their tests as proxies
for bottom water oxygenation. Manganese may be particu-
larly suitable as such a geochemical proxy because this re-
dox element is soluble in reduced form (Mn2+) and hence
can be incorporated into benthic foraminiferal tests under
low-oxygen conditions. Therefore, intra- and inter-test dif-
ferences in foraminiferal Mn/Ca ratios may hold impor-
tant information about short-term variability in pore water
Mn2+ concentrations and sediment redox conditions. Here,
we studied Mn/Ca intra- and inter-test variability in liv-
ing individuals of the shallow infaunal foraminifer Ammo-
nia tepida sampled in Lake Grevelingen (the Netherlands)
in three different months of 2012. The deeper parts of this
lake are characterized by seasonal hypoxia/anoxia with asso-
ciated shifts in microbial activity and sediment geochemistry,
leading to seasonal Mn2+ accumulation in the pore water. Earlier laboratory experiments with similar seawater Mn2+
concentrations as encountered in the pore waters of Lake
Grevelingen suggest that intra-test variability due to ontoge-
netic trends (i.e. size-related effects) and/or other vital effects
occurring during calcification in A. tepida (11–25 % relative
SD, RSD) is responsible for part of the observed variability in
Mn/Ca. Our present results show that the seasonally highly dynamic environmental conditions in the study area lead to
a strongly increased Mn/Ca intra- and inter-test variability
(average of 45 % RSD). Within single specimens, both in-
creasing and decreasing trends in Mn/Ca ratios with size are
observed. Our results suggest that the variability in succes-
sive single-chamber Mn/Ca ratios reflects the temporal vari-
ability in pore water Mn2+. Additionally, active or passive
migration of the foraminifera in the surface sediment may
explain part of the observed Mn/Ca variability. HAL Id: hal-02872796
https://univ-angers.hal.science/hal-02872796v1
Submitted on 3 Jun 2022 L’archive ouverte pluridisciplinaire HAL, est
destinée au dépôt et à la diffusion de documents
scientifiques de niveau recherche, publiés ou non,
émanant des établissements d’enseignement et de
recherche français ou étrangers, des laboratoires
publics ou privés. HAL is a multi-disciplinary open access
archive for the deposit and dissemination of sci-
entific research documents, whether they are pub-
lished or not. The documents may come from
teaching and research institutions in France or
abroad, or from public or private research centers. Distributed under a Creative Commons Attribution 4.0 International License Biogeosciences, 15, 331–348, 2018
https://doi.org/10.5194/bg-15-331-2018
© Author(s) 2018. This work is distributed under
the Creative Commons Attribution 4.0 License. 1
Introduction In many coastal ecosystems, high summer temperatures and
eutrophication lead to seasonally occurring hypoxia ([O2] <
63 µM; Rabalais et al., 2002; Diaz and Rosenberg, 2008),
linked to the emergence of water column stratification in
combination with lower oxygen solubility and higher respira-
tion rates in warmer waters (e.g. Keeling et al., 2010). On the
seafloor, oxygen is consumed by respiration of marine biota,
thereby coupling bottom water oxygenation (BWO), benthic
ecosystem functioning and organic carbon cycling (Altabet
et al., 1995; Levin et al., 2009; Koho et al., 2013). Most of the organic matter (OM) deposited on the seafloor
is mineralized in surface sediments by respiration processes, 3Department of Estuarine & Delta Systems, NIOZ Royal Netherlands Institute for Sea Research and Utrecht University,
Korringaweg 7, 4401 NT Yerseke, the Netherlands 3Department of Estuarine & Delta Systems, NIOZ Royal Netherlands Institute for Sea Research and Utrecht University,
Korringaweg 7, 4401 NT Yerseke, the Netherlands 3Department of Estuarine & Delta Systems, NIOZ Royal Netherlands Institute for Sea Research and Utrecht University,
Korringaweg 7, 4401 NT Yerseke, the Netherlands 4Department of Biology, University of Antwerp, Universiteitsplein 1, 2610 Wilrijk, Belgium
5Department of Earth Sciences (Geochemistry), Faculty of Geosciences, Utrecht University, 3508 TA Utrecht, the Netherland 4Department of Biology, University of Antwerp, Universiteitsplein 1, 2610 Wilrijk, Belgium
5Department of Earth Sciences (Geochemistry), Faculty of Geosciences, Utrecht University, 3508 TA Utrecht, the Netherlands Mn/Ca intra- and inter-test variability in the benthic foraminifer
Ammonia tepida Jassin Petersen1, Christine Barras1, Antoine Bézos1, Carole La1, Lennart J. de Nooijer2, Filip J. R. Meysman3,4,
Aurélia Mouret1, Caroline P. Slomp5, and Frans J. Jorissen1 1LPG UMR CNRS 6112, University of Nantes, University of Angers, UFR Sciences, 2 Boulevard Lavoisier,
49045 Angers CEDEX 01, France g
ment of Ocean Systems, NIOZ Royal Netherlands Institute for Sea Research and Utrecht University
g, the Netherlands 2Department of Ocean Systems, NIOZ Royal Netherlands Institute for Sea Research and Utrecht University,
Den Burg, the Netherlands J. Petersen et al.: Mn/Ca intra- and inter-test variability In such cases, single-chamber measure-
ments could provide information about the temporal variabil-
ity in the Mn dynamics. A first culturing study of Ammo-
nia tepida with different concentrations of seawater Mn2+
(in the range of naturally occurring concentrations in seawa-
ter, which is up to 3 orders of magnitude lower than con-
centrations found in pore waters of our coastal study area)
was performed by Munsel et al. (2010). A more recent cul-
turing study of Ammonia tepida and Bulimina marginata evi-
dences a species-specific linear relationship between Mn/Ca
ratios of the tests and a range of dissolved Mn2+ concen-
trations commonly found in pore waters of coastal areas
(Mn/Cacalcite = 0.086 ± 0.001 Mn/Caseawater for A. tepida,
R2 = 0.96, p < 0.001; Barras et al., 2018). This culturing
study found that Mn/Ca intrinsic variability varied between
11 and 25 % for A. tepida for a range of seawater Mn2+ con-
centrations similar to those found in the pore water of surface
sediments in our study area. )
Sedimentary records of manganese may reflect a variety of
past environmental conditions, including bottom water redox
state, continental runoff, surface water productivity and bot-
tom water current dynamics (Reichart et al., 1997; van der
Weijden et al., 2006; Lenz et al., 2015). The use of sedimen-
tary manganese as a proxy is complicated because of its com-
plex biogeochemical dynamics, including the remobilization
of once precipitated manganese oxides, and subsequent dia-
genetic overprinting (Schenau et al., 2002). Mn incorporated
into the calcite of benthic foraminifera potentially represents
a more reliable proxy of redox conditions, since these ma-
rine protists build their shells (tests) in the upper sediment
layer, where the presence of Mn2+ may be a function of bot-
tom water oxygenation, and Mn2+ can be permanently in-
corporated into the tests. However, the ecology of the various
foraminiferal species is crucial and can influence manganese
incorporation into the shell, since different adaptation strate-
gies to changes in the availability of OM and of oxygen lead
to different microhabitats and probably also to different cal-
cification depths and periods (Jorissen et al., 1995; Van der
Zwaan et al., 1999; Koho et al., 2015). So far, few studies
have investigated the potential of using Mn/Ca ratios in ben-
thic foraminifera as a proxy for bottom water redox state. J. Petersen et al.: Mn/Ca intra- and inter-test variability Dissolved Mn2+ can
then diffuse upwards in the pore water across the oxic–anoxic
boundary, where it precipitates again in the form of Mn ox-
ides, leading to a continuous cycling of Mn within the upper
sediment (Aller, 1994; Slomp et al., 1997). The sedimenta-
tion rate and the bioturbation intensity are important factors
controlling the cycling of various reactive compounds in sed-
iments, such as metal oxides (van de Velde and Meysman,
2016). Macrofaunal bioturbation may introduce Mn oxides
in the deeper, anoxic sediment, where these minerals are sub-
sequently reduced (Mouret et al., 2009; Thibault de Chan-
valon et al., 2016). When eutrophication and stratification of
the water column lead to (seasonally) hypoxic bottom water
conditions, the oxygen penetration depth is reduced, causing
an upward movement of the Mn redox front and diminishing
the possibility that pore water Mn2+ is oxidized. In extreme
cases, Mn2+ may diffuse from the pore water into the wa-
ter column (Sundby and Silverberg, 1985; Thamdrup et al.,
1994; Dellwig et al., 2007; Konovalov et al., 2007; Pakho-
mova et al., 2007; Kowalski et al., 2012). Barras et al., 2018). For example, Koho et al. (2015, 2017)
demonstrated that foraminiferal species occupying a rela-
tively deep microhabitat display higher Mn/Ca ratios than
those living near the sediment surface. Furthermore, by using
microanalytical techniques, capable of measuring elemental
concentrations in single foraminiferal chambers, short-term
variability in oxygenation may be unravelled. Recent studies
have shown the potential of intra-test variability in Mn/Ca
to resolve vertical migration in the sediment and/or seasonal
changes in oxygenation (Glock et al., 2012; McKay et al.,
2015). In order to apply Mn/Ca ratios in benthic foraminiferal
tests as a quantifiable proxy of palaeo-redox conditions, it is
necessary to assess the presence of ontogenetic trends (i.e. size-related effects) and/or other vital effects occurring dur-
ing calcification; the variability due to such biological pro-
cesses will hereafter be termed “intrinsic (intra-test) vari-
ability”. This intrinsic part of the total variability (which
can be assessed only in culturing studies) determines the
threshold value above which the variability within a sin-
gle specimen can be recognized successfully as a response
to environmental conditions. In the case of the total intra-
test variability being greater than the intrinsic variability, the
added part can be ascribed to environmental factors and/or
active or passive migration of the foraminifera in the sedi-
ment surface layer. J. Petersen et al.: Mn/Ca intra- and inter-test variability J. Petersen et al.: Mn/Ca intra- and inter-test variability involving oxygen, nitrate, manganese, and iron oxides or sul-
fate as electron acceptors (Froelich et al., 1979). If Fe and
Mn (hydr)oxides are sufficiently abundant, their reduction
can be relevant for the total OM decomposition in the sedi-
ment (Aller, 1990; Burdige, 1993; Canfield, 1993; Vandieken
et al., 2006). When the bottom waters of coastal water bod-
ies are oxygenated, Mn oxides are present in the oxic sur-
face layer of the sediment and are reduced to soluble Mn2+
in the deeper anoxic sediment layers. Dissolved Mn2+ can
then diffuse upwards in the pore water across the oxic–anoxic
boundary, where it precipitates again in the form of Mn ox-
ides, leading to a continuous cycling of Mn within the upper
sediment (Aller, 1994; Slomp et al., 1997). The sedimenta-
tion rate and the bioturbation intensity are important factors
controlling the cycling of various reactive compounds in sed-
iments, such as metal oxides (van de Velde and Meysman,
2016). Macrofaunal bioturbation may introduce Mn oxides
in the deeper, anoxic sediment, where these minerals are sub-
sequently reduced (Mouret et al., 2009; Thibault de Chan-
valon et al., 2016). When eutrophication and stratification of
the water column lead to (seasonally) hypoxic bottom water
conditions, the oxygen penetration depth is reduced, causing
an upward movement of the Mn redox front and diminishing
the possibility that pore water Mn2+ is oxidized. In extreme
cases, Mn2+ may diffuse from the pore water into the wa-
ter column (Sundby and Silverberg, 1985; Thamdrup et al.,
1994; Dellwig et al., 2007; Konovalov et al., 2007; Pakho-
mova et al., 2007; Kowalski et al., 2012). involving oxygen, nitrate, manganese, and iron oxides or sul-
fate as electron acceptors (Froelich et al., 1979). If Fe and
Mn (hydr)oxides are sufficiently abundant, their reduction
can be relevant for the total OM decomposition in the sedi-
ment (Aller, 1990; Burdige, 1993; Canfield, 1993; Vandieken
et al., 2006). When the bottom waters of coastal water bod-
ies are oxygenated, Mn oxides are present in the oxic sur-
face layer of the sediment and are reduced to soluble Mn2+
in the deeper anoxic sediment layers. Published by Copernicus Publications on behalf of the European Geosciences Union. Published by Copernicus Publications on behalf of the European Geosciences Union. 332 2.1
Study area Sediment samples were recovered in Lake Grevelingen at
a single location (51◦44.956′ N, 03◦53.826′ E; water depth
23.1 m; Fig. S1). The sampling site experiences seasonal hy-
poxia; for the year 2012, monthly recordings of temperature,
salinity and water column oxygen concentrations (Hagens
et al., 2015), sedimentary microbial community composition
(Seitaj et al., 2015), pore water geochemistry (Sulu-Gambari
et al., 2016a, b), and benthic O2 uptake rates (Seitaj et al.,
2017) are available. In 2012, BWO started to decrease in
April and attained a minimum of about 20 µM (∼8 % sat-
uration) in August (Fig. S2). After homogenization of the
water column in September/October, BWO quickly rose to
values of ∼200 µM (∼80 % saturation). Pore water Mn2+
showed highest concentrations (up to 310 µM) in winter and
early spring, followed by considerably lower concentrations
in summer and autumn (Sulu-Gambari et al., 2016b; Fig. S3). J. Petersen et al.: Mn/Ca intra- and inter-test variability These cable bacteria tend
to have a highly patchy distribution, which hence could com-
plicate the interpretation (Sulu-Gambari et al., 2016b; Seitaj
et al., 2017). Figure 1. Scanning electron microscope (SEM) image of A. tepida
specimen after laser ablation analysis of multiple chambers. Image
taken at SCIAM: Service Commun d’Imageries et d’Analyses Mi-
croscopiques, University of Angers. Our approach is to investigate the Mn/Ca intra- and inter-
test variability in A. tepida by measuring individual cham-
bers with a laser ablation inductively coupled plasma mass
spectrometer (LA-ICP-MS) for selected living specimens
sampled in three different months of 2012. We will com-
pare the obtained Mn/Ca intra-test variability with results
on strictly intrinsic intra-test variability for the same species,
measured for specimens from laboratory experiments by Bar-
ras et al. (2018), to find out whether the seasonal variability
in BWO and pore water Mn2+ concentrations has left an im-
print on benthic foraminiferal Mn/Ca ratios. 2.2
Samples of living benthic foraminifera Living specimens of Ammonia tepida were sampled in Lake
Grevelingen in March, July and September 2012. In all three
months, surface sediments of the sampling site were in-
habited by dense populations of A. tepida. Living speci-
mens were recognized by CellTracker Green (CTG; Bern-
hard et al., 2006). CTG was applied on board R/V Luctor
within 1 h of retrieval (Langlet, 2014). Adult specimens (size
fraction 150–315 µm) from the sediment depth 0–0.5 cm
were selected for LA-ICP-MS measurements. For March,
July and September, 10, 16 and 18 specimens were anal-
ysed, respectively. Prior to LA-ICP-MS analyses, all speci-
mens were cleaned to remove sediment adherences (Barker
et al., 2003) by rinsing them three times with ultra-pure wa-
ter in 200 µL tubes, followed by one rinse in methanol and
three final rinses in ultra-pure water. During each rinse, the
samples were gently agitated with a vortex machine. J. Petersen et al.: Mn/Ca intra- and inter-test variability 333 Figure 1. Scanning electron microscope (SEM) image of A. tepida
specimen after laser ablation analysis of multiple chambers. Image
taken at SCIAM: Service Commun d’Imageries et d’Analyses Mi-
croscopiques, University of Angers. et al., 2014; Nardelli et al., 2014; Thibault de Chanvalon
et al., 2015; Cesbron et al., 2016). The salty bottom waters of
Lake Grevelingen, an artificial lake created after the closure
of a branch of the Rhine–Meuse–Scheldt estuary, are char-
acterized by seasonal hypoxia ([O2] < 63 µM) and anoxia
([O2] < detection limit of 1 µM) (Hagens et al., 2015; Seitaj
et al., 2017). Therefore, this site provides a suitable location
to study short-term environmental variability (on a timescale
of weeks to months) in relation to elemental incorporation
into benthic foraminiferal tests. However, one complicating
factor is that the activity of cable bacteria strongly influences
the seasonal pattern of sediment geochemical cycles in Lake
Grevelingen. These electrogenic bacteria induce strong shifts
in the sediment geochemistry (Seitaj et al., 2015, 2017; Sulu-
Gambari et al., 2016a, b) and lead to substantial Mn2+ mobi-
lization in pore waters (Rao et al., 2016; Sulu-Gambari et al.,
2016b; van de Velde et al., 2016). These cable bacteria tend
to have a highly patchy distribution, which hence could com-
plicate the interpretation (Sulu-Gambari et al., 2016b; Seitaj
et al., 2017). et al., 2014; Nardelli et al., 2014; Thibault de Chanvalon
et al., 2015; Cesbron et al., 2016). The salty bottom waters of
Lake Grevelingen, an artificial lake created after the closure
of a branch of the Rhine–Meuse–Scheldt estuary, are char-
acterized by seasonal hypoxia ([O2] < 63 µM) and anoxia
([O2] < detection limit of 1 µM) (Hagens et al., 2015; Seitaj
et al., 2017). Therefore, this site provides a suitable location
to study short-term environmental variability (on a timescale
of weeks to months) in relation to elemental incorporation
into benthic foraminiferal tests. However, one complicating
factor is that the activity of cable bacteria strongly influences
the seasonal pattern of sediment geochemical cycles in Lake
Grevelingen. These electrogenic bacteria induce strong shifts
in the sediment geochemistry (Seitaj et al., 2015, 2017; Sulu-
Gambari et al., 2016a, b) and lead to substantial Mn2+ mobi-
lization in pore waters (Rao et al., 2016; Sulu-Gambari et al.,
2016b; van de Velde et al., 2016). J. Petersen et al.: Mn/Ca intra- and inter-test variability These studies show that benthic foraminifera register envi-
ronmental Mn2+ concentrations in their tests (Reichart et al.,
2003; Munsel et al., 2010; Glock et al., 2012; Groeneveld and
Filipsson, 2013; Koho et al., 2015, 2017; McKay et al., 2015; Here, we investigate Mn/Ca intra- and inter-test variabil-
ity in the same benthic foraminifer, A. tepida, as studied by
Barras et al. (2018; culturing experiments) from field sam-
ples of Lake Grevelingen (the Netherlands). The shallow in-
faunal A. tepida represents a species complex including sev-
eral pseudo-cryptic species (Hayward et al., 2004; Schweizer
et al., 2011; Saad and Wade, 2016). In Lake Grevelingen,
A. tepida is almost exclusively represented by the T6 geno-
type. Ammonia tepida is abundant in coastal areas of tem-
perate climate zones, tolerating diverse biological and en-
vironmental stress factors, including low-oxygen conditions
(Moodley and Hess, 1992; Sen Gupta et al., 1996; Geslin www.biogeosciences.net/15/331/2018/ Biogeosciences, 15, 331–348, 2018 J. Petersen et al.: Mn/Ca intra- and inter-test variability www.biogeosciences.net/15/331/2018/ 2.4
Data treatment a repetition rate of 4 Hz. To maximize the amount of ablated
material, spot sizes were adapted to the chamber size and
varied typically between 40 and 85 µm in diameter. The LA-
ICP-MS operating conditions are summarized in Table 1 and
the isotope masses selected for analyses were 24Mg, 27Al,
43Ca, 55Mn, 57Fe, 66Zn, 88Sr and 137Ba. All laser ablation profiles for each element and each sample
or reference material were carefully examined and processed
with the GLITTER software. Firstly, the integration interval
was based on constant raw counts of 44Ca as an indicator of
calcite ablation and on constant Sr/Ca ratios, as the intra-
shell heterogeneity of Sr/Ca in foraminifera in general and
of A. tepida in particular is known to be relatively small (e.g. Eggins et al., 2003; de Nooijer et al., 2014a). Secondly, each
laser ablation profile was screened for peaks in elements that
may indicate surface contamination (27Al, 57Fe, 66Zn). Typi-
cally, high concentrations of Mn on the outer and inner shell
surfaces are considered an indicator of contamination (Marr
et al., 2011; de Nooijer et al., 2014a; Leduc et al., 2014; Koho
et al., 2015). In our case, where Mn is the element of inter-
est, whenever peaks of Mn on outer and inner parts of pro-
files corresponded to peaks of other contaminant elements,
they were discarded from our data. All integration intervals
containing less than 10 data points or results with count rates
below the LOQ were removed from our dataset. Prior to each analytical session, the ICP-MS was tuned
with the NIST SRM 612 reference material to minimize ox-
ide formation (ThO+/Th+ < 0.5 %) and elemental fraction-
ation (U/Th close to 1), as well as to optimize the signal-
to-noise ratio for Mn. The typical laser ablation profile of
a foraminiferal chamber includes 30 s of data acquisition of
the background signal (laser turned off) followed by the ab-
lation of the chamber wall until it was completely pierced
(the laser was shut down after a visual control of the abla-
tion), and data acquisition was stopped after another mea-
suring interval of the background signal. J. Petersen et al.: Mn/Ca intra- and inter-test variability Analyte G2 laser ablation system (193 nm)
Laser fluence
0.91 Jcm−2
Laser pulse repetition rate
4 Hz
He flow rates for the HelEx II
0.7 and 0.3 Lmin−1
Two-volume cell
Projected spot size
40–85 µm
Varian Bruker 820-MS
Dwell time
20 ms
Radio frequency power
1.15 kW
Sheath gas flow
0.8–0.9 Lmin−1
232Th16O+/232Th+
< 0.5 %
U/Th
∼1 J. Petersen et al.: Mn/Ca intra- and inter-test variability were also normalized to the NIST SRM 612 glass to evalu-
ate the long-term reproducibility of our analyses (Table 2). For each element and analytical session we have calculated
the limit of detection (LOD = background signal +3.3 × σ;
σ = SD calculated on the background signal) and the limit
of quantification (LOQ = background signal +10 × σ) and
have discarded all data below the LOQ (Long and Wineford-
ner, 1983; Longerich et al., 1996; Bettencourt da Silva and
Williams, 2015). As a result of the low-energy laser ablation
conditions, 24Mg profiles of NIST SRM 612 were below the
LOQ for some datasets. However, the USGS MACS-3 has
high Mg concentrations (1880 ± 70 ppm), resulting in sig-
nals above the LOQ, and thus Mg/Ca ratios of foraminiferal
samples, NFHS and NIST SRM 610 were normalized to
USGS MACS-3 using the recommended values of Jochum
et al. (2012). were also normalized to the NIST SRM 612 glass to evalu-
ate the long-term reproducibility of our analyses (Table 2). For each element and analytical session we have calculated
the limit of detection (LOD = background signal +3.3 × σ;
σ = SD calculated on the background signal) and the limit
of quantification (LOQ = background signal +10 × σ) and
have discarded all data below the LOQ (Long and Wineford-
ner, 1983; Longerich et al., 1996; Bettencourt da Silva and
Williams, 2015). As a result of the low-energy laser ablation
conditions, 24Mg profiles of NIST SRM 612 were below the
LOQ for some datasets. However, the USGS MACS-3 has
high Mg concentrations (1880 ± 70 ppm), resulting in sig-
nals above the LOQ, and thus Mg/Ca ratios of foraminiferal
samples, NFHS and NIST SRM 610 were normalized to
USGS MACS-3 using the recommended values of Jochum
et al. (2012). Table 1. Summary of LA-ICP-MS operating conditions. Analyte G2 laser ablation system (193 nm)
Laser fluence
0.91 Jcm−2
Laser pulse repetition rate
4 Hz
He flow rates for the HelEx II
0.7 and 0.3 Lmin−1
Two-volume cell
Projected spot size
40–85 µm
Varian Bruker 820-MS
Dwell time
20 ms
Radio frequency power
1.15 kW
Sheath gas flow
0.8–0.9 Lmin−1
232Th16O+/232Th+
< 0.5 %
U/Th
∼1 Table 1. Summary of LA-ICP-MS operating conditions. Table 1. Summary of LA-ICP-MS operating conditions. 2.3
LA-ICP-MS operating conditions and instrument
calibration For all specimens several consecutive test chambers were
measured individually with a LA-ICP-MS (Fig. 1). The
analyses were performed with an ArF excimer laser
(193 nm, Analyte G2, Teledyne Photon Machines) coupled to
a quadrupole inductively coupled plasma mass spectrometer
(ICP-MS; Varian Bruker 820-MS) at the Laboratory of Plan-
etology and Geodynamics, Université de Nantes (France). Ablations were conducted in a HelEx II two-volume cell with
He as a carrier gas, a laser energy density of 0.91 Jcm−2 and www.biogeosciences.net/15/331/2018/ Biogeosciences, 15, 331–348, 2018 334 J. Petersen et al.: Mn/Ca intra- and inter-test variability www.biogeosciences.net/15/331/2018/ 2.5
Statistical analyses Statistical analyses were carried out using R (R Core Team,
2016) and the package ggplot2 for graphical representation
(Wickham, 2009). To verify whether data were normally dis-
tributed we used the Shapiro–Wilk test. For normally dis-
tributed data we used an analysis of variance (ANOVA) and
t tests with the Bonferroni adjustment as a post hoc test. In all
other cases we used a Kruskal–Wallis test and a Wilcoxon–
Mann–Whitney test as a post hoc test. When comparing
Mn/Ca ratios between chambers, in order to check for on-
togenetic trends, we used Spearman rank correlation. In all
cases, a p value below 0.05 was considered significant. J. Petersen et al.: Mn/Ca intra- and inter-test variability 335 Figure 2. Mn/Ca for each chamber in living specimens of A. tepida from Lake Grevelingen, collected in March (03/2012), July (07/2012)
and September (09/2012) 2012 (specimens labelled from 1 to 44 as in Table A1). One plot represents one specimen. Numbers on x axis
indicate chambers: 1 – penultimate chamber; 2 – antepenultimate chamber, etc.; 10 – central part of test. Error bars represent the Mn/Ca SD
of multiple analyses of the foraminiferal carbonate standard NFHS (see Sect. 2.3 and 2.4). Figure 2. Mn/Ca for each chamber in living specimens of A. tepida from Lake Grevelingen, collected in March (03/2012), July (07/2012)
and September (09/2012) 2012 (specimens labelled from 1 to 44 as in Table A1). One plot represents one specimen. Numbers on x axis
indicate chambers: 1 – penultimate chamber; 2 – antepenultimate chamber, etc.; 10 – central part of test. Error bars represent the Mn/Ca SD
of multiple analyses of the foraminiferal carbonate standard NFHS (see Sect. 2.3 and 2.4). 2.4
Data treatment The NIST SRM
612 glass reference material was analysed every 10 measure-
ments and the NIST SRM 610 (silicate glass reference ma-
terial), USGS MACS-3 (carbonate reference material) and
the NFHS (NIOZ, Netherlands Institute of Sea Research,
foraminifera in-house standard; Mezger et al., 2016) were
analysed every 20 analytical spots. All reference materials
have been analysed in raster mode, with the same laser en-
ergy as for the samples, a spot size of 65 µm and a scan speed
of 10 µms−1. The choice of raster mode for reference mate-
rials allows us to minimize elemental fractionation linked to
crater depth (Eggins et al., 1998). For the ablation of cham-
bers of A. tepida such fractionation is probably negligible be-
cause data acquisition was in general short (∼10 s for spot
analyses). All foraminiferal ablation profiles were normal-
ized to 43Ca as an internal standard, and element concen-
trations were calculated assuming 40 % wt for the CaCO3. The NIST SRM 612 glass served as calibration standard for
Mn/Ca and Sr/Ca of the foraminiferal samples, using the
recommended values of Jochum et al. (2011). The results
obtained from NIST SRM 610, USGS MACS-3 and NFHS Q
The external reproducibility of our Mn/Ca analyses was
2 % (calculated as 2 × SE, relative standard error) when
determined on the NFHS foraminiferal carbonate standard
(data for Mn/Ca, Mg/Ca and Sr/Ca in Table 2). This stan-
dard is adequate for this purpose since it has Mn concentra-
tions comparable to our samples and the same matrix. The
results for the USGS MACS-3 carbonate and NIST SRM
610 glass reference materials (Table 2) agree with the rec-
ommended values (published values have RSDs of 4 and 1 %,
respectively; Table 2). Additionally, we analysed the GJR in-
house calcite standard four times, and the calculated concen-
trations for Mn and Sr (107±1 ppm and 190±3 ppm, respec-
tively) compared well with published values (106 ± 7 ppm
and 184 ± 15 ppm, respectively) from Wit et al. (2010). Biogeosciences, 15, 331–348, 2018 www.biogeosciences.net/15/331/2018/ 3.1
Average Mn/Ca ratios per specimen For all specimens of A. tepida investigated, between four
and nine chambers per specimen fulfilled our profile se-
lection criteria for reliable Mn/Ca measurements (as de-
fined in Sect. 2.4). Mn/Ca ratios of individual chambers
ranged from 0.03 to 0.57 mmolmol−1 for the 10 speci-
mens sampled in March 2012 (Fig. 2 (03/2012); Table A1),
from 0.04 to 0.63 mmolmol−1 for the 16 specimens col-
lected in July 2012 (Fig. 2 (07/2012); Table A1) and from
0.04 to 0.48 mmolmol−1 for the 18 specimens sampled in
September 2012 (Fig. 2 (09/2012); Table A1). Average val-
ues for Mn/Ca per specimen were between 0.08 ± 0.02 and www.biogeosciences.net/15/331/2018/ J. Petersen et al.: Mn/Ca intra- and inter-test variability (2011)
(1997)
(mmolmol−1)
(ppm)
(ppm)
(ppm)
(ppm)
(ppm)
(ppm)
(ppm)
(ppm)
Mg/Ca
2.8 ± 0.2
Mg
688 ± 40
660
–
1756 ± 136
1880 ± 70
520 ± 12
432 ± 29
465 ± 27
(n = 44)
(n = 36)
(n = 19)
Mn/Ca
0.15 ± 0.02
Mn
85 ± 9
88
520 ± 26
536 ± 28
532 ± 23
445 ± 5
444 ± 6
433 ± 32
(n = 115)
(n = 115)
(n = 36)
(n = 35)
Sr/Ca
1.38 ± 0.04
Sr
1204 ± 37
1300
6590 ± 383
6760 ± 350
6570 ± 170
513 ± 16
515.5 ± 0.5
497 ± 18
(n = 115)
(n = 115)
(n = 36)
(n = 35)
Figure 3. Average Mn/Ca values and SD for all measured chambers of each analysed specimen of A. tepida for the three sampling campaigns
(a: March; b: July; c: September). (d) Population averages with SD calculated on the basis of mean values per specimen of all analysed
specimens for each of the three sampling months. NFHS
USGS MACS-3
NIST SRM 610
This
This
NIOZ (see
This
USGS
Jochum
This
Jochum
Pearce et al. study
study
caption)
study
et al. (2012)
study
et al. (2011)
(1997)
(mmolmol−1)
(ppm)
(ppm)
(ppm)
(ppm)
(ppm)
(ppm)
(ppm)
(ppm)
Mg/Ca
2.8 ± 0.2
Mg
688 ± 40
660
–
1756 ± 136
1880 ± 70
520 ± 12
432 ± 29
465 ± 27
(n = 44)
(n = 36)
(n = 19)
Mn/Ca
0.15 ± 0.02
Mn
85 ± 9
88
520 ± 26
536 ± 28
532 ± 23
445 ± 5
444 ± 6
433 ± 32
(n = 115)
(n = 115)
(n = 36)
(n = 35)
Sr/Ca
1.38 ± 0.04
Sr
1204 ± 37
1300
6590 ± 383
6760 ± 350
6570 ± 170
513 ± 16
515.5 ± 0.5
497 ± 18
(n = 115)
(n = 115)
(n = 36)
(n = 35) Figure 3. Average Mn/Ca values and SD for all measured chambers of each analysed specimen of A. tepida for the three sampling campaigns
(a: March; b: July; c: September). (d) Population averages with SD calculated on the basis of mean values per specimen of all analysed
specimens for each of the three sampling months. 49 % (±24 %) for Mg/Ca and 9 % (±4 %) for Sr/Ca. (Ta-
bles A1–A3). www.biogeosciences.net/15/331/2018/ Biogeosciences, 15, 331–348, 2018 J. Petersen et al.: Mn/Ca intra- and inter-test variability The specimens with the highest Mn/Ca RSD
(three specimens with RSD > 70 %) were sampled in July
(specimens 15, 17 and 18; Fig. 2). Out of these three speci-
mens, two had one chamber with much higher Mn/Ca com-
pared to the other chambers (specimens 15 and 17), whereas
one specimen had an increase in Mn/Ca ratios from older to
younger chambers (specimen 18). 0.28 ± 0.15 mmolmol−1 in March, between 0.08 ± 0.04 and
0.39±0.19 mmolmol−1 in July, and between 0.09±0.03 and
0.22±0.13 mmolmol−1 in September (Fig. 3a–c; Table A1). There are no significant differences in average Mn/Ca ratios
per specimen between the three sampled months (Fig. 3d;
ANOVA, Table 3). However, for the SDs per specimen, the
values from July 2012 are significantly larger than those of
the two other months (ANOVA, Table 3). To further investigate the variability in Mn/Ca ratios
within single specimens, we calculated the range (maxi-
mal minus minimal chamber value) for each measured in-
dividual. The so-defined Mn/Ca intra-test variability shows J. Petersen et al.: Mn/Ca intra- and inter-test variability 336 Table 2. Evaluation of external reproducibility of laser ablation ICP-MS analyses for Mg/Ca, Mn/Ca and Sr/Ca determined on the NFHS,
USGS MACS-3 and NIST SRM 610. For Mn/Ca and Sr/Ca, all values are calibrated against NIST SRM 612 and are given as average with
SDs from all sessions (in bold). For Mg/Ca, all values are calibrated against USGS MACS-3. Reference values with SD are given according
to Jochum et al. (2011), Jochum et al. (2012), USGS (S. Wilson, USGS, unpublished) and Pearce et al. (1997). Values for NFHS are in
mmolmol−1 for better comparison with results from samples (SD of standard applied as error bar on samples) as well as in ppm because
reference values are given in this unit (information obtained from the NIOZ, 2017) and for comparison with concentrations of other reference
materials. Values for USGS MACS-3 and NIST SRM 610 are in ppm for comparison with published reference values. NFHS
USGS MACS-3
NIST SRM 610
This
This
NIOZ (see
This
USGS
Jochum
This
Jochum
Pearce et al. study
study
caption)
study
et al. (2012)
study
et al. (2011)
(1997)
(mmolmol−1)
(ppm)
(ppm)
(ppm)
(ppm)
(ppm)
(ppm)
(ppm)
(ppm)
Mg/Ca
2.8 ± 0.2
Mg
688 ± 40
660
–
1756 ± 136
1880 ± 70
520 ± 12
432 ± 29
465 ± 27
(n = 44)
(n = 36)
(n = 19)
Mn/Ca
0.15 ± 0.02
Mn
85 ± 9
88
520 ± 26
536 ± 28
532 ± 23
445 ± 5
444 ± 6
433 ± 32
(n = 115)
(n = 115)
(n = 36)
(n = 35)
Sr/Ca
1.38 ± 0.04
Sr
1204 ± 37
1300
6590 ± 383
6760 ± 350
6570 ± 170
513 ± 16
515.5 ± 0.5
497 ± 18
(n = 115)
(n = 115)
(n = 36)
(n = 35)
Figure 3. Average Mn/Ca values and SD for all measured chambers of each analysed specimen of A. tepida for the three sampling campaigns
(a: March; b: July; c: September). (d) Population averages with SD calculated on the basis of mean values per specimen of all analysed
specimens for each of the three sampling months. NFHS
USGS MACS-3
NIST SRM 610
This
This
NIOZ (see
This
USGS
Jochum
This
Jochum
Pearce et al. study
study
caption)
study
et al. (2012)
study
et al. www.biogeosciences.net/15/331/2018/ J. Petersen et al.: Mn/Ca intra- and inter-test variability 337 Table 3. Statistical results for comparison of Mn/Ca measurements of A. tepida for different sampling months. Significant results (p value <
0.05) are indicated in bold. Average Mn/Ca
SD Mn/Ca
Range Mn/Ca
p value ANOVA
0.118
0.016
–
p value Kruskal–Wallis
–
–
0.044
p value post hoc test:
Mar 2012 vs. Jul 2012
0.720
0.045
0.041
Mar 2012 vs. Sep 2012
1.000
1.000
0.621
Jul 2012 vs. Sep 2012
0.120
0.036
0.033 Table 3. Statistical results for comparison of Mn/Ca measurements of A. tepida for different sampling months. Significant results (p value <
0.05) are indicated in bold. Table 3. Statistical results for comparison of Mn/Ca measurements of A. tepida for different sampling months. Significant results (p value <
0.05) are indicated in bold. /Ca
SD Mn/Ca
Range Mn/Ca
.118
0.016
–
–
–
0.044
.720
0.045
0.041
.000
1.000
0.621
.120
0.036
0.033
Figure 4. Histograms representing Mn/Ca intra-test variability as
a range (difference between maximum and minimum Mn/Ca per
specimen of A. tepida). (a) March, 10 specimens; (b) July, 16 spec-
imens; (c) September, 18 specimens. a different pattern for all specimens from each of the three
sampling campaigns (Fig. 4; data in Table A1). For in-
stance, in March, 7 out of 10 specimens had a range smaller
than 0.2 mmolmol−1, whereas the remaining 3 had a maxi-
mum Mn/Ca intra-test variability of up to 0.5 mmolmol−1. Yet, for the specimens sampled in July, 12 out of 16 had
a range larger than 0.2 mmolmol−1, with 1 specimen show-
ing a range of 0.5–0.6 mmolmol−1. Most specimens col-
lected in September had a Mn/Ca range within specimens
of 0.1–0.2 mmolmol−1, with no specimen exceeding a dif-
ference in Mn/Ca of 0.4 mmolmol−1. Because the range
histogram for March clearly shows a non-normal distribu-
tion (p value = 0.041) and the sample size per sampling date
is small (n between 10 and 18), we investigated the differ-
ences with a non-parametric test (Kruskal–Wallis test and a
Wilcoxon–Mann–Whitney test as a post hoc test) and ob-
tained a significant difference between July and the other
two months. This result confirms the higher Mn/Ca intra-
test variability in July compared to March and September
(Table 3). 3.2
Intra-test variability The intra-test variability defined as RSD per specimen is
on average 45 % (±15 % SD for all specimens) for Mn/Ca, www.biogeosciences.net/15/331/2018/ Biogeosciences, 15, 331–348, 2018 3.3
Ontogenetic trends of Mn/Ca ratios To investigate the presence of ontogenetic (i.e. size-related)
trends, we distinguished between four types: (a) specimens
with a trend towards lower values in later chambers (n = 7;
specimens 1, 2, 5, 6, 9, 14 and 31 in Fig. 2); (b) specimens
that tended to have higher Mn/Ca ratios in later chambers
(n = 6; specimens 18, 23, 28, 30, 34 and 39 in Fig. 2); (c)
specimens with only one or two values deviating from gen-
erally rather constant Mn/Ca ratios in all other chambers
(n = 2; specimens 17 and 19 in Fig. 2); and (d) specimens
with no apparent trend (the remaining 29 specimens). Most
specimens of the first group were sampled in March 2012;
in fact half of the specimens from March showed a tendency
towards lower Mn/Ca ratios in later chambers. Most indi-
viduals with increasing values towards later chambers were
collected in September 2012. In the sample of July 2012, no
dominant trend was observed. Figure 4. Histograms representing Mn/Ca intra-test variability as
a range (difference between maximum and minimum Mn/Ca per
specimen of A. tepida). (a) March, 10 specimens; (b) July, 16 spec-
imens; (c) September, 18 specimens. icant trend towards lower values in later chambers (Fig. 5a). In July 2012 no significant trend is found (Fig. 5b), whereas
in September 2012 there was a significant trend of slightly in-
creasing values towards later chambers (Fig. 5c). If all speci-
mens from the three sampled months are combined (Fig. 5d),
there is no significant relation with chamber stage. To check for the existence of persistent ontogenetic trends,
we combined for each successive chamber the values of all
measured specimens and tested for statistically significant
trends (Fig. 5). As expected, for March 2012 there is a signif- Biogeosciences, 15, 331–348, 2018 www.biogeosciences.net/15/331/2018/ 4.1
Comparison of Mn/Ca ratios between benthic
foraminiferal species from coastal and deep-sea
ecosystems 4.1
Comparison of Mn/Ca ratios between benthic
foraminiferal species from coastal and deep-sea
ecosystems Recently, Mn/Ca ratios in benthic foraminiferal tests have
been proposed as a potential palaeo-proxy for BWO. This
suggestion is based on several observations made in recent
continental margin and deep-sea ecosystems (Reichart et al.,
2003; Glock et al., 2012; Groeneveld and Filipsson, 2013;
Koho et al., 2015, 2017). Furthermore, the application of
foraminiferal Mn/Ca ratios as a proxy for dissolved Mn has
been tested in laboratory conditions, where calibration stud-
ies show a linear relation between seawater dissolved Mn2+
concentrations and foraminiferal Mn/Ca ratios, though with
species-dependent partition coefficients (Munsel et al., 2010;
Barras et al., 2018). The use of foraminiferal Mn/Ca ratios as
a palaeo-proxy for BWO has further been explored in several
studies of deep-sea sediment records (Klinkhammer et al.,
2009; Ní Fhlaithearta et al., 2010; McKay et al., 2015). J. Petersen et al.: Mn/Ca intra- and inter-test variability Figure 5. Mn/Ca of all measured specimens as a function of chamber succession. (a) March; (b) July; (c) September; (d) all combined. Numbers on x axis indicate chambers: 1 – penultimate chamber; 2 – antepenultimate chamber, etc.; 10 – central part of test. Small black
dots represent Mn/Ca of single measurements; larger white circles represent average values per chamber, when all specimens are combined. Trends were tested for statistical significance with a Spearman rank correlation (correlation coefficient rho, p value and n (total number of
data points) are indicated). Figure 5. Mn/Ca of all measured specimens as a function of chamber succession. (a) March; (b) July; (c) September; (d) all combined. Numbers on x axis indicate chambers: 1 – penultimate chamber; 2 – antepenultimate chamber, etc.; 10 – central part of test. Small black
dots represent Mn/Ca of single measurements; larger white circles represent average values per chamber, when all specimens are combined. Trends were tested for statistical significance with a Spearman rank correlation (correlation coefficient rho, p value and n (total number of
data points) are indicated). www.biogeosciences.net/15/331/2018/ www.biogeosciences.net/15/331/2018/ 338 4
Discussion ica, C. fimbriata; 0.0020 to 0.277 mmolmol−1; Koho et al.,
2017) but is elevated compared to Mn/Ca ratios measured
in single tests of living benthic foraminifera from the Pe-
ruvian oxygen minimum zone (OMZ) (B. spissa; 0.0021 to
0.010 mmolmol−1; Glock et al., 2012). This latter difference
can be explained by the much higher pore water Mn2+ con-
centrations in our study (< 310 µmolL−1) and in the study
of Koho et al. (2017; < 5 µmolL−1) compared to the study
of Glock et al. (2012; < 0.1 µmolL−1). Despite the differ-
ent pore water Mn2+ concentrations in our study compared
to Koho et al. (2017), we found rather similar foraminiferal
Mn/Ca ratios, and this contrast could be partly resolved by
higher partition coefficients in deep-sea species compared to
coastal species (Barras et al., 2018). For A. tepida, the range
measured in specimens from Lake Grevelingen in our study
compares well with measured average Mn/Ca ratios in the
study of Barras et al. (2018; 0.13 and 0.86 mmolmol−1) for
a similar range of pore water Mn2+ concentrations compared
to the concentration of dissolved Mn in the seawater of the
culturing experiments (10 to 100 µmolL−1). 4.2.2
Intrinsic intra-test variability in Mn/Ca and
ontogenetic trends Similarly to Mg/Ca ratios, it is possible to estimate the
amount of the intrinsic intra-test variability in Mn/Ca based
on results from cultured benthic foraminifera. Culture ex-
periments performed with A. tepida in controlled and sta-
ble conditions show 85 % of variability (RSD) for very low
Mn2+ concentrations of 2 µmolL−1 (all chambers calcified
under laboratory conditions, measured with a LA-ICP-MS,
Barras et al., 2018). However, for seawater Mn2+ concentra-
tions of 12 and 101 µmolL−1, comparable to the pore wa-
ter concentration at our site, the variability (RSD) in Mn/Ca
was 25 and 11 %, respectively (Barras et al., 2018). For even
higher seawater Mn2+ concentrations of 595 µmolL−1 the
Mn/Ca RSD was 17 % (Barras et al., 2018). Despite analyt-
ical considerations, the large RSD observed for the lowest
Mn2+ concentration was partly due to the presence of a clear
decrease in Mn/Ca during the ontogeny. At higher concen-
trations, no such ontogenetic tendencies were found (Barras
et al., 2018). In our material there were no systematic on-
togenetic trends either (for all specimens of all investigated
months combined; Figs. 2 and 5d), so that the 11–25 % RSD
range measured for Mn2+ concentrations of ∼10–100 µM
should be representative of the intrinsic intra-test variability
in our specimens. This range is much lower than the average
total intra-test variability of 45 % found for living A. tepida
in our study of Lake Grevelingen. The results for all specimens measured in this study
showed an average Mn/Ca intra-test variability for A. tep-
ida of 45±15 % (RSD average for all chambers measured on
a single specimen; Table A1), which is comparable to that re-
ported in some previous studies (30–50 %; Glock et al., 2012;
McKay et al., 2015; Koho et al., 2017). Our results from the
different sampling campaigns showed more specimens with
higher variability in July compared to March and September
(Figs. 3 and 4). In order to investigate if this significant dif-
ference can be attributed to environmental factors, our results
will be compared to variability reported in other elements. J. Petersen et al.: Mn/Ca intra- and inter-test variability for Mg/Ca, compared to ∼8 % for Sr/Ca (values for all
chambers analysed with a LA-ICP-MS; de Nooijer et al.,
2014a). This difference between intrinsic intra-test variabil-
ity in Mg/Ca and Sr/Ca was explained by the fact that during
biomineralization processes, unlike Sr, Mg is strongly dis-
criminated against in the calcifying fluid, and that these pro-
cesses may vary over time and/or position in the test (Ben-
tov and Erez, 2005, 2006; Nehrke et al., 2013; de Nooijer
et al., 2014b). In some field studies (Allison and Austin,
2003; Curry and Marchitto, 2008), values for Mg/Ca intra-
test variability were similar to those observed in the cultur-
ing experiment, whereas only a single field study (Raitzsch
et al., 2011) reported substantially higher Mg/Ca intra-test
variability, with an RSD of 51 %. In the case of our A. tepida
specimens, the intra-test variability in Mg/Ca was similarly
high (49 %). Contrastingly, the Sr/Ca RSD of A. tepida was
9 %, comparable to variability in Sr/Ca from culturing stud-
ies. effects) or environmental factors (e.g. seasonality of Mn cy-
cling in the surface sediments, microhabitat effects). In fact,
it is essential to know what degree of intrinsic variability can
be expected in a population having experienced the same en-
vironmental conditions and above what threshold changes
in Mn/Ca ratios can be ascribed to environmental factors. When using a LA-ICP-MS, or other microanalytical tech-
niques, Mn/Ca ratios are measured on small parts of the
foraminiferal test, making knowledge of intra-test variabil-
ity even more crucial for the interpretation of the measure-
ments. In the case of a high intrinsic variability, independent
of environmental parameters, more spot measurements will
be necessary to obtain a reliable mean value for one spec-
imen or several specimens from the same stable environ-
ment (Sadekov et al., 2005; de Nooijer et al., 2014a). Al-
though contamination is an important issue in Mn/Ca mea-
surements (Boyle, 1983; Barker et al., 2003; Pena et al.,
2005, 2008), pre-treatment cleaning and a precise targeting
of the measurement interval (when using secondary ion mass
spectrometry, SIMS, or a LA-ICP-MS), should largely elim-
inate the potential influence of contaminant phases and/or di-
agenetic overgrowths (Glock et al., 2012; Koho et al., 2015;
McKay et al., 2015). Therefore, the Mn/Ca intra-test vari-
ability should ideally not have a diagenetic contribution. www.biogeosciences.net/15/331/2018/ 4.2
Intra-test variability in elemental ratios in benthic
foraminiferal tests Our study is the first to investigate benthic foraminiferal
Mn/Ca ratios in a coastal ecosystem. The results for A. tep-
ida show an average Mn/Ca ratio of 0.17±0.08 mmolmol−1
and a range of 0.08 ± 0.04 to 0.39 ± 0.19 mmolmol−1 for
the average Mn/Ca per specimen (Table A1). This range is
comparable to that found in living specimens of some deep-
sea infaunal species from the NE Japan margin (E. batialis,
B. spissa, U. cf. graciliformis, U. akitaensis, N. labrador- Intra-test variability is less well documented for Mn/Ca ra-
tios than for other elemental ratios (el / Ca; e.g. Mg/Ca,
Sr/Ca), and it has not yet been established what portion of
the total variability can be attributed to either intrinsic (i.e. intra-test variability due to ontogenetic trends or other vital Biogeosciences, 15, 331–348, 2018 www.biogeosciences.net/15/331/2018/ 339 J. Petersen et al.: Mn/Ca intra- and inter-test variability 4.2.3
Seasonality of environmental factors as
explanation for Mn/Ca intra-test variability As explained in the previous paragraph, it appears that half or
more of the measured Mn/Ca intra-test variability in A. tep-
ida (total intra-test variability was 45 % RSD compared to
11–25 % RSD intrinsic variability) can be attributed to en-
vironmental parameters. Here, we will consider which spe-
cific factors may be responsible for this variability. First of
all, the sampling site of A. tepida in Lake Grevelingen at
23.1 m water depth shows strong seasonal fluctuations in bot-
tom water oxygenation (Sect. 2.1, Fig. S2; Hagens et al.,
2015). In theory, under oxic conditions the zone of man-
ganese reduction should lie below the microhabitat of A. tep-
ida, which is situated close to the sediment–water interface
(SWI) (Thibault de Chanvalon et al., 2015; Cesbron et al.,
2016), and chambers calcified in this condition should show
low Mn/Ca ratios. When bottom water oxygen concentra-
tions decrease (for instance due to enhanced fluxes of organic
matter to the seafloor or the stagnation of bottom waters),
the zone of manganese reduction will migrate upward; un-
der hypoxic conditions the Mn redox front will be situated
closer to the SWI and more Mn2+ should be incorporated
into chambers of A. tepida. However, in extreme cases, the
upward migration of the Mn redox front can lead to high
amounts of Mn2+ diffusing into the water column (Sundby
and Silverberg, 1985; Konovalov et al., 2007; Pakhomova
et al., 2007), and in the case of seafloor anoxia, Mn2+ may
almost entirely seep out of the sediment into the bottom wa-
ter (Slomp et al., 1997). Thus, foraminiferal Mn/Ca ratios
from chambers calcified under anoxic conditions should be
nominal. At our station in Lake Grevelingen, which experi-
ences hypoxic but not anoxic events in summer, these events
lead to the dissolution of Mn oxides in surface sediments, and
suspended-matter Mn in the water column (sampled ∼500 m
away from our station) shows maximum values in July 2012
(Sulu-Gambari et al., 2016b, 2017). Therefore, we would ex-
pect higher Mn2+ concentrations below the SWI in summer
than in winter, coinciding with higher Mn/Ca ratios in cham-
bers of A. tepida calcified in summer. However, contrary
to these theoretical expectations, observations showed max-
imum pore water Mn2+ concentrations in the topmost cen-
timetres from February to April 2012, whereas Mn2+ con-
centrations remained relatively low during the rest of 2012
(Fig. S3, Sulu-Gambari et al., 2016a, b). 4.2.3
Seasonality of environmental factors as
explanation for Mn/Ca intra-test variability Recently, similar
seasonal Mn2+ pore water patterns at a nearby but slightly
deeper station (∼500 m away, depth 34 m) in Lake Grev-
elingen were explained by the presence of cable bacteria in
winter and early spring, which are capable of performing
electrogenic sulfur oxidation resulting in the dissolution of
FeS (Seitaj et al., 2015; Sulu-Gambari et al., 2016a, b). Con-
tl
d diff
i
F 2+ i
idi
d b The foraminiferal lifespan generally varies from 3 months
to about 2 years (Boltovskoy and Lena, 1969; Murray, 1983). No precise data are available for A. tepida, although on the
basis of field evidence, Goldstein and Moodley (1993) and
Morvan et al. (2006) concluded that their longevity should
be at least 1 year. However, during the lifespan, chamber for-
mation is probably not a continuous process and could be
faster in juvenile stages (on the basis of lower metabolic rates
in later ontogenetic stages, inferred from increasing δ13C
ratios; e.g. Schmiedl et al., 2004; Mackensen, 2008; Schu-
macher et al., 2010; Raitzsch et al., 2011). Under experimen-
tal conditions, growth rates of A. tepida showed a decrease
with ontogeny from 1 chamber per day to 1 chamber per
week during a 3-week period (de Nooijer et al., 2014a). It
has been suggested (e.g. Jorissen, 1988) that at later onto-
genetic stages, calcification could be limited to short peri-
ods with favourable conditions, such as abundant food sup-
ply. In view of the rather scarce information about the timing
and duration of calcification, we assume that A. tepida from
Lake Grevelingen calcifies during several successive seasons
and that the specimens of our three samples (March, July
and September 2012) have each recorded a different part of
the yearly pore water Mn2+ cycle. The different degrees of
intra-test variability between sampled months (Figs. 3 and 4)
could be explained by the fact that some specimens mainly
calcify during periods with stable sediment Mn2+ concen-
trations (i.e. a large part of specimens sampled in March
and September), whereas others have calcified during periods
with rapid changes, such as in winter, when cable bacteria ac-
tivity and Mn2+ concentrations rapidly increase, or in spring,
when these two parameters decrease again (i.e. a large part of
specimens sampled in July). J. Petersen et al.: Mn/Ca intra- and inter-test variability J. Petersen et al.: Mn/Ca intra- and inter-test variability cal foraminiferal migration through different biogeochemical
micro-environments. oxides, which produces an accumulation of Mn2+ in the pore
water in winter (Sulu-Gambari et al., 2016a, b). This process
would be responsible for a temporal offset of ∼4–6 months
between minimum BWO and maximum Mn2+ concentra-
tions. Although cable bacteria were not detected at our sam-
pling station with fluorescence in situ hybridization (FISH),
which may be due to their patchy distribution (Sulu-Gambari
et al., 2016b; Seitaj et al., 2017), the similarity in pore water
data between the two stations strongly suggests that, also at
our sampling station, cable bacteria activity is responsible for
the observed strong seasonality in pore water Mn2+. There-
fore, it seems probable that A. tepida, sampled in March,
July and September 2012 were confronted with strongly dif-
ferent pore water Mn2+ concentrations, and may also have
experienced important fluctuations in Mn2+ in their micro-
environment during their lifespan. 4.2.1
Approach for estimating (intrinsic) intra-test
variability in Mn/Ca, Mg/Ca and Sr/Ca The average Mn/Ca intra-test variability observed for A. tep-
ida (45 %) is comparable to that measured for Mg/Ca (49 %;
Table A2) but is larger than for Sr/Ca (9 %; Table A3). Sim-
ilarly, in other field studies of benthic foraminifera, average
intra-test variability was ∼20–50 % RSD for Mg/Ca (Alli-
son and Austin, 2003; Curry and Marchitto, 2008; Raitzsch
et al., 2011) and 5 % RSD for Sr/Ca (Allison and Austin,
2003). In these studies it could not be determined how much
of the observed variability was due to intrinsic and environ-
mental factors. In contrast to field studies, the use of cul-
tured foraminifera offers the advantage that specimens have
grown under exactly the same stable physico-chemical con-
ditions (no environmental variability); thus, intra-test vari-
ability in elemental ratios is entirely due to biological pro-
cesses. These culture studies suggest that when only intrin-
sic intra-test variability is considered, the RSD is ∼30 % Consequently, it appears that at most, about half of the
total variability in Mn/Ca can be attributed to intrinsic fac-
tors. The remaining part of the variability in Mn/Ca should
be due to changing pore water Mn2+ concentrations in
the calcification environment of the foraminifera. This may
be due to environmental changes during the lifespan of
the individuals or, alternatively, to active or passive verti- www.biogeosciences.net/15/331/2018/ Biogeosciences, 15, 331–348, 2018 340 J. Petersen et al.: Mn/Ca intra- and inter-test variability J. Petersen et al.: Mn/Ca intra- and inter-test variability 5
Conclusions Tests of the coastal benthic foraminifer A. tepida, sampled
in three different months at 23.1 m depth in the seasonally
hypoxic/anoxic Lake Grevelingen show Mn/Ca ratios with
a range of 0.08±0.04 to 0.39±0.19 mmolmol−1 (for average
values per specimen), associated with a very large intra-test
variability (average RSD = 45 %). This high intra-test vari-
ability may partly represent intrinsic factors (due to biolog-
ical processes), although no systematic ontogenetic trends
could be identified (for all specimens from the three sam-
pled months combined). However, we ascribed the larger part
of the Mn/Ca intra-test variability to the substantial tempo-
ral variability in environmental parameters and to the dif-
ferent timing of the calcification of the analysed specimens. We suggest that the strong seasonal dynamics of pore water
Mn induced by seasonal hypoxia and cable bacteria activity,
leading to variations in absolute Mn2+ concentrations and/or
migration of the redox front in the sediment, is the main fac-
tor responsible for this large intra-test variability. Differences
in the timing of calcification could explain the different de-
grees of intra-test variability observed for the three sampled
months, whereas differences in individual life history (be-
tween the individuals found in a single sample) could even
explain inter-test differences observed in each of our three
samples. Some individual chambers with exceptionally high
Mn/Ca ratios could be due to active or passive migration
to, and calcification in, slightly deeper sediment layers. In
conclusion, in environments with strong seasonal changes
in redox conditions and microbial activity, Mn/Ca measure-
ments of successive chambers of individual tests may pro-
vide a powerful proxy to reconstruct the seasonal variability
in these parameters. Data availability. Datasets (including raw data) are given in Ta-
bles A1, A2 and A3. Additionally, the raw data can be accessed
under https://issues.pangaea.de/browse/PDI-16596. J. Petersen et al.: Mn/Ca intra- and inter-test variability different calcification histories and have not registered ex-
actly the same time window. Such information can only be obtained by spot measurements
of individual chambers. In fact, a comparison of Mn/Ca ra-
tios of successive individual chambers can potentially inform
us about the extent of these seasonal changes and, implic-
itly, about cable bacteria activity. If the calcification season
of the foraminifera could be determined more precisely (for
instance by using single specimen stable isotope analyses;
e.g. Diz et al., 2009), it would potentially even be possible to
reconstruct the annual cycle of pore water dynamics. 4.2.4
Foraminiferal vertical migration as an
explanation for Mn/Ca intra-test variability Ammonia tepida is a shallow infaunal taxon, preferring a mi-
crohabitat close to the SWI, implying that the chemical com-
position of its test in principle reflects conditions in the su-
perficial sediment layer (Thibault de Chanvalon et al., 2015;
Cesbron et al., 2016). Observations made in laboratory ex-
periments show that specimens of this species introduced
into deeper anoxic sediments rapidly decrease their activity
(Langlet et al., 2013; Maire et al., 2016), and, according to
a field study from the Loire estuary, only specimens from the
top 2 cm appear to be capable of regaining their preferred
niche at the sediment surface (Thibault de Chanvalon et al.,
2015). However, Geslin et al. (2014) and Barras et al. (2018)
have shown that several foraminiferal species (among which
are A. tepida) repeatedly calcify new chambers in hypoxic
conditions, whereas Nardelli et al. (2014) showed that even
in anoxic conditions, A. tepida is capable of producing at
least one new chamber. These data suggest that individu-
als which have actively (through vertical migration) or pas-
sively (being transported by burrowing macrofauna) moved
to deeper sediment layers, with higher Mn2+ levels (such as
in February and March 2012, when the oxygen penetration
depth was ∼0.2–0.3 cm below the SWI; Seitaj et al., 2017)
could incidentally calcify a single chamber with a much
higher Mn/Ca ratio. This could be the explanation for the
patterns observed, for instance, in specimens 15, 17 and 19
sampled in July 2012 (Fig. 2). This phenomenon could be re-
sponsible for a small contribution to increased Mn/Ca intra-
test variability (specimens 15 and 17 have high RSD values;
Table A1). In conclusion, we propose that a large part of
the overall Mn/Ca intra-test variability can be explained by
a different timing of calcification events with respect to the
seasonal cycle of pore water Mn2+ concentrations, whereas
extremely high values in single chambers could be due to oc-
casional chamber formation in deeper sediment layers. 4.2.3
Seasonality of environmental factors as
explanation for Mn/Ca intra-test variability This temporal aspect of Mn/Ca
variability in relation to environmental changes can be fur-
ther studied by considering all penultimate (and antepenul-
timate, etc.) chambers from the same sampling (Fig. 5a–c). However, the comparison of Mn/Ca ratios in the successive
chambers of different individuals of the same sample shows
important differences (Figs. 2 and 5a–c). This suggests that
even within the same sample, all individuals have slightly Biogeosciences, 15, 331–348, 2018 www.biogeosciences.net/15/331/2018/ 341 4.3
Relevance of benthic foraminiferal Mn/Ca ratios
for coastal environments and implications for
palaeoceanographic studies Concerning the reliability of single-chamber measurements,
in view of the large intra-test variability in the very dynamic
environment of Lake Grevelingen, it is necessary to measure
several chambers to obtain reliable average Mn/Ca ratios for
a single specimen. In this specific context, bulk measure-
ments could be a more practical solution to study the long-
term evolution of BWO. However, the large intra-test vari-
ability may include very useful information about the sea-
sonal variability in pore water dynamics driven by redox con-
ditions and sedimentation rates as well as microbial activity. Biogeosciences, 15, 331–348, 2018 www.biogeosciences.net/15/331/2018/ Appendix A Table A1. Results of Mn/Ca measurements for different samplings of living specimens of A. tepida from Lake Grevelingen, station ST2. All
data are normalized to NIST SRM 612. Values are calculated per specimen (numbers of specimens according to Fig. 2). The total number of
chambers included is indicated in column “n”. Calculations were performed using all decimals. Table A1. Results of Mn/Ca measurements for different samplings of living specimens of A. tepida from Lake Grevelingen, station ST2. All
data are normalized to NIST SRM 612. Values are calculated per specimen (numbers of specimens according to Fig. 2). The total number of
chambers included is indicated in column “n”. Calculations were performed using all decimals. mbers included is indicated in column n . Calculations were performed using all decimals. www.biogeosciences.net/15/331/2018/ www.biogeosciences.net/15/331/2018/ J. Petersen et al.: Mn/Ca intra- and inter-test variability 342 J. Petersen et al.: Mn/Ca intra- and inter-test variability Biogeosciences, 15, 331–348, 2018 Biogeosciences, 15, 331–348, 2018
www.biogeosciences.net/15/331/2018/ J. Petersen et al.: Mn/Ca intra- and inter-test variability 343 J. Petersen et al.: Mn/Ca intra- and inter-test variability J. Petersen et al.: Mn/Ca intra- and inter-test variability Table A2. Results of Mg/Ca measurements for different samplings of living specimens of A. tepida from Lake Grevelingen, station ST2. All
data are normalized to USGS MACS-3. Values are calculated per specimen (numbers of specimens according to Fig. 2). The total number of
chambers included is indicated in column “n”. Calculations were performed using all decimals. Table A2. Results of Mg/Ca measurements for different samplings of living specimens of A. tepida from Lake Grevelingen, station ST2. All
data are normalized to USGS MACS-3. Values are calculated per specimen (numbers of specimens according to Fig. 2). The total number of
chambers included is indicated in column “n”. Calculations were performed using all decimals. Appendix A Mn/Ca
Mn/Ca
Sampling
Specimen
n
Average
SD
RSD
Max
Min
Range
date
(molmol−1)
(molmol−1)
(%)
(molmol−1)
(molmol−1)
(molmol−1)
Mar 2012
1
8
0.08
0.02
27.4
0.10
0.03
0.07
Mar 2012
2
9
0.16
0.06
34.4
0.24
0.04
0.20
Mar 2012
3
7
0.28
0.15
56.3
0.57
0.11
0.45
Mar 2012
4
10
0.16
0.04
28.4
0.23
0.07
0.16
Mar 2012
5
5
0.26
0.09
32.9
0.35
0.15
0.20
Mar 2012
6
5
0.25
0.13
50.3
0.43
0.07
0.35
Mar 2012
7
4
0.14
0.04
28.3
0.17
0.08
0.09
Mar 2012
8
6
0.14
0.04
26.2
0.18
0.10
0.08
Mar 2012
9
5
0.12
0.05
44.3
0.20
0.06
0.14
Mar 2012
10
6
0.10
0.02
23.6
0.14
0.07
0.06
Jul 2012
11
8
0.13
0.03
27.1
0.17
0.08
0.08
Jul 2012
12
8
0.15
0.08
56.1
0.30
0.06
0.23
Jul 2012
13
8
0.17
0.09
56.1
0.32
0.06
0.26
Jul 2012
14
5
0.16
0.07
44.2
0.23
0.07
0.16
Jul 2012
15
7
0.19
0.16
85.0
0.53
0.06
0.48
Jul 2012
16
6
0.16
0.07
45.1
0.27
0.07
0.20
Jul 2012
17
9
0.20
0.14
72.7
0.54
0.08
0.46
Jul 2012
18
9
0.23
0.21
90.8
0.63
0.05
0.58
Jul 2012
19
7
0.28
0.14
49.0
0.50
0.19
0.31
Jul 2012
20
5
0.25
0.11
44.3
0.44
0.16
0.28
Jul 2012
21
8
0.21
0.08
39.2
0.33
0.10
0.24
Jul 2012
22
8
0.24
0.14
58.8
0.48
0.10
0.39
Jul 2012
23
4
0.39
0.19
49.1
0.63
0.17
0.46
Jul 2012
24
8
0.08
0.04
51.6
0.17
0.05
0.12
Jul 2012
25
5
0.09
0.03
33.9
0.11
0.04
0.07
Jul 2012
26
7
0.25
0.14
54.7
0.52
0.10
0.42
Sep 2012
27
5
0.10
0.02
19.3
0.12
0.07
0.05
Sep 2012
28
7
0.21
0.13
59.9
0.48
0.10
0.38
Sep 2012
29
9
0.11
0.06
51.2
0.23
0.05
0.18
Sep 2012
30
8
0.22
0.13
59.5
0.38
0.05
0.33
Sep 2012
31
6
0.12
0.04
37.9
0.18
0.06
0.12
Sep 2012
32
5
0.13
0.06
45.7
0.22
0.07
0.14
Sep 2012
33
4
0.16
0.05
33.3
0.24
0.11
0.13
Sep 2012
34
9
0.20
0.10
48.9
0.37
0.10
0.26
Sep 2012
35
8
0.09
0.04
43.4
0.14
0.04
0.10
Sep 2012
36
7
0.19
0.08
40.2
0.32
0.09
0.22
Sep 2012
37
5
0.17
0.08
46.3
0.24
0.07
0.18
Sep 2012
38
6
0.15
0.06
38.8
0.22
0.07
0.15
Sep 2012
39
6
0.21
0.09
43.7
0.30
0.10
0.20
Sep 2012
40
7
0.15
0.03
16.7
0.20
0.13
0.08
Sep 2012
41
7
0.16
0.10
60.3
0.32
0.04
0.28
Sep 2012
42
4
0.09
0.03
38.4
0.13
0.06
0.07
Sep 2012
43
8
0.17
0.09
53.4
0.30
0.07
0.24
Sep 2012
44
9
0.14
0.07
51.9
0.31
0.08
0.24
Average
6.8
0.17
0.08
45(±15)
0.31
0.08
0.22 Biogeosciences, 15, 331–348, 2018 Biogeosciences, 15, 331–348, 2018 www.biogeosciences.net/15/331/2018/ J. Petersen et al.: Mn/Ca intra- and inter-test variability Table A3. Results of Sr/Ca measurements for different samplings of living specimens of A. tepida from Lake Grevelingen, station ST2. All
data are normalized to NIST SRM 612. Values are calculated per specimen (numbers of specimens according to Fig. 2). The total number of
chambers included is indicated in column “n”. Calculations were performed using all decimals. Table A3. Results of Sr/Ca measurements for different samplings of living specimens of A. tepida from Lake Grevelingen, station ST2. All
data are normalized to NIST SRM 612. Values are calculated per specimen (numbers of specimens according to Fig. 2). The total number of
chambers included is indicated in column “n”. Calculations were performed using all decimals. www.biogeosciences.net/15/331/2018/ J. Petersen et al.: Mn/Ca intra- and inter-test variability p
g
Mg/Ca
Mg/Ca
Sampling
Specimen
n
Average
SD
RSD
Max
Min
Range
date
(molmol−1)
(molmol−1)
(%)
(molmol−1)
(molmol−1)
(molmol−1)
Mar 2012
1
8
1.62
0.20
12.2
2.06
1.43
0.63
Mar 2012
2
9
2.51
0.74
29.5
3.30
0.87
2.43
Mar 2012
3
7
4.16
2.48
59.6
8.21
1.67
6.54
Mar 2012
4
10
2.39
0.90
37.7
3.96
1.35
2.62
Mar 2012
5
5
3.78
0.89
23.5
4.52
2.55
1.97
Mar 2012
6
5
2.77
1.62
58.6
5.14
1.40
3.74
Mar 2012
7
4
2.47
1.42
57.4
4.56
1.41
3.16
Mar 2012
8
6
1.31
0.35
27.0
1.86
0.89
0.97
Mar 2012
9
5
1.83
0.66
36.1
2.52
0.78
1.73
Mar 2012
10
6
1.26
0.35
27.9
1.92
0.94
0.99
Jul 2012
11
8
1.19
0.42
35.5
1.97
0.82
1.15
Jul 2012
12
8
2.46
1.29
52.5
5.17
1.22
3.95
Jul 2012
13
8
1.99
1.55
78.3
4.73
0.73
3.99
Jul 2012
14
5
1.77
0.85
48.2
3.19
0.91
2.28
Jul 2012
15
7
3.12
2.75
88.3
7.21
0.74
6.47
Jul 2012
16
6
3.81
1.42
37.2
5.18
1.59
3.59
Jul 2012
17
9
4.92
1.29
26.2
6.44
2.55
3.89
Jul 2012
18
9
3.11
1.78
57.2
6.24
1.45
4.79
Jul 2012
19
7
5.69
2.32
40.7
9.47
2.40
7.07
Jul 2012
20
5
6.81
5.05
74.2
14.93
1.24
13.69
Jul 2012
21
8
3.34
2.24
67.0
7.36
1.12
6.23
Jul 2012
22
8
4.03
2.20
54.6
7.57
1.63
5.94
Jul 2012
23
4
7.71
3.58
46.4
10.42
2.71
7.71
Jul 2012
24
8
1.56
0.42
26.8
2.21
1.11
1.10
Jul 2012
25
5
1.71
1.13
65.9
2.95
0.73
2.22
Jul 2012
26
7
5.34
2.94
55.1
10.17
1.50
8.67
Sep 2012
27
5
2.24
1.74
77.6
5.32
1.09
4.23
Sep 2012
28
7
4.28
5.33
124.4
16.05
0.89
15.16
Sep 2012
29
9
1.84
0.60
32.7
2.51
0.93
1.58
Sep 2012
30
8
1.26
0.33
25.7
2.00
0.97
1.03
Sep 2012
31
6
1.64
0.31
19.1
1.99
1.11
0.88
Sep 2012
32
5
1.80
0.65
36.2
2.90
1.17
1.72
Sep 2012
33
4
2.29
0.14
6.3
2.45
2.14
0.30
Sep 2012
34
9
1.64
0.62
37.7
3.05
1.04
2.01
Sep 2012
35
8
1.73
0.69
40.0
3.10
1.12
1.98
Sep 2012
36
7
1.58
1.14
72.1
4.13
0.98
3.15
Sep 2012
37
5
1.29
0.59
45.8
2.08
0.72
1.36
Sep 2012
38
6
1.68
0.65
38.9
2.49
0.68
1.80
Sep 2012
39
6
3.46
1.53
44.2
5.58
1.40
4.18
Sep 2012
40
7
2.18
1.33
61.1
5.07
1.17
3.90
Sep 2012
41
7
4.32
4.71
109.0
13.88
0.91
12.98
Sep 2012
42
4
1.17
0.23
19.5
1.44
0.96
0.47
Sep 2012
43
8
2.69
1.86
68.8
7.08
1.33
5.75
Sep 2012
44
9
2.45
1.25
51.0
4.58
1.16
3.41
Average
6.8
2.78
1.47
49(±24)
5.20
1.26
3.94 www.biogeosciences.net/15/331/2018/
Biogeosciences, 15, 331–348, 2018 www.biogeosciences.net/15/331/2018/ 344 Biogeosciences, 15, 331–348, 2018 Biogeosciences, 15, 331–348, 2018
www.biogeosciences.net/15/331/2018/ References Aller, R. C.: Bioturbation and manganese cycling in hemipelagic
sediments, Philos. T. R. Soc. A, 331, 51–68, 1990. Dellwig, O., Bosselmann, K., Kölsch, S., Hentscher, M., and
Hinrichs, J.: Sources and fate of manganese in a tidal
basin of the German Wadden Sea, J. Sea Res., 57, 1–18,
https://doi.org/10.1016/j.seares.2006.07.006, 2007. Aller, R. C.: The sedimentary Mn cycle in Long Island Sound: Its
role as intermediate oxidant and the influence of bioturbation,
O2, and Corg flux on diagenetic reaction balances, J. Mar. Res.,
52, 259–295, https://doi.org/10.1357/0022240943077091, 1994. de
Nooijer,
L. J.,
Hathorne,
E. C.,
Reichart,
G. J.,
Langer, G., and Bijma, J.: Variability in calcitic Mg/Ca
and
Sr/Ca
ratios
in
clones
of
the
benthic
foraminifer
Ammonia
tepida,
Mar. Micropaleontol.,
107,
32–43,
https://doi.org/10.1016/j.marmicro.2014.02.002, 2014a. Allison, N. and Austin, W. E. N.: The potential of ion micro-
probe analysis in detecting geochemical variations across indi-
vidual foraminifera tests, Geochem. Geophy. Geosy., 4, 8403,
https://doi.org/10.1029/2002GC000430, 2003. Altabet, M. A., Francois, R., Murray, D. W., and Prell, W. L.:
Climate-related variations in denitrification in the Arabian Sea
from sediment 15N/14N ratios, Nature, 373, 506–509, 1995. de Nooijer, L. J., Spero, H. J., Erez, J., Bijma, J., and Reichart, G. J.:
Biomineralization in perforate foraminifera, Earth-Sci. Rev.,
135,
48–58,
https://doi.org/10.1016/j.earscirev.2014.03.013,
2014b. Barker,
S.,
Greaves,
M.,
and
Elderfield,
H.:
A
study
of
cleaning
procedures
used
for
foraminiferal
Mg/Ca
pa-
leothermometry,
Geochem. Geophy. Geosy.,
4,
8407,
https://doi.org/10.1029/2003GC000559, 2003. Diaz, R. J. and Rosenberg, R.: Spreading dead zones and con-
sequences for marine ecosystems, Science, 321, 926–929,
https://doi.org/10.1126/science.1156401, 2008. Barras, C., Mouret, A., Nardelli, M. P., Metzger, E., Petersen, J.,
La, C., Filipsson, H. L., and Jorissen, F. J.: Experimental cali-
bration of Mn incorporation in foraminiferal calcite, Geochim. Cosmochim. Ac., in review, 2018. Diz, P., Jorissen, F. J., Reichart, G. J., Poulain, C., Dehairs,
F., Leorri, E., and Paulet, Y.-M.: Interpretation of benthic
foraminiferal stable isotopes in subtidal estuarine environments,
Biogeosciences, 6, 2549–2560, https://doi.org/10.5194/bg-6-
2549-2009, 2009. Bentov, S. and Erez, J.: Novel observations on biomineralization
processes in foraminifera and implications for Mg/Ca ratio in the
shells, Geology, 33, 841–844, https://doi.org/10.1130/G21800.1,
2005. Eggins, S. M., Kinsley, L. P. J., and Shelley, J. M. G.: Deposition
and element fractionation processes during atmospheric pressure
laser sampling for analysis by ICP-MS, Appl. Surf. Sci., 127–
129, 278–286, https://doi.org/10.1016/S0169-4332(97)00643-0,
1998. Bentov, S. and Erez, J.: Impact of biomineralization pro-
cesses on the Mg content of foraminiferal shells: A bio-
logical perspective, Geochem. Geophy. J. Petersen et al.: Mn/Ca intra- and inter-test variability Supplement. The supplement related to this article is available
online at: https://doi.org/10.5194/bg-15-331-2018-supplement. surement, 1st edn., Eurachem and CITAC (Cooperation on Inter-
national Traceability in Analytical Chemistry), 2015. surement, 1st edn., Eurachem and CITAC (Cooperation on Inter-
national Traceability in Analytical Chemistry), 2015. Boltovskoy, E. and Lena, H.: Seasonal occurrences, standing crop
and production in benthic foraminifera of Puerto Deseado, Con-
trib. from Cushman Found. Foraminifer. Res., 20, Part 3, 87–95,
1969. Competing interests. The authors declare that they have no conflict
of interest. Boyle, E. A.: Manganese carbonate overgrowths on foraminifera
tests, Geochim. Cosmochim. Ac., 47, 1815–1819, 1983. Burdige, D. J.: The biogeochemistry of manganese and iron reduc-
tion in marine sediments, Earth-Sci. Rev., 35, 249–284, 1993. Acknowledgements. This research was financially supported by
the European Research Council under the European Union’s
Seventh Framework Programme (FP/2007-2013) through ERC
Grant 306933 to Filip J. R. Meysman. The Region Pays de la
Loire is thanked for financing the MADONA project, including the
PhD allocation of Jassin Petersen. We are grateful to the SCIAM
laboratory (University of Angers) for the foraminiferal SEM image. Canfield, D. E.: Organic matter oxidation in marine sediments, in:
Interactions of C, N, P and S Biogeochemical Cycles and Global
Change, edited by: Wollast, R., Mackenzie, F. T., Chou, L.,
Springer, Berlin Heidelberg, Germany, 333–363, 1993. Cesbron, F., Geslin, E., Jorissen, F. J., Delgard, M. L., Charrieau, L.,
Deflandre, B., Jézéquel, D., Anschutz, P., and Metzger, E.: Verti-
cal distribution and respiration rates of benthic foraminifera: con-
tribution to aerobic remineralization in intertidal mudflats cov-
ered by Zostera noltei meadows, Estuar. Coast. Shelf S., 179,
23–38, https://doi.org/10.1016/j.ecss.2015.12.005, 2016. Edited by: Markus Kienast
Reviewed by: two anonymous referees Edited by: Markus Kienast
Reviewed by: two anonymous referees Curry, W. B. and Marchitto, T. M.: A secondary ionization
mass spectrometry calibration of Cibicidoides pachyderma
Mg/Ca with temperature, Geochem. Geophy. Geosy., 9, Q04009,
https://doi.org/10.1029/2007GC001620, 2008. J. Petersen et al.: Mn/Ca intra- and inter-test variability Sr/Ca
Sr/Ca
Sampling
Specimen
n
Average
SD
RSD
Max
Min
Range
date
(molmol−1)
(molmol−1)
(%)
(molmol−1)
(molmol−1)
(molmol−1)
Mar 2012
1
8
1.36
0.17
12.6
1.59
1.00
0.59
Mar 2012
2
9
1.45
0.07
4.6
1.51
1.35
0.16
Mar 2012
3
7
1.43
0.08
5.5
1.55
1.33
0.22
Mar 2012
4
10
1.42
0.15
10.6
1.61
1.12
0.50
Mar 2012
5
5
1.43
0.08
5.9
1.55
1.32
0.23
Mar 2012
6
5
1.28
0.30
23.4
1.73
0.93
0.81
Mar 2012
7
4
1.37
0.14
10.1
1.46
1.16
0.30
Mar 2012
8
6
1.31
0.11
8.4
1.42
1.18
0.24
Mar 2012
9
5
1.43
0.10
6.9
1.51
1.26
0.25
Mar 2012
10
6
1.23
0.07
5.9
1.37
1.15
0.22
Jul 2012
11
8
1.19
0.04
3.7
1.25
1.12
0.12
Jul 2012
12
8
1.42
0.14
10.2
1.64
1.20
0.44
Jul 2012
13
8
1.42
0.17
12.3
1.64
1.17
0.46
Jul 2012
14
5
1.52
0.08
5.4
1.62
1.40
0.22
Jul 2012
15
7
1.44
0.09
6.4
1.59
1.34
0.26
Jul 2012
16
6
1.48
0.14
9.8
1.66
1.31
0.36
Jul 2012
17
9
1.52
0.10
6.5
1.64
1.38
0.27
Jul 2012
18
9
1.29
0.09
6.7
1.41
1.18
0.23
Jul 2012
19
7
1.31
0.14
10.5
1.51
1.16
0.35
Jul 2012
20
5
1.29
0.13
9.8
1.44
1.09
0.34
Jul 2012
21
8
1.38
0.13
9.1
1.63
1.23
0.40
Jul 2012
22
8
1.44
0.12
8.6
1.58
1.26
0.32
Jul 2012
23
4
1.31
0.23
17.4
1.65
1.18
0.47
Jul 2012
24
8
1.43
0.15
10.4
1.74
1.29
0.45
Jul 2012
25
5
1.20
0.17
13.8
1.43
1.01
0.42
Jul 2012
26
7
1.32
0.22
16.4
1.73
1.07
0.66
Sep 2012
27
5
1.21
0.15
12.2
1.44
1.06
0.37
Sep 2012
28
7
1.38
0.16
11.8
1.63
1.09
0.55
Sep 2012
29
9
1.31
0.10
7.4
1.52
1.19
0.33
Sep 2012
30
8
1.23
0.08
6.8
1.38
1.10
0.28
Sep 2012
31
6
1.45
0.06
4.0
1.54
1.39
0.16
Sep 2012
32
5
1.47
0.13
8.8
1.70
1.40
0.30
Sep 2012
33
4
1.25
0.12
9.5
1.36
1.09
0.27
Sep 2012
34
9
1.42
0.10
7.0
1.60
1.29
0.31
Sep 2012
35
8
1.39
0.08
5.4
1.54
1.27
0.27
Sep 2012
36
7
1.39
0.15
11.2
1.64
1.14
0.50
Sep 2012
37
5
1.27
0.12
9.1
1.38
1.10
0.29
Sep 2012
38
6
1.33
0.12
9.3
1.46
1.10
0.35
Sep 2012
39
6
1.37
0.06
4.6
1.42
1.26
0.17
Sep 2012
40
7
1.40
0.09
6.6
1.52
1.25
0.27
Sep 2012
41
7
1.59
0.29
18.3
2.09
1.30
0.79
Sep 2012
42
4
1.42
0.07
4.9
1.50
1.34
0.16
Sep 2012
43
8
1.21
0.13
10.5
1.38
1.06
0.33
Sep 2012
44
9
1.30
0.16
12.2
1.56
1.10
0.46
Average
6.8
1.36
0.13
9(±4)
1.55
1.20
0.35 Biogeosciences, 15, 331–348, 2018 345 Edited by: Markus Kienast
Reviewed by: two anonymous referees J. Petersen et al.: Mn/Ca intra- and inter-test variability J. Petersen et al.: Mn/Ca intra- and inter-test variability J. Petersen et al.: Mn/Ca intra- and inter-test variability 346 oxic diagenesis, Geochim. Cosmochim. Ac., 43, 1075–1090,
https://doi.org/10.1016/0016-7037(79)90095-4, 1979. oxic diagenesis, Geochim. Cosmochim. Ac., 43, 1075–1090,
https://doi.org/10.1016/0016-7037(79)90095-4, 1979. Koho, K. A., Nierop, K. G. J., Moodley, L., Middelburg, J. J.,
Pozzato, L., Soetaert, K., van der Plicht, J., and Reichart, G.-
J.: Microbial bioavailability regulates organic matter preser-
vation in marine sediments, Biogeosciences, 10, 1131–1141,
https://doi.org/10.5194/bg-10-1131-2013, 2013. Geslin, E., Barras, C., Langlet, D., Nardelli, M. P., Kim, J.-H., Bon-
nin, J., Metzger, E., and Jorissen, F. J.: Survival, reproduction and
calcification of three benthic foraminiferal species in response to
experimentally induced hypoxia, in: Approaches to Study Liv-
ing Foraminifera: Collection, Maintenance and Experimentation,
edited by: Kitazato, H. and Bernhard, J. M., Springer, Japan,
163–193, 2014. Koho, K. A., de Nooijer, L. J., and Reichart, G. J.: Com-
bining
benthic
foraminiferal
ecology
and
shell
Mn/Ca
to deconvolve past bottom water oxygenation and paleo-
productivity,
Geochim. Cosmochim. Ac.,
165,
294–306,
https://doi.org/10.1016/j.gca.2015.06.003, 2015. Glock, N., Eisenhauer, A., Liebetrau, V., Wiedenbeck, M., Hensen,
C., and Nehrke, G.: EMP and SIMS studies on Mn/Ca and
Fe / Ca systematics in benthic foraminifera from the Peruvian
OMZ: a contribution to the identification of potential redox prox-
ies and the impact of cleaning protocols, Biogeosciences, 9, 341–
359, https://doi.org/10.5194/bg-9-341-2012, 2012. Koho, K. A., de Nooijer, L. J., Fontanier, C., Toyofuku, T., Oguri,
K., Kitazato, H., and Reichart, G.-J.: Benthic foraminiferal
Mn/Ca ratios reflect microhabitat preferences, Biogeosciences,
14, 3067–3082, https://doi.org/10.5194/bg-14-3067-2017, 2017. Konovalov, S. K., Luther, G. W., and Yücel, M.: Porewater redox
species and processes in the Black Sea sediments, Chem. Geol.,
245, 254–274, https://doi.org/10.1016/j.chemgeo.2007.08.010,
2007. Goldstein, S. T. and Moodley, L.: Gametogenesis and the life cy-
cle of the foraminifer Ammonia beccarii (Linné) forma tepida
(Cushman), J. Foramin. Res., 23, 213–220, 1993. Groeneveld, J. and Filipsson, H. L.: Mg/Ca and Mn/Ca ratios in
benthic foraminifera: the potential to reconstruct past variations
in temperature and hypoxia in shelf regions, Biogeosciences, 10,
5125–5138, https://doi.org/10.5194/bg-10-5125-2013, 2013. Kowalski, N., Dellwig, O., Beck, M., Grunwald, M., Dürselen, C.,
Badewien, T. H., Brumsack, H.-J., van Beusekom, J. E. E.,
and Böttcher, M. E.: A comparative study of manganese dy-
namics in the water column and sediments of intertidal sys-
tems of the North Sea, Estuar. Coast. Shelf S., 100, 3–17,
https://doi.org/10.1016/j.ecss.2011.03.011, 2012. Hagens, M., Slomp, C. P., Meysman, F. J. J. Petersen et al.: Mn/Ca intra- and inter-test variability R., Seitaj, D., Harlay, J.,
Borges, A. V., and Middelburg, J. J.: Biogeochemical processes
and buffering capacity concurrently affect acidification in a sea-
sonally hypoxic coastal marine basin, Biogeosciences, 12, 1561–
1583, https://doi.org/10.5194/bg-12-1561-2015, 2015. Langlet, D.: Réponse des foraminifères benthiques à l’anoxie dans
les milieux côtiers: études in situ de l’écologie des foraminifères
benthiques, expériences en laboratoire et analyse du métabolisme
anaérobie, PhD Thesis, Université Nantes Angers Le Mans
(UNAM), Université d’Angers, France, 2014. Hayward, B. W., Holzmann, M., Grenfell, H. R., Pawlowski, J., and
Triggs, C. M.: Morphological distinction of molecular types in
Ammonia – towards a taxonomic revision of the world’s most
commonly misidentified foraminifera, Mar. Micropaleontol.,
50, 237–271, https://doi.org/10.1016/S0377-8398(03)00074-4,
2004. Langlet, D., Geslin, E., Baal, C., Metzger, E., Lejzerowicz, F.,
Riedel, B., Zuschin, M., Pawlowski, J., Stachowitsch, M., and
Jorissen, F. J.: Foraminiferal survival after long-term in situ ex-
perimentally induced anoxia, Biogeosciences, 10, 7463–7480,
https://doi.org/10.5194/bg-10-7463-2013, 2013. Jochum, K. P., Weis, U., Stoll, B., Kuzmin, D., Yang, Q., Raczek, I.,
Jacob, D. E., Stracke, A., Birbaum, K., Frick, D. A., Gün-
ther, D., and Enzweiler, J.: Determination of reference values for
NIST SRM 610-617 glasses following ISO guidelines, Geostand. Geoanal. Res., 35, 397–429, https://doi.org/10.1111/j.1751-
908X.2011.00120.x, 2011. Leduc, G., Garbe-Schönberg, D., Regenberg, M., Contoux, C.,
Etourneau, J., and Schneider, R.: The late Pliocene Benguela
upwelling status revisited by means of multiple temper-
ature
proxies,
Geochem. Geophy. Geosy.,
15,
475–491,
https://doi.org/10.1002/2013GC004940, 2014. Lenz, C., Jilbert, T., Conley, D. J., and Slomp, C. P.: Hypoxia-
driven variations in iron and manganese shuttling in the Baltic
Sea over the past 8 kyr, Geochem. Geophy. Geosy., 16, 3754–
3766, https://doi.org/10.1002/2015GC005960, 2015. Jochum, K. P., Scholz, D., Stoll, B., Weis, U., Wilson, S. A.,
Yang, Q., Schwalb, A., Börner, N., Jacob, D. E., and An-
dreae, M. O.: Accurate trace element analysis of speleothems and
biogenic calcium carbonates by LA-ICP-MS, Chem. Geol., 318–
319,
31–44,
https://doi.org/10.1016/j.chemgeo.2012.05.009,
2012. Levin, L. A., Whitcraft, C. R., Mendoza, G. F., Gonzalez, J. P.,
and Cowie, G.: Oxygen and organic matter thresholds for
benthic faunal activity on the Pakistan margin oxygen mini-
mum zone (700–1100 m), Deep-Sea Res.-Pt. II, 56, 449–471,
https://doi.org/10.1016/j.dsr2.2008.05.032, 2009. Jorissen, F. J.: Benthic Foraminifera from the Adriatic Sea; Princi-
ples of Phenotypic Variation, Utrecht Micropaleontoligac Bullet-
tin, Utrecht University, vol. 37, 174 pp., 1988. Jorissen, F. J., de Stigter, H., and Widmark, J. G. V: A conceptual
model explaining benthic foraminiferal microhabitats, Mar. References Geosy., 7, Q01P08,
https://doi.org/10.1029/2005GC001015, 2006. Eggins, S. M., De Deckker, P., and Marshall, J.: Mg/Ca variation
in planktonic foraminifera tests: Implications for reconstruct-
ing palaeo-seawater temperature and habitat migration, Earth
Planet. Sc. Lett., 212, 291–306, https://doi.org/10.1016/S0012-
821X(03)00283-8, 2003. Bernhard, J. M., Ostermann, D. R., Williams, D. S., and
Blanks, J. K.: Comparison of two methods to identify
live
benthic
foraminifera:
A
test
between
Rose
Bengal
and CellTracker Green with implications for stable iso-
tope
paleoreconstructions,
Paleoceanography,
21,
PA4210,
https://doi.org/10.1029/2006PA001290, 2006. Froelich, P. N., Klinkhammer, G. P., Bender, M. L., Luedtke, N. A.,
Heath, G. R., Cullen, D., Dauphin, P., Hammond, D., Hart-
man, B., and Maynard, V.: Early oxidation of organic-matter
in pelagic sediments of the eastern equatorial Atlantic – sub- Bettencourt da Silva, R. and Williams, A. (Eds.): Eurachem/CITAC
Guide: Setting and Using Target Uncertainty in Chemical Mea- www.biogeosciences.net/15/331/2018/ Biogeosciences, 15, 331–348, 2018 J. Petersen et al.: Mn/Ca intra- and inter-test variability 347 Pearce, N. J. G., Perkins, W. T., Westgate, J. A., Gorton, M. P.,
Jackson, S. E., Neal, C. R., and Chenery, S. P.: A compilation
of new and published major and trace element data for NIST
SRM 610 and NIST SRM 612 glass reference materials, Geo-
standard. Newslett., 21, 115–144, https://doi.org/10.1111/j.1751-
908X.1997.tb00538.x, 1997. Maire, O., Barras, C., Gestin, T., Nardelli, M. P., Romero-
Ramirez, A., Duchêne, J.-C., and Geslin, E.: How does macro-
faunal bioturbation influence the vertical distribution of living
benthic foraminifera?, Mar. Ecol.-Prog. Ser., 561, 83–97, 2016. Marr, J. P., Baker, J. A., Carter, L., Allan, A. S. R., Dun-
bar, G. B., and Bostock, H. C.: Ecological and tempera-
ture controls on Mg/Ca ratios of Globigerina bulloides from
the southwest Pacific Ocean, Paleoceanography, 26, PA2209,
https://doi.org/10.1029/2010PA002059, 2011. Pena, L. D., Calvo, E., Cacho, I., Eggins, S., and Pelejero, C.:
Identification and removal of Mn-Mg-rich contaminant phases
on foraminiferal tests: Implications for Mg/Ca past temper-
ature reconstructions, Geochem. Geophy. Geosy., 6, Q09P02,
https://doi.org/10.1029/2005GC000930, 2005. McKay, C. L., Groeneveld, J., Filipsson, H. L., Gallego-Torres, D.,
Whitehouse, M. J., Toyofuku, T., and Romero, O. E.: A compar-
ison of benthic foraminiferal Mn/Ca and sedimentary Mn/Al as
proxies of relative bottom-water oxygenation in the low-latitude
NE Atlantic upwelling system, Biogeosciences, 12, 5415–5428,
https://doi.org/10.5194/bg-12-5415-2015, 2015. Pena, L. D., Cacho, I., Calvo, E., Pelejero, C., Eggins, S.,
and
Sadekov,
A.:
Characterization
of
contaminant
phases
in
foraminifera
carbonates
by
electron
micro-
probe
mapping,
Geochem. Geophy. Geosy.,
9,
Q07012,
https://doi.org/10.1029/2008GC002018, 2008. Mezger, E. M., de Nooijer, L. J., Boer, W., Brummer, G. J. A.,
and Reichart, G. J.: Salinity controls on Na incorporation in Red
Sea planktonic foraminifera, Paleoceanography, 31, 1562–1582,
https://doi.org/10.1002/2016PA003052, 2016. Rabalais,
N. N.,
Turner,
R. E.,
and
Wiseman,
W. J.:
Gulf
of
Mexico
Hypoxia,
a.k.a. “The
Dead
Zone,”
Annu. Rev. Ecol. Syst.,
33,
235–263,
https://doi.org/10.1146/annurev.ecolsys.33.010802.150513,
2002. Moodley,
L. and
Hess,
C.:
Tolerance
of
infaunal
benthic
foraminifera for low and high oxygen concentrations, Biol. Bull.,
183, 94–98, https://doi.org/10.2307/1542410, 1992. R Core Team: R: A Language and Environment for Statistical Com-
puting, R Foundation for Statistical Computing, Vienna, Austria,
2016. Morvan, J., Debenay, J. P., Jorissen, F., Redois, F., Bénéteau, E.,
Delplancke, M., and Amato, A. S.: Patchiness and life cycle of
intertidal foraminifera: Implication for environmental and pale-
oenvironmental interpretation, Mar. Micropaleontol., 61, 131–
154, https://doi.org/10.1016/j.marmicro.2006.05.009, 2006. Raitzsch, M., Hathorne, E. J. Petersen et al.: Mn/Ca intra- and inter-test variability C., Kuhnert, H., Groeneveld, J.,
and Bickert, T.: Modern and late pleistocene B/Ca ra-
tios of the benthic foraminifer Planulina wuellerstorfideter-
mined with laser ablation ICP-MS, Geology, 39, 1039–1042,
https://doi.org/10.1130/G32009.1, 2011. Mouret, A., Anschutz, P., Lecroart, P., Chaillou, G., Hyacinthe, C.,
Deborde, J., Jorissen, F. J., Deflandre, B., Schmidt, S., and Jouan-
neau, J. M.: Benthic geochemistry of manganese in the Bay of
Biscay, and sediment mass accumulation rate, Geo-Mar. Lett.,
29, 133–149, https://doi.org/10.1007/s00367-008-0130-6, 2009. Rao, A. M. F., Malkin, S. Y., Hidalgo-Martinez, S., and
Meysman, F. J. R.: The impact of electrogenic sulfide ox-
idation on elemental cycling and solute fluxes in coastal
sediment,
Geochim. Cosmochim. Ac.,
172,
265–286,
https://doi.org/10.1016/j.gca.2015.09.014, 2016. Munsel, D., Kramar, U., Dissard, D., Nehrke, G., Berner, Z., Bi-
jma, J., Reichart, G.-J., and Neumann, T.: Heavy metal incorpo-
ration in foraminiferal calcite: results from multi-element enrich-
ment culture experiments with Ammonia tepida, Biogeosciences,
7, 2339–2350, https://doi.org/10.5194/bg-7-2339-2010, 2010. Reichart, G. J., den Dulk, M., Visser, H. J., van Der Weijden, C. H.,
and Zachariasse, W. J.: A 225 kyr record of dust supply, pale-
oproductivity and the oxygen minimum zone from the Murray
Ridge (Northern Arabian Sea), Palaeogeogr. Palaeocl., 134, 149–
169, https://doi.org/10.1016/S0031-0182(97)00071-0, 1997. Murray, J. W.: Population dynamics of benthic foraminifera; re-
sults from the Exe Estuary, England, J. Foramin. Res., 13, 1–12,
https://doi.org/10.2113/gsjfr.13.1.1, 1983. Reichart, G. J., Jorissen, F., Anschutz, P., and Mason, P. R. D.:
Single foraminiferal test chemistry records the marine envi-
ronment, Geology, 31, 355–358, https://doi.org/10.1130/0091-
7613(2003)031<0355:SFTCRT>2.0.CO;2, 2003. Nardelli, M. P., Barras, C., Metzger, E., Mouret, A., Filipsson,
H. L., Jorissen, F., and Geslin, E.: Experimental evidence for
foraminiferal calcification under anoxia, Biogeosciences, 11,
4029–4038, https://doi.org/10.5194/bg-11-4029-2014, 2014. Saad, S. A. and Wade, C. M.: Biogeographic distribution and
habitat association of Ammonia genetic variants around the
coastline of Great Britain, Mar. Micropaleontol., 124, 54–62,
https://doi.org/10.1016/j.marmicro.2016.01.004, 2016. Nehrke, G., Keul, N., Langer, G., de Nooijer, L. J., Bijma, J.,
and Meibom, A.: A new model for biomineralization and trace-
element signatures of Foraminifera tests, Biogeosciences, 10,
6759–6767, https://doi.org/10.5194/bg-10-6759-2013, 2013. Sadekov, A. Y., Eggins, S. M., and De Deckker, P.: Characteriza-
tion of Mg/Ca distributions in planktonic foraminifera species
by electron microprobe mapping, Geochem. Geophy. Geosy., 6,
Q12P06, https://doi.org/10.1029/2005GC000973, 2005. Ní Fhlaithearta, S., Reichart, G. J., Jorissen, F. J., Fontanier, C.,
Rohling, E. J., Thomson, J., and De Lange, G. J. Petersen et al.: Mn/Ca intra- and inter-test variability Mi-
cropaleontol., 26, 3–15, 1995. Long, G. L. and Winefordner, J. D.: Limit of detection: a closer look
at the IUPAC definition, Anal. Chem., 55, 712A–724A, 1983. Longerich, H. P., Jackson, S. E., and Günther, D.: Inter-laboratory
note. Laser ablation inductively coupled plasma mass spec-
trometric transient signal data acquisition and analyte concen-
tration calculation, J. Anal. Atom. Spectrom., 11, 899–904,
https://doi.org/10.1039/JA9961100899, 1996. Keeling, R. F., Körtzinger, A., and Gruber, N.: Ocean deoxygena-
tion in a warming world, Annu. Rev. Mar. Sci., 2, 199–229,
https://doi.org/10.1146/annurev.marine.010908.163855, 2010. Klinkhammer, G. P., Mix, A. C., and Haley, B. A.: Increased
dissolved terrestrial input to the coastal ocean during the
last deglaciation, Geochem. Geophy. Geosy., 10, Q03009,
https://doi.org/10.1029/2008GC002219, 2009. p
g
Mackensen, A.: On the use of benthic foraminiferal δ13C
in
palaeoceanography:
constraints
from
primary
proxy
relationships,
Geol. Soc. Spec. Publ.,
303,
121–133,
https://doi.org/10.1144/SP303.9, 2008. Biogeosciences, 15, 331–348, 2018 www.biogeosciences.net/15/331/2018/ J. Petersen et al.: Mn/Ca intra- and inter-test variability www.biogeosciences.net/15/331/2018/ J. Petersen et al.: Mn/Ca intra- and inter-test variability 348 ern Mediterranean Sea, Mar. Micropaleontol., 51(1–2), 129–152,
https://doi.org/10.1016/j.marmicro.2003.10.001, 2004. Thamdrup, B., Fossing, H., and Jørgensen, B. B.: Manganese, iron,
and sulfur cycling in a coastal marine sediment, Aarhus Bay,
Denmark, Geochim. Cosmochim. Ac., 58, 5115–5129, 1994. Schumacher,
S.,
Jorissen,
F. J.,
Mackensen,
A.,
Goo-
day, A. J., and Pays, O.: Ontogenetic effects on stable
carbon and oxygen isotopes in tests of live (Rose Ben-
gal stained) benthic foraminifera from the Pakistan con-
tinental
margin,
Mar. Micropaleontol.,
76(3–4),
92–103,
https://doi.org/10.1016/j.marmicro.2010.06.002, 2010. Thibault de Chanvalon, A., Metzger, E., Mouret, A., Cesbron, F., Thibault de Chanvalon, A., Metzger, E., Mouret, A., Cesbron, F.,
Knoery, J., Rozuel, E., Launeau, P., Nardelli, M. P., Jorissen,
F. J., and Geslin, E.: Two-dimensional distribution of living
benthic foraminifera in anoxic sediment layers of an estuarine
mudflat (Loire estuary, France), Biogeosciences, 12, 6219–6234,
https://doi.org/10.5194/bg-12-6219-2015, 2015. Schweizer, M., Polovodova, I., Nikulina, A., and Schönfeld, J.:
Molecular identification of Ammonia and Elphidium species
(Foraminifera, Rotaliida) from the Kiel Fjord (SW Baltic
Sea) with rDNA sequences, Helgoland Mar. Res., 65, 1–10,
https://doi.org/10.1007/s10152-010-0194-3, 2011. Thibault de Chanvalon, A., Mouret, A., Knoery, J., Geslin, E.,
Péron, O., and Metzger, E.: Manganese, iron and phosphorus cy-
cling in an estuarine mudflat, Loire, France, J. Sea Res., 118,
92–102, https://doi.org/10.1016/j.seares.2016.10.004, 2016. van de Velde, S. and Meysman, F. J. R.: The influence of
bioturbation on iron and sulphur cycling in marine sedi-
ments: a model analysis, Aquat. Geochem., 22, 469–504,
https://doi.org/10.1007/s10498-016-9301-7, 2016. Seitaj, D., Schauer, R., Sulu-Gambari, F., Hidalgo-Martinez, S.,
Malkin, S. Y., Burdorf, L. D. W., Slomp, C. P., and
Meysman, F. J. R.: Cable bacteria generate a firewall against eu-
xinia in seasonally hypoxic basins, P. Natl. Acad. Sci. USA, 112,
13278–13283, https://doi.org/10.1073/pnas.1510152112, 2015. van de Velde, S., Lesven, L., Burdorf, L. D. W., Hidalgo-
Martinez, S., Geelhoed, J. S., Van Rijswijk, P., Gao, Y.,
and Meysman, F. J. R.: The impact of electrogenic sul-
fur oxidation on the biogeochemistry of coastal sediments:
A field study, Geochim. Cosmochim. Ac., 194, 211–232,
https://doi.org/10.1016/j.gca.2016.08.038, 2016. Seitaj, D., Sulu-Gambari, F., Burdorf, L. D. W., Romero-
Ramirez, A., Maire, O., Malkin, S. Y., Slomp, C. P., and
Meysman, F. J. R.: Sedimentary oxygen dynamics in a sea-
sonally hypoxic basin, Limnol. Oceanogr., 62, 452–473,
https://doi.org/10.1002/lno.10434, 2017. Sen Gupta, B. K., Turner, R. E., and Rabalais, N. N.: Seasonal oxy-
gen depletion in continental-shelf waters of Louisiana?: Histori-
cal record of benthic foraminifers, Geology, 24, 227–230, 1996. J. Petersen et al.: Mn/Ca intra- and inter-test variability J.: Re-
constructing the seafloor environment during sapropel forma-
tion using benthic foraminiferal trace metals, stable isotopes,
and sediment composition, Paleoceanography, 25, PA4225,
https://doi.org/10.1029/2009PA001869, 2010. Schenau, S. J., Reichart, G. J., and De Lange, G. J.: Oxygen min-
imum zone controlled Mn redistribution in Arabian Sea sedi-
ments during the late Quaternary, Paleoceanography, 17, 1058,
https://doi.org/10.1029/2000PA000621, 2002. Pakhomova, S. V, Hall, P. O. J., Yu, M., Rozanov, A. G.,
Tengberg, A., and Vershinin, A. V: Fluxes of iron and
manganese
across
the
sediment
–
water
interface
un-
der various redox conditions, Mar. Chem., 107, 319–331,
https://doi.org/10.1016/j.marchem.2007.06.001, 2007. Schmiedl, G., Pfeilsticker, M., Hemleben, C., and Mackensen, A.:
Environmental and biological effects on the stable isotope com-
position of recent deep-sea benthic foraminifera from the west- www.biogeosciences.net/15/331/2018/ Biogeosciences, 15, 331–348, 2018 J. Petersen et al.: Mn/Ca intra- and inter-test variability van der Weijden, C. H., Reichart, G.-J., and van Os, B. J. H.:
Sedimentary
trace
element
records
over
the
last
200 kyr
from
within
and
below
the
northern
Arabian
Sea
oxygen
minimum
zone,
Mar. Geol.,
231,
69–88,
https://doi.org/10.1016/j.margeo.2006.05.013, 2006. Slomp, C. P., Malschaert, J. F. P., Lohse, L., and van Raaphorst, W.:
Iron and manganese cycling in different sedimentary environ-
ments on the North Sea continental margin, Cont. Shelf Res.,
17, 1083–1117, 1997. Van der Zwaan, G. J., Duijnstee, I. A. P., den Dulk, M., Ernst, S. R.,
Jannink, N. T., and Kouwenhoven, T. J.: Benthic foraminifers:
proxies or problems? A review of paleocological concepts,
Earth-Sci. Rev., 46, 213–236, 1999. Sulu-Gambari, F., Seitaj, D., Meysman, F. J. R., and Schauer, R.:
Cable bacteria control iron-phosphorus dynamics in sediments of
a coastal hypoxic basin, Environ. Sci. Technol., 50, 1227–1233,
https://doi.org/10.1021/acs.est.5b04369, 2016a. Vandieken, V., Nickel, M., and Jørgensen, B. B.: Carbon
mineralization
in
Arctic
sediments
northeast
of
Sval-
bard: Mn(IV) and Fe(III) reduction as principal anaerobic
respiratory pathways, Mar. Ecol.-Prog. Ser., 322, 15–27,
https://doi.org/10.3354/meps322015, 2006. Sulu-Gambari,
F.,
Seitaj,
D.,
Behrends,
T.,
Banerjee,
D.,
Meysman, F. J. R., and Slomp, C. P.: Impact of cable bacteria
on sedimentary iron and manganese dynamics in a seasonally-
hypoxic marine basin, Geochim. Cosmochim. Ac., 192, 49–69,
https://doi.org/10.1016/j.gca.2016.07.028, 2016b. Wickham, H.: ggplot2: Elegant Graphics for Data Analysis,
Springer New York, New York, 2009. Wit, J. C., Reichart, G. J., Jung, S. J. A., and Kroon, D.:
Approaches to unravel seasonality in sea surface temper-
atures
using
paired
single-specimen
foraminiferal
δ18O
and
Mg/Ca
analyses,
Paleoceanography,
25,
PA4220,
https://doi.org/10.1029/2009PA001857, 2010. Sulu-Gambari,
F.,
Roepert,
A.,
Jilbert,
T.,
Hagens,
M.,
Meysman,
F. J. R.,
and
Slomp,
C. P.:
Molybde-
num
dynamics
in
sediments
of
a
seasonally-hypoxic
coastal
marine
basin,
Chem. Geol.,
466(July),
627–640,
https://doi.org/10.1016/j.chemgeo.2017.07.015, 2017. Sundby,
B. and
Silverberg,
N.:
Manganese
fluxes
in
the
benthic boundary layer, Limnol. Oceanogr., 30, 372–381,
https://doi.org/10.4319/lo.1985.30.2.0372, 1985. Biogeosciences, 15, 331–348, 2018 www.biogeosciences.net/15/331/2018/
|
https://openalex.org/W2775226242
|
https://www.mdpi.com/2079-4991/7/12/446/pdf?version=1513263005
|
English
| null |
Impedance Spectroscopy Analysis and Equivalent Circuit Modeling of Graphene Oxide Solutions
|
Nanomaterials
| 2,017
|
cc-by
| 7,514
|
Impedance Spectroscopy Analysis and Equivalent
Circuit Modeling of Graphene Oxide Solutions Youngbin Yoon 1, Jeonghoo Jo 2, Seungdu Kim 1, In Gyu Lee 1, Byung Jin Cho 3,
Myunghun Shin 2,* and Wan Sik Hwang 1,* Youngbin Yoon 1, Jeonghoo Jo 2, Seungdu Kim 1, In Gyu Lee 1, Byung Jin Cho 3,
Myunghun Shin 2,* and Wan Sik Hwang 1,* 1
Department of Materials Engineering, Korea Aerospace University, Goyang 10540, Korea;
ybyoon93@gmail.com (Y.Y.); seungdukim@gmail.com (S.K.); leeig@kau.ac.kr (I.G.L.)
2
School of Electronics and Information Engineering, Korea Aerospace University, Goyang 10540, Korea ybyoon93@gmail.com (Y.Y.); seungdukim@gmail.com (S.K.); leeig@kau.ac.kr (I.G.L.)
2
School of Electronics and Information Engineering, Korea Aerospace University, Goyang 10540, Kore
jhjo@kau.kr ybyoon93@gmail.com (Y.Y.); seungdukim@gmail.com (S.K.); leeig@kau.ac.kr (I.G.L.)
2
School of Electronics and Information Engineering, Korea Aerospace University, Goyang 10540, Korea
jhjo@kau.kr y y
g
g
g
g
2
School of Electronics and Information Engineering, Korea Aerospace University, Goyang 10540, Ko
jhjo@kau.kr 3
Department of Electrical Engineering, KAIST, Daejeon 34141, Korea; elebjcho81@kaist.ac.kr
*
Correspondence: mhshin@kau.ac.kr (M.S.); whwang@kau.ac.kr (W.S.H.); Tel.: +82-2-300-0145 (M.S.);
Tel.: +82-2-300-0293 (W.S.H.) 3
Department of Electrical Engineering, KAIST, Daejeon 34141, Korea; elebjcho81@kaist.ac.kr
*
Correspondence: mhshin@kau.ac.kr (M.S.); whwang@kau.ac.kr (W.S.H.); Tel.: +82-2-300-0145 (M.S.);
Tel.: +82-2-300-0293 (W.S.H.) Received: 11 October 2017; Accepted: 9 December 2017; Published: 14 December 2017 Abstract: The optical and electrical characteristics of a graphene oxide solution (GS) with different
graphene oxide (GO) concentrations in de-ionized water are investigated via the electrochemical
impedance spectroscopy (EIS) method. The measurement results produced by the EIS for the GS
are represented with both Bode and Nyquist plots in a frequency range from 1 kHz to 10 MHz. Using these results, we develop an equivalent circuit model as a function of the GO concentration,
representing the GS as a mixed circuit of two-dimensional (2D) GO dispersed in parallel in de-ionized
(DI) water. The underlying physics of the current-flowing behavior in the GS are explained and
interpreted using empirical circuit models; the circuit model also shows that highly resistive GO
becomes conductive in GS form in the DI water. The findings in this work should draw new
attention toward GSes and related applications, including functional composite materials, catalysts,
and filter membranes. Keywords: graphene oxide solution; impedance spectroscopy analysis; equivalent circuit mode nanomaterials nanomaterials nanomaterials nanomaterials www.mdpi.com/journal/nanomaterials Nanomaterials 2017, 7, 446; doi:10.3390/nano7120446 2. Experiment
2. Experim The GO was obtained from graphite (−200 mesh, 99.9999%, Alfa Aesar, Ward Hill, MA, USA)
via the Hummers method [21,22]. The residual metal ions and acids in the GO were removed using a
filter press system, which was well described in the previous report [23]. After the purification process,
the residual impurities in the GO were measured via an X-ray photoelectron spectroscopy (XPS)
analysis. The results showed that K and Mn were not detected on the GO, but 1.1 percent of atomic
S was observed. Based on this XPS analysis, a very small amount of S remained on the GO, and the
S was not removed, even after the washing process was carried out five times. This indicated that
the residual S was strongly attached to the GO flakes, and the impedance effect would be negligible. Next, the purified GO was freeze-dried, turned into powder form, and subsequently dispersed in DI
water via sonication. Finally, a stable, homogeneous, dark-brown GS formed. This type of GS is known
to be stable, even after almost two years of aging [24]. For the experiment, various GS samples with
different GO concentrations were produced by mixing the predetermined 10 g/L GS (10 g GO per 1 L
DI water) with DI water at different ratios. The GO was obtained from graphite (−200 mesh, 99.9999%, Alfa Aesar, Ward Hill, MA, USA)
via the Hummers method [21,22]. The residual metal ions and acids in the GO were removed using
a filter press system, which was well described in the previous report [23]. After the purification
process, the residual impurities in the GO were measured via an X-ray photoelectron spectroscopy
(XPS) analysis. The results showed that K and Mn were not detected on the GO, but 1.1 percent of
atomic S was observed. Based on this XPS analysis, a very small amount of S remained on the GO,
and the S was not removed, even after the washing process was carried out five times. This indicated
that the residual S was strongly attached to the GO flakes, and the impedance effect would be
negligible. Next, the purified GO was freeze-dried, turned into powder form, and subsequently
dispersed in DI water via sonication. Finally, a stable, homogeneous, dark-brown GS formed. This
type of GS is known to be stable, even after almost two years of aging [24]. 2. Experiment
2. Experim For the experiment, various
GS samples with different GO concentrations were produced by mixing the predetermined 10 g/L GS
(10 g GO per 1 L DI water) with DI water at different ratios. Figure 1a shows the transmittance of the different GSes (different GO concentrations) as a Figure 1a shows the transmittance of the different GSes (different GO concentrations) as a function
of wavelength in a visible wavelength range from 400 to 700 nm. The GS with a low GO concentration of
0.05 g/L was somehow transparent in the visible wavelength, but the GS became obviously opaque as the
GO concentration increased over 0.5 g/L. Collectively, the transmittance of the GS at the 454- and 656-nm
wavelengths is shown as a function of GO concentration in Figure 1b. The transmittance of the GS
decreased exponentially as the GO concentration increased, and the GS became completely opaque when
the GO concentration exceeded 0.5 g/L at 454 nm (2 g/L at 656 nm). The transmittance of the GS with the
10-g/L GO concentration was 105 times smaller than that of the DI water. Figure 1a shows the transmittance of the different GSes (different GO concentrations) as a
function of wavelength in a visible wavelength range from 400 to 700 nm. The GS with a low GO
concentration of 0.05 g/L was somehow transparent in the visible wavelength, but the GS became
obviously opaque as the GO concentration increased over 0.5 g/L. Collectively, the transmittance of
the GS at the 454- and 656-nm wavelengths is shown as a function of GO concentration in Figure 1b. The transmittance of the GS decreased exponentially as the GO concentration increased, and the GS
became completely opaque when the GO concentration exceeded 0.5 g/L at 454 nm (2 g/L at 656 nm). The transmittance of the GS with the 10-g/L GO concentration was 105 times smaller than that of the
DI water. Figure 1. Transmittance of different GSes (a) as a function of wavelength and (b) as a function of
graphene oxide (GO) concentration in the graphene oxide solution (GS) at 454- and 656-nm
wavelengths. The inset shows a typical scanning electron microscope (SEM) image of the GO flakes
on the SiO2/Si wafer. Figure 1. Transmittance of different GSes (a) as a function of wavelength and (b) as a function
of graphene oxide (GO) concentration in the graphene oxide solution (GS) at 454- and 656-nm
wavelengths. 1. Introduction Graphene has been intensively investigated for various applications due to its excellent electrical,
mechanical, thermal, and optical properties [1–4], since its re-discovery using Scotch Tape in 2004. Though this early-stage graphene was initially produced via the mechanical exfoliation approach
(Scotch Tape method), at present, most graphene is prepared via the chemical vapor deposition (CVD)
method, which provides substantial benefits in terms of the large-scale processing of wafer size. However, CVD graphene requires a very high temperature and transferring processes when it is
deposited onto the desired substrate, which is not suitable for commercialized graphene-embedded
products. Alternatively, graphene oxide (GO) dispersed in a solvent could be a promising method
for wafer-size mass production at low costs [5,6]. GO can be easily coated on a target substrate
using spin-coating and/or spray methods as a pre-cursor, and then turned into graphene via
thermal, chemical, or photocatalytic reduction [7]. Meanwhile, GO is not only used as a precursor for
grapheme, but also performs its own functions with a hydroxyl/epoxide group on the basal plane and
carbonyl/carboxyl group on the edge, which is unlike the finally-produced grapheme [8,9]. Due to
such oxygen-containing functional groups, GO shows excellent chemical sensing properties [10]
and can be dispersed uniformly and stably in the de-ionized (DI) water (hereinafter referred to
as a graphene oxide solution (GS)) [11]. At present, GSes are commercially available for various
applications, such as transparent conductive coatings, polymer synthesis, ultra-strong graphene oxide www.mdpi.com/journal/nanomaterials Nanomaterials 2017, 7, 446; doi:10.3390/nano7120446 Nanomaterials 2017, 7, 446 2 of 8 paper, supercapacitors, solar cells, electro-static dissipation (ESD) films, chemical and bio sensors, and
the purification of wastewater [12–20]. Despite these numerous advantages, the detailed electrical
characteristics of GSes have not been fully reported yet for different GO concentrations. oxide paper, supercapacitors, solar cells, electro-static dissipation (ESD) films, chemical and bio
sensors, and the purification of wastewater [12–20]. Despite these numerous advantages, the detailed
electrical characteristics of GSes have not been fully reported yet for different GO concentrations This study investigated GS electrical characteristics for different GO concentrations. In detail,
electrochemical impedance spectroscopy (EIS) was performed, which was known to be a suitable
technique to investigate the electrical properties of liquid materials. The observed experimental results
were correlated with the equivalent circuit modeling at different GSes, which enabled the physical
parameters to be extracted to explain and interpret the current-flowing mechanism of the various GSes. 1. Introduction The findings in this work should draw new attention toward GSes and their related applications,
including functional composite materials, catalysts, and filter membranes. electrical characteristics of GSes have not been fully reported yet for different GO concentrations. This study investigated GS electrical characteristics for different GO concentrations. In detail,
electrochemical impedance spectroscopy (EIS) was performed, which was known to be a suitable
technique to investigate the electrical properties of liquid materials. The observed experimental
results were correlated with the equivalent circuit modeling at different GSes, which enabled the
physical parameters to be extracted to explain and interpret the current-flowing mechanism of the
various GSes. The findings in this work should draw new attention toward GSes and their related
applications, including functional composite materials, catalysts, and filter membranes. 2. Experiment
2. Experim The inset shows a typical scanning electron microscope (SEM) image of the GO flakes on
the SiO2/Si wafer. Figure 1. Transmittance of different GSes (a) as a function of wavelength and (b) as a function of
graphene oxide (GO) concentration in the graphene oxide solution (GS) at 454- and 656-nm
wavelengths. The inset shows a typical scanning electron microscope (SEM) image of the GO flakes
on the SiO2/Si wafer. Figure 1. Transmittance of different GSes (a) as a function of wavelength and (b) as a function
of graphene oxide (GO) concentration in the graphene oxide solution (GS) at 454- and 656-nm
wavelengths. The inset shows a typical scanning electron microscope (SEM) image of the GO flakes on
the SiO2/Si wafer. 3 of 8
f Nanomaterials 2017, 7, 446 Nanomaterials 2017, 7, 446 For convenience in the discussion below, a sample number of GSes was assigned according to
GO concentrations, as shown in Table 1; GS1 and GS20 were dispersed in the 0.25 g GO and 7.00 g GO,
respectively, in 1 L of DI water. In addition, the GS samples were classified into two groups—a relatively low
GO concentration (GS1 to GS11) and relatively high GO concentration (GS12 to GS20), which is subjective. The electrical impedances of the different GSes (from the DI water and GS1 to GS20) were measured using
24 × 24-mm2 parallel copper electrodes 20 mm apart. The copper plates were used as electrodes due to
the ease of integration into the homemade measurement system for this study. The measuring frequency
range was 1 kHz to 10 MHz, with an AC voltage of 30 mV using a Keithley 4200-SCS parameter analyzer,
indicating that interface information between the electrode and solution was excluded. For convenience in the discussion below, a sample number of GSes was assigned according to
GO concentrations, as shown in Table 1; GS1 and GS20 were dispersed in the 0.25 g GO and 7.00 g
GO, respectively, in 1 L of DI water. In addition, the GS samples were classified into two groups—a
relatively low GO concentration (GS1 to GS11) and relatively high GO concentration (GS12 to GS20),
which is subjective. The electrical impedances of the different GSes (from the DI water and GS1 to
GS20) were measured using 24 × 24-mm2 parallel copper electrodes 20 mm apart. 3. Results and Discussion
3. Results and Discussion The measured electrical impedance (Z) was rendered in a complex Cartesian form (Z = ZRe + jZIm),
where ZRe and ZIm were the real (resistance) and imaginary (reactance) portions of the impedance. The real portion in the low frequency range strongly depended on the resistor value measured in
DC mode while the imaginary portion was related to the reactance of the equivalent circuit formed
between the parallel copper electrodes, and it may have varied greatly with respect to the frequency
depending on the equivalent circuit structure. The measured electrical impedance (Z) was rendered in a complex Cartesian form (Z = ZRe
jZIm), where ZRe and ZIm were the real (resistance) and imaginary (reactance) portions of the
impedance. The real portion in the low frequency range strongly depended on the resistor value
measured in DC mode while the imaginary portion was related to the reactance of the equivalent
circuit formed between the parallel copper electrodes, and it may have varied greatly with respect to
the frequency depending on the equivalent circuit structure. b
h
f
f
h
l
d
f
h Figure 2a,b presents the frequency responses of the real and imaginary portions of the impedances
of the different GSes listed in Table 1. These results showed that both the real and imaginary
components of the impedances decreased as the GO concentration increased in the GS. As shown in
Figure 2a, the resistance of the GS decreased as the GO increased. Figure 2a,b presents the frequency responses of the real and imaginary portions of the
impedances of the different GSes listed in Table 1. These results showed that both the real and
imaginary components of the impedances decreased as the GO concentration increased in the GS. As
shown in Figure 2a, the resistance of the GS decreased as the GO increased. Figure 2. Frequency response of the (a) real and (b) imaginary portions of the GS impedances at
different GSes. Figure 2. Frequency response of the (a) real and (b) imaginary portions of the GS impedances at
different GSes. Figure 2. Frequency response of the (a) real and (b) imaginary portions of the GS impedances at
different GSes. Figure 2. Frequency response of the (a) real and (b) imaginary portions of the GS impedances at
different GSes. 2. Experiment
2. Experim The copper plates
were used as electrodes due to the ease of integration into the homemade measurement system for
this study. The measuring frequency range was 1 kHz to 10 MHz, with an AC voltage of 30 mV using
a Keithley 4200-SCS parameter analyzer, indicating that interface information between the electrode Table 1. Various GO concentration in GS samples. Low-GO Samples
GS1
GS2
GS3
GS4
GS5
GS6
GS7
GS8
GS9
GS10
GS11
Concentration(g/L)
0.25
0.49
0.73
0.96
1.19
2.24
3.16
3.93
4.55
4.73
4.91
High-GO samples
GS12
GS13
GS14
GS15
GS16
GS17
GS18
GS19
GS20
n/a
n/a
Concentration (g/L)
5.08
5.24
5.40
5.56
5.75
6.10
6.42
6.72
7.00
and solution was excluded. Table 1. Various GO concentration in GS samples. Low-GO Samples
GS1
GS2
GS3
GS4
GS5
GS6
GS7
GS8
GS9
GS10
GS11
Concentration(g/L)
0.25
0.49
0.73
0.96
1.19
2.24
3.16
3.93
4.55
4.73
4.91
High-GO samples
GS12
GS13
GS14
GS15
GS16
GS17
GS18
GS19
GS20
n/a
n/a
Concentration (g/L)
5.08
5.24
5.40
5.56
5.75
6.10
6.42
6.72
7.00 Table 1. Various GO concentration in GS samples. d. Table 1. Various GO concentration in GS samples. . 3. Results and Discussion
3. Results and Discussion Frequency response of GS1 to GS11: (a) real and (b) imaginary portions of the impedance,
and frequency response of GS12 to GS20: (c) real and (d) imaginary portions of the impedance. Figure 3. Frequency response of GS1 to GS11: (a) real and (b) imaginary portions of the impedance,
and frequency response of GS12 to GS20: (c) real and (d) imaginary portions of the impedance. This impedance changes with frequencies can be also interpreted with a Nyquist plot where the
impedance is plotted with Cartesian coordinates of the real portion on the X-axis and the imaginary
portion on the Y-axis. Figure 4 shows the Nyquist impedance plots of the two GS groups with the
low (Figure 4a) and high (Figure 4b) GO concentrations. In Figure 4a, for the GSes with low GO
concentrations, the semicircular frequency behavior was clearly observed, and the diameters of the
circles rapidly reduced as the GO concentration increased. This indicates that the charge conductivity
(due to the resistor components in the GS) significantly increased with the higher GO concentrations. However, for the GSes with high GO concentrations, as shown in Figure 4b, the reactance changes
with frequencies (toward the inductive direction) were smaller than the decreases in resistance,
according to the GO concentration, meaning that the conductive current was dominant (very small
resistance) in the GSes with high GO concentrations, and the phase delay of the conductive current
increased in proportion to the frequency as if the inductors were connected in series. This impedance changes with frequencies can be also interpreted with a Nyquist plot where the
impedance is plotted with Cartesian coordinates of the real portion on the X-axis and the imaginary
portion on the Y-axis. Figure 4 shows the Nyquist impedance plots of the two GS groups with the
low (Figure 4a) and high (Figure 4b) GO concentrations. In Figure 4a, for the GSes with low GO
concentrations, the semicircular frequency behavior was clearly observed, and the diameters of the
circles rapidly reduced as the GO concentration increased. This indicates that the charge conductivity
(due to the resistor components in the GS) significantly increased with the higher GO concentrations. 3. Results and Discussion
3. Results and Discussion In detail, the frequency responses of the real and imaginary portions of the GS impedance are
shown with a linear scale in Figure 3. (Figure 3a,b is for GS1 to GS11, and Figure 3c,d is for GS11 to
GS20.) As discussed in Figure 2, the resistance of the various GSes (GS1 to GS20) decreased as the GO
concentration in the GS increased, and the frequency improved for the GSes with both the low and
high GO concentrations. This implies that both the resistive and capacitive paths for the current-flow
existed in the GS. In Figure 3b, the reactance of the GSes with low GO concentrations showed the
In detail, the frequency responses of the real and imaginary portions of the GS impedance are
shown with a linear scale in Figure 3. (Figure 3a,b is for GS1 to GS11, and Figure 3c,d is for GS11
to GS20.) As discussed in Figure 2, the resistance of the various GSes (GS1 to GS20) decreased as the GO
concentration in the GS increased, and the frequency improved for the GSes with both the low and high
GO concentrations. This implies that both the resistive and capacitive paths for the current-flow existed 4 of 8 Nanomaterials 2017, 7, 446 in the GS. In Figure 3b, the reactance of the GSes with low GO concentrations showed the rollover for
the frequency, representing a capacitive-mode behavior. However, the reactance of the GSes with high
GO concentrations steadily declined at high frequencies, indicating that the capacitive-mode behavior
shifted to inductive-mode behavior, as shown in Figure 3d. This reveals that the current-flowing
mechanism in the GS involved inductor components for GSes with high GO concentrations and high
frequency regions. Nanomaterials 2017, 7, 446
4 of 8
capacitive-mode behavior shifted to inductive-mode behavior, as shown in Figure 3d. This reveals
that the current-flowing mechanism in the GS involved inductor components for GSes with high GO
concentrations and high frequency regions. Figure 3. Frequency response of GS1 to GS11: (a) real and (b) imaginary portions of the impedance,
and frequency response of GS12 to GS20: (c) real and (d) imaginary portions of the impedance. Figure 3. Frequency response of GS1 to GS11: (a) real and (b) imaginary portions of the impedance,
and frequency response of GS12 to GS20: (c) real and (d) imaginary portions of the impedance. Figure 3. 3. Results and Discussion
3. Results and Discussion At first, the circuit model was developed
for a stacked three-dimensional (3D) GO with an inductor (LGO), resistor (RGO), and two constant
phase elements (CPE: QGO1 and QGO2), as shown in Figure 5a [10]. Generally, when circuits are not
expressed as simple RC circuits, the CPE can be introduced [25] with a frequency independent Q-value,
an imperfective resistive capacitance, and index α (0 < α < 1; α = 0 for a pure resistor and α = 1 for an
ideal capacitor), as given in Equation (1). Based on the impedance analysis of Figures 2–4 and our previous work [10], we proposed an
equivalent GS circuit model, as shown in Figure 5. At first, the circuit model was developed for a
stacked three-dimensional (3D) GO with an inductor (LGO), resistor (RGO), and two constant phase
elements (CPE: QGO1 and QGO2), as shown in Figure 5a [10]. Generally, when circuits are not expressed
as simple RC circuits, the CPE can be introduced [25] with a frequency independent Q-value, an
imperfective resistive capacitance, and index α (0 < α < 1; α = 0 for a pure resistor
and α = 1 for an ideal capacitor) as given in Equation (1) ܼொൌ
1
ܳሺ݆ݓሻ
(1)
ZQ =
1
Q(jw)a
(1)
q
( )
ܼொൌ
1
ܳሺ݆ݓሻ
(1) (1)
(1)
(1) Figure 5. Equivalent circuit models: (a) stacked three-dimensional (3D) GO10; (b) de-ionized (DI), and
(c) GS. In the previous work [10] the stacked 3D graphite showed an inductive conductor proper
Figure 5. Equivalent circuit models: (a) stacked three-dimensional (3D) GO10; (b) de-ionized (DI), and
(c) GS. Figure 5. Equivalent circuit models: (a) stacked three-dimensional (3D) GO10; (b) de-ionized (DI),
and (c) GS. g
q
( )
(
)
; ( )
(
),
(c) GS. In the previous work [10] the stacked 3D graphite showed an inductive conductor proper
Figure 5. Equivalent circuit models: (a) stacked three-dimensional (3D) GO10; (b) de-ionized (DI), and
(c) GS. Figure 5. Equivalent circuit models: (a) stacked three-dimensional (3D) GO10; (b) de-ionized (DI),
and (c) GS. while the fully oxidized GO stack showed highly-resistive properties consisting of a large resistor
and a CPE pair with a long phase delay (an inductor in series), as shown in Figure 5a. Once the
oxidation was sufficiently completed, QGO1 became similar to an ideal capacitor with an α of ~1. 3. Results and Discussion
3. Results and Discussion While
the GOs were dispersed in DI water in this work, both of the circuit diagrams could be similar in
principle. From the EIS measurements that are shown in Figure 2, the DI water was identified as the
circuit of an RC pair, as shown in Figure 5b. In the circuit model of the DI water, even a very small
quantity of aqua ions could contribute to the conductive current flow, and the non-conducting water
served as the capacitor’s dielectric material via the water that was de-ionized for that purpose, which
was shown as the EIS results of Figure 2. When considering the GO in the GS as a conductive network
component and the DI water as a matrix material, we hypothesized that the equivalent circuit model
of the GS could be similar to that of the GO. However, one of the GS’s CPEs (Figure 5a) was closer to
an ideal capacitor because of the capacitor component in the DI water as shown in Figure 5c
In the previous work [10], the stacked 3D graphite showed an inductive conductor property,
while the fully oxidized GO stack showed highly-resistive properties consisting of a large resistor
and a CPE pair with a long phase delay (an inductor in series), as shown in Figure 5a. Once the
oxidation was sufficiently completed, QGO1 became similar to an ideal capacitor with an α of ~1. While
the GOs were dispersed in DI water in this work, both of the circuit diagrams could be similar in
principle. From the EIS measurements that are shown in Figure 2, the DI water was identified as the
circuit of an RC pair, as shown in Figure 5b. In the circuit model of the DI water, even a very small
quantity of aqua ions could contribute to the conductive current flow, and the non-conducting water
served as the capacitor’s dielectric material via the water that was de-ionized for that purpose, which
was shown as the EIS results of Figure 2. When considering the GO in the GS as a conductive network
component and the DI water as a matrix material, we hypothesized that the equivalent circuit model
of the GS could be similar to that of the GO. 3. Results and Discussion
3. Results and Discussion However, for the GSes with high GO concentrations, as shown in Figure 4b, the reactance changes with
frequencies (toward the inductive direction) were smaller than the decreases in resistance, according
to the GO concentration, meaning that the conductive current was dominant (very small resistance)
in the GSes with high GO concentrations, and the phase delay of the conductive current increased in
proportion to the frequency as if the inductors were connected in series. 5 of 8
5 of 8
5
f 8 Nanomaterials 2017, 7, 446
Nanomaterials 2017, 7, 446 Figure 4. Nyquist plot of the GS samples: (a) for GS1 to GS11, which have low GO concentrations,
and (b) for GS12 to GS20, which have high GO concentrations. Figure 4. Nyquist plot of the GS samples: (a) for GS1 to GS11, which have low GO concentrations,
and (b) for GS12 to GS20, which have high GO concentrations. Figure 4. Nyquist plot of the GS samples: (a) for GS1 to GS11, which have low GO concentrations,
and (b) for GS12 to GS20, which have high GO concentrations. Figure 4. Nyquist plot of the GS samples: (a) for GS1 to GS11, which have low GO concentrations,
and (b) for GS12 to GS20, which have high GO concentrations. Figure 4. Nyquist plot of the GS samples: (a) for GS1 to GS11, which have low GO concentrations,
and (b) for GS12 to GS20, which have high GO concentrations. Figure 4. Nyquist plot of the GS samples: (a) for GS1 to GS11, which have low GO concentrations, Based on the impedance analysis of Figures 2–4 and our previous work [10], we proposed an
equivalent GS circuit model, as shown in Figure 5. At first, the circuit model was developed for a
stacked three-dimensional (3D) GO with an inductor (LGO), resistor (RGO), and two constant phase
elements (CPE: QGO1 and QGO2), as shown in Figure 5a [10]. Generally, when circuits are not expressed
as simple RC circuits, the CPE can be introduced [25] with a frequency independent Q-value, an
imperfective resistive capacitance, and index α (0 < α < 1; α = 0 for a pure resistor
and α = 1 for an ideal capacitor), as given in Equation (1). Based on the impedance analysis of Figures 2–4 and our previous work [10], we proposed
an equivalent GS circuit model, as shown in Figure 5. 3. Results and Discussion
3. Results and Discussion However, one of the GS’s CPEs (Figure 5a) was closer to
In the previous work [10], the stacked 3D graphite showed an inductive conductor property,
while the fully oxidized GO stack showed highly-resistive properties consisting of a large resistor and
a CPE pair with a long phase delay (an inductor in series), as shown in Figure 5a. Once the oxidation
was sufficiently completed, QGO1 became similar to an ideal capacitor with an α of ~1. While the GOs
were dispersed in DI water in this work, both of the circuit diagrams could be similar in principle. From the EIS measurements that are shown in Figure 2, the DI water was identified as the circuit of
an RC pair, as shown in Figure 5b. In the circuit model of the DI water, even a very small quantity of
aqua ions could contribute to the conductive current flow, and the non-conducting water served as the
capacitor’s dielectric material via the water that was de-ionized for that purpose, which was shown as
the EIS results of Figure 2. When considering the GO in the GS as a conductive network component
and the DI water as a matrix material, we hypothesized that the equivalent circuit model of the GS 6 of 8 Nanomaterials 2017, 7, 446 could be similar to that of the GO. However, one of the GS’s CPEs (Figure 5a) was closer to an ideal
capacitor because of the capacitor component in the DI water, as shown in Figure 5c. Nanomaterials 2017, 7, 446
6 of 8
Nanomaterials 2017, 7, 446
6 of 8 Figure 6 shows the impedance characteristics of the GS samples depending on GO concentrations
with frequencies, which are correlated with the modeling values. This showed that the measured
magnitude (Figure 6a) and phase shift (Figure 6b) of the frequency response were well matched
with the calculated responses of the circuit model of Figure 5c using the best-fitted parameters. This agreement indicates that the accuracy of the model was further verified. It could be inferred
that the well-suspended GO flakes (a few hundred µm in size) in the DI water contributed to the
charge conduction as the GO concentration increased in the GS regions with low GO concentrations. 3. Results and Discussion
3. Results and Discussion On the other hand, in the GS regions with high GO concentrations, as the GO concentration increased,
the randomly-connected GO network and the complexity of the current-flowing mechanism retarded
the phase delay of the current to the external voltage. Thus, the inductive behavior of the current
could be observed at high frequencies, as shown in Figure 4b. Based on this premise, several physical
parameters, including the LGS, RGS, and QGS, with their own α and CGS values, were also extracted. Their typical values are listed in Table 2 and Figure 7 in detail. Figure 6 shows the impedance characteristics of the GS samples depending on GO
concentrations with frequencies, which are correlated with the modeling values. This showed that
the measured magnitude (Figure 6a) and phase shift (Figure 6b) of the frequency response were well
matched with the calculated responses of the circuit model of Figure 5c using the best-fitted
parameters. This agreement indicates that the accuracy of the model was further verified. It could be
inferred that the well-suspended GO flakes (a few hundred μm in size) in the DI water contributed
to the charge conduction as the GO concentration increased in the GS regions with low GO
concentrations. On the other hand, in the GS regions with high GO concentrations, as the GO
concentration increased, the randomly-connected GO network and the complexity of the current-
flowing mechanism retarded the phase delay of the current to the external voltage. Thus, the
inductive behavior of the current could be observed at high frequencies, as shown in Figure 4b. Based
on this premise, several physical parameters, including the LGS, RGS, and QGS, with their own α and
CGS values, were also extracted. Their typical values are listed in Table 2 and Figure 7 in detail. Figure 6 shows the impedance characteristics of the GS samples depending on GO
concentrations with frequencies, which are correlated with the modeling values. This showed that
the measured magnitude (Figure 6a) and phase shift (Figure 6b) of the frequency response were well
matched with the calculated responses of the circuit model of Figure 5c using the best-fitted
parameters. This agreement indicates that the accuracy of the model was further verified. 3. Results and Discussion
3. Results and Discussion GS circuit component parameters: (a) constant phase elements (QGS) with the index of α, and
(b) LGS and RGS as a function of frequency. Figure 7. GS circuit component parameters: (a) constant phase elements (QGS) with the index of α,
and (b) LGS and RGS as a function of frequency. (b) LGS and RGS as a function of frequency. Table 2 Component values of various GSes
Figure 7. GS circuit component parameters: (a) constant phase elements (QGS) with the index of α, and
(b) LGS and RGS as a function of frequency. Figure 7. GS circuit component parameters: (a) constant phase elements (QGS) with the index of α,
and (b) LGS and RGS as a function of frequency. 3. Results and Discussion
3. Results and Discussion It could be
inferred that the well-suspended GO flakes (a few hundred μm in size) in the DI water contributed
to the charge conduction as the GO concentration increased in the GS regions with low GO
concentrations. On the other hand, in the GS regions with high GO concentrations, as the GO
concentration increased, the randomly-connected GO network and the complexity of the current-
flowing mechanism retarded the phase delay of the current to the external voltage. Thus, the
inductive behavior of the current could be observed at high frequencies, as shown in Figure 4b. Based
on this premise, several physical parameters, including the LGS, RGS, and QGS, with their own α and
CGS values, were also extracted. Their typical values are listed in Table 2 and Figure 7 in detail. Figure 6. GS impedance characteristics for various GSes and GO concentrations; (a) magnitude and
(b) phase of the measured and calculated impedances. Figure 6. GS impedance characteristics for various GSes and GO concentrations; (a) magnitude and
(b) phase of the measured and calculated impedances. Figure 6. GS impedance characteristics for various GSes and GO concentrations; (a) magnitude and
(b) phase of the measured and calculated impedances. Figure 6. GS impedance characteristics for various GSes and GO concentrations; (a) magnitude and
(b) phase of the measured and calculated impedances. Figure 6. GS impedance characteristics for various GSes and GO concentrations; (a) magnitude and
(b) phase of the measured and calculated impedances. Figure 6. GS impedance characteristics for various GSes and GO concentrations; (a) magnitude and
(b) phase of the measured and calculated impedances. Figure 7. GS circuit component parameters: (a) constant phase elements (QGS) with the index of α, and
(b) LGS and RGS as a function of frequency. Table 2. Component values of various GSes. Figure 7. GS circuit component parameters: (a) constant phase elements (QGS) with the index of α, and
(b) LGS and RGS as a function of frequency. Figure 7. GS circuit component parameters: (a) constant phase elements (QGS) with the index of α,
and (b) LGS and RGS as a function of frequency. onstant phase elements (QGS) with the index of α and Figure 7 GS circuit component parameters: (a) con uit o
o e t
a a
ete (b) LGS and RGS as a function of frequency. Table 2 Component values of various GSes
Figure 7. it Component
DI Water
GS4
GS8
G
Table 2. Component values of various GSes.
Table 2. Component values of various GSes. it Component
DI Water
GS4
GS8
G
Table 2. Component values of various GSes. Table 2. Component values of various GSes. LGS (μH)
n/a
5.05
4.56
4.64
RGS (kΩ)
102
2.05
0.53
0.36
QGS (mF·S(α−1))
n/a
0.22
1.56
2.91
α
n/a
0.41
0.23
0.17
CGS (pF)
28.8
30.4
28.9
30.1
nts the extracted circuit parameters of the GS samples, LGS, RGS, and
o
e t atio
i
the GS We
ote that the C
alue a e
30
F fo
Circuit Component
DI Water
GS4
GS8
GS18
LGS (μH)
n/a
5.05
4.56
4.64
RGS (kΩ)
102
2.05
0.53
0.36
QGS (mF·S(α−1))
n/a
0.22
1.56
2.91
α
n/a
0.41
0.23
0.17
CGS (pF)
28.8
30.4
28.9
30.1
Circuit Component
DI Water
GS4
GS8
GS18
LGS (µH)
n/a
5.05
4.56
4.64
RGS (kΩ)
102
2.05
0.53
0.36
QGS (mF·S(α−1))
n/a
0.22
1.56
2.91
α
n/a
0.41
0.23
0.17
CGS (pF)
28.8
30.4
28.9
30.1 7 of 8 Nanomaterials 2017, 7, 446 Figure 7 presents the extracted circuit parameters of the GS samples, LGS, RGS, and QGS, and α,
as a function of GO concentration in the GS. We note that the CGS values are ~30 pF for all the GSes
regardless of GO concentration, and we presumed that the CGS value could be related to the volume
ratio of the DI water in the GS. The results show that both RGS and LGS consistently decreased as the
GO increased in the GS, as described for Figures 2–4. Of note, the resistor, RGO, of the stacked 3D
GO [10] was almost 10 times higher than that of the GS in this work. The stacked-3D GOs were closer to
the capacitor than the resistor; i.e., α1 and α2 were 1.0 and 0.85, respectively. On the other hand, the GS
in this work was closer to resistive (the α of the QGS was less than 0.42 in Figure 7a), even though
a small amount of GO was dispersed in the DI water. We presumed that the hydroxyl/epoxide or
carbonyl/carboxyl groups of the GO in the GS could be laterally connected as the GO concentration
increased, and thereby contribute to the resistive path. Based on this fundamental understanding,
the adsorption and desorption of molecules to the GO surfaces and their influence on the impedance
of the system would be worthy of studying via a titration experiment in further work [26]. 4. Conclusions Oxidizing graphite and dispersing it in DI water, we formed a GS and investigated its optical and
electrical properties for a range of GO concentrations (0.25 to 7.00 g/L). The transmittance of the GS
become totally opaque in the visible range of 300 to 700 nm at GO concentrations exceeding 2 g/L. The electrical impedance of the GS was measured using the EIS method in the frequency range of 1 kHz
to 10 MHz, and presented via both Bode and Nyquist plots. The capacitance component approached a
certain constant value, but the resistance component continuously decreased as the GO concentration
increased. Based on the results, we developed the equivalent GS circuit model for the various GO
concentrations. The circuit model showed that the highly resistive GO became very conductive in
the DI water. The findings in this work should draw new attention toward GSes and their related
applications, including functional composite materials, catalysts, and filter membranes, by providing
an improved electrochemical active area and improving the charge-flowing transfer properties. Acknowledgments: This work was supported by the Center for Advanced Soft-Electronics funded by the
Ministry of Science, ICT and Future Planning as a Global Frontier Project (CASE-2011-0031638), and by the Pioneer
Research Center Program though the National Research Foundation of Korea, which is funded by the Ministry of
Science, ICT, and Future Planning (NRF-2013M3C1A3065040). Author Contributions: W.S.H. and M.S. conceived and designed the research. Y.Y. and S.K. performed the
experiments. J.J. performed the circuit model. I.G.L. interpreted the experimental results. B.J.C. contributed the
circuit model development. All coauthors analyzed the data and wrote the paper. Conflicts of Interest: The authors declare no conflict of interest. References 1. Neto, A.C.; Guinea, F.; Peres, N.M.; Novoselov, K.S.; Geim, A.K. The electronic properties of graphene. Rev. Mod. Phys. 2009, 81, 109. [CrossRef] 1. Neto, A.C.; Guinea, F.; Peres, N.M.; Novoselov, K.S.; Geim, A.K. The electronic properties of graphene. Rev. Mod. Phys. 2009, 81, 109. [CrossRef] . Geim, A.K.; Novoselov, K.S. The rise of graphene. Nat. Mater. 2007, 6, 183–191. [CrossRef] [PubMed] 2. Geim, A.K.; Novoselov, K.S. The rise of graphene. Nat. Mater. 2007, 6, 183 191. [CrossRef] [PubMed]
3. Lee, C.; Wei, X.; Kysar, J.W.; Hone, J. Measurement of the elastic properties and intrinsic strength of monolayer
graphene. Science 2008, 321, 385–388. [CrossRef] [PubMed] 3. Lee, C.; Wei, X.; Kysar, J.W.; Hone, J. Measurement of the elastic properties and intrinsic strength of monolayer
graphene. Science 2008, 321, 385–388. [CrossRef] [PubMed] 4. Bonaccorso, F.; Sun, Z.; Hasan, T.; Ferrari, A.C. Graphene photonics and optoelectronics. Nat. Photonics 2010,
4, 611–622. [CrossRef] 4. Bonaccorso, F.; Sun, Z.; Hasan, T.; Ferrari, A.C. Graphene photonics and optoelectronics. Nat. Photonics 2010,
4, 611–622. [CrossRef] 5. Park, S.; Ruoff, R.S. Chemical methods for the production of graphenes. Nat. Nanotechnol. 2009, 4, 217–224. [CrossRef] [PubMed] 5. Park, S.; Ruoff, R.S. Chemical methods for the production of graphenes. Nat. Nanotechnol. 2009, 4, 217–224. [CrossRef] [PubMed] 6. Zhu, Y.; Murali, S.; Cai, W.; Li, X.; Suk, J.W.; Potts, J.R.; Ruoff, R.S. Graphene and graphene oxide: Synthesis,
properties, and applications. Adv. Mater. 2010, 22, 3906–3924. [CrossRef] [PubMed] 6. Zhu, Y.; Murali, S.; Cai, W.; Li, X.; Suk, J.W.; Potts, J.R.; Ruoff, R.S. Graphene and graphene oxide: Synthesis,
properties, and applications. Adv. Mater. 2010, 22, 3906–3924. [CrossRef] [PubMed] 7. Stankovich, S.; Dikin, D.A.; Piner, R.D.; Kohlhaas, K.A.; Kleinhammes, A.; Jia, Y.; Ruoff, R.S. Synthesis of
graphene-based nanosheets via chemical reduction of exfoliated graphite oxide. Carbon 2007, 45, 1558–1565. [CrossRef] 7. Stankovich, S.; Dikin, D.A.; Piner, R.D.; Kohlhaas, K.A.; Kleinhammes, A.; Jia, Y.; Ruoff, R.S. Synthesis of
graphene-based nanosheets via chemical reduction of exfoliated graphite oxide. Carbon 2007, 45, 1558–1565. [CrossRef] 8 of 8 Nanomaterials 2017, 7, 446 8. Dreyer, D.R.; Park, S.; Bielawski, C.W.; Ruoff, R.S. The chemistry of graphene oxide. Chem. Soc. Rev. 2010, 39,
228–240. [CrossRef] [PubMed] 9. Mkhoyan, K.A.; Contryman, A.W.; Silcox, J.; Stewart, D.A.; Eda, G.; Mattevi, C.; Chhowalla, M. Atomic and
l
t
i
t
t
f
h
id
N
L tt 2009 9 1058 1063 [C
R f] [P bM d] 9. References Mkhoyan, K.A.; Contryman, A.W.; Silcox, J.; Stewart, D.A.; Eda, G.; Mattevi, C.; Chhowalla, M. Atomic and
electronic structure of graphene-oxide. Nano Lett. 2009, 9, 1058–1063. [CrossRef] [PubMed] khoyan, K.A.; Contryman, A.W.; Silcox, J.; Stewart, D.A.; Eda, G.; Mattevi, C.; Chhowalla, M. Atomic and
ctronic structure of graphene-oxide. Nano Lett. 2009, 9, 1058–1063. [CrossRef] [PubMed] 10. Han, K.I.; Kim, S.D.; Yang, W.S.; Kim, H.S.; Shin, M.; Kim, J.P.; Hwang, W.S. Material characteristics and
equivalent circuit models of stacked graphene oxide for capacitive humidity sensors. AIP Adv. 2016, 6,
035203. [CrossRef] 11. Kim, J.; Cote, L.J.; Kim, F.; Yuan, W.; Shull, K.R.; Huang, J. Graphene oxide sheets at interfaces. J. Am. Chem. Soc. 2010, 132, 8180–8186. [CrossRef] [PubMed] 12. Potts, J.R.; Dreyer, D.R.; Bielawski, C.W.; Ruoff, R.S. Graphene-based polymer nanocomposites. Polymer
2011, 52, 5–25. [CrossRef] 13. Stankovich, S.; Dikin, D.A.; Dommett, G.H.; Kohlhaas, K.M.; Zimney, E.J.; Stach, E.A.; Ruoff, R.S. Graphene-based composite materials. Nature 2006, 442, 282–287. [CrossRef] [PubMed] 14. Chen, H.; Müller, M.B.; Gilmore, K.J.; Wallace, G.G.; Li, D. Mechanically strong, electrically conductive,
and biocompatible graphene paper. Adv. Mater. 2008, 20, 3557–3561. [CrossRef] 15. Becerril, H.A.; Mao, J.; Liu, Z.; Stoltenberg, R.M.; Bao, Z.; Chen, Y. Evaluation of solution-processed reduced
graphene oxide films as transparent conductors. ACS Nano 2008, 2, 463–470. [CrossRef] [PubMed] 16. Zhu, Y.; Murali, S.; Stoller, M.D.; Ganesh, K.J.; Cai, W.; Ferreira, P.J.; Su, D. Carbon-based supercapacitors
produced by activation of graphene. Science 2011, 332, 1537–1541. [CrossRef] [PubMed] 17. Li, S.S.; Tu, K.H.; Lin, C.C.; Chen, C.W.; Chhowalla, M. Solution-processable graphene oxide as an efficient
hole transport layer in polymer solar cells. ACS Nano 2010, 4, 3169–3174. [CrossRef] [PubMed] 18. Hu, M.; Mi, B. Layer-by-layer assembly of graphene oxide membranes via electrostatic interaction. J. Membr. Sci. 2014, 469, 80–87. [CrossRef] 19. Liu, Y.; Dong, X.; Chen, P. Biological and chemical sensors based on graphene materials. Chem. Soc. Rev. 2012, 41, 2283–2307. [CrossRef] [PubMed] 20. Chandra, V.; Park, J.; Chun, Y.; Lee, J.W.; Hwang, I.C.; Kim, K.S. Water-dispersible magnetite-reduced
graphene oxide composites for arsenic removal. ACS Nano 2010, 4, 3979–3986. [CrossRef] [PubMed]
21. Hummers, W.S., Jr.; Offeman, R.E. Preparation of graphitic oxide. J. Am. Chem. Soc. 1958, 80, 1339. [CrossRef] 22. Dimiev, A.M.; Tour, J.M. Mechanism of graphene oxide formation. ACS Nano 2014, 8, 3060–3068. [CrossRef]
[PubMed] 23. Park, W.K.; Kim, H.; Kim, T.Y.; Kim, Y.; Yoo, S.; Kim, S.; Yoon, D.H.; Yang, W.S. References Facile synthesis of graphene
oxide in a Couette–Taylor flow reactor. Carbon 2015, 83, 217–223. [CrossRef] 24. Shih, C.J.; Lin, S.; Sharma, R.; Strano, M.S.; Blankschtein, D. Understanding the pH-dependent behavior of
graphene oxide aqueous solutions: A comparative experimental and molecular dynamics simulation study. Langmuir 2011, 28, 235–241. [CrossRef] [PubMed] 25. Sluyters-Rehbach, M. Impedances of electrochemical systems:
Terminology,
nomenclature and
representation-Part I: Cells with metal electrodes and liquid solutions (IUPAC Recommendations 1994). Pure Appl. Chem. 1994, 66, 1831–1891. [CrossRef] 26. Graf, K.; Kappl, M. Physics and Chemistry of Interfaces, 2nd ed.; John Wiley & Sons: Hoboken, NJ, USA, 2006;
ISBN 978-3-527-60640-5. © 2017 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access
article distributed under the terms and conditions of the Creative Commons Attribution
(CC BY) license (http://creativecommons.org/licenses/by/4.0/). © 2017 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access
article distributed under the terms and conditions of the Creative Commons Attribution
(CC BY) license (http://creativecommons.org/licenses/by/4.0/).
|
https://openalex.org/W4220855551
|
https://roj.igb.ru/jour/article/download/885/433
|
Russian
| null |
The correlation of acoustic and morphometric parameters of the optic nerve and anthropometric data in healthy young people
|
Rossijskij oftalʹmologičeskij žurnal
| 2,022
|
cc-by
| 5,953
|
КЛИНИЧЕСКИЕ ИССЛЕДОВАНИЯ/CLINICAL STUDIES КЛИНИЧЕСКИЕ ИССЛЕДОВАНИЯ/CLINICAL STUDIES 39
Нероев В.В., Киселева Т.Н., Баева А.В., Елисеева Е.К., Журавлева А.Н., Ушаков А.И., Луговкина К.В., Судовская Т.В., Мышко И.В., 2022 Âçàèìîñâÿçü àêóñòè÷åñêèõ è ìîðôîìåòðè÷åñêèõ
õàðàêòåðèñòèê çðèòåëüíûõ íåðâîâ
ñ àíòðîïîìåòðè÷åñêèìè ïàðàìåòðàìè ó ìîëîäûõ
ëèö â íîðìå Â.Â. Íåðîåâ1, 2, Ò.Í. Êèñåëåâà1, À.Â. Áàåâà2, Å.Ê. Åëèñååâà1 , À.Í. Æóðàâëåâà1, À.È. Óøàêîâ1, Ê.Â. Ëóãîâêèíà1,
Ò.Â. Ñóäîâñêàÿ1, È.Â. Ìûøêî1 1 ФГБУ «НМИЦ глазных болезней им. Гельмгольца» Минздрава России, ул. Садовая-Черногрязская, д. 14/19, Москва,
105062, Россия ,
2 ФГБОУ ВО «Московский государственный медико-стоматологический университет им. А.И. Евдокимова» Минздрава
России, ул. Делегатская, д. 20, стр. 1, Москва, 127473, Россия Цель работы — изучение взаимосвязи между акустическими, морфометрическими параметрами зрительного нерва (ЗН)
и антропометрическими данными лиц молодого возраста в норме. Материал и методы. Обследовано 24 здоровых добровольца
(48 глаз), из них 9 мужчин и 15 женщин, в возрасте 25,0 ± 1,9 года. Антропометрия включала определение роста, веса и индекса
массы тела (ИМТ). Ультразвуковое исследование ретробульбарной части ЗН включало измерение толщины ЗН с оболочками
и без оболочек, эходенситометрию ЗН и биометрию глаза. С помощью оптической когерентной томографии измеряли морфомет-
рические параметры интраокулярной части ЗН с регистрацией показателей Bruch’s membrane opening и индекса овальности. Результаты. Средние показатели толщины ЗН с оболочками составили 4,6 ± 0,3 мм, без оболочек — 2,6 ± 0,2 мм. У мужчин
толщина ЗН с оболочками была выше, чем у женщин (р = 0,001). Акустическая плотность паренхимы орбитальной части ЗН
составила 101,2 ± 11,4 у. е., у женщин средние значения этого параметра ЗН были достоверно выше, чем у мужчин. Корреляцион-
ный анализ позволил выявить статистически достоверную корреляцию между ростом и толщиной ЗН с оболочками (r = 0,480). Наиболее сильная достоверная корреляционная связь толщины ЗН установлена с массой тела (r = 0,712) и ИМТ (r = 0,509)
(р < 0,05). Статистически достоверной корреляции между морфометрическими параметрами ДЗН, данными антропометрии
и акустическими показателями ЗН не обнаружено. Заключение. Исследование акустических и морфометрических характеристик
интраокулярного и интраорбитального отделов ЗН у молодых лиц в норме с учетом антропометрических данных и гендерных
особенностей позволит разработать более четкие диагностические критерии оценки состояния ЗН. Ключевые слова: ультразвуковое исследование; толщина оболочек зрительного нерва; эходенситометрия; оптическая
когерентная томография; морфометрические параметры диска зрительного нерва; антропометрические параметры
Конфликт интересов: отсутствует. Прозрачность финансовой деятельности: никто из авторов не имеет финансовой заинтересованности в представленных
материалах или методах. Прозрачность финансовой деятельности: никто из авторов не имеет финансовой заинтересованности в представленных
материалах или методах. Для цитирования: Нероев В.В., Киселева Т.Н., Баева А.В., Елисеева Е.К., Журавлева А.Н., Ушаков А.И., Луговкина К.В.,
Судовская Т.В., Мышко И.В. Взаимосвязь акустических и морфометрических характеристик зрительного нерва
с антропометрическими параметрами у молодых лиц в норме. Российский офтальмологический журнал. 2022; 15 (1) : 39-45. Для цитирования: Нероев В.В., Киселева Т.Н., Баева А.В., Елисеева Е.К., Журавлева А.Н., Ушаков А.И., Луговкина К.В.,
Судовская Т.В., Мышко И.В. Взаимосвязь акустических и морфометрических характеристик зрительного нерва
с антропометрическими параметрами у молодых лиц в норме. Российский офтальмологический журнал. 2022; 15 (1) : 39-45.
https://doi.org/10.21516/2072-0076-2022-15-1-39-45 The correlation of acoustic and morphometric
parameters of the optic nerve and anthropometric
data in healthy young people Vladimir V. Neroev1, 2, Tatiana N. Kiseleva1, Alena V. Baeva2, Elena K. Eliseeva1 , Anastasiya N. Zhuravleva1,
Alexandr I. Ushakov1, Kseniya V. Lugovkina1, Tatiana V. Sudovskaya1, Ilya V. Myshko1
1 Helmholtz National Medical Research Center of Eye Diseases, 14/19, Sadovaya-Chernogryazskaya St., Moscow, 105062,
Russia 2 Moscow Evdokimov State Medical Stomatological University of Medicine and Dentistry, 20/1, Delegatskaya St., Moscow,
127473, Russia
tkisseleva@yandex.ru Purpose: to study the correlation between the acoustic and morphometric parameters of the optic nerve (ON) and anthropometric data
in healthy young people. Material and methods. The research involved 24 healthy volunteers (48 eyes), including 9 men and 15 women,
the average age was 25.0 ± 1.9 years. Anthropometry included measurements of height, weight, and body mass index (BMI). Ultrasound
examination of the retrobulbar part of the ON included measurement of the thickness of the ON with and without sheath, echodensitometry
of the ON, and ocular biometry. Using optical coherence tomography, we measured the intraocular part of the ON, including the registration
of Bruch's membrane opening and ovality index. Results. The average ON thickness with the sheath was 4.6 ± 0.3 mm, and without sheath,
2.6 ± 0.2 mm. In men, ON with the sheath was thicker than in women (p = 0.001). The acoustic density of the parenchyma of the orbital part
of ON was 101.2 ± 11.4 r.u. in women, the mean acoustic density of the ON was significantly higher than that in men. Correlation analysis
revealed a statistically significant correlation between the height and the thickness of the ON with the sheath (r = 0.480). The strongest reliable
correlation was established between the index of the ON thickness and body weight (r = 0.712) and BMI (r = 0.509) (p < 0.05). No statisti-
cally significant correlation was found between the morphometric parameters of the optic disc, anthropometry data and acoustic parameters
of the ON. Conclusion. Studying the acoustic and morphometric characteristics of the intraocular and intraorbital parts of the ON in healthy
young people, which included anthropometric data and gender characteristics, will contribute to the development more accurate diagnostic
criteria of ON state evaluations. Keywords: ultrasound examination; optic nerve sheath thickness; echodensitometry; optical coherence tomography; morphometric
parameters of optic disk; anthropometric parameters Conflict of interests: there is no conflict of interests. Financial disclosure: The authors have no financial or property interest in any material or method mentioned. The correlation of acoustic and morphometric parameters of the optic nerve
and anthropometric data in healthy young people
40 Russian ophthalmological journal. 2022; 15 (1): 39-45 Âçàèìîñâÿçü àêóñòè÷åñêèõ è ìîðôîìåòðè÷åñêèõ
õàðàêòåðèñòèê çðèòåëüíûõ íåðâîâ
ñ àíòðîïîìåòðè÷åñêèìè ïàðàìåòðàìè ó ìîëîäûõ
ëèö â íîðìå https://doi.org/10.21516/2072-0076-2022-15-1-39-45 Для цитирования: Нероев В.В., Киселева Т.Н., Баева А.В., Елисеева Е.К., Журавлева А.Н., Ушаков А.И., Луговкина К.В.,
Судовская Т.В., Мышко И.В. Взаимосвязь акустических и морфометрических характеристик зрительного нерва
с антропометрическими параметрами у молодых лиц в норме. Российский офтальмологический журнал. 2022; 15 (1) : 39-45. https://doi.org/10.21516/2072-0076-2022-15-1-39-45 Vladimir V. Neroev1, 2, Tatiana N. Kiseleva1, Alena V. Baeva2, Elena K. Eliseeva1 , Anastasiya N. Zhuravleva1,
Alexandr I. Ushakov1, Kseniya V. Lugovkina1, Tatiana V. Sudovskaya1, Ilya V. Myshko1
1 Helmholtz National Medical Research Center of Eye Diseases, 14/19, Sadovaya-Chernogryazskaya St., Moscow, 105062,
Russia
2 M
E d ki
St t
M di
l St
t l
i
l U i
it
f M di i
d D
ti t
20/1 D l
t k
St
M МАТЕРИАЛ И МЕТОДЫ фического взаимодействия с окружающими орбитальными
структурами и тканями. Развитие современных компьютер-
ных технологий и применение ультразвуковых диагности-
ческих сканеров последнего поколения дают возможность
наиболее точно оценить состояние ретробульбарного участка
ЗН, его структурные особенности и кровоснабжение [6, 8]. Однако с помощью ультразвукового сканирования удается
визуализировать лишь две трети интраорбитального отдела
нерва, что является ограничением этого метода. Обследовано 24 здоровых добровольца (48 глаз), из них
9 (34,5%) мужчин и 15 (65,5%) женщин в возрасте 23–30 лет
(средний возраст — 25,0 ± 1,9 года). Критериями включения
в исследование являлись: подписанное информированное
согласие на проведение исследований, отсутствие в анамнезе
воспалительных, дегенеративно-дистрофических заболева-
ний сетчатки и ЗН, травмы глаз и аномалий рефракции за
исключением миопии слабой степени. исключением миопии слабой степени. Антропометрические измерения (рост, масса тела)
проводились с помощью стандартных приборов (напольный
ростометр и медицинские весы). На основе выполненных
измерений рассчитывали индекс массы тела (ИМТ) по
формуле: ИМТ= М/Р2, где М — масса тела, кг; Р — рост, м. Помимо стандартного офтальмологического обсле-
дования, включавшего визометрию, биомикроскопию, то-
нометрию, офтальмоскопию, компьютерную периметрию,
проведены УЗИ ретробульбарной части ЗН, эходенсито-
метрия паренхимы и оболочек ЗН, биометрия глаза и ОКТ
интраокулярного отдела ЗН. Для визуализации интраокулярного отдела ЗН в оф-
тальмологической практике используется оптическая
когерентная томография (ОКТ), которая относится к вы-
сокоинформативным методам определения морфометри-
ческих параметров диска зрительного нерва (ДЗН) [9, 10]. По мнению большинства авторов, ОКТ представляет собой
наиболее адекватный бесконтактный метод исследования,
позволяющий получить максимально полную информацию
о состоянии ДЗН. Томографические характеристики, вклю-
чающие площадь и объем нейроретинального пояска, размер
диска, глубину, площадь и объем экскавации, а также анализ
толщины перипапиллярного слоя нервных волокон сетчатки
(СНВС), являются важными для ранней диагностики опти-
ческих нейропатий различного генеза. Антропометрические измерения (рост, масса тела)
проводились с помощью стандартных приборов (напольный
ростометр и медицинские весы). На основе выполненных
измерений рассчитывали индекс массы тела (ИМТ) по
формуле: ИМТ= М/Р2, где М — масса тела, кг; Р — рост, м. Помимо стандартного офтальмологического обсле-
дования, включавшего визометрию, биомикроскопию, то-
нометрию, офтальмоскопию, компьютерную периметрию,
проведены УЗИ ретробульбарной части ЗН, эходенсито-
метрия паренхимы и оболочек ЗН, биометрия глаза и ОКТ
интраокулярного отдела ЗН. Эхография глаза и орбиты в В-режиме выполнялась
на приборе VOLUSON E8 с использованием высокочас-
тотного линейного датчика с частотой 10–18,5 MГц. Для
визуализации ЗН в орбите осуществлялось аксиальное
сканирование по горизонтальной и вертикальной плоскос-
тям в 3 мм от заднего полюса глаза (рис. 1, А). Определе-
ние длины передне-задней оси (ПЗО) глаза проводилось
с использованием оптической биометрии глаза (AL-scan,
Nidek). The correlation of acoustic and morphometric
parameters of the optic nerve and anthropometric
data in healthy young people For citation: Neroev V.V., Kiseleva T.N., Baeva A.V., Eliseeva E.K., Zhuravleva A.N., Ushakov A.I., Lugovkina K.V., Sudovskaya T.V.,
Myshko I.V. The correlation of acoustic and morphometric parameters of the optic nerve and anthropometric data in healthy young
people. Russian ophthalmological journal. 2022; 15 (1): 39-45 (In Russian). https://doi.org/10.21516/2072-0076-2022-15-1-39-45 Одной из актуальных проблем офтальмологии является
ранняя диагностика заболеваний зрительного нерва (ЗН),
включая воспалительные, сосудистые и дегенеративные на-
рушения [1]. С учетом высокой частоты выявления органи-
ческих поражений ЗН среди лиц молодого работоспособного
возраста (от 2 до 10 на 100 тыс. населения) поиск наиболее
информативных диагностических методик для оценки сос-
тояния ЗН остается социально значимым вопросом [2]. лизации становятся орбитальный, интраканаликулярный
и интракраниальный отделы ЗН [5]. Однако, несмотря на
очевидные преимущества и диагностическую информатив-
ность КТ и МРТ, к проведению этих исследований имеется
ряд противопоказаний. При КТ они связаны в большинстве
случаев с лучевой нагрузкой, при МРТ существует достаточно
широкий диапазон как абсолютных, так и относительных
противопоказаний (металлические имплантаты, электро-
стимуляторы, искусственные клапаны сердца, инсулиновые
помпы, клаустрофобия и т. д.). Известно, что ЗН состоит из 4 отделов: интраокуляр-
ный, интраорбитальный, интраканаликулярный и интра-
краниальный. Наиболее часто встречаются поражения
интраокулярного и интраорбитального отделов нерва, по-
этому разработка алгоритма выявления патологии данной
локализации приобретает большую значимость в клиничес-
кой практике. Ультразвуковые диагностические методы не имеют
противопоказаний к применению. Основными достоинства-
ми эхографии являются высокая информативность, доступ-
ность, экономичность метода и возможность его многократ-
ного применения у пациентов любого возраста без опасения
возникновения каких-либо нежелательных последствий. Кроме того, отсутствует необходимость предварительной
подготовки больных к проведению исследования, мобиль-
ность аппаратуры позволяет использовать ультразвук в от-
делениях неотложной помощи и интенсивной терапии [6, 7]. Для объективной оценки состояния ретробульбарной
части ЗН в офтальмологии применяются ультразвуковое
исследование (УЗИ), магнитно-резонансная томография
(МРТ) и компьютерная томография (КТ) [3, 4]. КТ наряду
с МРТ являются золотым стандартом диагностики внутри-
черепной патологии и поражений ЗН. Благодаря исполь-
зованию этих методов, полностью доступными для визуа- В настоящее время эхография широко используется для
определения анатомических характеристик ЗН и его топогра- Russian ophthalmological journal. 2022; 15 (1): 39-45 МАТЕРИАЛ И МЕТОДЫ С помощью эходенситометрии оценивали акусти-
ческую плотность паренхимы и оболочек ЗН (рис. 1, Б)
с построением двухмерных гистограмм и расчетом среднего
значения (А) в условных единицах (у. е.) цифрового анализа
изображения [11]. Другим значимым вопросом остается изучение со-
поставимости результатов измерения морфометрических
параметров ДЗН и анатомо-структурных характеристик
ретробульбарной части ЗН и зависимость этих признаков
от ряда факторов (антропометрические данные, раса, пол и
другие показатели). До сих пор продолжается разработка протоколов ис-
следования, предусматривающих количественную оценку
параметров ДЗН и ретробульбарного отдела ЗН у пациентов
с патологией зрительных путей. Поэтому существует не-
обходимость сравнения полученных результатов измере-
ния анатомических и морфометрических параметров ЗН
с данными нормативной базы с учетом влияния различных
анатомических и физиологических факторов на указанные
характеристики ЗН. Морфометрическое измерение интраокулярной части
ЗН выполнялось с использованием технологии ОКТ (OCT
Spectralis SD-OCT, Heidelberg Engineering) c программным
модулем GMPE (Glaucoma Module Premium Edition) по стан-
дартному протоколу Optic Disc / Optic Nerve Head. Для оцен-
ки показателей вертикального (vertical disc diameter, VDD)
и горизонтального (horizontal disc diameter, HDD) диаметров
ДЗН определялись границы анатомической области диска ЦЕЛЬ исследования — изучить взаимосвязь между аку-
стическими, морфометрическими параметрами ЗН и антро-
пометрическими данными у лиц молодого возраста в норме. Рис. 1. Эхограмма глаза и орбиты с изображением ЗН (В-режим, аксиальное сканирование). А — измерение диаметра поперечного
сечения (толщины) ЗН с оболочками и без оболочек. Б — измерение акустической плотности ЗН в у. е. цифрового анализа
Fig. 1. Echogram of the eye and orbit with the image of the optic nerve (B-mode, axial scan). A — measurement of the cross-sectional diameter
(thickness) of the optic nerve with and without sheaths. Б — measurement of the acoustic density of the optic nerve relative units of digital analysis
А
Б Б А Б Рис. 1. Эхограмма глаза и орбиты с изображением ЗН (В-режим, аксиальное сканирование). А — измерение диаметра поперечного
сечения (толщины) ЗН с оболочками и без оболочек. Б — измерение акустической плотности ЗН в у. е. цифрового анализа
Fig. 1. Echogram of the eye and orbit with the image of the optic nerve (B-mode, axial scan). A — measurement of the cross-sectional diameter
(thickness) of the optic nerve with and without sheaths. Б — measurement of the acoustic density of the optic nerve relative units of digital analysis Взаимосвязь акустических и морфометрических характеристик зрительных
нервов с антропометрическими параметрами у молодых лиц в норме
41
Ðîññèéñêèé îôòàëüìîëîãè÷åñêèé æóðíàë. МАТЕРИАЛ И МЕТОДЫ 2022; 15 (1): 39-45 Взаимосвязь акустических и морфометрических характеристик зрительных
нервов с антропометрическими параметрами у молодых лиц в норме
41
Ðîññèéñêèé îôòàëüìîëîãè÷åñêèé æóðíàë. 2022; 15 (1): 39-45 41 нормального распределения, для обработки данных исполь-
зовались непараметрические критерии, в статистической
обработке учитывали медиану (интерквартильный размах). Корреляционный анализ показателей осуществлялся по
методу Спирмена. по открытию мембраны Бруха (Bruch’s membrane opening,
BMO). В случае если границы открытия, определенные
автоматически, не соответствовали фактическому положе-
нию BMO, их измерение корректировалось мануальным
способом. Для изучения параметров овальности BMO изоб-
ражения конфокального лазерного офтальмоскопа (cSLO)
экспортировали в базу данных программного обеспечения
ImageJ (National Institute of Health, Bethesda, MD, США),
далее выполнялось измерение минимального и максималь-
ного диаметров полученного эллипсоида в мануальном
режиме (рис. 2). Соотношение значений диаметров (мень-
ший к большему) данного эллипсоида было определено как
индекс овальности (ИО). Russian ophthalmological journal. 2022; 15 (1): 39-45 The correlation of acoustic and morphometric parameters of the optic nerve
and anthropometric data in healthy young people
42 РЕЗУЛЬТАТЫ 2022; 15 (1): 39-45 наиболее сильная достоверная корреляционная связь
толщины ЗН с массой тела (r = 0,712) и ИМТ (r = 0,509;
р < 0,05) (рис. 3). чин толщина ЗН с оболочками оказалась выше, чем у жен-
щин (р = 0,001). чин толщина ЗН с оболочками оказалась выше, чем у жен-
щин (р = 0,001). При проведении эходенситометрии наименьшие зна-
чения акустической плотности регистрировали в области
паренхимы орбитальной части ЗН 101,2 ± 11,4 у. е. (от 95,2
до 108 у. е.). Сравнительная оценка акустической плот-
ности оболочек ЗН показала достоверное увеличение этих
парамет ров с медиальной стороны — 155,6 ± 15,6 у. е. [146,2;
169] по сравнению с латеральной — 139,6 ± 18,5 у. е. [125,2;
156,7] (р < 0,05). Кроме того, у женщин средние значения
акустической плотности ЗН были достоверно выше, чем
у мужчин (табл. 2). Третий этап исследования включал статистический
анализ данных морфометрии ДЗН по результатам ОКТ, со-
гласно которым средний вертикальный диаметр ЗН (VDD)
составил 1667,5 мкм (1508,0–1799,2), горизонтальный
диаметр ЗН (HDD) — 1559 мкм (1475,0–1701,5), медиана
площади BMO (BMO area) — 1,98 мм2 (1,7–2,2), медиана
наименьшей ширины нейроретинального пояска (Minimal
rim width, MRW) — 332 мкм (291,2–386,7), медиана ИО
BMO — 0,94 (0,87–0,98) (табл. 2). Следует отметить, что, в от-
личие от акустических параметров ЗН, достоверных различий
морфометрических показателей ДЗН мужчин и женщин не
выявлено (р > 0,05). Корреляционный анализ позволил выявить статисти-
чески достоверную связь (р < 0,05) между ростом и толщи-
ной ЗН с оболочками (r = 0,480). Кроме того, установлена Таблица 2. Сравнительная оценка акустических и морфометрических характеристик у женщин и мужчин
Table 2. Comparative assessment of acoustic and morphometric characteristics in men and women
Параметры
Parameters
Мужчины
Men
Женщины
Women
Достоверность различий
p — value
Толщина ЗН, мм
Thickness ОN, mm
Без оболочек
Without sheaths
2,6 ± 0,2 (2,5; 2,9)
2,6 ± 0,1 (2,5; 2,7)
0,729
С оболочками
With sheaths
4,8 ± 0,3 (4,5; 5,0)
4,4 ± 0,2 (4,3; 4,7)
0,001
Акустическая плотность, у. е. Acoustic density, r. u. РЕЗУЛЬТАТЫ На первом этапе нашего исследования проводи-
лась оценка антропометрических данных. Средний рост
и вес испытуемых составили 169,5 ± 1,8 см (165–177 см)
и 64,5 ± 13,0 кг (56,7–73,5 кг) соответственно. Второй этап исследования включал анализ резуль-
татов определения аксиальной длины глаза и параметров
ЗН, полученных при помощи эхографии. ПЗО составила
23,66 ± 0,70 мм, толщина ЗН без оболочек — 2,6 ± 0,2 мм,
толщина ЗН с оболочками — 4,6 ± 0,3 мм (табл. 1). У муж- Статистический анализ результатов выполняли на
персональном компьютере с использованием программы
SPSS 19.0 (IBM, Нью-Йорк, США). Все параметры не имели Рис. 2. ОКТ. Изображение cSLO и радиальный скан ДЗН по заданным точкам BMO
Fig. 2. OCT. cSLO image and radial scan of the ONH at the given BMO points Рис. 2. ОКТ. Изображение cSLO и радиальный скан ДЗН по заданным точкам BMO
Fig. 2. OCT. cSLO image and radial scan of the ONH at the given BMO points i
h h l
l
i
l j
l
( )
42
Таблица 1. Средние показатели акустических и морфометрических характеристик ЗН
Table 1. Mean indices of acoustic and morphometric characteristics of optic nerve (ON)
Параметры
Parameters
Медиана (интерквартильный размах)
Median (quartile range)
Минимальное значение
Minimum
Максимальное значение
Maximum
Толщина ЗН, мм
ОN thickness, mm
Без оболочек
Without sheaths
2,6 ± 0,2 (2,5; 2,8)
2,2
3,2
С оболочками
With sheaths
4,6 ± 0,3 (4,4; 4,8)
3,9
5,5
Акустическая плотность, у. е. Acoustic density, r. u. Паренхима
Parenchyma
101,2 ± 11,4
(95,2; 108,0)
77
124
Оболочка c медиальной стороны
Medial sheath side
155,6 ± 15,6
(146,2; 169,0)
118
184
Оболочка с латеральной стороны
Lateral sheath side
139,6 ± 18,5
(125,2; 156,7)
103
174
Морфометрические параметры
Morphometric parameters
Горизонтальный размер ДЗН, мкм
Horizontal OND size, μm
1597,6 ± 188,6
(1475; 1701,5)
1175
2094
Вертикальный размер ДЗН, мкм
Vertical OND size, μm
1692,7 ± 262,5
(1508; 1799)
1213
2424
Площадь ДЗН, мм2
OND square, mm2
2,0 ± 0,5
(1,7; 2,2)
1,19
3,6
Толщина НРП, мкм
NRR thickness, μm
340,0 ± 59,2
(291,2; 386,7)
249
509
ИО
0,90 ± 0,05
(0,87; 0,98)
0,78
0,99
Примечание. НРП — нейроретинальный поясок; ИО — индекс овальности ДЗН. Note. NRR — neuroretinal rim; IO — ovality index of the optic nerve disc. Таблица 1. Средние показатели акустических и морфометрических характеристик ЗН
Table 1. Mean indices of acoustic and morphometric characteristics of optic nerve (ON) Russian ophthalmological journal. РЕЗУЛЬТАТЫ Паренхима
Parenchyma
97,4 ± 12,6
(98,5; 103,7)
103,4 ± 10,2
(98,5; 110,5)
0,059
Оболочка c медиальной стороны
Medial sheath side
148,5 ± 14,3
(134,7; 160,2)
159,9 ± 15,0
(148,5; 172)
0,013
Оболочка с латеральной стороны
Lateral sheath side
133,1 ± 17,4
(120,7; 144,5)
143,6 ± 18,2
(127,5; 158,2)
0,067
Морфометрические параметры
Morphometric parameters
Горизонтальный размер ДЗН, мкм
Horizontal OND size, μm
1566,7 ± 199,4
(1409,7; 1691,2)
1616,0 ± 182,7
(1483,0; 1738,5)
0,544
Вертикальный размер ДЗН, мкм
Vertical OND size, μm
1613,1 ± 219,9
(1466,0; 1779,7)
1740,4 ± 277,5
(1517; 1846)
0,312
Площадь ДЗН, мм2
OND square, mm2
1,9 ± 0,4
(1,6; 2,2)
2,1 ± 0,5
(1,8; 2,2)
0,686
Толщина НРП, мкм
NRR thickness, μm
350,5 ± 62,9
(301,2; 394,0)
334,4 ± 57,0
(288,5; 369,2)
0,338
ИО
IO
0,95 ± 0,40
(0,92; 0,99)
0,91 ± 0,06
(0,86; 0,97)
0,027
Примечание. НРП — нейроретинальный поясок; ИО — индекс овальности ДЗН, р — достоверность различий между показателями
мужчин и женщин. Note. NRR — neuroretinal rim; IO — ovality index of the optic nerve disc, р — significant differences between indices of men and women. Таблица 2. Сравнительная оценка акустических и морфометрических характеристик у женщин и мужчин
Table 2. Comparative assessment of acoustic and morphometric characteristics in men and women Примечание. НРП — нейроретинальный поясок; ИО — индекс овальности ДЗН, р — достоверность различий между показателями
мужчин и женщин. Note NRR
neuroretinal rim; IO
ovality index of the optic nerve disc р
significant differences between indices of men and women мужчин и женщин. Note. NRR — neuroretinal rim; IO — ovality index of the optic nerve disc, р — significant differences between indices of men and women. Взаимосвязь акустических и морфометрических характеристик зрительных
нервов с антропометрическими параметрами у молодых лиц в норме
43
Ðîññèéñêèé îôòàëüìîëîãè÷åñêèé æóðíàë. 2022; 15 (1): 39-45
Рис. 3. Корреляционная взаимосвязь между антропометрическими данными и толщиной ЗН с оболочками
Fig. 3. Correlation relationship between anthropometric data (weight, height, body mass index) and thickness of ОN with sheath Рис. 3. Корреляционная взаимосвязь между антропометрическими данными и толщиной ЗН с оболочками
Fig. 3. Correlation relationship between anthropometric data (weight, height, body mass index) and thickness of ОN with sheath . 3. Корреляционная взаимосвязь между антропометрическими данными и толщиной ЗН с оболочками
3. ЗАКЛЮЧЕНИЕ Проведено комплексное инструментальное (акусти-
ческое, морфометрическое) исследование интраокулярного
и интраорбитального отделов ЗН у молодых здоровых лиц
с учетом антропометрических параметров и гендерных
особенностей с целью повышения точности обследования
и разработки более четких диагностических критериев оцен-
ки состояния ЗН. Выявлена достоверная взаимосвязь между
акустическими характеристиками толщины ретробульбарной
части ЗН и антропометрическими данными у молодых людей
одного возрастного диапазона при отсутствии взаимосвязи
с морфометрическими параметрами интраокулярной части
ЗН. Полученные данные свидетельствуют о необходимости
дальнейшего поиска новых стратегий персонифицирован-
ного подхода к диагностике патологии ЗН. Литература/References Литература/References р
ур /
f
1. Елисеева Е.К., Нероев В.В., Зуева М.В., Цапенко И.В., Захарова М.Н. Оптический неврит на фоне рассеянного склероза (обзор литературы
и результаты собственного исследования). Точка зрения. Восток — За-
пад. 2018; 2: 112–5. [Eliseeva E.K., Neroev V.V., Zueva M.V., Tsapenko I.V.,
Zakharova M.N. Optic neuritis with multiple sclerosis (review of literature and
own data). Point of View. East — West. 2018; 2: 112–5 (in Russian)]. https://
doi.org/10.25276/2410- 1257-2018-2-112-115 р
ур /
f
1. Елисеева Е.К., Нероев В.В., Зуева М.В., Цапенко И.В., Захарова М.Н. Оптический неврит на фоне рассеянного склероза (обзор литературы
и результаты собственного исследования). Точка зрения. Восток — За-
пад. 2018; 2: 112–5. [Eliseeva E.K., Neroev V.V., Zueva M.V., Tsapenko I.V.,
Zakharova M.N. Optic neuritis with multiple sclerosis (review of literature and
own data). Point of View. East — West. 2018; 2: 112–5 (in Russian)]. https://
doi.org/10.25276/2410- 1257-2018-2-112-115 2. Katz D.M., Trobe J.D. Is there treatment for nonarteritic anterior ischemic
optic neuropathy. Curr. Opin. Ophthalmol. 2015; 26 (6): 458–63. https://doi. org/10.1097/ICU.0000000000000199 Многие публикации посвящены изучению морфо-
метрических параметров ДЗН и их взаимосвязи с другими
биометрическими показателями. Выявлена зависимость
между размерами ДЗН, толщиной СНВС и аксиальной
длиной глаза у взрослых. Достоверную положительную кор-
реляцию между ПЗО и размерами ДЗН обнаружили у здоро-
вых лиц E. Chihara и K. Chihara [20]. C. Oliveira и соавт. [21]
показали, что у взрослых в глазах с увеличенной аксиальной
длиной больше размеры ДЗН. 3. B uerle J., Schuchardt F., Schroeder L., et al. Reproducibility and accuracy
of optic nerve sheath diameter assessment using ultrasound compared to
magnetic resonance imaging. BMC neurology. 2013; 13:1–6. https://doi. org/10.1186/1471-2377-13-187 4. Lagr ze W.A., Lazzaro A., Weigel M., et al. Morphometry of the retrobulbar
human optic nerve: comparison between conventional sonography and ultrafast
magnetic resonance sequences. Invest. Ophthalmol. Vis. Sci. 2007. 48 (5):
1913–7. https://doi.org/10.1167/iovs.06-1075 д
р
р
Д
В исследовании H. Nakanishi и соавт. The correlation of acoustic and morphometric parameters of the optic nerve
and anthropometric data in healthy young people
44 РЕЗУЛЬТАТЫ Correlation relationship between anthropometric data (weight, height, body mass index) and thickness of О ду
р
р
д
щ
p between anthropometric data (weight, height, body mass index) and thickness of ОN with sheath Взаимосвязь акустических и морфометрических характеристик зрительных
нервов с антропометрическими параметрами у молодых лиц в норме
43
Ðîññèéñêèé îôòàëüìîëîãè÷åñêèé æóðíàë. 2022; 15 (1): 39-45 Статистически достоверной корреляции между морфо-
метрическими параметрами ДЗН, данными антропометрии
и толщиной ретробульбарного отдела ЗН не обнаружено
(значения коэффициента корреляции — менее 0,4). установлена взаимосвязь между величиной сферического
эквивалента аномалии рефракции с морфометрическими
показателями [25–27]. Кроме того, по нашим данным, достоверная взаимо-
связь между морфометрическими параметрами ДЗН, ростом
и массой тела отсутствует, что совпадает с результатами
L.Wang и соавт. [16]. ОБСУЖДЕНИЕ Анализ данных литературы свидетельствует о высокой
информативности УЗИ в оценке состояния ретробульбар-
ного отдела ЗН [6, 7, 12]. В то же время остаются противоре-
чивыми сведения о средних значениях толщины ЗН в норме
у взрослых. Так, по данным K. Oluseyi, I. Ukamaka [11],
средний диаметр ретробульбарного отдела ЗН с оболочками
составляет 4,20 ± 0,38 мм. Похожие данные получили D. Kim
и соавт. [13] при обследовании 585 здоровых добровольцев
в Южной Корее (4,11 мм, 4,09–4,14, 95%-ный доверительный
интервал). H. Chen и соавт. [14] при проведении УЗИ орбиты
у 519 здоровых добровольцев определили медианы показа-
телей ЗН без оболочек и с оболочками — 3,2 мм (2,9–3,4) и
5,1 мм (4,7–5,4) соответственно. Ряд авторов не обнаружили
взаимосвязи между антропометрическими параметрами
(полом, возрастом, ростом и весом) и эхографическими
характеристиками ЗН [13–15]. Однако L. Wang и соавт. [16]
после комплексного ультразвукового исследования 230 здо-
ровых лиц установили достоверную взаимосвязь толщины
ЗН с ИМТ, что соответствует нашим данным. Большин-
ство авторов не выявили корреляцию между толщиной ЗН
с оболочками и возрастом, что может свидетельствовать об
отсутствии изменений биометрических параметров нерва
в течение жизни человека [14–19]. В нашем исследовании
определена статистически достоверная корреляционная
взаимосвязь между показателями толщины ЗН (с оболоч-
ками и без), ростом и весом. Кроме того, нами впервые
проведена оценка состояния внутренней структуры ЗН с
помощью эходенситометрии и установлены гендерные раз-
личия акустической плотности ЗН и показателей толщины
интраорбитального отдела нерва. Полученные в работе данные показывают теорети-
ческую и практическую значимость индивидуального под-
хода к пациенту, когда учитываются не только локальные
проявления заболевания, но и совокупность вариативных
фенотипических признаков [27]. Russian ophthalmological journal. 2022; 15 (1): 39-45 ЗАКЛЮЧЕНИЕ [22] при опреде-
лении стереометрических параметров ЗН у 17 здоровых лиц
(17 глаз) показатель ИО BMO в среднем составил 0,92 (0,88;
0,96), что соответствует нашим данным. Однако в работе ав-
торов средняя площадь BMO была на 28,5% выше (2,77 мм2)
и средний показатель BMO-MRW на 37,4% ниже (207,7 мкм)
по сравнению с соответствующими параметрами, получен-
ными в нашем исследовании (1,98 мм2 и 332 мкм). Вероятно,
наличие различий между указанными морфо метрическими
параметрами в двух исследованиях обусловлено этничес-
кими и демографическими особенностями. Кроме того,
остается дискуссионным вопрос о влиянии биометрических
показателей глаза на морфометрические параметры ДЗН. Наши данные не выявили прямой достоверной корреля-
ции между показателями ИО BMO и ПЗО, что согласуется
с результатами ряда зарубежных авторов [22–24]. В нашем
исследовании критерием исключения являлась миопическая
рефракция выше 3,0 дптр. В зарубежной литературе имеются
сведения о достоверной корреляции между показателями
ПЗО и ИО ВМО у лиц с миопией средней и высокой степени, 5. Liu H., Zhou H.-F., Zong L.-X, et al. China MRI histogram texture feature
analysis of the optic nerve in the patients with optic neuritis. Chin. Med. Sci. J. March. 2019; 34 (1): 18–23. https://doi.org/10.24920/003507 6. Siebler M. Neuro-orbital ultrasound. Manual of Neurosonology. 2016: 300. https://doi.org/10.1017/cbo9781107447905.031 7. Green R.L., Byrne S.F. Diagnostic ophthalmic ultrasound. Basic science,
inherited retinal disease and tumors. Retina. 2006; 1 (4): 265. https://doi. org/10.1016/b978-0-323-02598-0.50020-3 8. Goeres P., Zeiler F.A., Unger B., Karakitsos D., Gillman L.M. Ultrasound
assessment of optic nerve sheath diameter in healthy volunteers. J. Crit. Care. 2016; 31 (1): 168–71. doi: 10.1016/j.jcrc.2015.10.009 9. Нероев В.В., Елисеева Е.К., Зуева М.В. и др. Демиелинизирующий
оптичес кий неврит: корреляция данных оптической когерентной
томографии и мультифокальной электроретинографии. Анналы
клинической и экспериментальной неврологии. 2014; 8 (2): 22–6. [Neroev V.V., Eliseeva E.K., Zueva M.V., et al. Demyelinating optical
neuritis: correlation of data of optical coherence tomography and multifocal
electroretinography. Annals of clinical and experimental neurology. 2014. 8 (2): 22–6 (in Russian)]. 10. 10. Захарова М.А., Куроедов А.В. Оптическая когерентная томография:
технология, ставшая реальностью. РМЖ. Клиническая офтальмология. 2015. 4: 204–11. [Zakharova M.A., Kuroedov A.V. Optic coherent tomography –
technology which became a reality. RMJ. Clinical ophthalomology. 2015; 4:
204–11 (in Russian)]. Russian ophthalmological journal. 2022; 15 (1): 39-45 11. Oluseyi K.Y.H., Ukamaka I. Ultrasonograhic measurement of optic nerve
sheath diameter in normal adults. Annals of International Medical and Dental
Research. 2017; 3 (2): 30–4. https://doi.org/10.21276/aimdr.2017.3.2.RD9 20. Chihara E., Chihara K. ЗАКЛЮЧЕНИЕ Covariation of optic disc measurements and ocular
parameters in the healthy eye. Graefe’s Arch. Clin. Exp. Ophthalmol. 1994;
232 (2): 265–71. https://doi.org/10.1007/BF00194475 12. Нероев B.B., Киселевa Т.Н., ред. Ультразвуковые исследования в офтальмо-
логии: руководство для врачей. 1-е изд. Москва: ИКАР; 2019. [Neroev V.V.,
Kiseleva T.N., eds. Ultrasound in ophthalmology: a guide for doctors. 1st ed. Moscow: IKAR; 2019 (in Russian)]. 21. Oliveira C. Harizman N., Girkin C. A., et al. Axial length and optic size in
normal eyes. Br. J. Ophthalmol. 2007; 91 (1): 37–9. http://dx.doi.org/10.1136/
bjo.2006.102061 22. Nakanishi H., Suda K., Yoshikawa M., et al. Association of Bruch’s membrane
opening and optic disc morphology to axial length and visual field defects in
eyes with primary open-angle glaucoma. Graefes Arch. Clin. Exp. Ophthalmol. 2018; 256 (3): 599–610. https://doi.org/10.1007/s00417-017-3874-8 13. Kim D.H., Jun J.S., Kim R. Ultrasonographic measurement of the optic
nerve sheath diameter and its association with eyeball transverse diameter
in 585 healthy volunteers. Scientific reports. 2017; 7 (1): 1–6. https://doi. org/10.1038/s41598-017-16173-z 23. Nangia V., Matin A., Bhojwani K., et al. Optic disc size in a population-based
study in central India: the Central India Eye and Medical Study (CIEMS). Acta ophthalmologica. 2008; 86 (1): 103–4. https://doi.org/10.1111/j.1600-
0420.2007.00964.x 14. Chen H., Ding G.S., Zhao Y.C., Yu R.G., Zhou J.X. Ultrasound measurement of
optic nerve diameter and optic nerve sheath diameter in healthy Chinese adults. BMC Neurol. 2015; 15 (1): 1–6. https://doi.org/10.1186/s12883-015-0361-x 24. Chauhan B.C., Danthurebandara V.M., Sharpe G.P., et al. Bruch's membrane
opening minimum rim width and retinal nerve fiber layer thickness in a normal
white population: a multicenter study. Ophthalmology. 2015; 122 (9): 1786–94. https://doi.org/10.1016/j.ophtha.2015.06.001 15. Goeres P., Zeiler F. A., Unger B., Karakitsos D., Gillman L. M. Ultrasound
assessment of optic nerve sheath diameter in healthy volunteers. J. Crit. Care. 2016; 31 (1): 168–71. https://doi.org/10.1016/j.jcrc.2015.10.009 2016; 31 (1): 168–71. https://doi.org/10.1016/j.jcrc.2015.10.00 16. Wang L., Feng L., Yao Y., .et al. Ultrasonographic evaluation of optic nerve
sheath diameter among healthy Chinese adults. Ultrasound Med. Biol. 2016;
42 (3): 683–8. https://doi.org/10.1016/j.ultrasmedbio.2015.11.020 25. Zhao X.J., Jiang H.-Y., Li Y.-H., et al. Correlations between the optic nerve
head morphology and ocular biometrics in highly myopic eyes. International
journal of ophthalmology. 2018; 11 (6): 997–1001. https://doi.org/10.18240/
ijo.2018.06.17 17. Bauerle J., Lochner P., Kaps M., Nedelmann M. Intra- and interobserver
reliability of sonographic assessment of the optic nerve sheath diameter in
healthy adults. J. Neuroimaging. 2012; 22 (1): 42–5. https://doi.org/10.1111/
j.1552-6569.2010.00546.x 26. ЗАКЛЮЧЕНИЕ Guo Y., Liu L.J., Xu L., et al. Optic disc ovality in primary school children
in Beijing. Invest. Ophthalmol. Vis. Sci. 2015; 56 (8): 4547–53. https://doi. org/10.1167/iovs.15-16590 18. Maude R.R., Hossain M.A., Hassan M.U., et al. Transorbital sonographic
evaluation of normal optic nerve sheath diameter in healthy volunteers in
Bangladesh. PLoS One. 2013; 8 (12): e 81013. https://doi.org/10.1371/journal. pone.0081013 27. Журавлева А.Н., Киселева О.А., Кириллова М.О. Персонализированная
медицина в решении проблемы глаукомы. Российский офтальмоло-
гический журнал. 2019; 12 (3): 95–100. [Zhuravleva A.N., Kiseleva O.A.,
Kirillova M.O. Personalized medicine in glaucoma management. 2019;
12 (3): 95–100 (in Russian)]. https://doi.org/10.21516/2072-0076-2019-
12-3-95-100 19. Lochner P., Cantello R., Brigo F., et al. Transorbital sonography in acute optic
neuritis: a case-control study. Am. J. Neuroradiol. 2014; 35 (12): 2371–5. https://doi.org/10.3174/ajnr.A4051 Вклад авторов в работу: В.В. Нероев — руководство проектом; Т.Н. Киселева — научное редактирование
А.В. Баева — сбор и анализ литературных источников, проведение исследований, написание текста ст Вклад авторов в работу: В.В. Нероев
руководство проектом; Т.Н. Киселева
научное редактирование текста статьи, курирование данных;
А.В. Баева — сбор и анализ литературных источников, проведение исследований, написание текста статьи; Е.К. Елисеева — сбор и анализ
литературных источников, редактирование текста статьи; А.Н. Журавлева — проведение исследований, анализ результатов; А.И. Ушаков
— статистическая обработка данных; К.В. Луговкина — проведение исследований, анализ результатов; Т.В. Судовская — проведение ис-
следований, анализ результатов; И.В. Мышко — валидация данных. литературных источников, редактирование текста статьи; А.Н. Журавлева — проведение исследований, анализ результатов; А.И. Ушаков
— статистическая обработка данных; К.В. Луговкина — проведение исследований, анализ результатов; Т.В. Судовская — проведение ис-
следований, анализ результатов; И.В. Мышко — валидация данных. Authors' contribution: V.V. Neroev — project administration; T.N. Kiseleva — science editing, data curation; A.V. Baeva — literature data collection and
analysis, patients examination, writing of the article; E.K. Eliseeva — literature data collection and analysis, editing of the article; A.N. Zhuravleva —
patients examination, data analysis; A.I. Ushakov — data processing; K.V. Lugovkina — patients examination, data analysis; T.V. Sudovskaya — patients
examination, data analysis; I.V. Myshko — data validation. Поступила: 19.04.2021. Переработана: 27.04.2021. Принята к печати: 29.04.2021
Originally received: 19.04.2021. Final revision: 27.04.2021. Accepted: 29.04.2021 ИНФОРМАЦИЯ ОБ АВТОРАХ/INFORMATION ABOUT THE AUTHORS 1 Helmholtz National Medical Research Center of Eye Diseases, 14/19,
Sadovaya-Chernogryazskaya St., Moscow, 105062, Russia 1 ФГБУ «НМИЦ глазных болезней им. Гельмгольца» Минздрава России,
ул. Садовая-Черногрязская, д. 14/19, Москва, 105062, Россия у
р
р
/
2 ФГБОУ ВО «Московский государственный медико-стоматологи-
ческий университет им. А.И. Евдокимова» Минздрава России, ул. Делегатская, д. 20, стр. 1, Москва, 127473, Россия 2 Evdokimov Moscow State Medical Stomatological University of Medicine
and Dentistry, 20/1, Delegatskaya St., Moscow, 127473, Russia 2 Evdokimov Moscow State Medical Stomatological University of Medicine
and Dentistry, 20/1, Delegatskaya St., Moscow, 127473, Russia y,
/ ,
g
y
,
,
,
Vladimir V. Neroev — Academician of the Russian Academy of Sciences,
Dr. of Med. Sci., professor, director1, head of chair of ophthalmology2
Tatiana N. Kiseleva — Dr. of Med. Sci., professor, head of ultrasound
diagnostic department1 Владимир Владимирович Нероев — академик РАН, д-р мед. наук,
профессор, директор1, заведующий кафедрой глазных болезней2 Tatiana N. Kiseleva — Dr. of Med. Sci., professor, head of ultrasound
diagnostic department1 Tatiana N. Kiseleva — Dr. of Med. Sci., professor, head of ultrasound
diagnostic department1 Татьяна Николаевна Киселева — д-р мед. наук, профессор, началь-
ник отдела, главный научный сотрудник отдела ультразвуковых
исследований1 Татьяна Николаевна Киселева — д-р мед. наук, профессор, началь-
ник отдела, главный научный сотрудник отдела ультразвуковых
исследований1 Alena V. Baeva — PhD student, of chair of ophthalmology2 Elena K. Eliseeva — Cand. of Med. Sci., researcher of ultrasound diagnostic
department1 Алена Витальевна Баева — аспирант кафедры глазных болезней2
Е е а Ко с а
о
а Е
сее а
й Алена Витальевна Баева — аспирант кафедры глазных болезней2
Елена Константиновна Елисеева — канд. мед. наук, научный сотруд-
ник отдела ультразвуковых исследований1 Алена Витальевна Баева — аспирант кафедры глазных болезней2
Елена Константиновна Елисеева — канд. мед. наук, научный сотруд-
й1 Anastasia N. Zhuravleva — Cand. of Med. Sci., researcher of glaucoma
department1 Anastasia N. Zhuravleva — Cand. of Med. Sci., researcher of glaucoma
department1 ник отдела ультразвуковых исследований1 Alexandr I. Ushakov — resident doctor1 Анастасия Николаевна Журавлева — канд. мед. наук, научный со-
трудник отдела глаукомы1 Анастасия Николаевна Журавлева — канд. мед. наук, научный со-
трудник отдела глаукомы1 Kseniya V. Lugovkina — Cand. of Med. Sci., researcher of ultrasound
diagnostic department1 Александр Игоревич Ушаков — врач-ординатор1 Александр Игоревич Ушаков — врач-ординатор1 Tatiana V. Sudovskaya — Dr. of Med. Sci., doctor of highest category1
Ilya V. Myshko — resident doctor1 Tatiana V. Sudovskaya — Dr. of Med. Взаимосвязь акустических и морфометрических характеристик зрительных
нервов с антропометрическими параметрами у молодых лиц в норме
4
Ðîññèéñêèé îôòàëüìîëîãè÷åñêèé æóðíàë. 2022; 15 (1): 39-45 ИНФОРМАЦИЯ ОБ АВТОРАХ/INFORMATION ABOUT THE AUTHORS Sci., doctor of highest category1
Ilya V. Myshko — resident doctor1 Ксения Вадимовна Луговкина — канд. мед. наук, научный сотрудник Ксения Вадимовна Луговкина — канд. мед. наук, н Ксения Вадимовна Луговкина — канд. мед. наук, научный сотрудник
1 Ксения Вадимовна Луговкина — канд. мед. наук, научный сотрудник
отдела ультразвуковых исследований1 отдела ультразвуковых исследований1 Татьяна Викторовна Судовская — д-р мед. наук, врач высшей кате-
гории1 Татьяна Викторовна Судовская — д-р мед. наук, врач высшей кате-
гории1 Contact information: Elena K. Eliseeva,
eliseevaek@ya.ru Илья Владимирович Мышко — врач-ординатор1 Илья Владимирович Мышко — врач-ординатор1 Для контактов: Елена Константиновна Елисеева,
eliseevaek@ya.ru Для контактов: Елена Константиновна Елисеева 45 Взаимосвязь акустических и морфометрических характеристик зрительных
нервов с антропометрическими параметрами у молодых лиц в норме
45
Ðîññèéñêèé îôòàëüìîëîãè÷åñêèé æóðíàë. 2022; 15 (1): 39-45 имосвязь акустических и морфометрических характеристик зрительных
нервов с антропометрическими параметрами у молодых лиц в норме
|
https://openalex.org/W2031784543
|
https://www.nature.com/articles/bcj20111.pdf
|
English
| null |
The tyrosine kinase inhibitor dasatinib (SPRYCEL) inhibits chondrocyte activity and proliferation
|
Blood cancer journal
| 2,011
|
cc-by
| 3,751
|
The tyrosine kinase inhibitor dasatinib (SPRYCEL) inhibits chondrocyte activity and
proliferation Blood Cancer Journal (2011) 1, e2; doi:10.1038/bcj.2011.1;
published online 4 February 2011 in two of five animals after 12 weeks. The width of the growth
plate was significantly lower in the dasatinib-treated group than
in vehicle-treated controls after 12 weeks of treatment (Po0.05;
Figure 1). in two of five animals after 12 weeks. The width of the growth
plate was significantly lower in the dasatinib-treated group than
in vehicle-treated controls after 12 weeks of treatment (Po0.05;
Figure 1). Dasatinib
(BMS-354825,
SPRYCEL,
Bristol-Myers
Squibb,
New York, NY, USA) is an ATP-competitive protein tyrosine
kinase inhibitor (TKI), which was originally identified as a potent
inhibitor of Src family kinases (including Src, Lck, Hck, Yes, Fgr,
Lyn and Fyn) and was subsequently found to have activity against
Abl, Kit, the macrophage colony stimulating factor receptor (Fms),
the platelet-derived growth factor (PDGF) receptor (PDGFR)-a
and -b and the Eph receptor family members EphB1, EphB2 and
EphB4.1–3 Dasatinib is an effective therapy for chronic myeloid
leukaemia (CML) in patients who are resistant to front-line
imatinib mesylate therapy due to its increased affinity for the CML
oncoprotein BCR-Abl and its insensitivity to mutations in the
BCR-Abl kinase domain. Furthermore, recent data suggest that
dasatinib may be more effective than imatinib as a front-line
therapy for chronic phase CML.4 g
Postnatal longitudinal bone growth in the axial and appendi-
cular limb skeleton is controlled by the regulated proliferation
and
activity
of
chondrocytes
and
osteoblasts
within
the
epiphyseal
growth
plate. This
process
of
endochondral
ossification involves the chondrocyte-mediated production of
a cartilagenous template, which is later mineralised and
remodelled to form mature lamellar bone. We postulate that
the accelerated growth-plate narrowing observed in dasatinib-
treated rats may be due, in part, to the inhibition of chondrocyte
proliferation and activity. To determine whether dasatinib directly affects chondrocyte
proliferation and activity, the effects of dasatinib on the murine
pre-chondrocyte cell line ATDC5 were investigated in vitro. ATDC5 cells were cultured for up to 6 days with dasatinib or
vehicle
and
the
relative
number of
cells per
well
was
determined by WST-1 assay. In vehicle-treated cultures, cell
numbers increased fourfold during the 6 days of culture
(Figure 2a). Treatment of ATDC5 cells with dasatinib signifi-
cantly inhibited cell proliferation at concentrations of 2.5 nM
and higher after 2, 4 or 6 days, with an IC50 of 6.2 nM on day 6. The tyrosine kinase inhibitor dasatinib (SPRYCEL) inhibits chondrocyte activity and
proliferation The success of TKIs for the treatment of CML has resulted in
the investigation of the use of these drugs in an increasing
number
of
paediatric
haematological
and
solid
tumours. However, although TKI treatment is generally well tolerated,
there is emerging data to suggest that TKI therapy may result in
decreased growth in children. Three recently published case-
studies report decelerated growth in juvenile CML patients
undergoing imatinib therapy.5–7 Additionally, a French phase IV
trial has recently reported decreased growth in a cohort of 22
children and adolescents (age: 10 months–17 years) receiving
imatinib therapy for chronic-phase CML.8 Growth rates in this
group were significantly lower following imatinib treatment,
with a significant decrease in height z-score (median, 0.37;
range, 1.09 to þ 0.14) after 12 months of treatment, compared
with baseline.8 In keeping with this, we have previously
reported that imatinib treatment caused growth plate closure
in normal rats in vivo and inhibit chondrocyte proliferation and
activity in vitro,9 providing a possible explanation for the
reduced longitudinal growth observed in juvenile patients
treated with imatinib. The effects of dasatinib on chondrocyte activity were next
investigated using a GAG-synthesis assay. ATDC5 cultures were
treated with 10 ng/ml rhTGF-b1, an inducer of chondrocyte
differentiation and activity (Figure 2b). Dasatinib treatment for 200
150
*
100
Growth plate thickness (μm)
50
Vehicle
Dasatinib
0
wk 4
wk 8
wk 12 200
150
*
100
Growth plate thickness (μm)
50
Vehicle
Dasatinib
0
wk 4
wk 8
wk 12
Figure 1
The proximal tibial growth plate is thinned in dasatinib-
treated animals. Nine-month-old female Sprague–Dawley rats were
treated for 12 weeks with dasatinib (5 mg/kg per day) or vehicle
(10% DMSO/90% PEG 300). Six animals per group were humanely
killed after 4, 8 and 12 weeks of treatment and right tibiae were
collected for histological analysis. Samples were embedded in paraffin
and 5-mm sections were stained with safranin O/fast green. Growth
plate thickness was measured at five equidistant points across the
centre of the growth plate. Box plots depict median, 25th and 75th
percentiles±range (n ¼ 4–6). nPo0.05 (Student’s t-tests). In paediatric cases that are resistant or intolerant to imatinib
and where allogeneic transplants are not possible, dasatinib
is recommended as a second-line therapy. In light of this
increased investigation of the use of dasatinib in the treatment
of paediatric cancers, we investigated whether dasatinib, like
imatinib, affected the growth plate in vivo. Citation: Blood Cancer Journal (2011) 1, e2; doi:10.1038/bcj.2011.1
& 2011 Macmillan Publishers Limited All rights reserved 2044-5385/11 Citation: Blood Cancer Journal (2011) 1, e2; doi:10.1038/bcj.2011.1
& 2011 Macmillan Publishers Limited All rights reserved 2044-5385/11 www.nature.com/bcj The tyrosine kinase inhibitor dasatinib (SPRYCEL) inhibits chondrocyte activity and
proliferation The cells were then treated with dasatinib or vehicle
(0.05% DMSO) for the indicated periods of time before the relative number of cells per well were determined by WST-1 assay. Graphs depict
mean±range of two representative experiments. nPo0.05 relative to vehicle control at each time-point (one-way ANOVA with Dunnett’s post-
tests). (b) ATDC5 cells (1.56 105 cells per cm2) were treated with rhTGF-b1 (10 ng/ml) and the indicated doses of dasatinib or 0.05% DMSO
vehicle for 48 h. GAG levels were then quantitated and normalised to DNA content to determine relative GAG production per cell. Graphs
depict mean±s.e.m. of triplicate wells of a representative experiment, normalised to the rhTGF-b1-treated vehicle control. nPo0.05 relative to the
rhTGF-b1-treated vehicle control (one-way ANOVA with Dunnett’s post tests). (c) ATDC5 cells were cultured overnight at 1.56 105 cells per cm2
in six-well plates in c-a-MEM. The cells were serum starved by overnight incubation in serum-free a-MEM. The cells were pre-treated with
dasatinib or 0.05% DMSO for 2 h before stimulation with rhPDGF-BB (10 ng/ml) for 5 min. Cell lysates (30 mg per lane) were resolved through a
10% sodium dodecyl sulphate-PAGE gel and the phosphorylated forms of Akt (p-Akt) and ERK1/2 (p-ERK1/2), as well as total HSP90, were
detected using specific antibodies. Images from a representative of two experiments are shown. (d) Graphs indicate the pixel intensity for
p-Akt and p-ERK1/2, relative to HSP90, normalised to vehicle controls. (e) ATDC5 cells were treated with 40 nM dasatinib or vehicle (0.05%
DMSO) supplemented, where indicated, with 10 ng/ml rhPDGF-BB. After 6 days, the relative number of cells per well was determined by WST-1
assay. Graphs depict mean±range of two representative experiments, normalised to vehicle control. (f) ATDC5 cells (cultured as in b) were then
treated with vehicle (V) or 40 nM dasatinib (D) with or without 100 ng/ml rhPDGF-BB. After 48 h, GAG levels and DNA content were quantitated
and GAG production per cell was calculated. Graphs depict mean±s.e.m. of triplicate wells of a representative experiment, normalised to vehicle
controls. nPo0.05, as indicated (Student’s t-tests). the decreased growth plate thickness observed in the dasatinib-
treated normal rats. 48 h significantly decreased TGF-b1-induced GAG synthesis at
40 nM concentrations and higher, with a 40% reduction in GAG
synthesis at 40 and 80 nM dasatinib, relative to vehicle controls
(Figure 2b). The tyrosine kinase inhibitor dasatinib (SPRYCEL) inhibits chondrocyte activity and
proliferation In this study, it was found that dasatinib treatment resulted in
a significant decrease in cartilagenous growth plate thickness in
normal rats. Sprague–Dawley rats were treated with dasatinib
(5 mg/kg) or vehicle (10% DMSO/90% PEG 300) by daily oral
gavage for up to 12 weeks and the proximal tibial growth plate
was examined by histology. In vehicle-treated controls, the
width of the cartilagenous growth plate at the proximal tibia
remained constant at all time-points examined (P ¼ 0.070,
one-way ANOVA; Figure 1). In contrast, the mean growth plate
thickness decreased significantly over time in dasatinib-treated
animals, with complete closure at the centre of the growth plate Figure 1
The proximal tibial growth plate is thinned in dasatinib-
treated animals. Nine-month-old female Sprague–Dawley rats were
treated for 12 weeks with dasatinib (5 mg/kg per day) or vehicle
(10% DMSO/90% PEG 300). Six animals per group were humanely
killed after 4, 8 and 12 weeks of treatment and right tibiae were
collected for histological analysis. Samples were embedded in paraffin
and 5-mm sections were stained with safranin O/fast green. Growth
plate thickness was measured at five equidistant points across the
centre of the growth plate. Box plots depict median, 25th and 75th
percentiles±range (n ¼ 4–6). nPo0.05 (Student’s t-tests). Figure 1
The proximal tibial growth plate is thinned in dasatinib-
treated animals. Nine-month-old female Sprague–Dawley rats were
treated for 12 weeks with dasatinib (5 mg/kg per day) or vehicle
(10% DMSO/90% PEG 300). Six animals per group were humanely
killed after 4, 8 and 12 weeks of treatment and right tibiae were
collected for histological analysis. Samples were embedded in paraffin
and 5-mm sections were stained with safranin O/fast green. Growth
plate thickness was measured at five equidistant points across the
centre of the growth plate. Box plots depict median, 25th and 75th
percentiles±range (n ¼ 4–6). nPo0.05 (Student’s t-tests). The tyrosine kinase inhibitor dasatinib (SPRYCEL) inhibits chondrocyte activity and
proliferation 6
5
*
Vehicle
150%
100%
GAG synthesis
(35S counts/μg DNA)
50%
*
*
*
0%
V
V
5
10
+ TGF-β1
20 40 80
2.5 nM
5 nM
10 nM
20 nM
40 nM
*
*
**
4
3
Cells/well
(relative to day 0)
Cells/well
(relative to vehicle)
2
1
0
0
2
4
Days of treatment
+ PDGF
64 kDa
50 kDa
94 kDa
p-Akt
100%
50%
p-Akt
p-ERK1/2
0%
V
0.1
400%
200%
*
*
*
150%
100%
50%
0%
*
*
*
300%
200%
GAG synthesis
(35S counts/μg DNA)
100%
0%
V
D
V
+ PDGF
D
1
10
Dasatinib (nM)
100
1000
p-ERK1/2
HSP90
Vehicle
Vehicle
2.5 nM
5 nM
10 nM
20 nM
40 nM
6
8
V
D
V
+ PDGF
D
a
b
c
e
f
d
Figure 2
Dasatinib inhibits cell proliferation and activity in ATDC5 cultures. (a) ATDC5 cells (1.56 104 cells per cm2) were seeded in 96-well
plates in a-MEM with 10% fetal bovine serum, 2 mM L-glutamine, 1 mM sodium pyruvate, 15 mM HEPES, 50 U/ml penicillin, 50 mg/ml streptomycin
and 100 mM ascorbate (c-a-MEM) and were cultured overnight to allow the cells to adhere. The cells were then treated with dasatinib or vehicle
(0.05% DMSO) for the indicated periods of time before the relative number of cells per well were determined by WST-1 assay. Graphs depict
mean±range of two representative experiments. nPo0.05 relative to vehicle control at each time-point (one-way ANOVA with Dunnett’s post-
tests). (b) ATDC5 cells (1.56 105 cells per cm2) were treated with rhTGF-b1 (10 ng/ml) and the indicated doses of dasatinib or 0.05% DMSO
vehicle for 48 h. GAG levels were then quantitated and normalised to DNA content to determine relative GAG production per cell. Graphs
depict mean±s.e.m. of triplicate wells of a representative experiment, normalised to the rhTGF-b1-treated vehicle control. nPo0.05 relative to the
rhTGF-b1-treated vehicle control (one-way ANOVA with Dunnett’s post tests). (c) ATDC5 cells were cultured overnight at 1.56 105 cells per cm2
in six-well plates in c-a-MEM. The cells were serum starved by overnight incubation in serum-free a-MEM. The cells were pre-treated with
dasatinib or 0.05% DMSO for 2 h before stimulation with rhPDGF-BB (10 ng/ml) for 5 min. The tyrosine kinase inhibitor dasatinib (SPRYCEL) inhibits chondrocyte activity and
proliferation Cell lysates (30 mg per lane) were resolved through a
10% sodium dodecyl sulphate-PAGE gel and the phosphorylated forms of Akt (p-Akt) and ERK1/2 (p-ERK1/2), as well as total HSP90, were
detected using specific antibodies. Images from a representative of two experiments are shown. (d) Graphs indicate the pixel intensity for
p-Akt and p-ERK1/2, relative to HSP90, normalised to vehicle controls. (e) ATDC5 cells were treated with 40 nM dasatinib or vehicle (0.05%
DMSO) supplemented, where indicated, with 10 ng/ml rhPDGF-BB. After 6 days, the relative number of cells per well was determined by WST-1
assay. Graphs depict mean±range of two representative experiments, normalised to vehicle control. (f) ATDC5 cells (cultured as in b) were then
treated with vehicle (V) or 40 nM dasatinib (D) with or without 100 ng/ml rhPDGF-BB. After 48 h, GAG levels and DNA content were quantitated
and GAG production per cell was calculated. Graphs depict mean±s.e.m. of triplicate wells of a representative experiment, normalised to vehicle
controls. nPo0.05, as indicated (Student’s t-tests). Letter to the Editor Letter to the Editor 2 6
5
*
Vehicle
2.5 nM
5 nM
10 nM
20 nM
40 nM
*
*
**
4
3
Cells/well
(relative to day 0)
2
1
0
0
2
4
Days of treatment
6
8
a 150%
100%
GAG synthesis
(35S counts/μg DNA)
50%
*
*
*
0%
V
V
5
10
+ TGF-β1
20 40 80
b β
Akt
100%
50%
p-Akt
p-ERK1/2
0%
V
0.1
1
10
Dasatinib (nM)
100
1000
ERK1/2
P90
d + PDGF
64 kDa
50 kDa
94 kDa
p
p
H
Vehicle
Vehicle
2.5 nM
5 nM
10 nM
20 nM
40 nM
c d c Dasatinib (nM) 400%
*
*
*
300%
200%
GAG synthesis
(35S counts/μg DNA)
100%
0%
V
D
V
+ PDGF
D
Dasatinib (nM)
f Cells/well
(relative to vehicle)
200%
*
*
*
150%
100%
50%
0%
V
D
V
+ PDGF
D
e f f e Figure 2
Dasatinib inhibits cell proliferation and activity in ATDC5 cultures. (a) ATDC5 cells (1.56 104 cells per cm2) were seeded in 96-well
plates in a-MEM with 10% fetal bovine serum, 2 mM L-glutamine, 1 mM sodium pyruvate, 15 mM HEPES, 50 U/ml penicillin, 50 mg/ml streptomycin
and 100 mM ascorbate (c-a-MEM) and were cultured overnight to allow the cells to adhere. References 1 Lombardo LJ, Lee FY, Chen P, Norris D, Barrish JC, Behnia K et al. Discovery of N-(2-chloro-6-methyl-phenyl)-2-(6-(4-(2-hydroxyethyl)-
piperazin-1-yl)-2-methylpyrimidin-4- ylamino)thiazole-5-carboxamide
(BMS-354825), a dual Src/Abl kinase inhibitor with potent antitumor
activity in preclinical assays. J Med Chem 2004; 47: 6658–6661. 2 Melnick JS, Janes J, Kim S, Chang JY, Sipes DG, Gunderson D
et al. An efficient rapid system for profiling the cellular activities
of molecular libraries. Proc Natl Acad Sci USA 2006; 103:
3153–3158. 3 Vandyke K, Dewar AL, Farrugia AN, Fitter S, Bik To L, Hughes TP
et al. Therapeutic concentrations of dasatinib inhibit in vitro
osteoclastogenesis. Leukemia 2009; 23: 994–997. g
4 Kantarjian H, Shah NP, Hochhaus A, Cortes J, Shah S, Ayala M
et al. Dasatinib versus imatinib in newly diagnosed chronic-phase
chronic myeloid leukemia. N Engl J Med 2010; 362: 2260–2270. In
addition
to
inhibiting
chondrocyte
proliferation
and
activity, dasatinib can affect osteoblast activity, inhibiting
stromal-cell proliferation while, under at least some conditions,
promoting osteoblast activity.16–18 Growth plate thickness is
determined by rate of chondrocyte proliferation and hyper-
trophy and by the rate of replacement of the cartilagenous matrix
with mature bone. Therefore, the stimulation of osteoblast
activity by dasatinib may have additional effects on growth plate
thickness by accelerating growth plate mineralisation. However,
although published data suggest that dasatinib can increase
Sprague–Dawley osteoblast activity in vitro, we have previously
reported that treatment of normal Sprague–Dawley rats with
5 mg/kg dasatinib had no effect on osteoblast activity.16 These
findings suggest that the effects of dasatinib on the growth plate
in this study were independent of effects on osteoblast activity. 5 Mariani S, Giona F, Basciani S, Brama M, Gnessi L. Low bone
density and decreased inhibin-B/FSH ratio in a boy treated with
imatinib during puberty. Lancet 2008; 372: 111–112. g p
y
6 Kimoto T, Inoue M, Kawa K. Growth deceleration in a girl treated
with imatinib. Int J Hematol 2009; 89: 251–252. 7 Schmid H, Jaeger BA, Lohse J, Suttorp M. Longitudinal growth
retardation in a prepubertal girl with chronic myeloid leukemia
on long-term treatment with imatinib. Haematologica 2009; 94:
1177–1179. 8 Millot F, Baruchel A, Guilhot J, Petit A, Leblanc T, Bertrand Y
et al. Imatinib is efficient but has a negative impact on growth
in children with previously untreated chronic myelogenous
leukaemia (CML) in early chronic phase (CP): results of the
French national phase IV trial [abstract]. Blood 2009; 114: 863. Conflict of interest proliferation and activity through inhibition of PDGFR-b. We examined whether inhibition of PDGFR-b contributed
to the inhibitory effects of dasatinib on ATDC5 proliferation
and GAG production. First, the effects of dasatinib on PDGFR
receptor signalling through Akt and ERK1/2 were examined
by western blot. Consistent with the known effects of dasatinib
on the PDGFR, dasatinib inhibited the rhPDGF-BB-induced
phosphorylation of Akt and ERK1/2 in a dose-dependent
manner, with 40 nM dasatinib completely abrogating PDGFR
signalling (Figure 2c). Dasatinib inhibited PDGFR signalling
through
Akt
and
ERK1/2
at
IC50
of
13.0
and
16.0 nM,
respectively (Figure 2d). The authors declare no conflict of interest. The tyrosine kinase inhibitor dasatinib (SPRYCEL) inhibits chondrocyte activity and
proliferation These data demonstrate that dasatinib inhibits
proliferation and extracellular matrix synthesis in cultures of
chondrocyte-like cells in vitro, suggesting that inhibition of
chondrocyte proliferation and activity may be responsible for The dasatinib target PDGFR is an important regulator of
chondrocyte proliferation and activity. PDGF-BB is a known
mitogen for chondrocytes and has been reported to promote
chondrocyte activity in at least some cell types,10,11 suggesting
that dasatinib treatment may have direct effects on chondrocyte Blood Cancer Journal Letter to the Editor 3 Acknowledgements We are grateful to Behzad Baradaran and the staff at Veterinary
Services, IMVS, for their assistance with the animal experiments. We wish to thank Lee Anne Griffiths, Francis Lee, Richard Smykla
and Kate Church from Bristol-Myers Squibb for the provision of
dasatinib and for helpful discussions. This work was supported by
a grant from the Leukemia and Lymphoma Society Translational
Research Program. Consistent with the known effects of PDGF on chondrocyte
proliferation, treatment of ATDC5 with PDGF-BB for 6 days
increased the number of cells per well by 1.5-fold, compared
with vehicle-treated controls (Figure 2e). This increase in cell
proliferation was abrogated by dasatinib treatment (Figure 2e). Additionally, treatment with rhPDGF-BB for 48 h induced a
threefold increase in GAG production, on a per cell basis,
relative to untreated controls (Figure 2f). This stimulatory effect
of PDGF-BB was partially inhibited by co-treatment with 40 nM
dasatinib, although levels did not reach those of unstimulated
dasatinib-treated cultures (Figure 2f). K Vandyke1,2, S Fitter1 and ACW Zannettino1,2 1Myeloma Research Laboratory, Department of Haematology,
Centre for Cancer Biology Institute of Medical and Veterinary 1Myeloma Research Laboratory, Department of Haematology,
Centre for Cancer Biology, Institute of Medical and Veterinary
Science, Adelaide, Australia and Centre for Cancer Biology, Institute of Medical and Veterinary
Science, Adelaide, Australia and 2School of Medicine, University of Adelaide, Adelaide,
Australia E-mail: andrew.zannettino@health.sa.gov.au In this study, dasatinib treatment partially reversed the
activating effects of PDGFR on the proliferation and GAG-
synthetic properties of the murine chondrocyte cell line ATDC5,
suggesting that inhibition of PDGFR signalling is likely to
contribute to the inhibitory effects of dasatinib on chondrocytes. However, given the broad target specificities of dasatinib, other
tyrosine kinases may also have a role in the effects of dasatinib
on
chondrogenesis. Inhibition
of
Src-family
kinases
may
contribute to the anti-proliferative effects of dasatinib on
chondrocytes, as inhibition of Src-family kinases with PP2 has
previously been found to inhibit chondrocyte proliferation
in vitro.12–14 However, although Src-/- mice have retarded long
bone growth, the growth plates of Src-deficient mice are thicker
than normal15 suggesting that inhibition of Src alone cannot
explain the growth plate thinning observed in this study. References p
9 Vandyke K, Dewar AL, Fitter S, Menicanin D, To LB, Hughes TP
et al. Imatinib mesylate causes growth plate closure in vivo. Leukemia 2009; 23: 2155–2159. 10 Kieswetter K, Schwartz Z, Alderete M, Dean DD, Boyan BD. Platelet
derived growth factor stimulates chondrocyte proliferation but
prevents endochondral maturation. Endocrine 1997; 6: 257–264. The putative effects of dasatinib on the growth plate may have
implications for the use of dasatinib in the paediatric setting. Although these studies suggest that TKI therapy in pre-pubertal
individuals may retard growth, there are currently no reports
suggesting that dasatinib may also have effects on growth in
paediatric patients. Our results suggest that growth plate
changes should be investigated in paediatric patients who are
undergoing treatment with dasatinib. The relative benefit of
using dasatinib as a front-line treatment for diseases affecting
children
and
adolescents
may
need
to
be
re-evaluated,
taking into account the potential effects of dasatinib on the
growth plate. 11 Weiser L, Bhargava M, Attia E, Torzilli PA. Effect of serum and
platelet-derived growth factor on chondrocytes grown in collagen
gels. Tissue Eng 1999; 5: 533–544. 12 Calalb MB, Polte TR, Hanks SK. Tyrosine phosphorylation of
focal adhesion kinase at sites in the catalytic domain regulates
kinase activity: a role for Src family kinases. Mol Cell Biol 1995;
15: 954–963. 13 Gemba T, Valbracht J, Alsalameh S, Lotz M. Focal adhesion kinase
and mitogen-activated protein kinases are involved in chondrocyte
activation by the 29-kDa amino-terminal fibronectin fragment. J Biol Chem 2002; 277: 907–911. Blood Cancer Journal Letter to the Editor and enhances differentiation of osteoblasts. Oncogene 2010; 29:
3196–3207. and enhances differentiation of osteoblasts. Oncogene 2010; 29:
3196–3207. 14 Gill KS, Beier F, Goldberg HA. Rho-ROCK signaling differentially
regulates chondrocyte spreading on fibronectin and bone sialo-
protein. Am J Physiol Cell Physiol 2008; 295: C38–C49. 18 Id Boufker H, Lagneaux L, Najar M, Piccart M, Ghanem G, Body JJ
et al. The Src inhibitor dasatinib accelerates the differentiation of
human bone marrow-derived mesenchymal stromal cells into
osteoblasts. BMC Cancer 2010; 10: 298. p
y
y
15 Soriano P, Montgomery C, Geske R, Bradley A. Targeted disruption
of the c-src proto-oncogene leads to osteopetrosis in mice. Cell
1991; 64: 693–702. 16 Vandyke K, Dewar AL, Diamond P, Fitter S, Schultz CG, Sims NA
et al. The tyrosine kinase inhibitor dasatinib dysregulates bone
remodeling through inhibition of osteoclasts in vivo. References J Bone Miner
Res 2010; 25: 1759–1770. This work is licensed under the Creative Commons
Attribution-NonCommercial-No Derivative Works
3.0 Unported License. To view a copy of this license, visit http://
creativecommons.org/licenses/by-nc-nd/3.0/ 17 Lee YC, Huang CF, Murshed M, Chu K, Araujo JC, Ye X et al. Src family kinase/abl inhibitor dasatinib suppresses proliferation Blood Cancer Journal
|
https://openalex.org/W3025751266
|
https://www.jnwpu.org/articles/jnwpu/pdf/2020/01/jnwpu2020381p147.pdf
|
Chinese
| null |
A Link Distance Division Based Time Division Multiple Access Protocol for Directional Aeronautical Relay Networks
|
Xibei gongye daxue xuebao
| 2,020
|
cc-by
| 1,073
|
西北工业大学学报 西北工业大学学报 西北工业大学学报
ＪｏｕｒｎａｌｏｆＮｏｒｔｈｗｅｓｔｅｒｎＰｏｌｙｔｅｃｈｎｉｃａｌＵｎｉｖｅｒｓｉｔｙ Ｆｅｂ．
Ｖｏｌ．３８
２０２０
Ｎｏ．１ Ｆｅｂ．
Ｖｏｌ．３８
２０２０
Ｎｏ．１ ２０２０年２月
第３８卷第１期 ＪｏｕｒｎａｌｏｆＮｏｒｔｈｗｅｓｔｅｒｎＰｏｌｙｔｅｃｈｎｉｃａｌＵｎｉｖｅｒｓｉｔｙ ｈｔｔｐｓ：／／ｄｏｉ．ｏｒｇ／１０．１０５１／ｊｎｗｐｕ／２０２０３８１０１４７ 收稿日期：２０１９⁃０１⁃１２ 基金项目：国家自然科学基金（６１７７１３９２，６１７７１３９０，６１８７１３２２，６１５０１３７３，６１２７１２７９）、国家科技重
大专项（２０１５ＺＸ０３００２００６⁃００４，２０１６ＺＸ０３００１０１８⁃００４）与航空科学基金（２０１８５５５３０３５）资助
作者简介：闫中江（１９８３—），西北工业大学副教授，主要从事宽带无线网络组网协议设计、片上网络协议设计与实现研究。 收稿日期：２０１９⁃０１⁃１２ 基金项目：国家自然科学基金（６１７７１３９２，６１７７１３９０，６１８７１３２２，６１５０１３７３，６１２７１２７９）、国家科技重
大专项（２０１５ＺＸ０３００２００６⁃００４，２０１６ＺＸ０３００１０１８⁃００４）与航空科学基金（２０１８５５５３０３５）资助 定向航空中继网络中一种基于链路距离
分环的多址接入协议
闫中江，李倩倩，李波，杨懋
（西北工业大学电子信息学院，陕西西安 ７１００７２） 摘 要：在定向航空中继网络中，中继飞机利用定向天线通信距离远、发射功率小、无线信号干扰范围
小等优点，辅助地面节点克服地形遮挡、信号衰弱大、通信距离受限等缺点，以较少的跳数进行通信，
有效降低端到端数据传输时延、提高网络吞吐量。针对定向航空中继网络中由于通信链路长、无线信
号传播时延与数据传输时延相近，所引起的通信时延扩展、多址接入协议效率低等问题，提出了一种
基于通信链路距离进行分环的时分多址接入协议（ｌｉｎｋｄｉｓｔａｎｃｅｄｉｖｉｓｉｏｎｂａｓｅｄｔｉｍｅｄｉｖｉｓｉｏｎｍｕｌｔｉｐｌｅａｃ⁃
ｃｅｓｓｐｒｏｔｏｃｏｌ，ＬＤＤ⁃ＴＤＭＡ）。与传统的ＴＤＭＡ多址接入协议中所有链路均使用相同的时隙长度不同，
ＬＤＤ⁃ＴＤＭＡ根据通信链路的距离远近使用不同的时隙长度。进一步为了简化协议实现，提出通信覆
盖范围分环的概念，使得处于同一个环内链路距离相近的节点使用相同的时隙长度。接着，建模分析
并推导出了最大化多址接入效率的分环的个数、分环半径与节点最大通信距离之间的闭合表达式。
最后，仿真结果表明，当中继飞机最大通信距离为２００ｋｍ、分环个数为４时，ＬＤＤ⁃ＴＤＭＡ的多址接入效
率相较于传统的ＴＤＭＡ可提高１３．３７％。 关 键 词：定向航空中继网络；多址接入协议；链路距离；通信时延扩展；分环
中图分类号：ＴＮ９２５ 文献标志码：Ａ 文章编号：１０００⁃２７５８（２０２０）０１⁃０１４７⁃０８ 关 键 词：定向航空中继网络；多址接入协议；链路距离；通信时延扩展；分环
中图分类号：ＴＮ９２５ 文献标志码：Ａ 文章编号：１０００⁃２７５８（２０２０）０１⁃０１４７⁃０８ 定向航空中继网络（ｄｉｒｅｃｔｉｏｎａｌａｅｒｏｎａｕｔｉｃａｌｒｅｌａｙ
ｎｅｔｗｏｒｋ，ＤＡＲＮ）也被称为定向机载距离扩展网络
（ｄｉｒｅｃｔｉｏｎａｌａｉｒｂｏｒｎｅｒａｎｇｅｅｘｔｅｎｓｉｏｎｎｅｔｗｏｒｋｓ）［１⁃５］，近
年来引起了人们的广泛关注。在ＤＡＲＮ中有２类节
点，分别为空中飞机和地面节点；其中，空中飞机也
被称为空中定向中继节点，它具有飞行高度高（可
达数公里）、与大部分地面节点均具有视距通信路
径（ｌｉｎｅｏｆｓｉｇｈｔ，ＬＯＳ）、通信覆盖范围大（可达数百
公里）等优点，可为大量的、距离较远的地面节点中
继数据。然而，与传统的全向中继飞机［６］不同，定
向中继飞机挂载定向天线辅助地面节点进行通信。
定向天线通过采用定向波束成型技术，可将无
线信号集中到某个特定方向上；在相同发射功率的
情况下，与全向天线相比，定向天线的无线信号通信
距离更远；此外，定向天线还具有无线信号干扰范围 小、通信保密性强等优点，因此受到了国内外研究者
的广泛关注。然而，随着定向中继飞机与地面节点
的通信链路距离增大，也带来了通信时延扩展问
题［２⁃３］，降低了多址接入协议的效率。 通信时延扩展，指的是由于无线通信链路距离
的变长，导致原本可被忽略的无线信号传播时延变
大，与数据传输时延相比不再可被忽略；从而，完成
一次数据传输与应答（ＤＡＴＡ⁃ＡＣＫ）的通信元操作所
需要的时隙长度变长，导致多址接入协议效率降低
的问题。比如，令通信链路的数据速率为２Ｍｂ／ｓ，
则传输一个２ｋｂ的数据包所需要的数据传输时延
为１ｍｓ；当该通信链路长度为３００ｍ时，无线信号传
播时延约为１μｓ（无线信号以光速传播），此时完成
一次ＤＡＴＡ⁃ＡＣＫ所需时隙长度约为１．００２ｍｓ，约为
１ｍｓ（信号传播时延可忽略）；而当通信链路长度为 收稿日期：２０１９⁃０１⁃１２ 基金项目：国家自然科学基金（６１７７１３９２，６１７７１３９０，６１８７１３２２，６１５０１３７３，６１２７１２７９）、国家科技重
大专项（２０１５ＺＸ０３００２００６⁃００４，２０１６ＺＸ０３００１０１８⁃００４）与航空科学基金（２０１８５５５３０３５）资助
作者简介：闫中江（１９８３—），西北工业大学副教授，主要从事宽带无线网络组网协议设计、片上网络协议设计与实现研究。 ），西北工业大学副教授，主要从事宽带无线网络组网协议设计、片上网络协议设计与实现研究。 西 北 工 业 大 学 学 报 第３８卷 第３８卷 ８
４
１ 一类、划分到一个环中，使用相同的时隙长度；而不
同的环对应的时隙长度不同。建模分析并推导出了
分环的个数以及每个环的大小等ＬＤＤ⁃ＴＤＭＡ协议
的关键参数，仿真结果验证了所提协议的有效性。 ３００ｋｍ时，无线信号传播时延约为１ｍｓ，此时完成
一次ＤＡＴＡ⁃ＡＣＫ所需要的时隙长度约为３ｍｓ，约为
链路距离３００ｍ时的３倍。 为了解决通信链路距离变长而带来的通信时延
扩展问题、提高多址接入效率，国内外研究者对多波
束定向航空中继网络［３⁃５］、传感器网络［７⁃８］、卫星通
信网络［９］、移动自组织网络［１０⁃１１］、蜂窝网［１２］等进行
了大量研究。文献［４］指出通信时延扩展导致的信
号传播时延过大，使得基于虚拟或物理载波侦听的
随机多址接入协议失效，因此，现有研究工作可被分
为２类：基于随机接入和基于ＴＤＭＡ的工作。 为了解决通信链路距离变长而带来的通信时延
扩展问题、提高多址接入效率，国内外研究者对多波
束定向航空中继网络［３⁃５］、传感器网络［７⁃８］、卫星通
信网络［９］、移动自组织网络［１０⁃１１］、蜂窝网［１２］等进行
了大量研究。文献［４］指出通信时延扩展导致的信
号传播时延过大，使得基于虚拟或物理载波侦听的
随机多址接入协议失效，因此，现有研究工作可被分
为２类：基于随机接入和基于ＴＤＭＡ的工作。 １．２ 通信时延扩展问题分析 设网络中所有节点是时间同步的，令完成一次 设网络中所有节点是时间同步的，令完成一次 设网络中所有节点是时间同步的，令完成一次 ＤＡＴＡ⁃ＡＣＫ数据收发时隙（ｄａｔａｓｌｏｔ，ＤＳ）长度为Δ。
下面，通过协议模型分析并给出影响ＤＳ的主要因 素，并提出通信时延扩展问题。 １．１ 系统模型 图１给出了定向航空中继网络ＤＡＲＮ的示
意图。 图１ 定向航空中继网络ＤＡＲＮ示意图 图１ 定向航空中继网络ＤＡＲＮ示意图 为
类：基于随机接入和基于
的工作。
在基于随机接入的工作中，文献［３］首先提出
了一种新颖的非时隙、无协调的类ＡＬＯＨＡ随机多
址接入策略，并指出该协议更为适用于那些协调开
销较大的多址接入协议。文献［４］基于文献［３］的
工作，提出了一种适用于多波束定向网络的随机接
入方法。但是，在该方法中每个节点不仅需要跟踪
邻居节点，包括节点方位和节点距离等信息，而且还
需要跟踪邻居节点的波束收发时隙状态，从而确定
波束发射的工作时间，协议实现的复杂度大；此外，
与基于ＴＤＭＡ的ＭＡＣ协议相比，随机接入的效率
较低。在基于ＴＤＭＡ的工作中，文献［７］利用传感
器与被测目标的距离检测任务和空闲时隙，给网络
中的节点分配通信时隙。文献［８］提出了一种根据
探测节点与目标节点距离远近进行自适应分配的动
态规划方法。文献［９］研究了卫星与不同浮标链路
的传播时延差异大问题，通过时隙离散分配，提出了
一种根据浮标位置信息计算时隙长度的时分多址接
入协议。文献［１０］提出了一种分布式动态时隙算
法，根据业务需求动态分配时隙个数。文献［１１］也
提出了一种基于节点位置信息的多信道多址接入协
议，在位置信息的辅助下对数据传输时隙进行调整。
文献［１２］将蜂窝网中的用户根据与基站的距离远
近划分为不同级别的邻居，通过计算干扰水平得到
用户选择所有最小化干扰的接入时隙。综上所述，
现有研究工作已初步探索了基于节点位置信息和链
路距离的ＴＤＭＡ协议设计思路，但是未对链路距离
分类，也没有优化数据传输时隙的长度，缺乏对关键
参数设计的理论指导。 图１ 定向航空中继网络ＤＡＲＮ示意图 设在ＤＡＲＮ网络中，有Ｍ个空中定向中继节
点、Ｎ个地面节点，均装置了定向天线。令空中定向
中继节点的高度为ｈ，最大通信距离为Ｒ，定向波束
的半功率角为θ；设定向中继飞机在指定高度（数公
里）做盘旋运动（移动速度１００余公里每小时）；相
对于地面节点的部署区域或通信距离（数百公里）
来说，在单位时帧（数秒）内，由于飞机盘旋半径（数
公里）较小，可被近似认为是静止的［６］。同理，由于
定向中继飞机的高度一般约为数公里，与通信距离
相比较小，因此，定向中继网络的三维网络拓扑可被
降维为二维的网络拓扑。令Ｎ个地面节点以泊松点
过程的方式随机分布，被部署在圆形区域内；设Ｍ＝
１且位于该圆形区域圆心的正上方。 （１） Δ ＝２ＴＰ＋ＴＡ＋ｍδ ＋ｎδ。
（１） Δ ＝２ＴＰ＋ＴＡ＋ｍδ ＋ｎδ。 从（１）式可以看出：①在给定所需要传输的数
据情况下，ＤＡＴＡ占用的基本时隙个数ｍ和ＡＣＫ占
用的基本时隙ｎ，可被认为是确定值；②天线收发转
换时延ＴＡ，主要是由天线的硬件实现方式等因素决
定的（一般只有数微秒），可被忽略；③无线信号传
播时延ＴＰ则主要是由收发节点之间的距离（即无线
通信链路的距离）决定的；而在ＤＡＲＮ网络中，定向
中继飞机到地面节点的链路距离差异很大，最小为
定向中继飞机的高度ｈ（数公里），最大为定向中继
飞机的通信距离Ｒ（数百公里），因此ＴＰ不可被忽
略。综上所述，公式（１）可被近似为 图３给出了对源节点Ｓ的通信区域Ａ进行分环
的示意图；此时Ｋ＝３，可以看出源节点Ｓ的通信区
域Ａ，被划分为一个以ｄ１为半径的圆形，和３个分别
以ｄ２，ｄ３和Ｒ为半径的环形。 图３ 源节点Ｓ的通信区域分环示意图 Δ ≈２ＴＰ＋Ｃ＝２ｄＳＤ
ｃ
＋Ｃ
（２） （２） 式中，Ｃ＝（ｍ＋ｎ）δ 为常数。 式中，Ｃ＝（ｍ＋ｎ）δ 为常数。 当采用传统的ＴＤＭＡ方法（ｔｒａｄｉｔｉｏｎａｌＴＤＭＡ
ｓｃｈｅｍｅ，ＴＴＳ）设计时隙长度时，则应考虑定向中继
飞机的最大通信距离Ｒ，此时无线信号传播时延为
ＴＴＴＳ
Ｐ
＝Ｒ
ｃ，ＤＡＴＡ⁃ＡＣＫ的时隙长度为 ＴＴＴＳ
Ｐ
＝Ｒ
ｃ，ＤＡＴＡ⁃ＡＣＫ的时隙长度为 图３ 源节点Ｓ的通信区域分环示意图 ΔＴＴＳ＝２Ｒ
ｃ＋Ｃ ΔＴＴＳ＝２Ｒ
ｃ＋Ｃ ２．２ 协议流程 在ＬＤＤ⁃ＴＤＭＡ中，基本时隙是最小的时长粒
度，多个基本时隙构成一个ＤＳ。与传统的ＴＤＭＡ协
议不同，在ＴＴＳ中所有的ＤＳ的长度是相同的，而在
ＬＤＤ⁃ＴＤＭＡ中，ＤＳ的长度被划分为Ｋ＋１档，每一档
的ＤＳ长度根据对应的收发节点之间的距离确定。
ＬＤＤ⁃ＴＤＭＡ协议的工作流程包括３个阶段，分
别为邻节点发现阶段、源节点与目的节点通信距离
的测量阶段、数据传输阶段。其中前２个阶段是为 在ＬＤＤ⁃ＴＤＭＡ中，基本时隙是最小的时长粒
度，多个基本时隙构成一个ＤＳ。与传统的ＴＤＭＡ协
议不同，在ＴＴＳ中所有的ＤＳ的长度是相同的，而在
ＬＤＤ⁃ＴＤＭＡ中，ＤＳ的长度被划分为Ｋ＋１档，每一档
的ＤＳ长度根据对应的收发节点之间的距离确定。 在本文提出的ＬＤＤ⁃ＴＤＭＡ中，我们将降低所有
链路对应的距离Ｒ，从而减小数据收发时隙的长度。 素，并提出通信时延扩展问题。 设源节点Ｓ以ＤＡＴＡ⁃ＡＣＫ模式发送ＤＡＴＡ给
目的节点Ｄ，节点Ｄ收到ＤＡＴＡ之后立即应答
ＡＣＫ。令δ 为基本时隙的时长，设ＤＡＴＡ占用ｍ个
基本时隙，ＡＣＫ占用ｎ个基本时隙；无线信号传播时
延为ＴＰ，天线收发转换时延为ＴＡ，如图２所示。其 设源节点Ｓ以ＤＡＴＡ⁃ＡＣＫ模式发送ＤＡＴＡ给
目的节点Ｄ，节点Ｄ收到ＤＡＴＡ之后立即应答
ＡＣＫ。令δ 为基本时隙的时长，设ＤＡＴＡ占用ｍ个
基本时隙，ＡＣＫ占用ｎ个基本时隙；无线信号传播时
延为ＴＰ，天线收发转换时延为ＴＡ，如图２所示。其 因此，本文提出了一种基于通信链路距离进行
分环的时分多址接入协议（ｌｉｎｋｄｉｓｔａｎｃｅｄｉｖｉｓｉｏｎ
ｂａｓｅｄｔｉｍｅｄｉｖｉｓｉｏｎｍｕｌｔｉｐｌｅａｃｃｅｓｓｐｒｏｔｏｃｏｌ，ＬＤＤ⁃ＴＤ⁃
ＭＡ）。其基本思想是将通信距离相近的链路归为 闫中江，等：定向航空中继网络中一种基于链路距离分环的多址接入协议 第１期 ９
４
１ 图２ ＤＡＴＡ⁃ＡＣＫ模式数据传输示意图 ２．１ 基本思想 ＬＤＤ⁃ＴＤＭＡ的基本思想如下所述：首先，令圆
心为数据源节点Ｓ（即定向中继飞机），半径为定向
天线的通信范围Ｒ，确定该源节点的圆形通信区域
Ａ；然后，仍以节点Ｓ为圆心，再令半径为ｄ１，ｄ２，…，
ｄＫ，０＝ｄ０＜ｄ１＜ｄ２＜… ＜ｄＫ＜ｄＫ＋１＝Ｒ，则可将区
域Ａ划分为１个半径为ｄ１的圆形区域，和半径分别
为ｄ２，…，ｄＫ，ｄＫ＋１的Ｋ个环形区域；接着，计算并得
到链路距离ｄ１，ｄ２，…，ｄＫ，ｄＫ＋１所对应的无线信号传
播时延Ｔ１
Ｐ，Ｔ２
Ｐ，…，ＴＫ
Ｐ，ＴＫ＋１
Ｐ
，其中Ｔｋ
Ｐ＝ｄｋ／ｃ，１≤ｋ≤Ｋ
＋１；将Ｔ１
Ｐ，Ｔ２
Ｐ，…，ＴＫ
Ｐ代入公式（１）可得该传播时延
对应的ＤＳ时间长度Ｍ１δ，Ｍ２δ，…，ＭＫδ，ＭＫ＋１δ，其中
Ｍｋδ，１≤ｋ≤Ｋ＋１，对应链路距离大于ｄｋ－１而小于
ｄｋ的环形（或圆形，当ｋ＝１时）区域；最后，如果目
的节点Ｄ位于第ｋ个环形，则源节点Ｓ使用时间长度
为Ｍｋδ 的时隙与目的节点Ｄ进行数据传输。 图２ ＤＡＴＡ⁃ＡＣＫ模式数据传输示意图 中传播时延为ＴＰ＝ｄＳＤ
ｃ，ｄＳＤ为Ｓ与Ｄ之间的距离，ｃ
为无线信号传播的速度（近似为光速）。因此，可得
ＤＳ的时长 中传播时延为ＴＰ＝ｄＳＤ
ｃ，ｄＳＤ为Ｓ与Ｄ之间的距离，ｃ
为无线信号传播的速度（近似为光速）。因此，可得
ＤＳ的时长 ２ 基于链路距离分环的多址接入协议 ＬＤＤ⁃ＴＤＭＡ协议的工作流程包括３个阶段，分
别为邻节点发现阶段、源节点与目的节点通信距离
的测量阶段、数据传输阶段。其中前２个阶段是为 针对通信时延扩展问题，在传统的ＴＤＭＡ多址
接入协议基础上，本文提出一种基于链路距离分环
的多址接入协议ＬＤＤ⁃ＴＤＭＡ。 西 北 工 业 大 学 学 报 第３８卷 ０
５
１ 路距离相当于通信距离Ｒ。而在ＬＤＤ⁃ＴＤＭＡ中，由
于采用了分环的思想，当源与目的节点通信时，所采
用的ＤＳ长度为该目的节点所位于的环形的半径
（环形的圆心为源节点所处于的位置）。因此，从平
均ＤＳ长度的角度看，每个目的节点与源节点的链
路距离相当于平均分环距离，即每个环形内的节点
个数乘以对应环形半径的平均值。当该平均环形距
离最小时分环增益最大，相应的ＬＤＤ⁃ＴＤＭＡ协议的
效率也最大。 数据传输阶段服务的，它们所占用的时间，将作为
ＬＤＤ⁃ＴＤＭＡ协议的开销；而在数据传输阶段，定向
中继节点则可以根据与每个邻节点的距离，采用对
应分环的ＤＳ长度与该邻节点进行数据传输。 数据传输阶段服务的，它们所占用的时间，将作为
ＬＤＤ⁃ＴＤＭＡ协议的开销；而在数据传输阶段，定向
中继节点则可以根据与每个邻节点的距离，采用对
应分环的ＤＳ长度与该邻节点进行数据传输。 下面，给出ＬＤＤ⁃ＴＤＭＡ协议在不同阶段所完成
的功能，及其具体的工作流程。 在邻节点发现阶段，源节点可采用轮询扫描、令
牌传递等算法发现邻节点，具体的邻节点发现算法
不在本文的研究范围内。之后，源节点根据其缓存
数据的目的地址，与目的节点进行测距，得到通信链
路的距离；该阶段也可与邻节点发现阶段相结合，在
邻节点发现的同时，通过采用基于信号到达时间、基
于信号能量等的方法测量与目的节点的距离。最
后，在数据传输阶段，根据定向中继节点与邻节点的
距离，确定所采用的ＤＳ的长度，成功进行握手之
后，源节点将数据发送给目的节点。 因此，本证明的基本思路是：在给定环形个数Ｋ
的情况下，首先求出由于分环所带来的目的节点与
源节点的平均分环距离；然后，通过最小化该平均分
环距离，推导求得分环参数的最优值。下面给出详
细的证明过程。 由于Ｎ个地面节点在半径为Ｒ的圆形内被随
机均匀部署，因此每个环形区域内的节点个数为 Ｎ１＝Ａ１
ＡＮ＝πｄ２
１
πＲ２Ｎ＝ｄ２
１
Ｒ２Ｎ
Ｎ２＝Ａ２
ＡＮ＝πｄ２
２－πｄ２
１
πＲ２
Ｎ＝ｄ２
２－ｄ２
１
Ｒ２
Ｎ
︙
Ｎｋ＝Ａｋ
ＡＮ＝πｄ２
ｋ－πｄ２
ｋ－１
πＲ２
Ｎ＝ｄ２
ｋ－ｄ２
ｋ－１
Ｒ２
Ｎ
︙
ＮＫ＋１＝ＡＫ＋１
ＡＮ＝πＲ２－πｄ２
Ｋ
πＲ２
Ｎ＝Ｒ２－ｄ２
Ｋ
Ｒ２
Ｎ
ì
î
í
ï
ï
ï
ï
ï
ï
ïï
ï
ï
ï
ï
ï
ï
ï
（４） ３ 理论性能分析 本节对ＬＤＤ⁃ＴＤＭＡ协议进行建模，并推导出在
给定Ｋ的情况下，最优的分环大小以及最优的ＬＤＤ⁃
ＴＤＭＡ协议的效率，即最大的分环增益，对ＬＤＤ⁃
ＴＤＭＡ协议中关键参数的设定提供理论指导。 （４） 定理１ 设定向中继飞机为源节点Ｓ位于二维
坐标系的原点处，Ｎ个地面节点被随机均匀部署在
以Ｓ为圆心Ｒ为半径的圆内。令分环半径为ｄ１，ｄ２，
…，ｄＫ，０＝ｄ０＜ｄ１＜ｄ２＜… ＜ｄＫ＜ｄＫ＋１＝Ｒ，对地
面节点的被部署区域进行划分，得到１个以ｄ１为半
径的圆形和Ｋ个环形。当分环参数为 设ｄ′ｋ为第ｋ个环形内的节点Ｄ与源节点Ｓ的距 离，则ｄｋ－１≤ｄ′ｋ≤ｄｋ。由于在ＬＤＤ⁃ＴＤＭＡ中，每个
环形所对应的ＤＳ长度，是按照该环形内距离源节点
Ｓ最远的邻节点计算的，因此，源节点Ｓ与其周边所
有邻节点的平均距离可被计算为 ｄ１＝２
３ｄ２＝ａ１ｄ２
ｄ２＝
２
３－ａ２
１
ｘ３＝ａ２ｄ３
︙
ｄｋ＝
２
３－ａ２
ｋ－１
ｄｋ＋１＝ａｋｄｋ＋１
︙
ｄＫ＝
２
３－ａ２
Ｋ－１
Ｒ＝ａＫＲ
ì
î
í
ï
ï
ï
ï
ï
ïï
ï
ï
ï
ï
ï
ïï ｆ（ｄ１，ｄ２，…，ｄＫ）＝
Ｎ１× ｄ１＋Ｎ２× ｄ２＋… ＋Ｎｋ× ｄｋ＋… ＋ＮＫ＋１× Ｒ
Ｎ
＝
ｄ２
１
Ｒ２ｄ１＋ｄ２
２－ｄ２
１
Ｒ２
ｄ２＋… ＋ｄ２
ｋ－ｄ２
ｋ－１
Ｒ２
ｄｉ＋… ＋
Ｒ２－ｄ２
Ｋ
Ｒ２
Ｒ＝１
Ｒ２∑
Ｋ＋１
ｋ＝１
（ｄ３
ｋ－ｄｋｄ２
ｋ－１）
（５） ｆ（ｄ１，ｄ２，…，ｄＫ）＝
Ｎ１× ｄ１＋Ｎ２× ｄ２＋… ＋Ｎｋ× ｄｋ＋… ＋ＮＫ＋１× Ｒ
Ｎ
＝ （３） Ｎ
ｄ２
１
Ｒ２ｄ１＋ｄ２
２－ｄ２
１
Ｒ２
ｄ２＋… ＋ｄ２
ｋ－ｄ２
ｋ－１
Ｒ２
ｄｉ＋… ＋
Ｒ２－ｄ２
Ｋ
Ｒ２
Ｒ＝１
Ｒ２∑
Ｋ＋１
ｋ＝１
（ｄ３
ｋ－ｄｋｄ２
ｋ－１）
（５） ｄ２
１
Ｒ２ｄ１＋ｄ２
２－ｄ２
１
Ｒ２
ｄ２＋… ＋ｄ２
ｋ－ｄ２
ｋ－１
Ｒ２
ｄｉ＋… ＋ （５） 为了计算最优的环形参数ｄ１，ｄ２，…，ｄＫ，首先求
ｆ（ｄ１，ｄ２，…，ｄＫ）对ｄ１，ｄ２，…，ｄＫ的偏导，可得 为了计算最优的环形参数ｄ１，ｄ２，…，ｄＫ，首先求
ｄ１，ｄ２，…，ｄＫ）对ｄ１，ｄ２，…，ｄＫ的偏导，可得 时，ＬＤＤ⁃ＴＤＭＡ的平均时隙长度达到最小值。 证明：在传统的ＴＤＭＡ协议中，所有节点的ＤＳ
长度均按照距离最远的节点进行计算。因此，从每
个ＤＳ长度的角度看，每个目的节点与源节点的链 证明：在传统的ＴＤＭＡ协议中，所有节点的ＤＳ 将表１中不同分环个数对应的平均距离ｆ（ｄ１， 分
平
ｆ（
１，
ｄ２，…，ｄＫ）代入公式（２），可得到对应ＬＤＤ⁃ＴＤＭＡ
协议的平均时隙长度
ΔＬＤＤ⁃ＴＤＭＡ
Ｋ
＝２ｆ（ｄ１，ｄ２，…，ｄＫ）
ｃ
＋Ｃ 从而可计算得到ＬＤＤ⁃ＴＤＭＡ协议相对于ＴＴＳ协议
的增益 闫中江，等：定向航空中继网络中一种基于链路距离分环的多址接入协议 第１期 １
５
１ ｆ′ｄ１（ｄ１，ｄ２，…，ｄＫ）＝１
Ｒ２（３ｄ２
１－２ｄ２ｄ１）
ｆ′ｄ２（ｄ１，ｄ２，…，ｄＫ）＝１
Ｒ２（３ｄ２
２－ｄ２
１－２ｄ２ｄ３）
︙
ｆ′ｄｋ（ｄ１，ｄ２，…，ｄＫ）＝１
Ｒ２（３ｄ２
ｋ－ｄ２
ｋ－１－２ｄｋｄｋ＋１）
︙
ｆ′ｄＫ（ｄ１，ｄ２，…，ｄＫ）＝１
Ｒ２（３ｄ２
Ｋ－ｄ２
Ｋ－１－２ＲｄＫ）
ì
î
í
ï
ï
ï
ï
ï
ïï
ï
ï
ï
ï
ï
ïï 将表１中不同分环个数对应的平均距离ｆ（ｄ１， 从而可计算得到ＬＤＤ⁃ＴＤＭＡ协议相对于ＴＴＳ协议
的增益 ηＫ＝１－
ΔＬＤＤ⁃ＴＤＭＡ
Ｋ
ΔＴＴＳ 此外，从表１可得随分环个数的增大，链路平均距离
减小。定理２给出了当分环个数Ｋ趋于∞时，链路
距离的最小值。 （６） 然后，令上述偏导函数全部等于０，求解上述联 然后，令上述偏导函数全部等于０，求解上述联 距离的最小值。 立的方程组可得公式（３）。为了证明公式（３）可使
得ｆ（ｄ１，ｄ２，…，ｄＫ）取得最小值，对其求二次偏导可
得（７）式 定理２
设在半径为Ｒ的圆形内，以泊松点过
程分布有Ｎ个地面节点；如果以每个节点到圆心的
距离作为一个环形半径，则当环形个数趋于无穷时， 定理２
设在半径为Ｒ的圆形内，以泊松点过 程分布有Ｎ个地面节点；如果以每个节点到圆心的
距离作为一个环形半径，则当环形个数趋于无穷时， ｆ″ｄ１（ｄ１，ｄ２，…，ｄＫ）＝１
Ｒ２（６ｄ１－２ｄ２）
ｆ″ｄ２（ｄ１，ｄ２，…，ｄＫ）＝１
Ｒ２（６ｄ２－２ｄ３）
︙
ｆ″ｄｋ（ｄ１，ｄ２，…，ｄＫ）＝１
Ｒ２（６ｄｋ－２ｄｋ＋１）
︙
ｆ″ｄＫ（ｄ１，ｄ２，…，ｄＫ）＝１
Ｒ２（６ｄＫ－２Ｒ）
ì
î
í
ï
ï
ï
ï
ï
ïï
ï
ï
ï
ï
ï
ïï
（７） 即Ｎ→∞时，所有节点的平均距离为２
３Ｒ。 即Ｎ→∞时，所有节点的平均距离为２
３Ｒ。 即Ｎ→∞时，所有节点的平均距离为２
３Ｒ。 证明
设节点部署区域的圆心位置为极坐标
的原点，令Ｎ个节点的位置为（Ｒ１，θ１），（Ｒ２，θ２），
…，（ＲＮ，θＮ）。由于Ｎ个节点以泊松点过程被部署，
因此可知θ１，θ２，…，θｎ与Ｒ１，Ｒ２，…ＲＮ独立，且服从
Ｕ［０，２π］的均匀分布；其中，Ｒ１，Ｒ２，…ＲＮ的密度函 （７） 数为ｆ（ｚ）＝２ｚ
Ｒ２，ｚ∈［０，Ｒ］。 对ｆ（ｚ）求积分，可得Ｎ个节点到源节点Ｓ距离
的期望为Ｅ（Ｒｉ）＝∫
Ｒ
０ｚｆ（ｚ）ｄｚ＝∫
Ｒ
０ｚ· ２ｚ
Ｒ２ｄｚ＝２Ｒ
３。 将（３）式代入（７）式可知ｆ″ｄｋ（·）＞０，１≤ｋ≤ 将（３）式代入（７）式可知ｆ″ｄｋ（·）＞０，１≤ｋ≤
Ｋ，因此，公式（３）给出的环形参数，可最小化平均距
离ｆ（ｄ１，ｄ２，…，ｄＫ），从而最大化ＬＤＤ⁃ＴＤＭＡ协议的
效率。证毕。 证毕。 证毕。 基于定理２，可知当分环个数Ｋ趋于∞时，
ＬＤＤ⁃ＴＤＭＡ协议中ＤＳ的平均长度为 基于定理１给出的公式（３），可计算得到当源
节点Ｓ所覆盖的圆形区域被分环时，每个环形的最
优半径；之后，基于式（５）可计算出分环后的链路平
均距离。表１给出了当圆形区域被划分为２～５个
环形时，每个环形的半径和分环后的平均链路半径。 ＬＤＤ⁃ＴＤＭＡ协议中ＤＳ的平均长度为
ΔＬＤＤ⁃ＴＤＭＡ＝４Ｒ
３ｃ＋Ｃ
（８） （８） 当Ｃ为常数时，即源节点Ｓ的ＤＡＴＡ数据传输 时间与目的节点Ｄ的ＡＣＫ传输时间之和为常数时，
ＬＤＤ⁃ＴＤＭＡ协议与ＴＴＳ协议相比，在每个数据收发
时隙上的平均增益为 表１ 理论计算得到的环形半径
分环
个数
平均
距离ｆ
环形１
半径
环形２
半径
环形３
半径
环形４
半径
环形５
半径
２
０．８５Ｒ
０．６７Ｒ
Ｒ
３
０．８０Ｒ
０．５２Ｒ
０．７８Ｒ
Ｒ
４
０．７７Ｒ
０．４４Ｒ
０．６６Ｒ
０．８４Ｒ
Ｒ
５
０．７５Ｒ
０．３９Ｒ
０．５７Ｒ
０．７３Ｒ
０．８７Ｒ
Ｒ η ＝１－
ΔＬＤＤ⁃ＴＤＭＡ
ΔＴＴＳ
＝
２Ｒ
６Ｒ＋３Ｃ× ｃ
（９） （９） 从（９）式可知，在给定Ｒ和ｃ时，随Ｃ的增大每个数
据收发时隙上的平均增益越来越小。 西 北 工 业 大 学 学 报 第３８卷 ２
５
１ 表２ ＬＤＤ⁃ＴＤＭＡ协议与ＴＴＳ相比的理论性能增益
（Ｒ＝２００ｋｍ，Ｃ＝０．００１ｓ）
分环个数Ｋ
平均距离ｆ／ｋｍ
理论增益ηＫ／％
２
１７０．３７
８．４７
３
１５９．１７
１１．６７
４
１５３．２２
１３．３７
５
１４９．５１
１４．４３
∞
１３３．３３
２１．０５ 表２ ＬＤＤ⁃ＴＤＭＡ协议与ＴＴＳ相比的理论性能增益
（Ｒ＝２００ｋｍ，Ｃ＝０．００１ｓ） １０Ｂ，则ＤＡＴＡ数据传输时间与目的节点的ＡＣＫ传
输时间之和为１× １０
－３ｓ，即Ｃ＝１ｍｓ。
图４给出了不同节点密度（λ ＝１０
－８和λ ＝
１０
－９）情况下，在给定Ｒ＝２００ｋｍ和Ｃ＝０．００１ｓ时，
不同分环个数对应的平均链路距离以及最小分环平
均链路距离，其中“不分环距离” 即为传统的ＴＴＳ协
议。从图４可得，仿真结果与理论分析结果（参见表
２）相吻合，验证了理论分析的正确性；此外，与理论
分析一致，网络中的节点密度λ 和节点个数Ｎ对分
环平均链路距离几乎没有影响；随着分环个数的增
加，链路的平均距离逐渐减小，并趋近于理论极
限。 １０Ｂ，则ＤＡＴＡ数据传输时间与目的节点的ＡＣＫ传
输时间之和为１× １０
－３ｓ，即Ｃ＝１ｍｓ。 １０Ｂ，则ＤＡＴＡ数据传输时间与目的节点的ＡＣＫ传
输时间之和为１× １０
－３ｓ，即Ｃ＝１ｍｓ。
图４给出了不同节点密度（λ ＝１０
－８和λ ＝
１０
－９）情况下，在给定Ｒ＝２００ｋｍ和Ｃ＝０．００１ｓ时，
不同分环个数对应的平均链路距离以及最小分环平
均链路距离，其中“不分环距离” 即为传统的ＴＴＳ协
议。从图４可得，仿真结果与理论分析结果（参见表
２）相吻合，验证了理论分析的正确性；此外，与理论
分析一致，网络中的节点密度λ 和节点个数Ｎ对分
环平均链路距离几乎没有影响；随着分环个数的增
加，链路的平均距离逐渐减小，并趋近于理论极
限。 由公式（９）可求出分环所带来的理论性能增 由公式（９）可求出分环所带来的理论性能增
益，如表２所示。表２给出了当Ｒ＝２００ｋｍ，Ｃ＝
０．００１ｓ时，分环个数分别为２～５，以及分环个数趋
于∞时的分环增益。从表２可知，随分环个数的增
大，分环所带来的理论增益增大。当Ｒ＝２００ｋｍ，Ｃ
＝０．００１ｓ时，分环个数为４时所带来的分环增益，与
分环个数为５时所带来的分环增益已经接近，表明
进一步的增加分环个数，所带来的分环增益增量却
变小。 图５给出了节点密度λ ＝１０
－９和Ｒ＝２００ｋｍ，不
同数据传输时间Ｃ（Ｃ＝０．０００５ｓ，Ｃ＝０．００１ｓ，Ｃ＝
０．００１５ｓ）下，不同分环个数所带来的ＬＤＤ⁃ＴＤＭＡ
协议与ＴＴＳ相比的性能增益曲线。从图５可知，随
着分环个数的增大，ＬＤＤ⁃ＴＤＭＡ协议所带来的分环
增益也越来越大，随着Ｃ的增大分环增益越来越
小，该结论与公式（９）得出的结论一致。从图５可
知，当分环个数达到４之后，ＬＤＤ⁃ＴＤＭＡ协议的分环
增益可达１３．３７％（Ｃ＝０．００１ｓ时），且随着分环个数
的增加，增益的增量部分变小；该结论意味着在实际
的ＬＤＤ⁃ＴＤＭＡ协议中，尽管分环个数越多协议增益
越大，但一般只需要将网络划分为３～４个分环即
可，从而降低了协议实现的复杂度。 闫中江，等：定向航空中继网络中一种基于链路距离分环的多址接入协议 第１期 第１期 ３
５
１ 距离，可求出不同分环个数对应的ＬＤＤ⁃ＴＤＭＡ协议
的理论分环增益。仿真结果验证了理论模型分析的
正确性，结果表明当最大通信半径Ｒ为２００ｋｍ，数
据传输时延为０．００１ｓ时，在ＬＤＤ⁃ＴＤＭＡ协议的实
际应用中，只需要将节点的通信区域划分为４个分
环即可获得约１３．３７％增益。 ５ 结 论 针对定向航空中继网络中，定向通信链路长、无
线信号传播时延与数据传输时延相近，所引起的通
信时延扩展问题，本文提出了一种基于通信链路距
离分环的时分多址接入协议ＬＤＤ⁃ＴＤＭＡ。理论推
导并给出了不同分环个数对应的最优环形半径的闭
合表达式，通过计算不同分环个数下平均环形链路 未来研究中将考虑如何在复杂的三维网络拓扑
中，解决本文所提出的通信时延扩展问题。 参考文献： ４ 仿真性能分析 在ＭＡＴＬＡＢ中搭建了仿真平台，并与传统的
ＴＤＭＡ协议ＴＴＳ进行了仿真对比，验证了分环算法
所带来的性能增益。设网络中共有Ｎ个节点，以密
度为λ 的泊松点过程，被部署在半径Ｒ＝２００ｋｍ的
圆形内；定向飞机中继节点被部署在圆心，最大通信
半径为Ｒ。
此时，无线信号的往返传播延迟为
２× ２００ｋｍ
３× １０８ｍ／ｓ
＝１．３３ｍｓ。设ＤＡＴＡ与ＡＣＫ的数据传
输速率为２Ｍｂ／ｓ，ＤＡＴＡ长度为２４０Ｂ，ＡＣＫ长度为 图６给出了节点密度λ ＝１０
－９和Ｃ＝０．００１ｍｓ，
不同分环个数（３，４，５）下，不同节点最大传输距离Ｒ
下的链路平均距离。从图６可知，随链路距离Ｒ增
大，给定分环所对应的链路平均距离逐渐增大；且分
环个数越多链路的平均距离越小。 输速率为２Ｍｂ／ｓ，ＤＡＴＡ长度为２４０Ｂ，ＡＣＫ长度为 图５ 不同数据传输时间Ｃ下 图６ 不同节点最大传输距离Ｒ
ＬＤＤ⁃ＴＤＭＡ协议增益
对平均链路距离的影响 图５ 不同数据传输时间Ｃ下 图６ 不同节点最大传输距离Ｒ
ＬＤＤ⁃ＴＤＭＡ协议增益
对平均链路距离的影响 图４ 不同分环个数下平均距 图５ 不同数据传输时间Ｃ下 图６ 不同节点最大传输距离Ｒ
离与理论最小平均距离
ＬＤＤ⁃ＴＤＭＡ协议增益
对平均链路距离的影响 图４ 不同分环个数下平均距
离与理论最小平均距离 图５ 不同数据传输时间Ｃ下
ＬＤＤ⁃ＴＤＭＡ协议增益 图４ 不同分环个数下平均距
离与理论最小平均距离 不同节点最大传输距离Ｒ
对平均链路距离的影响 闫中江，等：定向航空中继网络中一种基于链路距离分环的多址接入协议 参考文献： ［１］ ＭＡＣＬＥＯＤＲＢ，ＭＡＲＧＥＴＴＳＡ．ＮｅｔｗｏｒｋｅｄＡｉｒｂｏｒｎｅＣｏｍｍｕｎｉｃａｔｉｏｎｓＵｓｉｎｇＡｄａｐｔｉｖｅＭｕｌｔｉ⁃ＢｅａｍＤｉｒｅｃｔｉｏｎａｌＬｉｎｋｓ［Ｃ］∥２０１６
ＩＥＥＥＡｅｒｏｓｐａｃｅＣｏｎｆｅｒｅｎｃｅ，ＢｉｇＳｋｙ，ＭＴ，ＵＳＡ，２０１６：１⁃７ ［２］ ＳＨＡＫＥＴ，ＡＭＩＮＲ．ＭａｘｉｍｉｚｉｎｇＩｎｔｅｒｃｏｎｎｅｃｔｅｄｎｅｓｓａｎｄＡｖａｉｌａｂｉｌｉｔｙｉｎＤｉｒｅｃｔｉｏｎａｌＡｉｒｂｏｒｎｅＲａｎｇｅＥｘｔｅｎｓｉｏｎＮｅｔｗｏｒｋｓ［Ｃ］∥
２０１７ＩＥＥＥＭｉｌｉｔａｒｙＣｏｍｍｕｎｉｃａｔｉｏｎｓＣｏｎｆｅｒｅｎｃｅ，Ｂａｌｔｉｍｏｒｅ，ＭＤ，ＵＳＡ，２０１７：２７３⁃２７８ ［３］ ＫＵＰＥＲＭＡＮＧ，ＭＡＲＧＯＬＩＥＳＲ，ＪＯＮＥＳＮＭ，ｅｔａｌ．ＵｎｃｏｏｒｄｉｎａｔｅｄＭＡＣｆｏｒＡｄａｐｔｉｖｅＭｕｌｔｉ⁃ＢｅａｍＤｉｒｅｃｔｉｏｎａｌＮｅｔｗｏｒｋｓ：Ａ⁃
ｎａｌｙｓｉｓａｎｄＥｖａｌｕａｔｉｏｎ［Ｃ］∥２０１６２５ｔｈＩｎｔｅｒｎａｔｉｏｎａｌＣｏｎｆｅｒｅｎｃｅｏｎＣｏｍｐｕｔｅｒＣｏｍｍｕｎｉｃａｔｉｏｎａｎｄＮｅｔｗｏｒｋｓ，Ｗａｉｋｏｌｏａ，ＨＩ，
ＵＳＡ，２０１６：１⁃１０ ＯＬＩＥＳＲ，ＪＯＮＥＳＮＭ，ｅｔａｌ．ＵｎｃｏｏｒｄｉｎａｔｅｄＭＡＣｆｏｒＡｄａｐｔｉｖｅＭｕｌｔｉ⁃ＢｅａｍＤｉｒｅｃｔｉｏｎａｌＮｅｔｗｏｒｋｓ：Ａ⁃ ｎａｌｙｓｉｓａｎｄＥｖａｌｕａｔｉｏｎ［Ｃ］∥２０１６２５ｔｈＩｎｔｅｒｎａｔｉｏｎａｌＣｏｎｆｅｒｅｎｃｅｏｎＣｏｍｐｕｔｅｒＣｏｍｍｕｎｉｃａｔｉｏｎａｎｄＮｅｔｗｏｒｋｓ，Ｗａｉｋｏｌｏａ，ＨＩ，
ＵＳＡ，２０１６：１⁃１０ ［４］ ＰＲＯＵＬＸＢ，ＫＵＰＥＲＭＡＮＧ，ＪＯＮＥＳＮＭ，ｅｔａｌ．ＳｉｍｕｌａｔｉｏｎａｎｄＭｏｄｅｌｉｎｇｏｆａＮｅｗＭｅｄｉｕｍＡｃｃｅｓｓＣｏｎｔｒｏｌＳｃｈｅｍｅｆｏｒＭｕｌｔｉ⁃
ＢｅａｍＤｉｒｅｃｔｉｏｎａｌＮｅｔｗｏｒｋｉｎｇ［Ｃ］∥２０１７ＩＥＥＥＡｅｒｏｓｐａｃｅＣｏｎｆｅｒｅｎｃｅ，ＢｉｇＳｋｙ，ＭＴ，ＵＳＡ，２０１７：１⁃９ ［４］ ＰＲＯＵＬＸＢ，ＫＵＰＥＲＭＡＮＧ，ＪＯＮＥＳＮＭ，ｅｔａｌ．ＳｉｍｕｌａｔｉｏｎａｎｄＭｏｄｅｌｉｎｇｏｆａＮｅｗＭｅｄｉｕｍＡｃｃｅｓｓＣｏｎｔｒｏｌＳｃｈｅｍｅｆｏｒＭｕｌｔｉ⁃
Ｂ
Ｄｉ
ｉ
ｌＮ
ｋｉ
［Ｃ］∥２０１７ＩＥＥＥＡ
Ｃ
ｆ
ＢｉＳｋ
ＭＴ
ＵＳＡ
２０１７１９ ＤｉｒｅｃｔｉｏｎａｌＮｅｔｗｏｒｋｉｎｇ［Ｃ］∥２０１７ＩＥＥＥＡｅｒｏｓｐａｃｅＣｏｎｆｅｒｅｎｃｅ，ＢｉｇＳｋｙ，ＭＴ，ＵＳＡ，２０１７：１⁃９ ［５］ ＳＨＡＫＥＴ．ＴｏｐｏｌｏｇｙＤｅｓｉｇｎｆｏｒＤｉｒｅｃｔｉｏｎａｌＲａｎｇｅＥｘｔｅｎｓｉｏｎＮｅｔｗｏｒｋｓｗｉｔｈＡｎｔｅｎｎａＢｌｏｃｋａｇｅ［Ｃ］∥２０１７ＩＥＥＥＷｉｒｅｌｅｓｓＣｏｍ⁃
ｍｕｎｉｃａｔｉｏｎｓａｎｄＮｅｔｗｏｒｋｉｎｇＣｏｎｆｅｒｅｎｃｅ，ＳａｎＦｒａｎｃｉｓｃｏ，ＣＡ，ＵＳＡ，２０１７：１⁃６ 帆，等．一种面向航空集群的无人机中继网络部署策略［Ｊ］．计算机工程，２０１８，４４（５）：１０７⁃１１２ ［６］ 刘创，吕娜，陈柯帆，等．一种面向航空集群的无人机中继网络部署策略［Ｊ］．计算机工程，２０１８，４４（５）：１０７⁃１１２
ＬＩＵＣｈｕａｎｇ，ＬＹＵＮａ，ＣＨＥＮＫｅｆａｎ，ｅｔａｌ．ＡｎＵｎｍａｎｎｅｄＡｅｒｉａｌＶｅｈｉｃｌｅＲｅｌａｙＮｅｔｗｏｒｋＤｅｐｌｏｙｍｅｎｔＳｔｒａｔｅｇｙｆｏｒＡｅｒｏｎａｕｔｉｃ
Ｓｗａｒｍ［Ｊ］．ＣｏｍｐｕｔｅｒＥｎｇｉｎｅｅｒｉｎｇ，２０１８，４４（５）：１０７⁃１１２（ｉｎＣｈｉｎｅｓｅ） ＬＩＵＣｈｕａｎｇ，ＬＹＵＮａ，ＣＨＥＮＫｅｆａｎ，ｅｔａｌ．ＡｎＵｎｍａｎｎｅｄＡｅｒｉａｌＶｅｈｉｃｌｅＲｅｌａｙＮｅｔｗｏｒｋＤｅｐｌｏｙｍｅｎｔＳｔｒａｔｅｇｙｆｏｒＡｅｒｏｎａｕｔｉｃ
Ｓｗａｒｍ［Ｊ］．ＣｏｍｐｕｔｅｒＥｎｇｉｎｅｅｒｉｎｇ，２０１８，４４（５）：１０７⁃１１２（ｉｎＣｈｉｎｅｓｅ） ［７］ 周杰，刘元安，吴帆，等．基于混沌并行遗传算法的多目标无线传感器网络跨层资源分配［Ｊ］．物理学报，２０１１，６０（９）：
１４８⁃１５７ ［７］ 周杰，刘元安，吴帆，等．基于混沌并行遗传算法的多目标无线传感器网络跨层资源分配［Ｊ］．物理学报，２０１１，６０（９）：
１４８⁃１５７
ＺＨＯＵＪｉｅ，ＬＩＵＹｕａｎａｎ，ＷＵＦａｎ，ｅｔａｌ．ＡｌｌｏｃａｔｉｏｎｏｆＭｕｌｔｉ⁃ＯｂｊｅｃｔｉｖｅＣｒｏｓｓ⁃ＬａｙｅｒＷｉｒｅｌｅｓｓＳｅｎｓｏｒＮｅｔｗｏｒｋＲｅｓｏｕｒｃｅＢａｓｅｄｏｎ
ＣｈａｏｔｉｃＰａｒａｌｌｅｌＧｅｎｅｔｉｃＡｌｇｏｒｉｔｈｍ［Ｊ］．ＡｃｔａＰｈｙｓＳｉｎ，２０１１，６０（９）：１４８⁃１５７（ｉｎＣｈｉｎｅｓｅ） １４８⁃１５７
ＺＨＯＵＪｉｅ，ＬＩＵＹｕａｎａｎ，ＷＵＦａｎ，ｅｔａｌ．ＡｌｌｏｃａｔｉｏｎｏｆＭｕｌｔｉ⁃ＯｂｊｅｃｔｉｖｅＣｒｏｓｓ⁃ＬａｙｅｒＷｉｒｅｌｅｓｓＳｅｎｓｏｒＮｅｔｗｏｒｋＲｅｓｏｕｒｃｅＢａｓｅｄｏｎ
ＣｈａｏｔｉｃＰａｒａｌｌｅｌＧｅｎｅｔｉｃＡｌｇｏｒｉｔｈｍ［Ｊ］．ＡｃｔａＰｈｙｓＳｉｎ，２０１１，６０（９）：１４８⁃１５７（ｉｎＣｈｉｎｅｓｅ） ＣｈａｏｔｉｃＰａｒａｌｌｅｌＧｅｎｅｔｉｃＡｌｇｏｒｉｔｈｍ［Ｊ］．ＡｃｔａＰｈｙｓＳｉｎ，２０１１，６０（９）：１４８⁃１５７（ｉｎＣｈｉｎｅｓｅ） ［８］ ＢＯＵＫＥＲＣＨＥＡ，ＳＡＭＡＲＡＨＳ．ＡＮｏｖｅｌＡｌｇｏｒｉｔｈｍｆｏｒＭｉｎｉｎｇＡｓｓｏｃｉａｔｉｏｎＲｕｌｅｓｉｎＷｉｒｅｌｅｓｓＡｄＨｏｃＳｅｎｓｏｒＮｅｔｗｏｒｋｓ［Ｊ］．
ＩＥＥＥＴｒａｎｓｏｎＰａｒａｌｌｅｌ＆ＤｉｓｔｒｉｂｕｔｅｄＳｙｓｔｅｍｓ，２００８，１９（７）：８６５⁃８７７ ［９］ 栾鹏，朱江，高凯．低轨卫星接入系统中基于位置信息的时隙分配协议［Ｊ］．电讯技术，２０１６，５６（９）：９９０⁃９９４
ＬＵＡＮＰｅｎｇ，ＺＨＵＪｉａｎｇ，ＧＡＯＫａｉ．ＡＰｏｓｉｔｉｏｎ⁃ＢａｓｅｄＳｌｏｔＡｓｓｉｇｎｍｅｎｔＰｒｏｔｏｃｏｌｆｏｒＬＥＯＳａｔｅｌｌｉｔｅｓＮｅｔｗｏｒｋ［Ｊ］．Ｔｅｌｅｃｏｍｍｕｎｉｃａ⁃
ｔｉｏｎＥｎｇｉｎｅｅｒｉｎｇ，２０１６，５６（９）：９９０⁃９９４（ｉｎＣｈｉｎｅｓｅ） ＬＵＡＮＰｅｎｇ，ＺＨＵＪｉａｎｇ，ＧＡＯＫａｉ．ＡＰｏｓｉｔｉｏｎ⁃ＢａｓｅｄＳｌｏｔＡｓｓｉｇｎｍｅｎｔＰｒｏｔｏｃｏｌｆｏｒＬＥＯＳａｔｅｌｌｉｔｅｓＮｅｔｗｏｒｋ［Ｊ］．Ｔｅｌｅｃｏｍｍｕｎｉｃａ⁃
ｔｉｏｎＥｎｇｉｎｅｅｒｉｎｇ，２０１６，５６（９）：９９０⁃９９４（ｉｎＣｈｉｎｅｓｅ） ［１０］程楠，刘志敏，王继新．Ａｄｈｏｃ网络ＴＤＭＡ分布式动态时隙算法［Ｊ］．计算机应用研究，２００５，２２（１）：２２２⁃２２５
ＣＨＥＮＧＮａｎ，ＬＩＵＺｈｉｍｉｎ，ＷＡＮＧＪｉｘｉｎ．ＴＤＭＡ⁃ＢａｓｅｄＤｉｓｔｒｉｂｕｔｅｄＤｙｎａｍｉｃＳｌｏｔＡｌｇｏｒｉｔｈｍｆｏｒＡｄｈｏｃＮｅｔｗｏｒｋｓ［Ｊ］．Ａｐｐｌｉｃａｔｉｏｎ
ＲｅｓｅａｒｃｈｏｆＣｏｍｐｕｔｅｒｓ，２００５，２２（１）：２２２⁃２２５（ｉｎＣｈｉｎｅｓｅ） ＣＨＥＮＧＮａｎ，ＬＩＵＺｈｉｍｉｎ，ＷＡＮＧＪｉｘｉｎ．ＴＤＭＡ⁃ＢａｓｅｄＤｉｓｔｒｉｂｕｔｅｄＤｙｎａｍｉｃＳｌｏｔＡｌｇｏｒｉｔｈｍｆｏｒＡｄｈｏｃＮｅｔｗｏｒｋｓ［Ｊ］．Ａｐｐｌｉｃａｔｉｏｎ
ＲｅｓｅａｒｃｈｏｆＣｏｍｐｕｔｅｒｓ，２００５，２２（１）：２２２⁃２２５（ｉｎＣｈｉｎｅｓｅ） ［１１］闫少晨，马正新，石荣．基于位置信息的大区域移动自组网ＭＡＣ协议［Ｊ］．电视技术，２０１０，３４（增刊２）：６０⁃６２＋７２
ＹＡＮＳｈａｏｃｈｅｎ，ＭＡＺｈｅｎｇｘｉｎ，ＳＨＩＲｏｎｇ．ＡＰｏｓｉｔｉｏｎＢａｓｅｄＭＡＣＰｒｏｔｏｃｏｌｆｏｒＬａｒｇｅＳｃａｌｅＭｏｂｉｌｅＡｄｈｏｃＮｅｔｗｏｒｋｓ［Ｊ］．２０１０，
３４（ｓｕｐｐｌ２）：６０⁃６２＋７２（ｉｎＣｈｉｎｅｓｅ） ＹＡＮＳｈａｏｃｈｅｎ，ＭＡＺｈｅｎｇｘｉｎ，ＳＨＩＲｏｎｇ．ＡＰｏｓｉｔｉｏｎＢａｓｅｄＭＡＣＰｒｏｔｏｃｏｌｆｏｒＬａｒｇｅＳｃａｌｅＭｏｂｉｌｅＡｄｈｏｃＮｅｔｗｏｒｋｓ［Ｊ］．２０１０，
３４（ｓｕｐｐｌ２）：６０⁃６２＋７２（ｉｎＣｈｉｎｅｓｅ） ［１２］吴光宇，姚凯凌．一种广义干扰模型以及分布式ＡＢＳ时隙接入方法［Ｐ］．ＣＮ１０６７９２８０７Ａ，２０１７
ＷＵＧｕａｎｇｙｕ，ＹＡＯＫａｉｌｉｎｇ．ＡＧｅｎｅｒａｌＩｎｔｅｒｆｅｒｅｎｃｅＭｏｄｅｌａｎｄＤｉｓｔｒｉｂｕｔｅｄＡＢＳＳｌｏｔＡｃｃｅｓｓＭｅｔｈｏｄ［Ｐ］．Ｃｈｉｎａ，
ＣＮ１０６７９２８０７Ａ，２０１７（ｉｎＣｈｉｎｅｓｅ） 西 北 工 业 大 学 学 报 第３８卷 ４
５
１ ＡＬｉｎｋＤｉｓｔａｎｃｅＤｉｖｉｓｉｏｎＢａｓｅｄＴｉｍｅＤｉｖｉｓｉｏｎＭｕｌｔｉｐｌｅＡｃｃｅｓｓ
ＰｒｏｔｏｃｏｌｆｏｒＤｉｒｅｃｔｉｏｎａｌＡｅｒｏｎａｕｔｉｃａｌＲｅｌａｙＮｅｔｗｏｒｋｓ
ＹＡＮＺｈｏｎｇｊｉａｎｇ，ＬＩＱｉａｎｑｉａｎ，ＬＩＢｏ，ＹＡＮＧＭａｏ
（ＳｃｈｏｏｌｏｆＥｌｅｃｔｒｏｎｉｃｓａｎｄＩｎｆｏｒｍａｔｉｏｎ，ＮｏｒｔｈｗｅｓｔｅｒｎＰｏｌｙｔｅｃｈｎｉｃａｌＵｎｉｖｅｒｓｉｔｙ，Ｘｉ′ａｎ７１００７２，Ｃｈｉｎａ） Ａｂｓｔｒａｃｔ：Ｉｎｄｉｒｅｃｔｉｏｎａｌａｅｒｏｎａｕｔｉｃａｌｒｅｌａｙｎｅｔｗｏｒｋｓ，ｔｈｅａｉｒｐｌａｎｅｒｅｌａｙｅｘｐｌｏｒｅｓｔｈｅａｄｖａｎｔａｇｅｓｏｆｔｈｅｄｉｒｅｃｔｉｏｎａｌ
ａｎｔｅｎｎａ，ｉｎｔｅｒｍｓｏｆｌｏｎｇｔｒａｎｓｍｉｓｓｉｏｎｄｉｓｔａｎｃｅ，ｌｏｗｔｒａｎｓｍｉｓｓｉｏｎｐｏｗｅｒ，ｓｍａｌｌｉｎｔｅｒｆｅｒｅｎｃｅｒａｎｇｅａｎｄｓｏｏｎ，ｔｏ
ｈｅｌｐｔｈｅｇｒｏｕｎｄｎｏｄｅｓｔｒａｎｓｍｉｔｔｉｎｇｄａｔａ．Ｈｏｗｅｖｅｒ，ｄａｔａｔｒａｎｓｍｉｓｓｉｏｎｄｅｌａｙｅｘｔｅｎｓｉｏｎｐｒｏｂｌｅｍｏｃｃｕｒｓｗｈｅｎｔｈｅｄｉｓ⁃
ｔａｎｃｅｏｆｔｈｅｔｒａｎｓｍｉｓｓｉｏｎｌｉｎｋｅｘｔｅｎｄｓ，ｗｈｅｒｅｔｈｅｗｉｒｅｌｅｓｓｓｉｇｎａｌｔｒａｎｓｍｉｓｓｉｏｎｔｉｍｅａｐｐｒｏｘｉｍａｔｅｓｔｈｅｄａｔａｔｒａｎｓｍｉｓ⁃
ｓｉｏｎｔｉｍｅｓｕｃｈｔｈａｔｔｈｅｗｉｒｅｌｅｓｓｓｉｇｎａｌｔｒａｎｓｍｉｓｓｉｏｎｔｉｍｅｃａｎｎｏｔｂｅｏｍｉｔｔｅｄ．Ｔｏａｄｄｒｅｓｓｔｈｅｄａｔａｔｒａｎｓｍｉｓｓｉｏｎｄｅｌａｙ
ｅｘｔｅｎｓｉｏｎｐｒｏｂｌｅｍ，ａｌｉｎｋｄｉｓｔａｎｃｅｄｉｖｉｓｉｏｎｂａｓｅｄｔｉｍｅｄｉｖｉｓｉｏｎｍｕｌｔｉｐｌｅａｃｃｅｓｓｐｒｏｔｏｃｏｌ，ＬＤＤ⁃ＴＤＭＡ，ｉｓｐｒｏｐｏｓｅｄ
ｉｎｔｈｉｓｐａｐｅｒ．ＤｉｆｆｅｒｅｎｔｆｒｏｍｔｈｅｔｒａｄｉｔｉｏｎａｌＴＤＭＡｐｒｏｔｏｃｏｌ，ｗｈｅｒｅｔｈｅｔｉｍｅｓｌｏｔｓａｒｅｅｑｕａｌａｎｄａｒｅｄｅｔｅｒｍｉｎｅｄｂｙ
ｔｈｅｌｏｎｇｅｓｔｔｒａｎｓｍｉｓｓｉｏｎｌｉｎｋ，ｔｈｅｌｅｎｇｔｈｏｆｔｈｅｔｉｍｅｓｌｏｔｓａｒｅｄｉｆｆｅｒｅｎｔａｎｄａｒｅｄｅｔｅｒｍｉｎｅｄｂｙｄｉｆｆｅｒｅｎｔｔｒａｎｓｍｉｓｓｉｏｎ
ｌｉｎｋｓ．Ｆｕｒｔｈｅｒｍｏｒｅ，ｔｈｅｃｏｎｃｅｐｔｏｆｃｏｍｍｕｎｉｃａｔｉｏｎｃｏｖｅｒａｇｅｒｉｎｇｉｓｐｒｏｐｏｓｅｄｗｈｅｒｅｔｈｅｎｏｄｅｓｌｏｃａｔｅｄｉｎｔｈｅｓａｍｅ
ｒｉｎｇｃｏｍｍｕｎｉｃａｔｅｗｉｔｈｔｈｅｒｅｌａｙｕｔｉｌｉｚｉｎｇｔｈｅｓａｍｅｔｉｍｅｓｌｏｔｌｅｎｇｔｈ．Ｔｈｅｒｅｌａｔｉｏｎｓｈｉｐｓｂｅｔｗｅｅｎｔｈｅｎｕｍｂｅｒｏｆｒｉｎｇｓ，
ｔｈｅｒｉｎｇｒａｄｉｕｓａｎｄｔｈｅｔｒａｎｓｍｉｓｓｉｏｎｒａｎｇｅｏｆｔｈｅＬＤＤ⁃ＴＤＭＡａｒｅｍｏｄｅｌｌｅｄａｎｄｄｅｒｉｖｅｄａｓａｃｌｏｓｅｄｆｏｒｍｕｌａ，ｗｈｅｒｅ
ｔｈｅｒｉｎｇｒａｄｉｕｓａｒｅｏｐｔｉｍｉｚｅｄｓｕｃｈｔｈａｔｔｈｅｇａｉｎｉｓｍａｘｉｍｉｚｅｄ．ＳｉｍｕｌａｔｉｏｎｒｅｓｕｌｔｓｓｈｏｗｔｈａｔＬＤＤ⁃ＴＤＭＡｏｕｔｐｅｒｆｏｒｍｓ
ＴＤＭＡｂｙ１３．３７％ｗｈｅｎｔｈｅｔｒａｎｓｍｉｓｓｉｏｎｒａｎｇｅｉｓ２００ｋｍａｎｄｔｈｅｒｉｎｇｎｕｍｂｅｒｉｓ４． Ｋｅｙｗｏｒｄｓ：ｄｉｒｅｃｔｉｏｎａｌａｅｒｏｎａｕｔｉｃａｌｒｅｌａｙｎｅｔｗｏｒｋｓ；ｍｕｌｔｉｐｌｅａｃｃｅｓｓｐｒｏｔｏｃｏｌ；ｌｉｎｋｄｉｓｔａｎｃｅ；ｄａｔａｔｒａｎｓｍｉｓｓｉｏｎ
ｄｅｌａｙｅｘｔｅｎｓｉｏｎ；ｃｏｍｍｕｎｉｃａｔｉｏｎｒｉｎｇ ©２０１９ＪｏｕｒｎａｌｏｆＮｏｒｔｈｗｅｓｔｅｒｎＰｏｌｙｔｅｃｈｎｉｃａｌＵｎｉｖｅｒｓｉｔｙ． ©２０１９ＪｏｕｒｎａｌｏｆＮｏｒｔｈｗｅｓｔｅｒｎＰｏｌｙｔｅｃｈｎｉｃａｌＵｎｉｖｅｒｓｉｔｙ． ＴｈｉｓｉｓａｎＯｐｅｎＡｃｃｅｓｓａｒｔｉｃｌｅｄｉｓｔｒｉｂｕｔｅｄｕｎｄｅｒｔｈｅｔｅｒｍｓｏｆｔｈｅＣｒｅａｔｉｖｅＣｏｍｍｏｎｓＡｔｔｒｉｂｕｔｉｏｎＬｉｃｅｎｓｅ（ｈｔｔｐ：／／ｃｒｅａｔｉｖｅｃｏｍｍｏｎｓ．ｏｒｇ／ｌｉｃｅｎｓｅｓ／ｂｙ／４．０），ｗｈｉｃｈ
ｐｅｒｍｉｔｓｕｎｒｅｓｔｒｉｃｔｅｄｕｓｅ，ｄｉｓｔｒｉｂｕｔｉｏｎ，ａｎｄｒｅｐｒｏｄｕｃｔｉｏｎｉｎａｎｙｍｅｄｉｕｍ，ｐｒｏｖｉｄｅｄｔｈｅｏｒｉｇｉｎａｌｗｏｒｋｉｓｐｒｏｐｅｒｌｙｃｉｔｅｄ．
|
https://openalex.org/W2985661740
|
https://open.library.ubc.ca/media/download/pdf/52383/1.0385834/3
|
English
| null |
Determining a Threshold to Delimit the Amazonian Forests from the Tree Canopy Cover 2000 GFC Data
|
Sensors
| 2,019
|
cc-by
| 8,080
|
Received: 2 September 2019; Accepted: 13 November 2019; Published: 18 November 2019 Abstract: Open global forest cover data can be a critical component for Reducing Emissions from
Deforestation and Forest Degradation (REDD+) policies. In this work, we determine the best threshold,
compatible with the official Brazilian dataset, for establishing a forest mask cover within the Amazon
basin for the year 2000 using the Tree Canopy Cover 2000 GFC product. We compared forest cover
maps produced using several thresholds (10%, 30%, 50%, 80%, 85%, 90%, and 95%) with a forest
cover map for the same year from the Brazilian Amazon Deforestation Monitoring Project (PRODES)
data, produced by the National Institute for Space Research (INPE). We also compared the forest
cover classifications indicated by each of these maps to 2550 independently assessed Landsat pixels
for the year 2000, providing an accuracy assessment for each of these map products. We found that
thresholds of 80% and 85% best matched with the PRODES data. Consequently, we recommend using
an 80% threshold for the Tree Canopy Cover 2000 data for assessing forest cover in the Amazon basin. Keywords: forest mapping; Google Earth Engine; REDD+; remote sensing; forest degradation sensors sensors Sensors 2019, 19, 5020; doi:10.3390/s19225020 Determining a Threshold to Delimit the Amazonian
Forests from the Tree Canopy Cover 2000 GFC Data Kaio Allan Cruz Gasparini 1,†,*
, Celso Henrique Leite Silva Junior 1,†
,
Yosio Edemir Shimabukuro 1, Egidio Arai 1, Luiz Eduardo Oliveira Cruz e Aragão 1
,
Carlos Alberto Silva 2
and Peter L. Marshall 3 Kaio Allan Cruz Gasparini 1,†,*
, Celso Henrique Leite Silva Junior 1,†
, Kaio Allan Cruz Gasparini 1,†,*
, Celso Henrique Leite Silva Junior 1,†
,
Yosio Edemir Shimabukuro 1, Egidio Arai 1, Luiz Eduardo Oliveira Cruz e Aragão 1 Carlos Alberto Silva 2
and Peter L. Marshall 3 Carlos Alberto Silva 2
and Peter L. Marshall 3 1
Divisão de Sensoriamento Remoto, Instituto Nacional de Pesquisas Espaciais, São José dos Campos – SP,
Brazil; celso.junior@inpe.br (C.H.L.S.J.); yosio.shimabukuro@inpe.br (Y.E.S.); egidio.arai@inpe.br (E.A.);
luiz.aragao@inpe.br (L.E.O.C.eA.) 1
Divisão de Sensoriamento Remoto, Instituto Nacional de Pesquisas Espaciais, São José dos Campos – SP,
Brazil; celso.junior@inpe.br (C.H.L.S.J.); yosio.shimabukuro@inpe.br (Y.E.S.); egidio.arai@inpe.br (E.A.);
luiz.aragao@inpe.br (L.E.O.C.eA.) g
p
(
)
2
Department of Geographical Sciences, University of Maryland, College Park, Maryland, MD 20740, USA;
carlos_engflorestal@outlook.com g
3
Department of Forest Resources Management, The University of
Vancouver, BC V6T 1Z4, Canada; peter.marshall@ubc.ca g
3
Department of Forest Resources Management, The University of British Columbia, 2424 Main Mall,
Vancouver, BC V6T 1Z4, Canada; peter.marshall@ubc.ca *
Correspondence: kaioallangasparini@gmail.com †
These authors contributed equally to this work. sensors sensors 1. Introduction Tropical forests play an essential role in the carbon (C) cycle [1]. However, human actions through
deforestation and forest degradation can revert the natural tropical forest C sink. It is estimated that
deforestation is the second largest source of C emission into the atmosphere, emitting between 0.81
and 1.14 Pg C annually [2–5]. Degradation is another important driver of carbon emission within
tropical regions. In the Amazon, during drought years, C emissions from forest fires can exacerbate
old-growth forest deforestation emissions by more than half [6]. Brazil has been monitoring the extent of the Amazon rainforest since the late 1980s using
Landsat images available through the Brazilian Amazon Deforestation Monitoring Project (PRODES),
a project operated by the Brazilian National Institute of Space Research in collaboration with the
Ministry of the Environment and the Brazilian Institute of Environment and Natural Resources [7]. Since deforestation is the second largest CO2 emission driver [8], PRODES allowed Brazil to be a pioneer
in reducing emissions under the Reducing Emissions from Deforestation and Forest Degradation
(REDD+) policy [9]. For the correct monitoring of forest areas, there is a need to draw a baseline, Sensors 2019, 19, 5020; doi:10.3390/s19225020 www.mdpi.com/journal/sensors www.mdpi.com/journal/sensors 2 of 11 Sensors 2019, 19, 5020 or year of reference, to begin monitoring. Consequently, a map representing the forest area at the
beginning of forest monitoring is important [10]. With the baseline forest (a forest mask) established,
one can quantify new deforestation as well as degraded areas by identifying the affected areas on the
initial map. Recently, such forest monitoring initiatives have gained popularity [11,12], especially after
the emergence of remote sensing applications using cloud computing as a platform [13]. g
g pp
g
p
g
p
The Google Earth Engine platform has allowed studies on a planetary scale [14]. To date, the uses
of the Google Earth Engine have been diverse, such as mapping malaria risks [15], environmental
monitoring [13,16], and urban mapping [17,18]. Hansen et al. [13] were among the first researchers to
use the Google Earth Engine platform. They mapped forest changes worldwide from 2000 to 2013 using
Landsat imagery collection (30 m of spatial resolution). In the current version of the mapping engine
(v1.6), the authors processed images up to 2018 [19]. A highlight of their research was the construction
of a raster map of the 2000 baseline named ‘Tree Canopy Cover 2000 GFC Data’ (TreeCover2000). 1. Introduction TreeCover2000 represents tree cover for the year 2000 at a 30-meter spatial resolution. These data were
generated using multispectral satellite imagery from the Landsat 7 thematic mapper plus (ETM+)
sensor [13,20]. Cloud-free observations from over 600,000 images were analyzed using Google Earth
Engine to determine per-pixel tree cover using a supervised machine learning algorithm (regression
tree approach) [13,20]. The regression tree training data were obtained from Ikonos (4 m of spatial
resolution) and QuickBird (2.8 m of spatial resolution) image classifications, in which tree cover was
transformed to percent values [21,22]. The result was a map of percentage values of tree cover, ranging
from 0% to 100%, where 0% represented no tree cover and 100% represented maximum tree cover
for a pixel. These data require a user to choose a percentage threshold value to determine whether a
pixel is considered forest (i.e., cover value equal to or greater than the threshold). Several thresholds
have been empirically chosen in the literature to define forest cover with the TreeCover2000 data. For example: Grecchi et al. [23], Taubert et al. [24], Brinck et al. [25], and Esquivel-Muelbert et al. [26]
used 30%; Shimabukuro et al. [27] used 50%; Wagner et al. [28] used 80%; and Tyukavina et al. [29]
used 83%. However, as far as we know, no study has focused on testing the performance of these
thresholds in delimiting the Amazonian rainforests in order to match with other countries’ definition
of forest. Without a threshold standardization, comparisons among studies and with official statistics
become impossible. Bearing in mind that the Amazonian forest extends beyond Brazil’s borders and, apart from Brazil,
few other Amazonian countries have long-term, large-scale forest mapping projects in place, the use of
the TreeCover2000 dataset is essential for establishing baselines of intact forest areas. It is also valuable
to have measurements that are comparable across countries. However, it is not clear what threshold
would best delimit Amazonian forests. To answer this question, we compared the PRODES forest
cover map with maps produced using different thresholds of the TreeCover2000 data for an area within
the Brazilian portion of the Amazon basin. We also compared the classifications in all the maps to 2550
independently assessed Landsat pixels indicating forest coverage in the year 2000. 2.2. Datasets
2.2. Datasets The Tree Canopy Cover 2000 GFC data (TreeCover2000) were obtained from the Global Forest
Change website [19]. These data represent the percentage of all vegetation greater than 5 m in height,
not necessarily natural forest, with values ranging from 0% (no vegetation greater than 5 m in height)
to 100% (total cover) [13]. The global dataset is divided into 10-degree by 10-degree segments (tiles). Four such tiles were necessary to get complete coverage of Mato Grosso (00N_060W, 00N_070W,
10S_60S, and 10S_70S). The spatial resolution of the data is 0.00025 degrees, which is equivalent to 30
m in the Equator region
The Tree Canopy Cover 2000 GFC data (TreeCover2000) were obtained from the Global Forest
Change website [19]. These data represent the percentage of all vegetation greater than 5 m in height,
not necessarily natural forest, with values ranging from 0% (no vegetation greater than 5 m in height)
to 100% (total cover) [13]. The global dataset is divided into 10-degree by 10-degree segments (tiles). Four such tiles were necessary to get complete coverage of Mato Grosso (00N_060W, 00N_070W,
10S_60S, and 10S_70S). The spatial resolution of the data is 0.00025 degrees, which is equivalent to
30 m in the Equator region. m in the Equator region. The PRODES data were obtained from [34]. The data have a 1:250,000 mapping scale and
represent areas of old-growth forests mapped manually from radar and optical remote sensing data. We used the vector data from 2014. From that, we reconstructed the forest cover for the year 2000 by
converting all deforested polygons from the period 2001 to 2014 back to forest, with the remainder
labelled as non-forest. These data were then converted to a 30 m raster basis to be compatible with
T ee Ca o y Co e 2000 data
The PRODES data were obtained from [34]. The data have a 1:250,000 mapping scale and represent
areas of old-growth forests mapped manually from radar and optical remote sensing data. We used
the vector data from 2014. From that, we reconstructed the forest cover for the year 2000 by converting
all deforested polygons from the period 2001 to 2014 back to forest, with the remainder labelled as
non-forest. These data were then converted to a 30 m raster basis to be compatible with Tree Canopy
Cover 2000 data. Tree Canopy Cover 2000 data. 2.1. Study Area We used the Brazilian state of Mato Grosso (Figure 1) for this analysis due to its physiographic
heterogeneity. Mato Grosso includes flooded fields (Pantanal), natural fields and savannas (Cerrado),
and deciduous and evergreen forests (Amazonia) [30]. This region is also part of the “arc of
deforestation”, an area of intense deforestation that borders the Cerrado [31] and recently considered a
consolidated agricultural frontier [32]. Consequently, Mato Grosso is potentially the only place that
provides the full range of phytophysiognomies found in the Amazon basin as a whole. 3 of 11
3 f 12 Sensors 2019, 19, 5020 gure 1. Location of the state of Mato Grosso, representative biomes, and the spatial distribution of
e 10 km2 plots used as samples. The smaller map shows the extent of the Amazon basin [33]. gure 1. Location of the state of Mato Grosso, representative biomes, and the spatial distribution of
10 km2 plots used as samples. The smaller map shows the extent of the Amazon basin [33]. t Figure 1. Location of the state of Mato Grosso, representative biomes, and the spatial distribution of
the 10 km2 plots used as samples. The smaller map shows the extent of the Amazon basin [33]. Figure 1. Location of the state of Mato Grosso, representative biomes, and the spatial distribution of
the 10 km2 plots used as samples. The smaller map shows the extent of the Amazon basin [33]. Figure 1. Location of the state of Mato Grosso, representative biomes, and the spatial distribution of
the 10 km2 plots used as samples. The smaller map shows the extent of the Amazon basin [33]. Figure 1. Location of the state of Mato Grosso, representative biomes, and the spatial distribution of
the 10 km2 plots used as samples. The smaller map shows the extent of the Amazon basin [33]. 2.2. Datasets
2.2. Datasets An independent reference sample of forest and non-forest coverage in the study area was
obtained from Tyukavina et al. [29,35]. In their study, 10,000 Landsat (5, 7, and 8) 30 m pixels were
classified over the period from 1999 to 2013 across the entire Brazilian Amazon. Of these sample
pixels, 2550 were located in Mato Grosso. As this work is based on the year 2000, an external evaluator
assessed each of the 2550 pixels in Mato Grosso and labelled each as forest or non-forest based on the
An independent reference sample of forest and non-forest coverage in the study area was obtained
from Tyukavina et al. [29,35]. In their study, 10,000 Landsat (5, 7, and 8) 30 m pixels were classified over
the period from 1999 to 2013 across the entire Brazilian Amazon. Of these sample pixels, 2550 were
located in Mato Grosso. As this work is based on the year 2000, an external evaluator assessed each of
the 2550 pixels in Mato Grosso and labelled each as forest or non-forest based on the year 2000 data. year 2000 dat
2.3. Methods 2.3. Methods
About 9329 10 by 10 km plots (samples) were established across Mato Grosso (Figure 1). Of these
plots, 4994 (53%) were in the Amazon biome, 3703 (40%) were in the Cerrado biome, and 632 (7%)
were in the Pantanal biome. Initially, we calculated the forest cover percentage from the PRODES
data within each of the 9329 plots. Then, we calculated the forest cover proportion from the Tree
Canopy Cover 2000 maps, which used the following thresholds: (1) greater than 10%; (2) greater than
30%; (3) greater than 50%; (4) greater than 80%; (5) greater than 85%; (6) greater than 90%; and (7)
About 9329 10 by 10 km plots (samples) were established across Mato Grosso (Figure 1). Of these
plots, 4994 (53%) were in the Amazon biome, 3703 (40%) were in the Cerrado biome, and 632 (7%) were
in the Pantanal biome. Initially, we calculated the forest cover percentage from the PRODES data within
each of the 9329 plots. Then, we calculated the forest cover proportion from the Tree Canopy Cover
2000 maps, which used the following thresholds: (1) greater than 10%; (2) greater than 30%; (3) greater
than 50%; (4) greater than 80%; (5) greater than 85%; (6) greater than 90%; and (7) greater than 95%. Any pixel that exceeded these thresholds was considered as forest. We selected these thresholds based Sensors 2019, 19, 5020 4 of 11 on preliminary tests and the previous studies from the literature [23,24,26–29], representing a wide
range of thresholds. Hereafter, we refer to these thresholds as 10%, 30%, and so on. The proportions of forest cover within each of the 9329 plots defined by each threshold and
the PRODES forest cover were compared using linear regression, suggested by Shimabukuro et
al. [30]. This method is a robust and straightforward approach to validating remote sensing mapping. The coefficient of determination (R2), intercept, slope, and the root mean squared error (RMSE) were
used for comparison among the various regression equations to determine the best threshold to use. The ideal value for R2 was 1, 0 for the intercept, 1 for the slope, and 0 for the RMSE. We used the
bootstrap method for this analysis. In this approach, 10,000 interactions were performed, where
each interaction was randomly raffled 10% of the 9329 plots with replacement. Then, the means and
standard deviations of the 10,000 R2, intercept, slope, and the RMSE were calculated. year 2000 dat
2.3. Methods All analyses
were performed using the R statistical software package [36]. To calculate the forest proportion within
each of the 10 by 10 km plots, the “lsm_c_pland” function of the “landscapemetrics” package [37] was
used. Linear regressions were performed using the "ln" native function of the R software. Moreover, we built a confusion matrix that was used to tabulate the differences in classification
between the maps and the reference samples using the “Accuracy Assessment” tool [38], developed by
the Food and Agriculture Organization of the United Nations (FAO), from the 2550 Landsat sample
pixels (year 2000) provided by Tyukavina et al. [29]. As suggested by Olofsson et al. [39] and the
FAO [38], we determined a weighted matrix error using the ratio of the area of each class sampled over
the total area of each class (Equation (1)). This weighted error was then used to calculate the standard
error (Equation 2) and the confidence interval for the estimated area of each class, ˆpi,j =
Wi·ni,j
ni
(1)
S( ˆp) =
v
u
tX
i
Wi ˆ·pi,j −ˆp2
i,j
ni −1
(2) (1) (2) where ˆpi,j is the proportional area for each cell in the matrix, Wi corresponds to the class weights (the
proportional area of class i), ni,j is the sample count in cell i,j, ni is the total sample count in class i,
and S( ˆp) is the standard error of the area estimates assuming simple random sampling. where ˆpi,j is the proportional area for each cell in the matrix, Wi corresponds to the class weights (the
proportional area of class i), ni,j is the sample count in cell i,j, ni is the total sample count in class i,
and S( ˆp) is the standard error of the area estimates assuming simple random sampling. 3. Results Our analyses revealed that the lower thresholds (10%, 30%, and 50%) included forest cover in
areas covered by savanna vegetation within the Cerrado biome (see Figure 1) [40], in the south-central
portion of Mato Grosso (see Figures 2 and 3) [40]. Maps produced using the 80%, 85%, and 90%
thresholds showed greater visual similarity to the PRODES forest map (Figures 2 and 3). 5 of 11 Sensors 2019, 19, 5020 Figure 2. Spatial arrangement of each threshold assessed in the study compared to the Brazilian
Amazon Deforestation Monitoring Project (PRODES) data. Figure 2. Spatial arrangement of each threshold assessed in the study compared to the Brazilian
Amazon Deforestation Monitoring Project (PRODES) data. Figure 2. Spatial arrangement of each threshold assessed in the study compared to the Brazilian
Amazon Deforestation Monitoring Project (PRODES) data. Figure 2. Spatial arrangement of each threshold assessed in the study compared to the Brazilian
Amazon Deforestation Monitoring Project (PRODES) data. Figure 2. Spatial arrangement of each threshold assessed in the study compared to the Brazilian
Amazon Deforestation Monitoring Project (PRODES) data. Figure 2. Spatial arrangement of each threshold assessed in the study compared to the Brazilian
Amazon Deforestation Monitoring Project (PRODES) data. Figure 2. Spatial arrangement of each threshold assessed in the study compared to the Brazilian
Amazon Deforestation Monitoring Project (PRODES) data. Figure 3. Maps of differences between spatial arrangements of each threshold assessed in the study
compared to the PRODES data. Red (−1) represents pixels that were non-forest cover using PRODES
and forest cover using the Tree Canopy Cover 2000 data. White (0) represents pixels classified into
the same class using both datasets. Blue (1) represents pixels that were classified as forest cover using
PRODES and non-forest cover using Tree Canopy Cover 2000 data. Figure 2. Spatial arrangement of each threshold assessed in the study compared to the Brazilian
Amazon Deforestation Monitoring Project (PRODES) data. Figure 2. Spatial arrangement of each threshold assessed in the study compared to the Brazilian
Amazon Deforestation Monitoring Project (PRODES) data. Figure 3. Maps of differences between spatial arrangements of each threshold assessed in the study
compared to the PRODES data. Red (−1) represents pixels that were non-forest cover using PRODES
and forest cover using the Tree Canopy Cover 2000 data. White (0) represents pixels classified into
the same class using both datasets. 3. Results Blue (1) represents pixels that were classified as forest cover using
PRODES and non-forest cover using Tree Canopy Cover 2000 data. Figure 3. Maps of differences between spatial arrangements of each threshold assessed in the study
compared to the PRODES data. Red (−1) represents pixels that were non-forest cover using PRODES
and forest cover using the Tree Canopy Cover 2000 data. White (0) represents pixels classified into
the same class using both datasets. Blue (1) represents pixels that were classified as forest cover using
PRODES and non-forest cover using Tree Canopy Cover 2000 data. Figure 2. Spatial arrangement of each threshold assessed in the study compared to the Brazilian
Amazon Deforestation Monitoring Project (PRODES) data. Figure 2. Spatial arrangement of each threshold assessed in the study compared to the Brazilian
Amazon Deforestation Monitoring Project (PRODES) data. Figure 2. Spatial arrangement of each threshold assessed in the study compared to the Brazilian
Amazon Deforestation Monitoring Project (PRODES) data. Figure 2. Spatial arrangement of each threshold assessed in the study compared to the Brazilian
Amazon Deforestation Monitoring Project (PRODES) data. Figure 2. Spatial arrangement of each threshold assessed in the study compared to the Brazilian
Amazon Deforestation Monitoring Project (PRODES) data. Figure 2. Spatial arrangement of each threshold assessed in the study compared to the Brazilian
Amazon Deforestation Monitoring Project (PRODES) data. Figure 3. Maps of differences between spatial arrangements of each threshold assessed in the study
compared to the PRODES data. Red (−1) represents pixels that were non-forest cover using PRODES
and forest cover using the Tree Canopy Cover 2000 data. White (0) represents pixels classified into
the same class using both datasets. Blue (1) represents pixels that were classified as forest cover using
PRODES and non-forest cover using Tree Canopy Cover 2000 data. Figure 3. Maps of differences between spatial arrangements of each threshold assessed in the study
compared to the PRODES data. Red (−1) represents pixels that were non-forest cover using PRODES
and forest cover using the Tree Canopy Cover 2000 data. White (0) represents pixels classified into
the same class using both datasets. Blue (1) represents pixels that were classified as forest cover using
PRODES and non-forest cover using Tree Canopy Cover 2000 data. Figure 3. Maps of differences between spatial arrangements of each threshold assessed in the study
compared to the PRODES data. 3. Results Red (−1) represents pixels that were non-forest cover using PRODES
and forest cover using the Tree Canopy Cover 2000 data. White (0) represents pixels classified into
the same class using both datasets. Blue (1) represents pixels that were classified as forest cover using
PRODES and non-forest cover using Tree Canopy Cover 2000 data. Figure 3. Maps of differences between spatial arrangements of each threshold assessed in the study
compared to the PRODES data. Red (−1) represents pixels that were non-forest cover using PRODES
and forest cover using the Tree Canopy Cover 2000 data. White (0) represents pixels classified into
the same class using both datasets. Blue (1) represents pixels that were classified as forest cover using
PRODES and non-forest cover using Tree Canopy Cover 2000 data. Figure 3. Maps of differences between spatial arrangements of each threshold assessed in the study
compared to the PRODES data. Red (−1) represents pixels that were non-forest cover using PRODES
and forest cover using the Tree Canopy Cover 2000 data. White (0) represents pixels classified into
the same class using both datasets. Blue (1) represents pixels that were classified as forest cover using
PRODES and non-forest cover using Tree Canopy Cover 2000 data. Figure 3. Maps of differences between spatial arrangements of each threshold assessed in the study
compared to the PRODES data. Red (−1) represents pixels that were non-forest cover using PRODES
and forest cover using the Tree Canopy Cover 2000 data. White (0) represents pixels classified into
the same class using both datasets. Blue (1) represents pixels that were classified as forest cover using
PRODES and non-forest cover using Tree Canopy Cover 2000 data. 6 of 11 Sensors 2019, 19, 5020 As expected, the 95% threshold map indicated the smallest forest cover, about 37% of Mato Grosso,
and the 10% threshold map indicated the largest forest cover, about 66% of Mato Grosso (Table 1). The PRODES map indicated approximately 41% of forest cover within Mato Grosso in 2000, which
was most closely matched by the 90% threshold. Table 1. Forest cover assessment for the year 2000 in Mato Grosso, Brazil, according to the PRODES
map and maps produced using various thresholds with the Tree Canopy Cover 2000 data. Table 1. Forest cover assessment for the year 2000 in Mato Grosso, Brazil, according to the PRODES
map and maps produced using various thresholds with the Tree Canopy Cover 2000 data. 3. Results Forest cover assessment for the year 2000 in Mato Grosso, Brazil, according to the PRODES
maps produced using various thresholds with the Tree Canopy Cover 2000 data. Map
Forest (%)
Non-Forest (%)
PRODES map
41
59
10% Threshold
66
34
30% Threshold
63
37
50% Threshold
58
42
80% Threshold
49
51
85% Threshold
47
53
90% Threshold
40
60
95% Threshold
37
63 We found that forest cover maps produced from the 10%, 30%, and 50% thresholds tended to
overestimate forest cover based on the 9329 sample plots, as evidenced by the concentration of the
majority of the points above the 1:1 line (Figure 4). The maps produced using thresholds of 90% and
95% underestimated forest cover, as more points fell below the 1:1 line than above it. Finally, the forest
maps produced using the 80% and 85% thresholds showed a more even dispersion of points around
the regression line and the 1:1 line. Sensors 2019, 19, x FOR PEER REVIEW
7 of 12 Figure 4. Regression between the forest percentage within all of 9329 (10 by 10 km) samples cells using
different thresholds from the Tree Canopy Cover 2000 data and their corresponding percentages on
a reference map developed using the PRODES data. The dashed red line is the 1:1 line. The blue line
is the average regression line from 10,000 interactions for each threshold tested. All the 10,000 regressions were significant at a level of 1% (p-values less than 0.001). The R2
increased from an average of 0.612 ± 0.021 to 0.869 ± 0.013 for 10% and 85%, respectively. The intercept
Figure 4. Regression between the forest percentage within all of 9329 (10 by 10 km) samples cells using
different thresholds from the Tree Canopy Cover 2000 data and their corresponding percentages on a
reference map developed using the PRODES data. The dashed red line is the 1:1 line. The blue line is
the average regression line from 10,000 interactions for each threshold tested. Figure 4. Regression between the forest percentage within all of 9329 (10 by 10 km) samples cells using
different thresholds from the Tree Canopy Cover 2000 data and their corresponding percentages on
a reference map developed using the PRODES data. The dashed red line is the 1:1 line. The blue line
is the average regression line from 10,000 interactions for each threshold tested. 3. Results All the 10,000 regressions were significant at a level of 1% (p-values less than 0.001). The R2
Figure 4. Regression between the forest percentage within all of 9329 (10 by 10 km) samples cells using
different thresholds from the Tree Canopy Cover 2000 data and their corresponding percentages on a
reference map developed using the PRODES data. The dashed red line is the 1:1 line. The blue line is
the average regression line from 10,000 interactions for each threshold tested. Figure 4. Regression between the forest percentage within all of 9329 (10 by 10 km) samples cells using
different thresholds from the Tree Canopy Cover 2000 data and their corresponding percentages on
a reference map developed using the PRODES data. The dashed red line is the 1:1 line. The blue line
is the average regression line from 10,000 interactions for each threshold tested. All the 10,000 regressions were significant at a level of 1% (p-values less than 0.001). The R2
d f
f
f
d
l
h
Figure 4. Regression between the forest percentage within all of 9329 (10 by 10 km) samples cells using
different thresholds from the Tree Canopy Cover 2000 data and their corresponding percentages on a
reference map developed using the PRODES data. The dashed red line is the 1:1 line. The blue line is
the average regression line from 10,000 interactions for each threshold tested. 7 of 11 Sensors 2019, 19, 5020 All the 10,000 regressions were significant at a level of 1% (p-values less than 0.001). The R2
increased from an average of 0.612 ± 0.021 to 0.869 ± 0.013 for 10% and 85%, respectively. The intercept
coefficient values decreased from an average of 41.560 ± 1.058 for the 10% threshold to an average of
0.978 ± 0.473 for the 95% threshold. At the same time, the slope coefficient values increased from an
average of 0.591 ± 0.013 for the 10% threshold to an average of 0.888 ± 0.009 for the 85% threshold. Finally, the RMSE had a lower average of 13.060 ± 0.644 for the 85% threshold, and a higher average of
17.910 ± 0.498 for the 10% threshold. The accuracy assessment results showed that the poorest match (lowest overall accuracy) was
obtained with the 95% threshold (71%) and the best match (highest overall accuracy) was obtained
with the 50% threshold (87%) (Table 2). 3. Results The PRODES data had the third lowest match (74%), albeit
slightly higher than the 90% threshold. The 80% threshold best matched with the PRODES data in
terms of the estimated cover of non-forest and forest in Mato Grosso. Table 2. Result from 10,000 bootstrap interactions. Intercept, slope, coefficient of determination (R2),
root mean squared error (RMSE), and p-values of the linear regressions are used for comparing the
tested thresholds to the PRODES data. SD is the standard deviation. Threshold
R2 (± SD)
Intercept (± SD)
Slope (± SD)
RMSE (± SD)
p-value (± SD)
10%
0.612 ± 0.021
41.560 ± 1.058
0.591 ± 0.013
17.910 ± 0.498
0 ± 0
30%
0.669 ± 0.020
35.870 ± 1.029
0.655 ± 0.013
17.520 ± 0.542
0 ± 0
50%
0.752 ± 0.019
26.630 ± 0.953
0.755 ± 0.012
16.500 ± 0.620
0 ± 0
80%
0.860 ± 0.014
13.290 ± 0.693
0.874 ± 0.009
13.380 ± 0.646
0 ± 0
85%
0.869 ± 0.013
10.351 ± 0.693
0.888 ± 0.009
13.060 ± 0.644
0 ± 0
90%
0.857 ± 0.013
4.120 ± 0.513
0.872 ± 0.010
13.540 ± 0.550
0 ± 0
95%
0.852 ± 0.013
0.978 ± 0.473
0.879 ± 0.010
13.960 ± 0.010
0 ± 0
Not surprisingly, the highest user’s accuracy (85%) for non-forest cover was obtained with the 10%
threshold and declined with increasing threshold size (Table 3). The highest user’s accuracy (99%) for
forest cover was found with the 95% threshold and declined with decreasing thresholds. The pattern
for the producer’s accuracy was the opposite of that of the user’s accuracy. The highest producer’s
accuracy for non-forest cover (99%) was found for the 95% threshold and the highest producer’s
accuracy for forest cover (94%) was found for the 10% threshold. The thresholds that best resembled
those obtained by the PRODES data were 85% (user’s accuracy) and 50% (producer’s accuracy). Table 2. Result from 10,000 bootstrap interactions. Intercept, slope, coefficient of determination (R2),
root mean squared error (RMSE), and p-values of the linear regressions are used for comparing the
tested thresholds to the PRODES data. SD is the standard deviation. Table 2. Result from 10,000 bootstrap interactions. Intercept, slope, coefficient of determination (R2),
root mean squared error (RMSE), and p-values of the linear regressions are used for comparing the
tested thresholds to the PRODES data. SD is the standard deviation. Table 2. Result from 10,000 bootstrap interactions. 3. Results Intercept, slope, coefficient of determination (R2),
root mean squared error (RMSE), and p-values of the linear regressions are used for comparing the
tested thresholds to the PRODES data. SD is the standard deviation. Threshold
R2 (± SD)
Intercept (± SD)
Slope (± SD)
RMSE (± SD)
p-value (± SD)
10%
0.612 ± 0.021
41.560 ± 1.058
0.591 ± 0.013
17.910 ± 0.498
0 ± 0
30%
0.669 ± 0.020
35.870 ± 1.029
0.655 ± 0.013
17.520 ± 0.542
0 ± 0
50%
0.752 ± 0.019
26.630 ± 0.953
0.755 ± 0.012
16.500 ± 0.620
0 ± 0
80%
0.860 ± 0.014
13.290 ± 0.693
0.874 ± 0.009
13.380 ± 0.646
0 ± 0
85%
0.869 ± 0.013
10.351 ± 0.693
0.888 ± 0.009
13.060 ± 0.644
0 ± 0
90%
0.857 ± 0.013
4.120 ± 0.513
0.872 ± 0.010
13.540 ± 0.550
0 ± 0
95%
0.852 ± 0.013
0.978 ± 0.473
0.879 ± 0.010
13.960 ± 0.010
0 ± 0 Not surprisingly, the highest user’s accuracy (85%) for non-forest cover was obtained with the 10%
threshold and declined with increasing threshold size (Table 3). The highest user’s accuracy (99%) for
forest cover was found with the 95% threshold and declined with decreasing thresholds. The pattern
for the producer’s accuracy was the opposite of that of the user’s accuracy. The highest producer’s
accuracy for non-forest cover (99%) was found for the 95% threshold and the highest producer’s
accuracy for forest cover (94%) was found for the 10% threshold. The thresholds that best resembled
those obtained by the PRODES data were 85% (user’s accuracy) and 50% (producer’s accuracy). 8 of 11 Sensors 2019, 19, 5020 nce interval for each class (forest and non-forest) at each threshold compared to the PRODES data. Table 3. Error matrix, accuracy, and confidence interval for each class (forest and non-forest) at each threshold compared to the PRODES data. Table 3. Error matrix, accuracy, and confidence interval for each class (forest and non-forest) at each thres Table 3. Error matrix, accuracy, and confidence interval for each class (forest and non-forest) at each threshold compared to the PRODES data. 3. Results Predictions
Class
Observed
Estimated Area
(%)
Estimated Area ± 95%
Confidence (km2)
User’s Accuracy
(%)
Producer’s
Accuracy (%)
Overall
Accuracy (%)
Non-Forest
Forest
Non-forest
544
96
39.3
352,159 ± 12,849
85
64
10%
Forest
300
1601
60.7
543,914 ± 12,849
84
94
84
30%
Non-forest
600
118
39.4
353,006 ± 12,606
84
71
86
Forest
244
1579
60.6
543,067 ± 12,606
87
93
Non-forest
688
186
38.6
346,277 ± 12,568
79
82
50%
Forest
156
1511
61.4
549,796 ± 12,568
91
74
87
80%
Non-forest
787
389
36.0
322,227 ± 13,082
67
93
82
Forest
57
1308
64.0
573,846 ± 13,082
96
77
Non-forest
803
461
35.2
315,401 ± 13,256
64
95
85%
Forest
41
1236
64.8
580,672 ± 13,256
97
73
80
90%
Non-forest
821
651
34.3
307,281 ± 13,985
56
97
73
Forest
23
1046
65.7
588,792 ± 13,985
98
62
Non-forest
832
737
33.6
302,454 ± 14,090
53
99
95%
Forest
12
960
66.4
593,618 ± 14,090
99
57
71
PRODES
Non-forest
748
572
36.6
327,194 ± 15,134
57
89
74
Forest
96
1125
63.4
567.323 ± 15,134
92
66 Sensors 2019, 19, 5020 9 of 11 4. Discussion and Conclusions Funding: This study was financed in part by the Coordenação de Aperfeiçoamento de Pessoal de Nível Superior
- Brasil (CAPES) - Finance Code 001. L.E.O.C.A. thanks Conselho Nacional de Desenvolvimento Científico e
Tecnológico (CNPq) - Processo 305054/2016-3. 95/5000. Finally, the authors thank FAPESP (process 2016/19806-3)
for funding part of the research. Acknowledgments: K.A.C.G. thanks the support of the Government of Canada, by the ELAP scholarship. Finally,
we also thank the editor and reviewers whose helpful comments and suggestions helped to improve and clarify
this manuscript. Conflicts of Interest: The authors declare no conflicts of interest. 4. Discussion and Conclusions We compared Amazonian forest cover using Tree Canopy Cover 2000 and the PRODES data within
a portion of the Amazon basin in Brazil. By creating several forest maps from applying thresholds to the
Tree Canopy Cover 2000 data, we showed that there are relevant differences in the quantity and spatial
distribution of forest cover depending on the threshold chosen. For instance, we found that thresholds
of 80% and 85% for the Tree Canopy Cover 2000 data best matched the PRODES forest cover data
across Mato Grosso State in Brazil. For example, the lower thresholds evaluated—between 10% and
50%—overestimated forest cover, especially within savanna formations, where trees are sparser within
the landscape. Thus, the choice of less conservative thresholds may inflate the quantification of forest
disturbances, such as fires [27], deforestation [13], selective logging [23], forest fragmentation [24–26],
and, consequently, carbon emissions accounting. We emphasize that the choice of a threshold level for the Tree Canopy Cover 2000 data will depend
on the nature of the application. If the change in forest cover through time is being assessed, then,
obviously, consistency through time among threshold levels is more important than the threshold level
ultimately chosen for the assessment. However, in general, based on our analyses, we recommend
using an 80% threshold with the Tree Canopy Cover 2000 data for assessing forest cover in the
Amazon basin. The approach we presented may help countries currently involved in REDD+ in the Amazon
region. Forest reference levels can be based on Tree Canopy Cover 2000 data using our suggested
threshold (UNFCCC, 2010) [41]. Presently, Brazil, Colombia, Ecuador, Guyana, Peru, and Suriname are
members of REDD+ [42]. A standardization of methods across countries would be beneficial for the
evaluation of the effectiveness of countries’ efforts to mitigate emissions from land cover change. Author Contributions: K.A.C.G., C.H.L.S.J., and L.E.O.C.eA. designed the research. K.A.C.G., C.H.L.S.J., and
E.A. processed the data. K.A.C.G., C.H.L.S.J, Y.E.S., P.L.M., and L.E.O.C.A. analyzed the results. K.A.C.G. and
P.L.M. wrote the manuscript with inputs from C.H.L.S.J., C.A.S., and Y.E.S. Funding: This study was financed in part by the Coordenação de Aperfeiçoamento de Pessoal de Nível Superior
- Brasil (CAPES) - Finance Code 001. L.E.O.C.A. thanks Conselho Nacional de Desenvolvimento Científico e
Tecnológico (CNPq) - Processo 305054/2016-3. 95/5000. Finally, the authors thank FAPESP (process 2016/19806-3)
for funding part of the research. 5.
Baccini, A.; Goetz, S.J.; Walker, W.S.; Laporte, N.T.; Sun, M.; Sulla-Menashe, D.; Hackler, J.; Beck, P.S.A.;
Dubayah, R.; Friedl, M.A.; et al. Estimated carbon dioxide emissions from tropical deforestation improved
by carbon-density maps. Nat. Clim. Chang. 2012, 2, 182–185. [CrossRef] References Aragão, L.E.O.C.; Poulter, B.; Barlow, J.B.; Anderson, L.O.; Malhi, Y.; Saatchi, S.; Phillips, O.L.; Gloor, E. Environmental change and the carbon balance of Amazonian forests. Biol. Rev. 2014, 89, 913–931. [CrossRef] 8. Aragão, L.E.O.C.; Poulter, B.; Barlow, J.B.; Anderson, L.O.; Malhi, Y.; Saatchi, S.; Phillips, O.L.; Gloor, E. Environmental change and the carbon balance of Amazonian forests. Biol. Rev. 2014, 89, 913–931. [CrossRef]
9. Hargita, Y.;
Günter, S.;
Köthke, M. Brazil submitted the first REDD+ reference level to the
UNFCCC-Implications regarding climate effectiveness and cost-efficiency. Land Use Policy 2015, 55, 340–347. [CrossRef] 9. Hargita, Y.;
Günter, S.;
Köthke, M. Brazil submitted the first REDD+ reference level to the
UNFCCC-Implications regarding climate effectiveness and cost-efficiency. Land Use Policy 2015, 55, 340–347. [CrossRef] 10. Shimabukuro, Y.E.; Arai, E.; Duarte, V.; Jorge, A.; dos Santos, E.G.; Gasparini, K.A.C.; Dutra, A.C. Monitoring
deforestation and forest degradation using multi-temporal fraction images derived from Landsat sensor
data in the Brazilian Amazon. Int. J. Remote Sens. 2019, 40, 5475–5496. [CrossRef] 11. Potapov, P.V.; Turubanova, S.A.; Hansen, M.C.; Adusei, B.; Broich, M.; Altstatt, A.; Mane, L.; Justice, C.O. Quantifying forest cover loss in Democratic Republic of the Congo, 2000–2010, with Landsat ETM+ data. Remote Sens. Environ. 2012, 122, 106–116. [CrossRef] 2. Hansen, M.C.; Stehman, S.V.; Potapov, P.V. Quantification of global gross forest cover loss. Proc. Natl. A
Sci. USA 2010, 107, 8650–8655. [CrossRef] [PubMed] 13. Hansen, M.C.; Potapov, P.V.; Moore, R.; Hancher, M.; Turubanova, S.A.; Tyukavina, A.; Thau, D.; Stehman, S.V.;
Goetz, S.J.; Loveland, T.R.; et al. High-Resolution Global Maps of 21st-Century Forest Cover Change. Science
2013, 342, 850–853. [CrossRef] [PubMed] 14. Gorelick, N.; Hancher, M.; Dixon, M.; Ilyushchenko, S.; Thau, D.; Moore, R. Google Earth Engine:
Planetary-scale geospatial analysis for everyone. Remote Sens. Environ. 2017, 202, 18–27. [CrossRef] 15. Sturrock, H.J.; Cohen, J.M.; Keil, P.; Tatem, A.J.; Le Menach, A.; Ntshalintshali, N.E.; Hsiang, M.S.; Gosling, R.D. Fine-scale malaria risk mapping from routine aggregated case data. Malar. J. 2014, 13, 421. [CrossRef]
[PubMed] 16. Pekel, J.-F.; Cottam, A.; Gorelick, N.; Belward, A.S. High-resolution mapping of global surface water and its
long-term changes. Nature 2016, 540, 418–422. [CrossRef] [PubMed] 17. Patel, N.N.; Angiuli, E.; Gamba, P.; Gaughan, A.; Lisini, G.; Stevens, F.R.; Tatem, A.J.; Trianni, G. Multitemporal
settlement and population mapping from Landsat using Google Earth Engine. Int. J. Appl. Earth Obs. Geoinf. 2015, 35, 199–208. [CrossRef] 18. Zhang, Q.; Li, B.; Thau, D.; Moore, R. References 1. Le Quéré, C.; Raupach, M.R.; Canadell, J.G.; Marland, G.; Le Quéré, C.; Raupach, M.R.; Canadell, J.G.;
Marland, G.; Bopp, L.; Ciais, P.; et al. Trends in the sources and sinks of carbon dioxide. Nat. Geosci. 2009,
2, 831–836. [CrossRef] 1. Le Quéré, C.; Raupach, M.R.; Canadell, J.G.; Marland, G.; Le Quéré, C.; Raupach, M.R.; Canadell, J.G.;
Marland, G.; Bopp, L.; Ciais, P.; et al. Trends in the sources and sinks of carbon dioxide. Nat. Geosci. 2009,
2, 831–836. [CrossRef] 2. van der Werf, G.R.; Morton, D.C.; DeFries, R.S.; Olivier, J.G.J.; Kasibhatla, P.S.; Jackson, R.B.; Coll
Randerson, J.T. CO2 emissions from forest loss. Nat. Geosci. 2009, 2, 737–738. [CrossRef] 3. Houghton, R.A.; House, J.I.; Pongratz, J.; van der Werf, G.R.; DeFries, R.S.; Hansen, M.C.; Le Quéré, C.;
Ramankutty, N. Carbon emissions from land use and land-cover change. Biogeosciences 2012, 9, 5125–5142. [CrossRef] 3. Houghton, R.A.; House, J.I.; Pongratz, J.; van der Werf, G.R.; DeFries, R.S.; Hansen, M.C.; Le Quéré, C.;
Ramankutty, N. Carbon emissions from land use and land-cover change. Biogeosciences 2012, 9, 5125–5142. [CrossRef] 4. Harris, N.L.; Brown, S.; Hagen, S.C.; Saatchi, S.S.; Petrova, S.; Salas, W.; Hansen, M.C.; Potapov, P.V.; Lotsch, A. Baseline Map of Carbon Emissions from Deforestation in Tropical Regions. Science 2012, 336, 1573–1576. [CrossRef] [PubMed] 4. Harris, N.L.; Brown, S.; Hagen, S.C.; Saatchi, S.S.; Petrova, S.; Salas, W.; Hansen, M.C.; Potapov, P.V.; Lotsch, A. Baseline Map of Carbon Emissions from Deforestation in Tropical Regions. Science 2012, 336, 1573–1576. [CrossRef] [PubMed] 5. Baccini, A.; Goetz, S.J.; Walker, W.S.; Laporte, N.T.; Sun, M.; Sulla-Menashe, D.; Hackler, J.; Beck, P.S.A.;
Dubayah, R.; Friedl, M.A.; et al. Estimated carbon dioxide emissions from tropical deforestation improved
by carbon-density maps. Nat. Clim. Chang. 2012, 2, 182–185. [CrossRef] 5. Baccini, A.; Goetz, S.J.; Walker, W.S.; Laporte, N.T.; Sun, M.; Sulla-Menashe, D.; Hackler, J.; Beck, P.S.A.;
Dubayah, R.; Friedl, M.A.; et al. Estimated carbon dioxide emissions from tropical deforestation improved
by carbon-density maps. Nat. Clim. Chang. 2012, 2, 182–185. [CrossRef] Sensors 2019, 19, 5020 10 of 11 10 of 11 6. Aragão, L.E.O.C.; Anderson, L.O.; Fonseca, M.G.; Rosan, T.M.; Vedovato, L.B.; Wagner, F.H.; Silva, C.V.J.;
Silva Junior, C.H.L.; Arai, E.; Aguiar, A.P.; et al. 21st Century drought-related fires counteract the decline of
Amazon deforestation carbon emissions. Nat. Commun. 2018, 9, 536. [CrossRef] 7. Monitoramento do Desmatamento da Floresta Amazônica Brasileira por Satélite. Available online: h
//www.obt.inpe.br/OBT/assuntos/programas/amazonia/prodes (accessed on 1 January 2019). 8. References [CrossRef] [PubMed] 11 of 11 Sensors 2019, 19, 5020 26. Esquivel-Muelbert, A.; Bennett, A.C.; Sullivan, M.J.P.; Baker, J.C.A.; Gavish, Y.; Johnson, M.O.; Wang, Y.;
Chambers-Ostler, A.; Lisli Giannichi, M.; Gomes, L.; et al. A Spatial and Temporal Risk Assessment of the
Impacts of El Niño on the Tropical Forest Carbon Cycle: Theoretical Framework, Scenarios, and Implications. Atmosphere 2019, 10, 588. [CrossRef] 27. Shimabukuro, Y.E.; Arai, E.; Anderson, L.O.; de Aragão, L.E.O.eC.; Duarte, V. Mapping Degraded Forest
Areas Caused By Fires During the Year 2010 in Mato Grosso State, Brazilian Legal Amazon Using Landsat-5
Tm Fraction Images. Rev. Bras. Cartogr. 2017, 69, 23–32. g
g
28. Wagner, F.H.; Hérault, B.; Rossi, V.; Hilker, T.; Maeda, E.E.; Sanchez, A.; Lyapustin, A.I.; Galvão, L.S.; Wang, Y.;
Aragão, L.E.O.C. Climate drivers of the Amazon forest greening. PLoS ONE 2017, 12, e0180932. [CrossRef] g
,
;
,
;
,
;
,
;
,
;
,
; y p
,
;
,
;
g,
;
Aragão, L.E.O.C. Climate drivers of the Amazon forest greening. PLoS ONE 2017, 12, e0180932. [CrossRef]
29. Tyukavina, A.; Hansen, M.C.; Potapov, P.V.; Stehman, S.V.; Smith-Rodriguez, K.; Okpa, C.; Aguilar, R. Types
d
t
f f
t di t
b
i
B
ili
L
l A
2000 2013 S i Ad
2017 3 1 16 [C
R f] 29. Tyukavina, A.; Hansen, M.C.; Potapov, P.V.; Stehman, S.V.; Smith-Rodriguez, K.; Okpa, C.; Aguilar, R. Types
and rates of forest disturbance in Brazilian Legal Amazon, 2000–2013. Sci. Adv. 2017, 3, 1–16. [CrossRef] 30. Shimabukuro, Y.E.; Miettinen, J.; Beuchle, R.; Grecchi, R.C.; Simonetti, D.; Achard, F. Estimating Burned Area
in Mato Grosso, Brazil, Using an Object-Based Classification Method on a Systematic Sample of Medium
Resolution Satellite Images. IEEE J. Sel. Top. Appl. Earth Obs. Remote Sens. 2015, 8, 4502–4508. [CrossRef] 31. Monitoramento da cobertura florestal da Amazônia por satélites: Sistemas PRODES, DETER, DEGRAD e
Queimadas 2007–2008. Available online: https://www.google.com.hk/url?sa=t&rct=j&q=&esrc=s&source=
web&cd=8&ved=2ahUKEwisou2mjenlAhVmG6YKHYfqC_kQFjAHegQIAxAC&url=http%3A%2F%
2Fwww.obt.inpe.br%2Fprodes%2FRelatorio_Prodes2008.pdf&usg=AOvVaw2mYp1hD-ekxUCilvrALCRp
(accessed on 14 September 2019). 32. Kalamandeen, M.; Gloor, E.; Mitchard, E.; Quincey, D.; Ziv, G.; Spracklen, D.; Spracklen, B.; Adami, M.;
Aragão, L.E.O.C.; Galbraith, D. Pervasive Rise of Small-scale Deforestation in Amazonia. Sci. Rep. 2018,
8, 1600. [CrossRef] 33. Eva, H.; Huber, O.; Achard, F.; Balslev, H.; Beck, S.; Behling, H.; Belward, A.; Beuchle, R.; Cleef, A.;
Colchester, M.; et al. A Proposal for Defining the Geographical Boundaries of Amazonia; Publications Office:
Luxembourg, Belgium, 2005; ISBN 92-79-00012-8. 34. References Building a better Urban picture: Combining day and night remote
sensing imagery. Remote Sens. 2015, 7, 11887–11913. [CrossRef] 19. GFC—Global Forest Change. Available online: https://earthenginepartners.appspot.com/science-2013-
global-forest/download_v1.6.html (accessed on 1 January 2019). 20. GFW—Global Forest Watch Tree Cover 2000. Available online: http://data.globalforestwatch.org/datasets/
tree-cover-2000 (accessed on 1 January 2019). 21. Hansen, M.C.; DeFries, R.S.; Townshend, J.R.G.; Carroll, M.; Dimiceli, C.; Sohlberg, R.A. Global Percent
Tree Cover at a Spatial Resolution of 500 Meters: First Results of the MODIS Vegetation Continuous Fields
Algorithm. Earth Interact. 2003, 7, 1–15. [CrossRef] 22. Hansen, M.C.; Egorov, A.; Roy, D.P.; Potapov, P.; Ju, J.; Turubanova, S.; Kommareddy, I.; Loveland, T.R. Continuous fields of land cover for the conterminous United States using Landsat data: First results from the
Web-Enabled Landsat Data (WELD) project. Remote Sens. Lett. 2011, 2, 279–288. [CrossRef] 23. Grecchi, R.C.; Beuchle, R.; Shimabukuro, Y.E.; Aragão, L.E.O.C.; Arai, E.; Simonetti, D.; Achard, F. An integrated remote sensing and GIS approach for monitoring areas affected by selective logging: A case
study in northern Mato Grosso, Brazilian Amazon. Int. J. Appl. Earth Obs. Geoinf. 2017, 61, 70–80. [CrossRef] 23. Grecchi, R.C.; Beuchle, R.; Shimabukuro, Y.E.; Aragão, L.E.O.C.; Arai, E.; Simonetti, D.; Achard, F. An integrated remote sensing and GIS approach for monitoring areas affected by selective logging: A case
study in northern Mato Grosso, Brazilian Amazon. Int. J. Appl. Earth Obs. Geoinf. 2017, 61, 70–80. [CrossRef]
24. Taubert, F.; Fischer, R.; Groeneveld, J.; Lehmann, S.; Müller, M.S.; Rödig, E.; Wiegand, T.; Huth, A. Global
f
l f
f
N
554
1
[C
R f] [P bM d] An integrated remote sensing and GIS approach for monitoring areas affected by selective logging: A case
study in northern Mato Grosso, Brazilian Amazon. Int. J. Appl. Earth Obs. Geoinf. 2017, 61, 70–80. [CrossRef]
24. Taubert, F.; Fischer, R.; Groeneveld, J.; Lehmann, S.; Müller, M.S.; Rödig, E.; Wiegand, T.; Huth, A. Global
patterns of tropical forest fragmentation. Nature 2018, 554, 519–522. [CrossRef] [PubMed] y
pp
f
24. Taubert, F.; Fischer, R.; Groeneveld, J.; Lehmann, S.; Müller, M.S.; Rödig, E.; Wiegand, T.; Huth, A. Global
patterns of tropical forest fragmentation. Nature 2018, 554, 519–522. [CrossRef] [PubMed] 25. Brinck, K.; Fischer, R.; Groeneveld, J.; Lehmann, S.; Dantas De Paula, M.; Pütz, S.; Sexton, J.O.; Song, D.;
Huth, A. High resolution analysis of tropical forest fragmentation and its impact on the global carbon cycle. Nat. Commun. 2017, 8, 14855. References Projeto PRODES Monitoramento da Floresta Amazônica Brasileira por Satélite. Available online: http:
//terrabrasilis.dpi.inpe.br/downloads (accessed on 1 January 2019). 5. Sample interpretation results, Brazilian Legal Amazon. Available online: https://glad.umd.edu/brazil/in
php (accessed on 1 January 2019). p p (
y
)
36. R Core Team. R: A Language and Environment for Statistical Computing; R Foundation for Statistical Computing:
Vienna, Austria, 2019. 37. Hesselbarth, M.H.K.; Sciaini, M.; With, K.A.; Wiegand, K.; Nowosad, J. Landscapemetrics: An open-source R
tool to calculate landscape metrics. Ecography 2019, 42, 1648–1657. [CrossRef] 38. FAO. Map Accuracy Assessment and Area Estimation Map Accuracy Assessment and Area Estimation: A Practical
Guide; FAO: Roma, Italy, 2016; p. 69. 39. Olofsson, P.; Foody, G.M.; Herold, M.; Stehman, S.V.; Woodcock, C.E.; Wulder, M.A. Good practices for
estimating area and assessing accuracy of land change. Remote Sens. Environ. 2014, 148, 42–57. [CrossRef] 40. Picoli, M.C.A.; Camara, G.; Sanches, I.; Simões, R.; Carvalho, A.; Maciel, A.; Coutinho, A.; Esquerdo, J.;
Antunes, J.; Begotti, R.A.; et al. Big earth observation time series analysis for monitoring Brazilian agriculture. ISPRS J. Photogramm. Remote Sens. 2018, 145, 328–339. [CrossRef] 41. UNFCCC. FCCC/CP/2009/11/Add.1, Decision 4/CP.15: Methodological Guidance for Activities Relating to Reducing
Emissions from Deforestation and Forest Degradation and the Role of Conservation, Sustainable Management of
Forests and Enhancement of Forest Carbon Stock; UN: New York, NY, USA, 2010. 42. Reducing Emission From Deforestation And Forest Degradation in Develping Countries - Web Platform. Available online: https://redd.unfccc.int/submissions.html (accessed on 1 January 2019). © 2019 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access
article distributed under the terms and conditions of the Creative Commons Attribution
(CC BY) license (http://creativecommons.org/licenses/by/4.0/).
|
https://openalex.org/W2297164983
|
http://www.scielo.br/pdf/codas/v27n6/en_2317-1782-codas-27-06-00520.pdf
|
English
| null |
Avaliação dos aspectos temporais em cantores populares
|
CoDAS
| 2,015
|
cc-by
| 4,724
|
Study carried out at the Special Coordination of the Course of Speech-Language Pathology and Audiology,
Universidade Federal de Santa Catarina – UFSC – Florianópolis (SC), Brazil.
(1) Course of Speech-Language Pathology and Audiology, Universidade Federal de Santa Catarina – UFSC –
Florianópolis (SC), Brazil.
(2) Special Coordination of the Course of Speech-Language Pathology and Audiology, Universidade Federal
de Santa Catarina – UFSC – Florianópolis (SC), Brazil.
Conflict of interests: nothing to declare. Avaliação dos aspectos
temporais em cantores populares Ana Cláudia Mondini Ribeiro1
Renata Coelho Scharlach2
Maria Madalena Canina Pinheiro2 Ana Cláudia Mondini Ribeiro1
Renata Coelho Scharlach2
Maria Madalena Canina Pinheiro2 ABSTRACT Purpose: To evaluate the temporal processing of popular singers who do or do not play a musical instrument. Methods: The study population comprised 30 popular band singers. Of them, 15 play a musical instrument
(G1) and 15 do them do not play a musical instrument (G2). All of them were submitted to basic audiological
evaluation and temporal processing tests: test of frequency standard (TFS) and gaps in noise (GIN) detection. Results: Significant differences were observed in performance in the comparison between the groups with regard
to the temporal acuity threshold and percentage of correct responses in the GIN, as well as the performance
in the TFS. The results of the group of singers who play a musical instrument were found to be better than
those of the group that only sings. Conclusion: Popular singers that play musical instruments have a better
performance in resolution and temporal ordering auditory skills than singers who do not play an instrument. Auditory Perception
Music
Hearing Tests
Hearing
Singing DOI: 10.1590/2317-1782/20152014234 DOI: 10.1590/2317-1782/20152014234 Assessment of temporal aspects in popular singers Original Article Original Article
Artigo Original METHODS The project of this study was submitted, reviewed, and
approved by the Research Ethics Committee for Research
with Human Beings of the institution where it was carried out,
under protocol no. 711.422. All individuals invited to partici
pate signed an informed consent authorizing their voluntary
participation before the start of the assessments. Central auditory processing (CAP) is the domain from
which the central nervous system uses auditory information(2)
and includes auditory mechanisms that underlie the abilities
of sound location and lateralization, auditory discrimination,
recognition of auditory patterns, temporal aspects of hearing,
including temporal integration, temporal resolution, temporal
ordering, temporal masking, auditory performance with compet
ing acoustic signals, and auditory performance with degraded
acoustic signals(1,3). The study included 30 popular singers in ballroom bands
from the city of Florianópolis, SC-Brazil, who do and do
not play musical instruments, males and females, without
age restrictions. Participants were divided into two groups:
the G1 group, composed of 15 singers who also play musi
cal instruments, and the G2 group, composed of 15 singers
who only sing. Temporal processing refers to the processing of acoustic
stimuli over time, necessary for the ability to understand speech
in quiet and noisy environments, as well as speech stimuli and
other background sounds that vary over time. It can also be
considered as the basis for auditory processing, since many
characteristics of auditory information are somehow influ
enced by time(4). Inclusion criteria listed for the study population were
the following: absence of middle ear diseases; no evidence
of cognitive and neurological disorders; auditory thresholds
within normal limits; logoaudiometry with result greater than
or equal to 88%; type A tympanometric curve bilaterally and
contralateral stapedial reflexes present at normal levels; having
Brazilian Portuguese as a first language; acting only as a singer
in a ballroom band from the city of Florianópolis, SC-Brazil;
and being literate. Temporal processing is divided into two main components:
temporal ordering and temporal resolution. Temporal ordering
refers to the processing of two or more auditory stimuli in their
order of occurrence over time(5). Temporal resolution is defined
as the shortest time necessary for the central nervous system
to discriminate two acoustic stimuli(6). All participants underwent a battery of procedures consist
ing of anamnesis, basic audiologic evaluation and testing for
the evaluation of the temporal processing, described below. METHODS Music is the art of combining sounds either simultaneously
or successively with order, balance, and proportion within a
period. Studying musical perception allows the detailed anal
ysis of sound and can profoundly influence the formation of
the musician or other speaking and listening professionals, as
it acts as a qualitative process of its development(7). Speech-language/audiological anamnesis consisted of ques
tions whose purpose was to know individual musical aspects of
each participant, as well as sociodemographic characteristics,
such as gender and age. Each participant individually filled an
anamnesis instrument. After that, there was a basic audiological evaluation com
prised of meatoscopy, pure tone audiometry, speech audiometry,
and immittance, to prove the normality of hearing. Studies conducted with musicians suggest that daily musi
cal training, used by professional musicians, can functionally
induce the reorganization of the cerebral cortex. Thus, the con
tact with music before the age of seven could contribute to the
development of the CAP and, more precisely, of the temporal
processing(8,9). Another study conducted with professional and
amateur singers, signing in and out of key, showed the supe
riority of musicians who have received proper music theory
training over the years in the temporal ordering tests compared
to those who did not receive any theoretical training in music,
which showed that exposure to music theory is a contributing
factor to the auditory temporal processing(10). The temporal processing evaluation was performed using
the gaps in noise (GIN) test(12), in order to assess the temporal
resolution ability, and through the frequency pattern test (FPT)
(13,14),to assess the temporal ordering ability. The GIN test consists of the presentation of six sec
onds of white noise segments, with a five-second interval. Inserted in white noise stimuli, there are gaps in different
time positions and which can last 2, 3, 4, 5, 6, 10, 12, 15, and
20 ms. In the noise segments, there may be 1, 2, 3, or no gap(12). This test was applied at 50 dBSL, based on the tritone aver
age of 500, 1000, and 2000 Hz frequencies, as identified in the
basic audiological evaluation, and participants were instructed
to press the audiometer response button each time a gap inserted
into the noise was detected. RESUMO Percepção Auditiva
Música
Testes Auditivos
Audição
Canto Objetivo: Avaliar o processamento temporal de cantores populares que tocam ou não instrumento musical. Métodos: O estudo foi composto por 30 cantores populares de bandas baile,15 dos quais cantam e tocam
instrumento(s) musical(is) (G1) e 15 apenas cantam (G2). Todos os participantes foram submetidos à realização
da avaliação audiológica básica e dos testes do processamento temporal: teste de padrão de frequência (TPF)
e teste de detecção de gaps no ruído (GIN). Resultados: Houve diferença estatisticamente significante na
comparação do desempenho entre os grupos no que se refere ao limiar de acuidade temporal e percentual
de acertos do GIN, bem como no desempenho do TPF, sendo os resultados do grupo de cantores que tocam
instrumento musical melhores do que os obtidos pelo grupo que só canta. Conclusão: Cantores populares
que tocam instrumentos musicais apresentam melhor desempenho nas habilidades auditivas de resolução e
ordenação temporal quando comparados àqueles que só cantam. CoDAS 2015;27(6):520-5
Correspondence address:
Maria Madalena Canina Pinheiro
Rua Eurico Hosterno, 204, Santa Mônica,
Florianópolis (SC), Brasil, CEP-88035-400. E-mail: madapinheiro@hotmail.com
Received: 12/26/2014
Accepted: 04/23/2015 Correspondence address:
Maria Madalena Canina Pinheiro
Rua Eurico Hosterno, 204, Santa Mônica,
Florianópolis (SC), Brasil, CEP-88035-400. E-mail: madapinheiro@hotmail.com CoDAS 2015;27(6):520-5 521 Temporal aspects in popular singers INTRODUCTION due to a shortage of studies associating the practice of playing
musical instruments and temporal skills and the influence of
music in the performance of temporal skills. Many neurophysiological and cognitive mechanisms and
processes are needed for a perfect decoding, perception, rec
ognition, and interpretation of the audio signal. Thus, the act
of “hearing” does not refer simply to the mere detection of the
acoustic signal. Hearing plays an essential role in the proper
recognition and discrimination of auditory events, from the
simplest, such as nonverbal stimuli, to the more complex mes
sages, such as speech and language(1). Thus, this study aimed to evaluate the temporal processing
of popular singers who do and do not play a musical instrument. METHODS Distribution of sociodemographic data of subjects according
to their respective groups: G1 and G2 to t e
espect e g oups G
a d G
Group
Sex
Mean
Age (years)
n
Mean
Minimum
Maximum
G1
F
33.3 (%)
5
31.93
19
50
M
66.6 (%) 10
G2
F
60.0 (%)
9
29.86
19
55
M
40.0 (%)
6
Caption: G1 = Group 1; G2 = Group 2; Min = minimum; Max = maximum;
F = female; M = male 10
9
8
7
6
5
4
3
2
1
0
Acoustic guitar
Electric guitar
Piano
Bass guitar
Drums
Percussion instruments
Flute
; 10
9
8
7
6
5
4
3
2
1
0
Acoustic guitar
Electric guitar
Piano
Bass guitar
Drums
Percussion instruments
Flute To perform the meatoscopy, the otoscope used was the mini
3000 model, by HEINE. The pure tone and speech audiometry
were performed using the AC40 two-channel audiometer by
Interacoustics and TDH39 earphones. For the impedance tests,
the Interacoustic AT235h imitanciometer was used. The GIN
and FPT tests were also conducted through the Interacoustic
AC40 two-channel audiometer. The stimuli used were recorded
on CD, which required insertion in a computer connected to
the audiometer for the application of the tests. The analysis of variance parametric test was used for the
statistical analysis. The significance level was set at 0.05 (5%). METHODS The test was applied to each ear
separately and the lowest threshold sensed four times or more
was considered the temporal acuity threshold obtained by the Other studies with musicians who played a musical instru
ment and non-musicians showed that those with the greatest
musical aptitude show superiority in temporal processing when
compared to non-musicians(9,11). There is a theoretical gap in the literature in relation to stud
ies covering musicians and temporal processing. Knowing the
temporal processing of singers contributes to science mainly CoDAS 2015;27(6):520-5 522 Ribeiro ACM, Scharlach RC, Pinheiro MMC participant. It is noteworthy that thresholds lower or equal to
5 ms were considered within the normal range(15). percentages of gap recognition than singers who do not play
any instrument (G2). As for the FPT test, the comparison between the perfor
mances of G1 and G2 is shown in Table 4 and Figure 4. The FPT consists of sequences of three pure tones, two with
the same frequency and the third with a different frequency. The version used in this study(14) consists in the presentation of
tones of 880 Hz (low frequency) and 1122 Hz (high frequency),
with intervals of 150 and 200 ms between them. The tones are
presented in groups of three, with six possible sequences, as
follows: HHL, GAG LHL, GAA LHH, AGA HLH, ABB
HLL, AAG HHL, where A refers to the highest stimula
tion, the 1122 Hz tone, and G refers to the lowest, the 800
Hz tone. The proposal is to present 30 binaural sequences
in an intensity level of 50 dBSL. The test was performed
with an intensity of 50 dBNS, based on the tritone average
of frequencies of 500, 1000, and 2000 Hz, as identified in the
basic audiological evaluation, and participants were instructed
to orally repeat the sequence heard. The test was presented
to both ears simultaneously. As for the percentage of correct
answers, were considered within normal limits those that were
greater than or equal to 76%(16). There was a statistically significant difference when
comparing the groups, and once again, G1 had the best per
formance in the recognition of tones in different frequen
cies than G2. Table 1. Distribution of sociodemographic data of subjects according
to their respective groups: G1 and G2 Table 1. CoDAS 2015;27(6):520-5 DISCUSSION Description of the percentage of gap recognition in the gap in
noise test, obtained by ear and by study group Groups
Mean
(%)
Median
(%)
SD
(%)
CV
(%)
Min
(%)
Max
(%)
n
CI
(%)
p-value
Gin % – RE
G1
75.5
75
4.4
5.8
70
85
15
2.2
0.002*
G2
67.5
68
8.3
12.2
55
80
15
4.2
Gin % – LE
G1
75.9
75
5.5
7.2
66
88
15
2.8
0.001*
G2
67.4
68
6.6
9.8
56
78
15
3.3
Caption: G1 = Group 1; G2 = Group 2; GIN% = percentage of correct responses in
the GIN test; SD = standard deviation; RE = right ear; LE = left ear; CV = coefficient
of variation; Min = minimum; Max = maximum; CI = confidence interval; GIN =
gap in noise test; *ANOVA test sensory encoding of the temporal information, such as dura
tion, interval, and order of different stimulus patterns, is the
clue that guides temporal processing, being important for the
perception of music(18). The musical practice stimulates the development of melodic
and harmonic auditory perception through the perceptive train
ing of intervals, rhythm, and other acoustic parameters(19). In the GIN test, the singers in G1 obtained a lower tempo
ral acuity threshold (Table 1), as well as a higher percentage
of gap recognition (Table 2) than individuals in G2, mean
ing that popular singers who play musical instruments have
better temporal resolution ability when compared to those
who just sing. 76%
74%
72%
70%
68%
66%
64%
62%
RE
LE
G1
G2
75.00%
75.00%
67.40%
68.50%
Caption: RE = right ear; LE = left ear; G1 = Group 1; G2 = Group 2; GIN = gap
in noise test
Figure 3. Average values of gap recognition percentages obtained in
the gap in noise test, by ear and study group 76%
74%
72%
70%
68%
66%
64%
62%
RE
LE
G1
G2
75.00%
75.00%
67.40%
68.50% Mishra, Panda and Herbert(20)compared the auditory ability
of temporal resolution in musicians and non-musicians through
the gap detection threshold test. The authors found that the group
of musicians performed better, that is, had lower temporal acu
ity thresholds. Thus, musicians showed greater sensitivity to
detect the presence of gaps between two markers, confirming
the findings in this study. DISCUSSION frequency pattern test
Groups
Mean
(%)
Median
(%)
Standard
Deviation
(%)
CV
(%)
Min
(%)
Max
(%) n
CI
(%) p-value
FPT –
Binaural
G1
95.7
100
8.2
8.6
73
100 15 4.1
0.028*
G2
82.3
90
20.8
25.3
44
100 15 10.6
*ANOVA test
Caption: G1 = Group 1; G2 = Group 2; FPT – binaural = frequency pattern test
in both ears simultaneously; CV = coefficient of variation; Min = minimum; Max =
maximum; CI = confidence interval It is known that the auditory temporal processing is defined
as the perception of sound or modification of a sound within a
restricted period of time. In musical practice, singers need to fol
low musical melodies with harmony and represent them through
the voice, and a good performance of the temporal resolution
ability is fundamental to make it happen(10). Furthermore, the 6
5
4
3
1
0
RE
LE
G1
G2
2
4.47
5.20
4.40
5.53
Caption: RE = right ear; LE = left ear; G1 = Group 1; G2 = Group 2; GIN = gap in noise test
Figure 2. Average values of temporal acuity threshold, in milliseconds,
obtained through the gap in noise test, by ear and study group 6
5
4
3
1
0
RE
LE
G1
G2
2
4.47
5.20
4.40
5.53 Caption: RE = right ear; LE = left ear; G1 = Group 1; G2 = Group 2
Figure 4. Average values of the percentage of correct responses
obtained in the frequency pattern test, by study group
100%
95%
90%
85%
80%
75%
RE and LE
G1
G2
95.70%
82.30% Caption: RE = right ear; LE = left ear; G1 = Group 1; G2 = Group 2; GIN = gap in noise test
Figure 2. Average values of temporal acuity threshold, in milliseconds,
obtained through the gap in noise test, by ear and study group Caption: RE = right ear; LE = left ear; G1 = Group 1; G2 = Group 2
Figure 4. Average values of the percentage of correct responses
obtained in the frequency pattern test, by study group Table 3. RESULTS Comparison between the performances of G1 and G2 in the
frequency pattern test Table 4. Comparison between the performances of G1 and G2 in the
frequency pattern test RESULTS Table 1 shows the distribution of the participants’ sociode
mographic data according to their respective groups: G1 and G2. Figure 1 outlines the types of musical instruments played
by participants who constituted the G1. The guitar was the
instrument with the higher incidence, used by 80% participants. The time of professional activity for participants in G1
varies from 2 to 30 years, while in G2, this time ranged from
1 to 30 years. Figure 1. Distribution in absolute values of the types of musical
instruments played by participants who formed Group 1 Figure 1. Distribution in absolute values of the types of musical
instruments played by participants who formed Group 1 The results of temporal acuity thresholds in the GIN test,
obtained by ear and by group studied, are shown in Table 2 and
Figure 2. In the right ear, there was a trend toward statistical
significance (p=0.055) for individuals in G1, because they had
a lower temporal acuity threshold than individuals in G2. For
the left ear, the results were statistically significant (p=0.02),
that is, the singers who play musical instruments had a lower
temporal acuity threshold in the GIN test, which means they
have a better temporal resolution ability compared to those
who just sing. Table 2.Description of the gap in noise test regarding the temporal
acuity threshold in milliseconds (ms) obtained by ear and by study group
Groups
Mean Median SD
CV
Min Max n
CI
p-value
GIN TAT- RE
G1
4.47
4
0.52
12
4
5
15 0.26
0.055
G2
5.20
5
1.32
25
4
8
15 0.67
GIN TAT- LE
G1
4.40
4
0.63
14
3
5
15 0.32
0.002*
G2
5.53
5
1.13
20
4
8
15 0.57
*ANOVA test
Caption: G1 = Group 1; G2 = Group 2; GIN TAT = temporal acuity threshold;
RE = right ear; LE = left ear; CV = coefficient of variation; Min = minimum; Max =
maximum; CI = confidence interval; GIN = gap in noise test The percentage performance of gap recognition in the GIN
test was also evaluated by ear and study group. As can be seen
in Table 3 and Figure 3, there were statistically significant dif
ferences between the two groups in both ears, which means
singers who play a musical instrument (G1) showed better Temporal aspects in popular singers 523 Table 4. REFERENCES 1. Bellis TJ. Interpretation of central auditory assessment results. In: Bellis
TJ. Assessment and management of central auditory processing disorders
in the education setting: form science to practice. 2nd ed. San Diego:
Singular Publisching Group; 2003. p. 267-478. 2. ASHA: American Speech and Hearing Association [Internet].Central
Auditory Processing Disorders, 2005. [Acesso em 28/10/2013]. Disponível
em: http://www.asha.org/docs/html/tr2005-‘00043.html 3. Chermack GD, Musiek FE. Central auditory processing: New perspectives. San Diego: Singular Publishing; 1997. The gap recognition percentage in G1 approaches that
obtained in a study with normal-hearing adults(15), while in
G2 is lower, different from the comparison carried out with
another study, carried out with teenagers(28), in which the per
centage of correct responses is close to that obtained by the
two groups in this study. 4. Samelli AG, Schochat E. Processamento auditivo resolução temporal
e teste de detecção de Gap: revisão de literatura. Rev CEFAC. 2008;
10(3):369-77. 5. Balen SA, Massignani R, Schillo R. Aplicabilidade do software fast forword
na reabilitação dos distúrbios do processamento auditivo: resultados
iniciais. Rev CEFAC. 2008;10(4):572-87. 5. Balen SA, Massignani R, Schillo R. Aplicabilidade do software fast forword
na reabilitação dos distúrbios do processamento auditivo: resultados
iniciais. Rev CEFAC. 2008;10(4):572-87. 6. Fortes AB, Pereira LD, Azevedo MF. Resolução temporal: análise em
pré-escolares nascidos a termo e pré-termo. Pró-Fono R Atual Cient. 2007;19(1):90-6. 6. Fortes AB, Pereira LD, Azevedo MF. Resolução temporal: análise em
pré-escolares nascidos a termo e pré-termo. Pró-Fono R Atual Cient. 2007;19(1):90-6. Just as in the GIN test, in the FPT, the group of popular
singers who play musical instruments performed better and
thus showed a better temporal ordering ability than those who
only sing (Table 3). These findings corroborate those obtained
in previous studies(9,10). 7. Rodrigues FM. Sistema online de música e percepção, uma proposta de
auxílio à educação musical à distância: aprendizagem significativa e a
percepção musical [Trabalho de conclusão de curso]. Brasília: Universidade
de Brasília; 2008. A study compared the temporal ordering ability among
violinists and a control group using the FPT and found that
the performance of the group of musicians was better, with a
statistically significant difference(9). 8. Ohnishi T, Matsuda H, Asada T, Aruga M, Hirakata M, Nishikawa M,
et al. Functional anatomy of musical perception in musicians. Cerebr
Cortex. 2001;1(8):754-60. 9. Nascimento FM, Monteiro RAM, Soares CD, Ferreira MID. Temporal
Sequencing Abilities in Musicians Violinists and Non-Musicians. Arq. Intl Arch Otorhinolaryngology. 2010;14(2):217-24. DISCUSSION In contrast, in another study(10), professional and amateur
singers, who sing in and out of key, had the temporal resolution
auditory skills assessed by the RGDT test and statistically sig
nificant differences were detected between the groups formed. The authors also concluded that singing out of key was not asso
ciated with lack of efficiency in temporal resolution, contrary
to the conclusion reached in this study, which demonstrated the Caption: RE = right ear; LE = left ear; G1 = Group 1; G2 = Group 2; GIN = gap
in noise test Figure 3. Average values of gap recognition percentages obtained in
the gap in noise test, by ear and study group CoDAS 2015;27(6):520-5 Ribeiro ACM, Scharlach RC, Pinheiro MMC 524 fact that playing musical instruments contributes to the good
performance in the temporal resolution ability of musicians. correct responses found by the authors mentioned is quite close
to that found in this study, and the values for the groups com
posed of singers who do not play instruments was identical to
those for amateur singers. Other authors claim that musical training, as well as the
musical experiences, greatly increase the cortical and subcor
tical processing of sounds, which may have a direct influence
on the superior temporal perception obtained by musicians that
was also found in this study(21,22). The findings of this study have shown that auditory training
(referred to here as the practice of playing musical instruments)
contributes significantly to the performance of the temporal
processing abilities. Thus, playing musical instruments might
be used as a practical possibility for the rehabilitation of skills
involved in temporal processing. By analyzing Table 2, it was found that the average of the
temporal acuity thresholds obtained from the group consist
ing of singers that do not play any instrument is outside the
normal range in both the left and the right ear. Results outside
of the normal range were detected in five singers who do not
play instruments. The GIN test was sensitive to detect prob
lems in the temporal resolution ability of singers. In other
studies evaluating this skill through the RGDT for singers and
musicians, this did not occur(10,23). The comparison between
GIN and RGDT was not the subject of this study. However, it
was noted in the literature that other authors(24–26),in doing this
comparison, detected GIN to be the most sensitive to assess
the aforementioned ability. CONCLUSION Singers who play a musical instrument perform better on
tests that evaluate temporal processing than singers who do not
play any instruments. In this way, auditory abilities of temporal
resolution and ordering are more developed in popular singers
who play musical instruments. The standardization of the GIN test started in 2005, in order
to verify the diagnostic value of the test as a tool to assess the
temporal resolution ability. The first studies(27) assessed normal
and neurological patients in whom it was found that the tem
poral resolution ability was affected by injuries in the (central)
auditory nervous system. *ACMR was responsible for data collecting and preparation of the article;
RCS and MMCP were in charge for drafting and preparation of the article. In the control group, comprised of patients with neuro
logical disorders, detection thresholds found were 4.8 ms for
the left ear and 4.9 ms for the right ear. In Brazil, Samelli, in
2005(15) studied the test in normal-hearing young adults and
found that the overall average for the temporal acuity thresholds
was 3.98 ms bilaterally. In 2008(24), the average of the thresh
olds in adults with normal hearing were found, with 5.38 ms
for the right ear and 4.88 ms for the left ear, which explains
the results outside the normal ranges obtained by singers who
do not play musical instruments. DISCUSSION We suggest further research addressing the temporal aspects
of musicians playing different musical instruments to assess
whether there is an instrument that has greater influence on
the development of temporal ordering and resolution skills. CoDAS 2015;27(6):520-5 *ACMR was responsible for data collecting and preparation of the article;
RCS and MMCP were in charge for drafting and preparation of the article. REFERENCES In another study(10), the FPT was applied to professional
and amateur singers singing in and out of key, and statisti
cally significant differences were also detected regarding the
performance of the groups in that test, and the group of pro
fessional singers had the best performance (96.5%), followed
by the groups of amateur singers (82.3%). The percentage of 10. Ishii C, Arashiro PM, Pereira LD. Ordenação e resolução temporal em
cantores profissionais e amadores afinados e desafinados. Pró-Fono R
Atual Cient. 2006;18(3):285-92. 11. Rammsayer T, Altenmüller E. Temporal Information Processing in
musicians and non-musicians. Music Perception. 2006;24(1):37-48. CoDAS 2015;27(6):520-5 Temporal aspects in popular singers 525 12. Musiek FE, Zaidan EP, Baran JA, Shinn JB, Jirsa RE. Assessing temporal
process adult with LD: the GIN test. In: Convention of American Academy
of Audiology. Salt Lake City: AAA; 2004. p.203. 21. Habibi A, Wirantana V, Starr A. Cortical Activity during Perception
of Musical Rhythm; Comparing Musicians and Non-musicians. Psychomusicology. 2014;24(2):125-35. 13. Pinheiro M, Ptacek PH. Pattern reversal in auditory perception. J Acoust
Soc Am. 1971;49(2):493. 22. Mishra SK, Panda MR. Experience-dependent learning of auditory temporal
resolution: evidence from Carnatic-trained musicians. Neuroreport. 2014;25(2):134-7. 14. Musiek, F. Frequency (pitch) and duration pattern tests. J Am Acad Audiol. 1994;5:265-8. 23. Ascari MF, Deconto J. Avaliação do processamento auditivo temporal:
habilidade de resolução temporal em músicos. In: Anais da SIEPE – Semana
de Integração Ensino, Pesquisa e Extensão. Irati; 2009. 15. Samelli AG. O teste GIN (Gap in Noise): limiares de detecção de gap em
adultos com audição normal [doutorado]. São Paulo: Programa de Pós-
graduação em Fisiopatologia Experimental - Faculdade de Medicina da
Universidade de São Paulo; 2005. 24. Zaidan E, Garcia AP, Tedesco MLF, Baran JA. Desempenho de adultos
jovens normais em dois testes de resolução temporal. Pró-Fono R Atual
Cient. 2008;20(1):19-24. 16. Corazza MCA. Avaliação do Processamento Auditivo Central em Adultos:
Tese de Padrões Tonais Auditivos de Frequência e Teste de Padrões
Tonais Auditivos de Duração [doutorado]. São Paulo: Curso de Curso de
Pós-graduação em Distúrbios da Comunicação Humana - Universidade
Federal de São Paulo; 1998. 25. Amaral MIR, Martins PMF, Santos MFC. Temporal resolution:
assessment procedures and parameters for school-aged childrens. Braz J
Otorhinolaryngol. 2013;79(3):317-324. 26. Iliadou V, Bamiou DE, Chermak GD, Nimatoudis L. Comparison of two
tests of auditory temporal resolution in children with central auditory
processing disorder, adults with psychosis, and adult professional musicians. Int J Audiol. 2014;53:507-13. 17. REFERENCES Shinn JB, Chermak GD, Musiek FE. GIN (Gaps-In-Noise) Performance
in the pediatric population. J Am Acad Audiol. 2009;20:229-38. 18. Mulsow J, Reichmuth C. Electrophysiological assessment of temporal
resolution in pinnipeds. Aquatic Mammals. 2007;33(1):122-31. 27. Musiek FE, Shinn JB, Jirsa R, Bamiou DE, Baran, JA, Zaidan E. GIN
(Gaps-In-Noise) Test performance in subjects with confirmed central
auditory nervous system involvement. Ear Hear. 2005;26(6):608-18. 19. Costa MJ, Soncini F. Efeito da prática musical no reconhecimento da fala
no silêncio e no ruído. Pro-fono R Atual Cient. 2006;18(2):161-70. 20. Mishra SK, Panda MR, Herbert C. Enhanced auditory temporal gap detection
in listeners with musical training. J Acoust Soc Am. 2014;136(2):EL173-8. 28. Perez AP, Pereira LD. O teste gap in noise em crianças de 11 e 12 anos. Pró-Fono R Atual Cient. 2010;22(1):7-12. CoDAS 2015;27(6):520-5
|
Subsets and Splits
No community queries yet
The top public SQL queries from the community will appear here once available.